In order fly FPV quadcopters we strap one or two monitors to our head with all the associated electronics and in some case batteries. This is quite a bit of weight so having goggles that are comfortable to wear for a long period of time is important. Straps are a key part of this and because most of the material investment for goggles (rightly) goes into the screens and receiver, the strap is typically neglected and are often inconvenient, uncomfortable or both. I knew when I purchased my DJI FPV goggles that the included strap was poor and that fitting and aftermarket strap was made difficult thanks to the closed loop clips on the DJI model which is why I was so pleased to hear the release of the iFlight fpv goggle headstrap that included new open loop clips especially to fit DJI goggles and at a reasonable price. To be clear this means that they include everything needed to fit to the dji goggles without cutting or trimming anything

A closer look at what is included
For $12-$15 retail the following items are included:

• The strap itself with adjustment buckle and battery strap
• 2 x open loop clips to replace the closed loop clips on the DJI goggles.

Pictures are probably more useful than words so here you are - click to embiggen

Fitting to the DJI goggles
Fitment takes less than 2 minutes, all you need is a 1.5mm hex driver which I'm fairly certain will be in your toolkit:

So clearly very easy to fit to the DJI goggles. Easy to fit to any other goggle that accepts a 25mm wide strap too, just refer to your own goggle instructions.

How do they perform?
Good. The main strap being 40mm wide relieves a lot of pressure on your head, even for long sessions with a heavy goggle. Strap is easy to adjust with one hand while on your head and tends to stay put. If you like having a goggle mounted pack, this one is good for the twin 18650 liion style pack size (like the fatshark battery case here). Stretch on these is about right - not uncomfortably tight but tight enough to give you confidence they won't fall off. They may stretch over time but this hasn't been an issue with other straps for me in the past so don't expect that to happen here. Overall really pleased with them, particularly the clean mounting option for the DJI goggles.

Alternatives
Speaking more specifically for the DJI goggles these are the other aftermarket options I see as competition for the iFlight straps:

First of all the build needs to be based around a very light micro quad; 3 inches or smaller. This is because anything larger will make too much noise and will be travelling too fast to fly in you local neighbourhood and to avoid objects at night. Lightness (and the right choice of props) is key to making this quiet but more on that later.

Without putting to fine a point on it, it is the improvement in cameras that have made this build possible and practical: it used to be that the only good night cameras light the runcam owl were very large, even the more recent runcam eagle micro was too big for a lightweight quad. However now that we have the foxeer toothless nano 2 (and to a lesser extent the caddx baby ratel) we have MUCH better night vision. There are many good review on the toothless 2 nano at night, particularly the one by Nick Burns that I based this choice on. I've also included DVR from a day flight to show that even though you gain performance in low-light, it doesn't sacrifice daytime performance.

You can simply slap this cam on an existing build and have a night flier as is but to really make it discrete to fly at anytime read on...

The best way to minimise noise is less weight and more blade area. I think the best candidate therefore is a toothpick style build but changing twin blades out for triblades. In my case I've taken my modified Eachine red-devil (article here) and have used the gemfan 2540 props. In hindsight following completion of this article I think the Happymodel larva-x would be a reasonable although slightly heavy candidate - slightly larger 1103 motors, the right sized vtx and canopy to start plus a handy on-board DVR. Of course the camera has been swapped out for the Foxeer Toothless 2 nano. A word of warning, make sure you choose the starlight 2.1mm lens when ordering as this lets more light in that the 1.8mm lens I chose.

If you prefer to use a ducted quad you need to get a little more specific, namely a Happymodel Mobula 6 moved over to a mobula 7 frame with gemfan 1636 4 bladed props and the larva-x canopy. Here the mobula 6 running gear allows you to stay light, the mobula 7 frame allows for larger, quieter props and the larva-x canopy allows for the nano sized cam. I'd recommend against 2s as it starts getting too heavy even on the 40mm props if you want to keep quiet.

Although unnecessary for my needs, you may wish to increase the control range, If you don't want to spec any money you can use this hack but adding a discrete receiver will be the most reliable option. If FRSKY I strongly recommend the XM over the heavier diversity XM+.

Depending on the VTX you are using you may also want a nice safe transmit power, say 200mW. Not an option if you are using the onboard mobula 6 unit but if you want to upgrade the Rush Tank Tiny I reviewed here is excellent and can go up to 400mW.

Finding these little quads at night are actually easier at night but you need to be sure you have enough LEDs. Luckily the board used on my quad has enough LEDS to easily spot but if you need extras I'd recommend one of these little lightweight LED sets. You may wish to consider a buzzer as well since d-shot beacon on these little motors is not very loud. With that said I definitely think the lights are a lot more effective, easier in fact that finding a downed micro quad in daylight.

So there you have it, a fairly simple formula for making a little quad that can extend your flying hours. Good news of course is that this is not limited to the night - these camera also look great during the day and having a light, quiet quad makes flying around people more realistic. 

​The Happymodel Mobula 6 is proving to be one of the most popular drones of early 2020 just like the mobula 7 in 2019. What makes it so successful? It is small, light and agile and fast enough given the power source. And, with the introduction of esc firmware capable of running at 48khz, it seems to offer the flight time.

Even though it’s a great quad as is, I always like to tinker to see what can be squeezed out. In my opinion it is the super light weight of the Mobula6 that makes it so good in all the parameters above, so how to make it better? Make it EVEN lighter.

So how is it done?
Get a pack of zip ties (like these) that are at least 80mm long and 4mm wide. The smaller the zip mechanism the better. Use a new blade on a hobby knife to cut a slit along the length as below. The lens of the camera will be wedged in this slit and tightened at an angle that suits your flying.

Summary of findings
​In retrospect this mod changed the quad more than I’d expected, here are the advantages and disadvantages I picked up in rough order. For me you can see the advantages outweighed the disadvantages fairly heavily.

I’m really pleased with this as a mod but in no way am I bad-mouthing the stock canopy. Although I never liked the rounded look of the stock canopy, it is very functional – surprisingly light and certainly strong enough to take the hits given the weight of this quad. The camera mounting and adjustment is really good too. As you can tell though I do like this mod and will be keep it this way for myself.

