Got thrown off the GT, board took off without my command, petition to Evolve to fix remote

Evidence towards this is the fact that the Bamboo GTs seem to not have this as an issue at all yet they are using the same tech just different deck material.

 

I agree that the Carbon would present more of a challenge - however mine with the issue is actually a Bamboo GT

 

Great minds think alike - I JUST posted the same thing over on the receiver thread!

 

From the author:
============
I wanted a slim, strong, light case for my iPhone 3GS. What better material than carbon fibre?

The case (an OSIR O-shield) is great - except I have now learnt that carbon fibre not only inhibits RF signals, but also reduces the reception strength of wireless comms.

I guess this is why there's only 4 or 5 carbon fibre iPhone case manufacturers
============

 

I think a good idea for now is to not ride with any source of bluetooth or similar on your body:
so...
- no phone (or put in Airplane mode)
- no smartwatch
- no bluetooth or wifi or RF outputting anything (if possible)
- also these are much more prevalent in more built up areas - most of my riding at the moment is rural and I plan to keep it that way for now

I know others have ridden with BT headsets etc fine - but until we know more since now we now the radio can get messed up with BT (and other 2.4Ghz sources) Its likely a good idea

 

No problems so far with my Bamboo GT, but I've only been riding it for 4 or 5 hours so far. (knocking on wood)

 

Yea - I just went for a 20min 5KM ride (Ontario - grey and cold + 3degC) and mine was pretty dreamy - no glitches and smooth (rural road - few power lines and houses and no source of RF on me)

 

It'd be cool if we could get a percentage of owners of each (Bamboo & Carbon) that have had dropouts. If the Carbon cuts out a lot more than the Bamboo, then I may change my mind in swapping (looking for more range).

 

So here is a theory to what may be happening.

I think a bad (or lesser quality or actual defective) batch of YJ-13039 nRF24L01 radio daughter card modules (that are in both the remote and the board) MAY have worked into their production recently (for the early November 2016 release). (and it could actually just be the nRF24 chip itself - see my other thread for more details on the varying quality of these chips)

This would explain the intermittent nature of what we are all seeing with some additional points:

- EvolveinLA has two boards - a Bamboo GT which he received first and a Carbon GT. His Bamboo GT is perfect - his Carbon GT despite replacements on the board side controller (containing the YJ-13039) and remote side (same module) still has issues. If a bad batch of the radio modules made it into production after his Bamboo GT was made this would explain it. It also helps to explain why even after getting radio components on both sides replaced on his Carbon he still has issues because they could be replacing defective (or suspect or lesser quality boards) with other defective ones! (on either end)

- My board is a Bamboo GT - also has same problem as EvoleinLA's Carbon GT - so we can't say the Bamboo GT is fine and the Carbon is not.

Of course I am assuming that the same control system is in both Carbon and Bamboo boards which is logical (it looks like all the electronics with the exception of the battery are identical in both)

Now when I say bad radio boards - it could just be that some of these nRF24 chips are not as robust when it comes to dealing with outside interference or unstable power.

In addition to the above - here are a few more observations and points

All three of us: EvolveinLA, myself and Wiztecy had a problem that showed up well into our rides.


Explanations:
1 - the outside interference was not present when starting the ride. (but in my case I was mostly riding in a large loop on some nice pavement - same basic area)
2 - heat builds up in the board side (from motor controllers / esc) which causes more RF type noise around the esc affecting the radio more - and if you notice the way the board side controller riser card mounts to the esc board they are sandwiched together face to face. (no shielding is around the YJ-13039 on either side - its right next to the ESC)
3 - deeper into a ride the voltage drops and becomes more unstable (supplying the YJ-13039) - my problem surfaced at the last 30% of my pack - it was just before FAST mode auto-down-shifted to ECHO mode due to a low main battery (and did come on kinda gradually with what felt like a sticky remote with 1/2sec glitches until the longer 2 sec stuck throttle happened.
4 -the code (firmware) in the remote or the board controller side may have something like a resource leak (the code the ARM processor runs that is the master brain on the controller side).
5 - there is actually a bug in the firmware (that may manifest under certain conditions) - and this may have appeared with the introduction of the two button (left gear down / right gear up) feature on the newest remote
6 - its just the nature of the beast and we have to find a way to deal with it (not likely - some riders have ZERO issues)
7 - they use different suppliers for the YJ-13039 boards and or more likely there is enough variation in specific batches (die release) that some are better than others (remember the long long thread I found in my other post where techs were testing releases of this chip and how some were better than others.
8 - a combination of above possibilities or something else not even mentioned yet - we see this in IT (rarely but can happen) where a user reports a problem with a pc (like it won't boot) - so you replace the power supply (and it still doesn't boot) then you put the original back in (and replace the motherboard) - still won't boot - BUT if you are only replacing one thing at a time and there are TWO devices at fault simultaneously - its not fixed.