The Eachine Tyro79 is a low cost brushless high performance FPV quadcopter that is sold sold as a kit exclusively from Banggood. I have written a full review here and the popularity has prompted me to write this full and final upgrade guide after nearly a year of testing, tweaking and changing parts. Sadly my Tyro 79 has been lost in a dense bush area and I have not been able to recover so cannot share pictures of the upgraded on the quad. However, I have noted carefully what has worked for me and this is shared below split into the following systems:

• FPV (camera, vtx, vtx antennas, goggles)
• Motors and props
• Electronics (flight controller, ESC)
• Frame
• Batteries
• Control gear (receiver and transmitter)
• Add ons
• Betaflight

I'll make my strongest recommendations clear in each section and will recap in the summary as well.

FPV systems (Came, VTX, Antennas, Goggles)
I put this first as I think the most crucial improvements are here. Without mincing words the default camera is awful. Although it was working well for the first one or two flights you can see from this video below in slightly overcast conditions that it can be nearly impossible to fly.

The good news is that the tyro79 takes a standard micro 19mm camera and there are some good options available for not much money as below. This is the very first upgrade I would recommend.

I had no issues with the VTX personally but have seen several people who have online. The 20x20mm VTX that the tyro79 uses is very common and there are some excellent options available. My three recommendations are below as your budget allows but all three offer smar audio control and a minimum of 200mW maximum power which is plenty for this quad. All require soldering and/or splicing of wires. With that said I used the stock VTX right up until the end. If you do want the stock vtx it can be found here but I'd honestly recommend one of the below ones in its place.

The Tyro 79 uses a basic dipole antenna connect to the VTX by mirco UFL - another common standard. These are perfect fine when flying by yourself and as durable as it gets. I would recommend a MenaceRC Bandicoot Linear Patch Antenna if your goggles have diversity though for increased range. If you are flying with others though yoy will want a circularly polarised antenna. These three below are compact, reliable and connect directly to your VTX.

Budget plays a hug part on choosing goggles if you haven't done so already. I'm wary to recommend spending too much at the moment such is the impact of the HD DJI systems but bang for buck in any category is the Aomway commander V1s which is being cleared at less than $200. My full review of that goggle is here. Other goggles I can recommend are the Fatshark OLED Attitude V5 (review here) and of course the skyzone SKY03O OLED goggles (review here).

Motors and Props
If the camera is the first thing you should change, the props are most definitely the second and regardless, they are a consumable item anyway. The stock 4-blades are awful - loud, inefficient and full of vibrations. Prop choice can vary depending on flying style but the below options are all very durable, efficient and reliable

The stock motors are heavy, have bearings that are prone to premature wear and have awful plastic coated wires but they fly fine. If you damage or want to upgrade I'd recommend the options below. They are all physically smaller and lighter than the stock ones but make more power and are more durable

Electronics (Flight controller and ESC)
This is pretty much a case of "if it's working it's fine". If you do have issues I only have one recommendation which is just so much better value than anything else around:

Frame
I'd only recommend replacing the stock frame if it gets damaged. Given this is a 1-piece bottom plate this is a distinct possibility and I'd recommend the following option as a replacement. Not, some trimming of wires may be required with these frames but the components will fit.

Batteries
4s is the number of cells you will want for the tyro79 even if starting since maximum throttle can ealiy be limited if needed (see article here). Capacity will be a matter of preference though. I'd recommend 650-850mah in general - smaller than 650mah won't be able to supply enough current at full throttle and larger than 850 will start making it feel heavy and sluggish. I personally run CNHL ministar 4s 650mah but here are a few options:

Control gear (Receiver and Transmitter)
I run FRSKY gear and will look at only these options for the purpose of this review. I feel it is a good price/performance ratio but of course there are cheaper and more expensive options. For receivers:

Things are getting a little more tricky with transmitters now that FRSKY is going with the closed source ACCESS protocol and my choices below reflect this:

Add Ons
Probably the first add on you might consider is an HD cam. The Tyro79 can carry a stand-alone HD cam, particularly some of the lighter ones but performance does take a bit of a hit. Alternatively you can replace your FPV cam with one of the split options - a camera and board that transmits in regular analogue but can record in HD

Betaflight flight software
Betaflight is the free flight firmware that comes installed on the flight controller. By upgrading to 4.1.1 or above will see a big improvement as the stock tune is now excellent. You can go one step further and install JESC on the 4 in 1 ESC for around $5. This will allow you to run 48k frequency on your motors which add smoothness and extend flight time but more importantly this firmware adds bi-directional dshot meaning you can run RPM filtering that everyone is talking about. This seems to be the single biggest improvement to tuning I have seen in my time in the hobby. Sadly I did not get to write my guide for RPM filtering on the Tyro79 before I lost it but install is straightforward and there are a lot of guides out there.

Summary
Well that wraps up my Tyro79 upgrade guide. More detail above of course but the very first things I upgraded were props and camera because the stock items were miserable. VTX was fine as was ESC and FC but there are good budget options to replace these line the diatone mamba stack and the vtxs above. Upgrading motors will give you a good performance bump but I'd stick with the stock ones until they die. Having a number of good batteries is key, you'll need these regardless but definitely go for 4s. Radio gear is farily personal but I still like FRSKY because it is a reliable link and receivers are small and inexpensive. Adding an HD cam is a good way to capture you flight in high definition. If you have the energy, go for one of the split cam options because hauling a stand alone HD camera can tax performance.

The good and bad of the Sailfly-X and Red Devil
Happymodel have launched two of the better value toothpicks recently of which I have reviewed both - the Happymodel Sailfly and the Eachine Red Devil. Unfortunately both have minor niggles on the FPV side of things -

The objective of the mods:
Since the underlying powertrain of these quads is light and effective I set out to improve the above issues while trying to stay as light as possible. Also, in order to keep things simple I wanted to avoid soldering where possible to provide an easy and cost effective upgrade path for those who wanted to emulate the build. The result? I'm super pleased:

The Build.
The build is very straightforward, especially if you are coming from an Eachine Red Devil. I have however included the steps unique to the Sailfly-x in italics:
1. Change components to the Eachine Red Devil Frame
2. Remove canopy and unplug camera
3. Replace AIO camera from Sailfly-x with the Red Devil Nano VTX listed above
4. Replace 4x nylon screws for mounting crazybee board and canopy with the steel ones from the larva-x listed above
5. Mount the Runcam Nano 2 camera to the Larva-x canopy.

6. Plug the Runcam Nano 2 camera into the nano-vtx

The outcome
Super pleased with the outcome. Firstly all jello related to the awful red devil/trashcan canopy is gone. I love the larva-x canopy - it is strong yet lighter than any TPU canopy due to injection moulding. It is also super compact and has and excellent range for camera angle adjustment.

The camera is not a million miles better than the EOS2 in direct sunshine but it really comes into its own in low or less than perfect lighting conditions. The flight below was a late dusk and I has no issue with colour aberations or missing obstacles. The VTX from the Red Devil has proved sound as well - I like to run it on 100mW so as not to stress it.

Rubber bands are a much more practical mounting solution meaning I can quickly switch between mounting types. Anything with a battery less than about 35g should be using rubber bands.

HQ props are excellent. Personally I've had no issues with the regular 65mm but I hear some batches have had problems with balance whereas the t65mm have not. Also the t65mm cost slightly less and have the option for using screws if needed although I prefer no to on these motors.

Conclusion and recommendation
If you have a Eachine I'd definitely recommend upgrading to the canopy at minimum plus using rubber bands to mount your battery. From there the camera is a bit more expensive and perhaps not necessary but worth thinking about especially if you fly in less than perfect light.

For the Sailfly-x the decision is hard due to the extra parts needed - frame and vtx. Personally I'll be keeping my sailfly stock but if I only had that then the power bump from vtx and image bump from the better cam is well worth pondering as is the ability to adjust camera angle.

Introduction
That Happymodel Larva-x is a great value bind and fly ultralight toothpick-style quad which I reviewed and provided and improved tune for here. This is their first model to come without a battery which is due to the fact that many international shipping suppliers are banning lipo batteries from shipment. Stock propellers are the Emax Avan Rush 2.5 but due to supply issues HQ Prop 2.5x2.5x3 are also included as spares. My full review is here where I flew almost exclusively on the HQ 2.5x2.5x3, not because they were necessarily better but because they were less stiff and transmit less impact to what is a relatively weak motor (see my review for more details including a response from Happymodel).

If you're new here you'll find I'm never satisfied with status quo and like to tinker. Propellers and batteries are dispoable items so I've chosen to share the outcome of my tinkering with these. What I did find however is that this quad is very sensitive to these combinations and although some improvements can be had, there are also some very poor combinations that are inefficient and noisy with less than stellar performance.

HQ T65 - In practice
Firstly, as you can see from the images above, the HQ T65mm props are a full 2.5g lighter across the set compared to the HQ 2.5x2.5x3. This is a lot of weight (5% in fact) for such a small quad. If you are brave enough to use these without screws you can acutally save another 2g total.

As expected 2s was too slow on this prop: low motor kv, low prop pitch and only 2 blades means that this is not worthwhile on 2s unless you simply want maximum endurance.

On 3s these props are fast but they are not as good as I thought and here's why:

• The are louder
• Since the motor has much more authority over the propeller the tune needs to be much more tame
• They are very sensitive to weight by way of battery size.

450mah is simply too much weight for this to carry. It's goes from being very quiet to being suprisingly loud which is a sure indication that props are overloaded. Furthermore handling is pretty loose even when the tune is on point.

300mah 3s is about right - it's quieter (not as quiet as 3 blade) and it is easier to tune. You will need a good battery though and I strongly recommend the GNB HV 80c 3s 300mah rather than the Tattu 3s 300mah I was using. If you can perservere you'll be rewarded with some impressive speed. I only had a brief play but you can see my results in the DVR below.

HQ 2.5x2.5x3
I think these will be my prop of choice moving forward. They just seem to offer so much more versatility on this particalur setup and for me the advantages as listed below certainly outweigh the surprisngly small disadvantage in speed and efficiency:

• MUCH quieter (key when flying nearby people or in small parks)
• Wider range of useful batteries: 2s, 3s, 300-450mah all have their place
• Better grip through turns
• Less vibration as 3 blades are naturally better balanced

To show you that you can still have fun on 2s with these, here is some DVR I captured below running 2s 450mah for 5 minutes of flight:

Here's all the parts mentioned in my review:

Why? The OSD runs on it's own chip and although USB is enough to power the memory for the bulk of betaflight settings, it is a hit or miss affair whether the OSD chip gets enough power to store them. With the BEC powered by a lipo there is more than enough power and all of a sudden these little quirks disappear. Caution, with the battery plugged in the motors have access to more power should something go wrong. Please make sure you remove your props.

Recently I made my biggest ever FPV purchase - the Fatshark attitude V5 FPV goggles with the new 30 degree FOV OLED screens. I'm still field testing these for a review which will follow but first I wanted to look at the battery case, more to the point, how to fix the shortcomings.

In this tutorial I rectify this issue by showing you how to safely open the case and mod it to allow for real balance charging whilst losing noe of the rest of the functionality. With this mode you do not need a special charger for 18650 cells, you can use your regular balance charger and charge your pack safely and efficiently. If you need batteries a local vape shop will probably have the best prices but if not then I used these 3000mah ones that have been rebranded.

See below how to complete the work. This took me about 30 minutes. Most of the time was to figure out the best path for the wire and make the path for it accordingly. As always, pictures can be embiggened

Charging
Now for the easy part - charging. Since you will already have a balance charger for your lipos, this is the ideal charging unit for it. Just make sure you select Li Ion as the cell type for charging which means charging is typically limited to 4.1v per cell. Unlike LiPo batteries you should charge at a maximum of 0.5C e.g. 1.5A for 3000mah batteries or 1A for 2000mah batteries etc.

Did it work?
Yep! Beautifully - just as I'd hoped. Charging went flawlessly - my 3000mah calls took 2600mah charging from 3.4v which sound about right. The battery indicator works perfectly too. Given the capacity of these batteries they will probably never come out of the case - I will just charge them up in the case as they get low. I honestly don't know why Fatshark don't do this from the factory but then again there are a number of curious design decisions which I'll comment on more in my goggle review.

The Diatone GT R249+ is my favourite traditional 2.5 inch micro quad based on it's relatively lightweight in this class, motors which have a wonderful mix of power and efficiency, reliable electronics and excellent built quality. My review of this quad is here and it is still available across multiple retailers.

However over a short period of time I started getting a lot of Jello which is concerning given the camera is a CCD which should be much more stable. I did a lot of PID and filter tuning but gave up after a while and let it sit on my shelf. Now I've come back to it with a fresh set of ideas and believe I have fixed the Jello problem once and for all. Firstly I can categorically state that in my case it was not the tune nor the propellers that fixed the issue as I've seen stated elsewhere. The issue is around the camera/VTX and this fix should be relevant to all canopy versions of this quad that doesn't use the TBS vtx i.e.​

How do I fix it?
As an aside you may have noticed I went down the path of changing the camera to a Caddx Turbo F2 (CMOS) cam I had lying around. You don't need to change the cam but the fact that mine is working well now with a CMOS cam (that is more susceptible to jello) is a great sign.

The first way to fix this is to remove the camera and vtx and carefully re-tighten the screws holding the camera sensor board to the lens housing. This may be enough. In addition you can glue the housing to the sensory baord using a self leveling glue like e6000 (my boring review here) or even hot glue taking care not to get it on the sensor itself - I didn't need to go this far on mine.

The other way to fix it that directly adresses the mounting issue it to separate the camera from the vtx, replace the screws that hold the lens housing to the sensor boards with the original ones and put it all back together. That way the screws that hold the camera to the canopy only need to support the camera (and not the VTX) and there should be less demand on the screws that hold the board to the lens housing. There is enough room at the top of the stack and enough slack in the wires to move the vtx on top of the plastic cover on the stack. This would be my ideal fix as it also offers more movement in camera angle - presently the vtx board prohibits full movement.

So there it is, better late than never so hopefully it helps some of you with the jello issue on an otherwise excellent quad. The good news is that all the work I did on the pids meant that it flies really nicely! If you need these pids please see below or more detail in my original review.

Over the time that I have owned this quad I have improved and customised this quad to overcome some of the initial shortcomings and improve on what already makes this great. I have written several articles on this and have summarised these here with links to the full article plus any other improvements I've since made. Practically all of these upgrades are relevant for the Eachine Trashcan as well (which I reviewed here)

​In terms of which batteries to buy I've personally used the GNB and URUAV batteries which I find to be cost effective with good performance. 2s 300mah is my recommended battery size or if you are using 1s then 1s 450-550mah.

Here are the many of the parts options for the Mobula 7 including my recommendations.

The Happymodel Mobula7 uses a F3 based flight controller and since Betaflight 3.4, these boards have not had full functionality of F4 and F7 boards due to smaller memory size for firmware storage. However in this article I will cover how to upgrade to the performance edition of Betaflight 4.04 kindly compiled by UAVTech (his in depth YouTube Channel is here). This special version adds the key performance functions at the cost of typically irrelevant features like servo control, acro trainer, camera control etc. In this article I will show you where to access this firmware and then I'll show you the filters that suit brushless whoops best as well as some good starting PIDs. 

Betaflight 4 Performance edition for F3 boards
This is the final version of Betaflight you can use and I strongly recommend you use this version rather than the one on the Betaflight configurator to get full performance functionality otherwise you will miss out on features like:

• Gyro Lowpass2
• Launch Control (optional - OFF by default)
• Dynamic Lowpass Filters (Dyn_LPF) (ON by default)
• D-term Minimum (D_min) (ON by default)
• Throttle Boost (ON by default)
• iTerm Relax (ON by default)
• RC Smoothing - Filter (ON by default)
• TPA Mode - D-term ONLY option (ON by default)

Download the custom file below at the GITHUB source. SKIP THIS STEP FOR THE EACHINE TRASHCAN

Filter setup
I'll keep this simple and will recommend the community preset for 'brushless' whoop (credit again to UAV Tech). Filters are customised to the higher RPM (hence frequency that these small motors run and resonate at. This is more important than the PIDs Copy/paste below into your CLI:

set gyro_lowpass_type = PT1
set dyn_notch_min_hz = 150
set dyn_lpf_gyro_min_hz = 150
set dyn_lpf_gyro_max_hz = 750
set dyn_lpf_dterm_min_hz = 150
set dyn_lpf_dterm_max_hz = 250
set dterm_lowpass_type = BiQUAD
set dterm_lowpass2_hz = 0

Alternatively you can enter into the Betaflight GUI as per the screenshot below. Note: I recommend these settings for all brushless whoops including the Mobula7, Eachine Trashcan, Emax Tinyhawk etc.

Recommended PIDs
I've seen no need to change the PIDs that are also recommended in the community presets. Once oscillations are taken care of PIDs are more of a 'feel' thing. I've been happy with these PIDs, they are certainly better for me than default:

set vbat_pid_gain = ON
set anti_gravity_gain = 5000
set p_pitch = 33
set i_pitch = 85
set d_pitch = 35
set p_roll = 28
set i_roll = 78
set d_roll = 32
set d_min_pitch = 18
set d_min_roll = 16
set d_min_boost_gain = 30
set d_min_advance = 0

And again with the screenshot if you'd prefer (click to embiggen):

If you want to use or review betaflight exactly as I run it on my own Mobula7, I have attached my "Diff All" file below:

Summary of software improvements
Wow. Not since I've flown my NFE silverware brushed whoop have I enjoyed so much control. I'm not a great pilot but even flying at night in poor lighting conditions I felt fast and largely in control. Flight was predictable and I was able to put the quad where I wanted practically all of the time. Below is a DVR of me flying at night in my backyard. There's no wind which is a welcome relief. This flight was done with a 75% throttle limit on a 2s 300mah battery and low rates (about 300 deg/s max on yaw, pitch, roll). I hope you enjoy these settings as much as I do.

And this will round out my series of Mobula7 upgrades. It's been fun tinkering with this very popular brushless quad for the last few months and these software settings finally connect all the hardware improvements and customisations I've made over this series. Click here to see all of them

The Happymodel Mobula7 (reviewed here), Eachine Trashcan (reviewed here) and many other brushless whoops use generic 4 bladed 40mm props that are cheap and readily available but are they the best? Simply put, no. They are from a toy grade mould and we have accepted them because when these brushless whoops were first released there was simply nothing else available. Fast forward to the present day with the growth in popularity of these models there are a lot more options, some readily available internationally, some not so much. In this article I'll give you my experiences on what I found to be my favourite prop and the justification of why (what it is I look for in a propeller).

What props are available?
If you're in North America probably the full list of what is below. I'm not in North America though so I was limited to what was locally available and what had low (or no) international shipping cost. Therefore this is not an exhaustive test, my approach was more on of pragmatism. ​

My Findings
I'll compare all my findings back to a baseline result of the stock props for simplicity

Gemfan 1636 (4 blade)
I used these when I was running the v3/trashcan frame which I blogged about here. These were a direct improvement over the stock props in all aspects but were a little more amp-hungry. Grip and speed was improved and control felt better. Props are well balanced but heavier than stock. They are extremely durable especially when you consider the protection afforded by the ducts. You will achieve hover at very low throttle

These props are a good upgrade but not my personal favourite.

Note: These are more like 41mm than 40mm and frustratingly don't quite fit in the ducts of of the v3/trashcan frame - the ones you see above were painstakingly sanded down and even then still rubbed every now and then. I hear TBS props have the same issue

Happymodel cut down to 2 blade
Frustrated with the fit of the gemfan props and wanting to try something a bit faster an moe efficient I cut down the standard 4 bladed props to a 2-blade using nail clippers. Try to cut as close to the hub as possible but "good enough" is ok because any difference in cut has less impact the closer the variation is to the hub. Anyway I really liked the performance - they had a high top speed than the gemfan and a lower current draw but they did feel a bit looser than I would have liked given the substatial drop in useful blade area. More throttle was needed to hover for the same reason. I'd recommend giving these a try if you have a spare set floating around. I liked these for outdoor but felt a bit sloppy inside.

Gemfan 1635 (3 blade)
I got fed up with trimming the gemfan 1636 to fit the v3/trashcan frame I actually switched back to an original mobula frame. I know it's not as durable but I really wanted to try the Gemfan 1635 unadulterated. I'm glad I did as these are the props I've been looking for all along. They felt the fastest of all 4 props I've tried and I still managed a good 3:30 on my old turnigy nanotech 2s 300mah batteries. Control/grip felt at least as good as the stock 4-bladed props and only slightly less grip than the 4-bladed 1636. Like the 1636 they are well balanced and very durable. These are my favourite all round propeller for the Mobula7 and trashcan.

Alongside the reduced durability, the original mobula 7 frame is designed for 1s batteries. I have made a simple mod as below to allow any battery. It's not beautiful but is very functional and adds only a tiny amount of weight (would be less if I used a lighter rubber band).

Conclusions
Firstly none of these props get rid of yaw washout in dirty (propwash) air - this is just a limitation of ducted frames. Some say it can be tuned out but results I've seen have not be repeatable. The best fix is to adapt your flying style to suit.

Funnily the props I've reviewed handle different situations in different ways, meaning a horses for courses approach. My favourite is the Gemfan 1635 3-blade which is something of an allrounder. To summarise:

If I get some energy I'm quite keen to try the Gemfan 1636 as a 2-bladed prop - I think it would be a speed demon outdoor but then again I'd probably choose to fly the Sailfly-x which is faster, better control with more battery life.

The Happymodel Sailfly X is my favourite quad I've reviewed so fat this year (review is here) but as with any product, there are shortcomings especially when is it built to a price. Like I did with the Mobula7 I'm starting a new series of articles on upgrading and customising the Sailfly-x.

Step 1: Software only
The most simple mod to do is ensure the onboard receiver is given an armchair ride from a software standpoint. This means changing D16 to D8 mode in betaflight and turning off telemetry which has been known to lock up the flight controller. This can be done as below. 

Next you'll need to rebind with your trasnmitter set to D8 if you were previously bound in D16 mode. Same binding procedure as previously applies (here). Update: you don't actually need to rebind! The bind settings stored in betaflight are the same regardless of D16/D8

Lastly for this step use a small plastic too to fish the recevier antenna (single black wire) out from underneath the canopy to pop out in the opening by the VTX button per second image below circled in white. This took minimum effort but gave me a small but repeatable improvement in signal quality at a fixed range.

Step 2: Lengthen the stock antenna
If you find you want more range (and you probably do) the next step is to increase the receiver wire length from a 1/4 wavelength (31.23mm) to a half wavelength (62.46mm). This will allow you to pop it out of the canopy further and increase the size of the active element. You'll obviously need to solder and measure accurately but you can use any small light guage wire you have lying around. See below for how I did this:

Step 3: Add a discrete receiver
This is where you start spending money but not a lot. I recommend purchasing a XM or XM+ receiver at 1.0 or 1.5g respectively. The XM is smaller and lighter but the XM+ adds diversity. You could also use the r-xsr which has the weight of the XM with diversity AND telemmetry but is about twice the price. Personally I think the cheaper, lighter XM is best suited to this application because even this receiver will outstrip the performance of the 25mW VTX. 

Step 4: 900MHz (Crossfire or R9m)
If you want to go for ultimate range or you use only these systems you can go for Crossfire or R9m. If you've got these then likely you have the knowledge to wire yourself but for completeness see below for wiring and software setup:

So there you go, options to upgrade your reception range from free to expensive. I'm currently using the 1/2 wavelength antenna (step 2) and it is fine given the VTX output but have a FRSKY XM receiver in waiting.

Voltage sensing is critical in Betaflight not least of all because it is the best indicator of when to land in order to protect your batteries to make them last longer. I've always taken for granted that this reading is correct on every board from the factory but I can tell you it is not. On 2 recent quads I've set up it was well out (Happy model Sailfly-x and my 2019 freestyle build) to the point were it reported my 6s battery as being 24v instead of 22v.

First you'll need to connect your quad to betaflight and enter the "power and battery" tab. Next plug in your battery (please do it in this order). With you battery plugged in you'll see the voltage as shown by betaflight in the area highlighted below:

Next use the digital multimeter to test the battery voltage using the balance connector. Do this by touching the multimeter probes on the terminals at the extreme edges of the balance connector as below. Don't worry if on backwards, your multimeter will just report a negative voltage which is fine. It is important to do this whilst the battery is connected to the quad as the electronics (especially the VTX) will draw a small amount of current and thus cause the voltage to drop a little compared with having the battery disconnected.

Compare the voltage reported on betaflight to that on the multimeter. If it is the same then great, it needs no adjustment. If betaflight is reporting too high, reduce the number "scale" under Voltage meter in the screen below. Go up or down one point at a time and press save in the bottom right corner and the betaflight voltage will update instantly without having to reload. Continue this process up or down to get the closest number. You'll not likely get this right to 2 decimal places, 1 d.p. is fine. As you'll soon find the adjustment in beta flight is reasonably coarse so get as close as you can.

That's it - it seems innocuous but this is such an important measurement and calibration is so simple that I'd recommend it for every quad regardless of whether is seems right or not.

Both the Happymodel Mobula7 (review) and Eachine Trashcan (review) both come with 4 x 1s batteries that are designed to be used in series 2 at a time with a fiddly ph2.0 series connector. This introduces a lot of resistance in the power circuit and resultant voltage drop under load which shows as battery sag. It is easy to upgrade to a true 2s JST connector as I've shown in this blog or a XT30 if you'd prefer. Although 2s 300mah batteries are not expensive it does feel wasteful just leaving the old batteries to die. This article covers how to convert these or any other 1s cell into a 2s battery and is scaleable up to 6s!

Getting prepped
As well as having a well lit area and a good soldering iron, solder and flux (see my article on budget tools here) you'll nee the following items:

• 2s balance connectors with silicon wire like these: JST-XH 2S LiPo Balance Cable
• JST female plugs with silicon wires like these: 5PCS Amass 20AWG 10cm JST Male Plug Cable 
• or XT30 female plug with silicon wire like this: XT30 Connector To 18AWG Cable 100mm
• 30mm wide PVC heat shrink tubing like this: 5M Flat 29.5MM 18.5MM PVC Heat Shrink Tubing

So that's it! I've done about 5 cycles in this battery so far and other than the slightly lower capactiy it flies just fine compared to a stock 2s battery. This is a great way to use up the otherwise sound orphaned 1s batteries you may have floating around.

Although you are likely to already have one, if these are genuinely your first multicelled batteries you will need a balance charger so that you can charge more than one cell at a time. Although they can ben expensive, a good charger will stay with you forever and they are a LOT more accurate than the chargers these quads come with which at best undercharge or at worst overchage your cells. My recommendation is the Hobbymade D6 pro duo charger. Without going into detail I've done a full review here.

See below for weight comparisons.

Tools I recommend for this job:

• Small Phillips screwdriver e.g. something like this
• Good soldering iron with small tip. Strongly recommend the TS100
• Side cutters. I have been using these cheap Daniu ones for about a year now
• Hot glue gun. This is optional to insulated new solder joints and provide a stronger fix for the antenna UFL connection. I've been using this one for about 6 months.

So how does it go? It's works exactly as you would hope. You plug it in, it starts charging and it turns off when it's done. It charges up to 4.2v per cell which just right for a Li-Ion cell. It's a shame it didn't come like this from the factory but I'm glad fullspeed had the vision to come up with a solution.

Update: FRSKY have just announced 2 new x-lite variants (X-lite pro and X-lite s that do have USB charging as standard). These are more expensive than the base models of which there are a LOT in circulation. Now that mine has USB charging I personally see no reason to update for what I need from a controller.

I promise I'll keep this one short. We are forever taking antennae on and off our quads and goggles as we arrive and leave our flight locations. I'm guilty of being too eager to get in the sky and not tightening my antennae correctly thereby leaving a poor connection to the active element on both the quad and goggle ends. Using a spanner is not a fix I'd recommend as it is easy to over-torque and damage the antenna or mount.

The fix here is really easy and I got the tip from my favourite VAS antennas which all include a plastic knurled grip (now I know why). Using a knurled grip that sits over the hexagonal drive on an SMA-style connecter allows you to get a better grip and more torque on tightening whilst avoiding over torquing as above. To put this another way - using these grips allows you to get the right tightness every time - good for the best connection and it will also help them coming loose in flight.

I love the emax tinyhawk, even though it is 'only' 1s it is well suited as an indoor racer and they've recently started a spec class near me just for this. My full review of it is here. One one common furstration however is that the FPV camera can come loose in a crash and if you try to push it back in the wrong way you can easily snap the tabs on the frame that hold it in. I've prepared a simple step by step walkthrough heretht will allow you to fix it safely in less than 5 minutes (I timed it!).

First you need to know that the camera is attached to the frame by two pastic tabs on the frame that holds it in the right place at the right angle. in a crash the camera can easily pop out of one or both of these tabs. Below are the steps needed to correct this, click on any of them to embiggen.

So that is it really. While you are in there with the board out you also have to option of making the VTX antenna external (I tried, no noticeable improvment) and the rx antenna external which you can probably see in steps 9 and 10 when you expand the image. I haven't yet tried this out.

If you are looking for other improvements I'd recommend my step by step article on upgrading to the 4 bladed props that improves thrust and allows for turtle mode.

​Thanks for reading, if you found this article useful please feel free to like or share, the facebook links below directly link/like this article. Links are affiliated and help me buy the bits I need to produce this type of content.

Introduction
The Emax Tinyhawk is a 1s brushless whoop-style quadcopter that has steadily grown on me since I first reviewed it here. It's certainly not the fastest but is probably the most well designed ducted quad available at present. It comes stock with very good 3-bladed emax props that perform well except that lack thrust when running in reverse meaning "turtle mode" does not work. I've installed and reviewed their new 4-bladed props where turtle mode works and how they perform in general.

Why are there so few prop options?
Due to the motors being set up pointing down (pusher) rather than up as most traditional quads it means that, at this stage, you are limited to Emax tinyhawk props only. It's not bad news though, both the original and new props are light, well designed and balance and made of VERY durable material. 

Weight Comparison
Before I get into the flight performance, you can see below that the weight is very similar for the props meaning the sum total difference for all for props is just 0.3g. Now I'm a stickler for weight, especially on micros but 0.3g just doesn't factor for me.

Flight review
Performance is similar but I get more response down low which is consistent with having an extra blade. What was unexpected though was that I felt I had more thrust at the top end. My best guess this this has something to do with the different blade design and is a welcome feature. In term of efficiency I think I had a little less flight time, perhaps 10% when flying in the same location. Across 4 flight I averaged out at 190 seconds (3:10) on the 4-bladed props and about 208 seconds (3:28) on the 3 bladed props on the stock emax 450mah batteries.

Turtle mode?
Yes. It works. To embelish on this the reason this is so crucial is that it can can sit on your lazy arse if you are flying around home but more crucially you can get back in the air and racing if you crash to the ground. You may not win but you also won't get a DNF placing. This is well worth noting because tinyhawk spec racing classes are becoming more and more popular since the tinyhawk is well designed, relatively inexpensive, extremely durable and is also available as a cost-effective ready-to-flykit with goggles and transmitter.

Conclusions
Just get them and give them a try. They're only $2 for a set and given the durability of the frame and the props themselves they will last you a very long time. In new news I see emax have just launched in smokey grey. Beyond performance improvements I love the idea of customisation and think these would look great with the new black frame.

Reversing the rotation of all of your propellers (known as 'props out') is something you commonly do on larger quads to eliminate debris bring flung into your FPV camera and to push obstacles like branches clear. Obviously these don't apply when the props are gaurded like in a Mobula 7 but there is also a hypothesis out there that it can improve flight characteristics such as note on KababFPV's video here. Some of the beta fpv guys also believe it helps with yaw washout but the evidence appears inconclusive. Typically to do this correctly you need to

1. Remove and swap all of the props
2. Reverse each ESC direction in BL Heli Suite
3. Tick 'reverse props' in betflight.

The method documented here allows you to skip steps 1 and 2 which is especially useful because you can run the risk of damaging the rather fragile motors when removing and re-installing the propellers. Just a quick note the the Eachine trashcan comes default with reversed-props.

Below are the steps to take when changing to props out. I'd recommend doing this when switching frames. I did it when I upgraded to the v3/trashcan frame as I wrote about here. Use the little philips screwdriver that came with your quadcopter. Click images to embiggen.

That's it - easy eh? If the quad flips out when you throttle up, check the physical rotation is the same as shown in the betaflight image for prop rotation above.

Introduction
The crazybee F3 and F4 series of all in one boards that are the heart of the mobula7, m7hd, eachine trashcan, snapper7, ur65, uk/us65 and even beta75pro2 are an amazing piece of technology that incorporates a flight controller, 4xBL_HeliS ESCs and a receiver. Cramming all of this on one board does mean some compromises though and the biggest one for me is the poor receiver range. This brief article will show how to get a minor improvement with zero investment. Note: this mod is relevant for all the models listed above.

Current situation
The mobula7 (and other models listed in the intro) uses an onboard receiver - either frsky, flysky or dsm protocols all of which run on the 2.4GHz wavelength. The have a small antenna soldered to a through hole that is tucked safely underneath the main board and above the battery tray. This is is the most durable solution but not the best option for maximising control signal.

Optimising the range
The easiest way to improve the reception of the control signal without changing anything else is to move the antenna into a better location. This is to receive an unimpeded control signal as clearly as as possible which for me meant pointing the antenna up to clear of the rest of the quad as possible. Obviously the antenna pointing down is not ideal for this so feel free to follow the 5 minute guide below on how to solve this.

Has it improved?
Yes is has, but only a little. I often fly from my backyard and loop a tree about 50m away from where I stand. I'd always get telemetry lost warnings when I rounded it every time but now only about 20-30% of the time. A minor improvement but given the poor range I'll take what I can get.

Recommendations for further improvement
The next easiest improvement would be to use a piece of insulated wire that is 62.46mm long - 1/2 a wavelength. This is now to much to increase the wavelength as it is to lengthen the active element so that some of it can clear the body of the quadcopter. Credit to KababFPV for this recommendation here. Note if you want to do this I'd recommend using some wire you have lying but if you have none it is available in all gauges cheaply here (I recommend 26AWG for this application)

Thanks for reading, if you found this article useful please feel free to like or share, the facebook links below directly link/like this article. Links are affiliated and help me buy the bits I need to produce this type of content.

Introduction
In order to run 2s, the Mobula 7 comes standard with 2 x PH2.0 connectors that allow you to run 2 x 1s batteries in series. When running in series the strength of the connection is determined by the weakest link in that circuit. The strength of connection determines the maximum current that can pass and the voltage drop that results. Although it works this system has a very low current overhead from a design perspective - 2 x budget PH2.0 connectors with the dreaded 'folded' pin and very lightweight 26AWG wire. All this adds up to a lot of voltage drop when even reasonable current is pulled - something all will be familiar with on the mobula where you see voltage drop to scary levels (6v and below) during punches even when you have a full battery. Although it recovers almost immediately it does mean you don't get access to the battery's full voltage potential.

New Update: Read here how to convert your 1s batteries to a 2s battery

Why not XT30?
I'll address the elephant in the room - short answer is weight. The JST connector is lighter than the XT30 and has plenty of overhead on this quad considering it only has a 4x5a ESC (20a max). I've also had history of using both on my larger 1.9" ultralight build before too and the JST was not left wanting. The only reason I would go to XT30 over the JST in the future was if I needed more durability. (photos to follow shortly comparing weight)

New batteries
Going ahead with this mod means you'll no longer be able to use 2 x 1s batteries as before unless you make a harness of female JST to 2 x series PH2.0 I'd strongly recommend against this though as this adds weight and a lot of extra resistance. Assuming you have upgraded to the Eachine trashcan/mobula v3 frame as covered here (which is honestly the very first mod you should do) then there are 2 batteries I can strongly recommend:

If you do not have a balance charger, you will need one to charge these 2s batteries or ay other multicelled lipo battery. I use the hobbymate pro 2 duo (reviewed here) - a little more expensive but so versatile.

What to do with you 1s batteries you have left? I have an article coming to show you how to convert these to proper 2s batteries you can use with this new connector alongside the GNB or nanotech batteries.

How to install
Before I launch in with pics and steps I'll say although quick this is a fiddly install only because the solder pads are small in relation to the wires you will connect to them. I do have a method to confirm everything is safe but you have been warned!

Conclusions
I probably spoiled the suprise earlier but this mod is well worthwhile. I skeptical but really pleased to find the standard battery sag issue was instantly improved meaning I didn't have to guess at how much battery I had left as before finishing my flight. In addition I seem to have access to more power more quickly meaning response feels snappier. Also on long punches thrust seems to remain more consistent. If you have a good soldering iron and are a confident solderer I'd definitely recommend this one, particularly if you are running the v3/trashcan frame. On my year old 2s 300mah turnigy nanotech batteries I got just over 3 minutes of hard flying in a large area. During that time voltage did not dip below 6.8v and then only toward the end. Compare this to dips below 6.0v with the original setup in the first minute of light. This large voltage drop is less to do with the battery and more to do with the poor connection on the original setup hence the improvement to power. As a side bonus this should mean power will be more consistent to the VTX and receiver as well for more consistent performance. All in all I can recommend this mod but stick with the 22AWG wire and bear in mind the change in battery requirements.

Introduction and why I like the Mobula7 better.
It's been a few months since I reviewed the HappyModel Mobula 7 (here) and have more recently reviewed the Eachine Trashcan here. Since reviewing the Trashcan I've been flying the Mobula 7 more and find I actually prefer it. Why? Even thought the image quality is better I can't get used to the narrow vertical field of view on the traschan's EOS2 16:9 camera and I find the 0803 motors on the trashcan draw too much power on without delivering a whole lot more speed. The Mobula7's major downside is the frame which breaks easily even though the v2 frame is an improvement and e6000 glue (as tested here) is great for repairs.

For this reason I've written an article for newcomers on how to upgrade to the V3 (trashcan) frame which completely addresses the strength issues.

Instead of just writing about it, please see the steps below with captions. Click on the picture to embiggen

So as you can see a fairly basic change that takes about 5-20 minutes depending on your level of experience. You don't need to drill, cut, solder or modify anything - it is a direct change. You don't even need to update anything in betaflight.

After spending some time flying this frame it is so much more durable. The first night was very windy and I crashed a LOT. No damage to the frame whatsovever with crashed which would have destroyed the Mobula7 frame. 

A lot of comments have been made that the V3 frame is much heavier, I'm here to debunk that myth as you can see from the first 2 pics below. about 15 packs in I feel no difference in performance to the original frame given the weight is near identical and the ducts are no bigger. I can say my flying has improved though because I now feel more confident the frame will hold up better when crashing from trying more risky moves.

Final note on the battery bay
As you know in order to run 2s on the Mobula7 you needed to use 2 single 1s batteries and the square 39 printed mount so that the batteries could be piggy-backed. The V3/trashcan is different, and more awkward with 2 x 1s batteries where foam tape is used to wedge them in. The reason for this is that the frame is actually designed for a proper 2s battery as you can see below.

Introduction
​Recently I've reviewed the Eachine Trashcan and in many ways it is superior to the Mobula7 as I found in my review here. One omission from the trashcan though compared to the Mobula7 is the lack of a physical beeper. On the upside though is that the trashcan (like most other brushless quads) uses Dshot as the ESC protocol meaning you can use the motors as a beeper. This is in fact immediately available on the Trashcan with some simple software changes which I will outline below.

Betaflight setup
By now you should be at least a little familiar with betaflight having set up your trashcan for flight. You will need to enable DSHOT beeper in the configuration screen below.

That's it for Betaflight,on to BL Heli Suite.

BL Heli Suite
Assuming you have not yet used BL Heli Suite you will need to download it here:
www.mediafire.com/folder/dx6kfaasyo24l/BLHeliSuite
Once extracted, run the BL Heli Suite exe file and then follow the screens below:

And that is it! Now you should have a reasonably loud beeper that can be set to an auxilliary switch in betaflight so that when you crash you can better find your quad. Just remember, don't run the beeper for more than 10-20 seconds at a time - it does have the possibility to overheat your motors but not when used in short burst.

Hope this helps, if you found it useful please feel free to like or share on Facebook with the links below.

I did a thing. After reviewing the new FullSpeed TinyLeader 2s-3s brushless whoop (review here) I learnt to fly around the heavier weight which has come about because of it's larger motors and better FPV system. I still think that 1103 motors like on the TinyLeader and Beta75x are better suited to a lightweight 2" propeller like my ultralight build here rather than the 40mm props (1.6") as we see so commonly on the brushless whoops now. I decided to put my money where my mouth is so to speak and prove it rather than just hypothesing.

Since my objective was to retain that character of the Tinyleader I knew I wanted to keep the canopy and FPV system as this is one of the defining advantages of the TinyLeader. It would have been much easier to lose these in favour of an AIO cam andjust use one of the emerging frames from FlexRC (Mira95 to launch shortly), Tomoquads or even the Lil Bandito. The problem is however the canopy and camera mounting means there needs to be a lot of chassis in front of the centreline of the quad as you can see from the frame image below. Note - front of quad is at the right of the image.

Rather than tell you the frames I tried that didn't work I'll tell you what did - the trusty Eyas X2 Elf from Banggood which coincidentally was the first micro I built a year and a half ago. Rather than tell a bring story I've captioned the images below to let you know my though process and the outcome. FYI the X2 Elf frame is available from Banggood for less than $10 ($7 at time of writing). It requires minimal modifcation as you'll see below. Click on any image to embiggen.

As you can tell the Tinyleader frame has front motors spaced much further apart than rears which allows you to keep propellers out of view which was obviously a consideration for the TinyLeader HD in particuar. This frame shape is know as a 'deadcat'. In a stroke of skill (read:luck) The X2 Elf frame has a slight deadcat to it mean this will keep props out of view too. 

After test-fitting I found that full 2" props are bit of a push with connectors and cameras screws. Luckily the Gemfan Hulkie 1940 1.9" is an excellent propeller and fits perfectly with no fouling.

I love the new 2s brushless whoops - I have the TinyLeader (review here), the Mobula7 (review here) and the Eachine Traschcan (review here) but feel they are pushing the limits of this design and result in some odd flight behaviour in propwash - yaw twitches where betaflight seems to freak out. Without so much as touching the tune the move to the new chassis has completely solved this.

There is so much more thrust and efficiency on this quadcopter now which was not unexpected - this behaviour is typical when you go from an over loaded prop to one with the correct load. I'm up by about 30 seconds of flight time and have much more power and control now - I can pull out of dives much later and the 'point and shoot' flight characteristics of the ducted frame is gone - it is now much more like a conventional micro where you can bank and slide through corners rather than being more yaw-based.

Test flights so fare have been using a GNB 2s HV 300mah 80c which is now available at banggood. This is a really good battery and gives me around 3 minutes of flight as below coming down at around 3.75v per cell rested. Although the TinyLeader can run 3s I believe these motors are too high in kV for this with this larger prop size. Conversely I intend to try with 1s (for science!) but will realistically just run a betaflight throttle scale if I want to tame it down for small areas.

As you can see from the DVR footage above there are no props in view on this new frame which stays true to the TinyLeader Design (and all Fullspeed Leader designs actually). This is good news for the regular TonyLeader but great news for the HD TinyLeader in that HD footage will be very clean.

That's it for this mod session, it was easier than expected given the frame but is not for noobs. It has been very rewarding though and intend to keep the TinyLeader in this configuration. Thanks for reading and if you found this helpful please share this blog using the FB link below or with whatever platform you use. Product links are affiliated and help me keep tinkering and sharing like this article.