So what now?

I think we just have to keep getting the remote / controller boards replaced until this works through and it may take a while.
If it gets worse with heat build up, riders like myself may not even see problems until the spring. This happened with a new car I got a few years back where the CVT transmission would bump (hard) and feel like it was dropping out of gear - went away in the fall and came back in the spring. (I stayed on them and got a brand new car out of this - took 6 months). Also it goes without saying, keep wearing the helmet and other protective gear.

I'm not sure if Evolve watches this forum or if anyone on here has a more direct path to production but its likely worth mentioning all this.
(of course they may already suspect the nRF boards and trying to figure out a strategy to test for the good ones)
(and they may simply come premounted on the larger pcbs they are attached to - wonder if they actually make the main boards in AUS or if they come from Asia?)


Other ideas:

There are many things you can do with a remote control system based on the nRF24L01 chip set. You can provide a better or more tuned antenna to help eliminate outside noise. You can adjust the retry's during transmission. You can adjust the bandwidth you are using (2MBps down to something like 256Kbps) which makes the transmission more reliable. You can adjust the packet size from 32bytes down to something smaller which is easier to send/rec. You can enable or disable auto ack (acknowledgement of packet received). You can program in a Frequency Hopping system (FHSS) to automatically change channels if transmission breaks down. And many other things so the upside its just because a remote system is based on the nRF24L01 doesn't mean its crap - they may just need to add some code to make them more reliable due to the varying quality of these chips.

EvolveHQ posted in another thread about the re-pairing process picking another "channel" - huh - (also mentions them using more encrypted bluetooth channels in latest remote - which doesn't seem to make sense considering they don't appear to be using bt or ble in the first place - unless they just switched to this in latest remotes - which is possible but seems unlikely)

 

When did you get your board and where from ? (just wondering roughly when Evolve made it)

 

Great summary Julian. I was thinking the same thing about a bad batch of transceiver boards - it would explain why it seems to happen more with the newest remotes/new boards. And why Evolve doesn't seem to have diagnosed it yet - if they are replacing the controllers with new controllers, and the problem's still occurring, of course it'd be confounding for them too.

A definitive test on this would be to replace the motor controller and remote on my CGT, with the one from my perfectly working Bamboo GT. But I don't really want to do this! The Bamboo has been working so beautifully, I'm afraid to mess it up, too. But it would be a great test.

About my theory on happening only when warmed up, I think that's disproven. Added this to my info post on Reddit:

EDIT: Was doing some testing Saturday to see if I could a). see the Bluetooth signal if it's there, and b). cause Bluetooth interference. Bluetooth scanner apps on both iPhone and Android did not detect either board or either remote as having BT signal. I turned on/off the BT on both phones about 10 times. One time that I turned BT off from the iPhone 5, the display (battery % and antenna icon) DID cut out for 10 seconds. So this would seem to disprove the idea that it (dropouts/signal loss) only happens when hot/warmed up. (The board had been sitting unused for 12 hours before the test and board/remote were fully charged.)

Impossible to say for sure if the turning on/off of BT multiple times near the board caused the display to disappear - I tried replicating this, and could not get it do drop out again.

ALSO - took it for a brief ride Sunday (yesterday), about 10 minutes only. Had a 1/2 sec. power dropout about 3 minutes into the ride- so it's definitely not from heat.

---------------------------------------------------------
EDIT: As an addendum to your summary, I'd like to add possible explanation #9:

9 - When the code was updated to rectify the 'instant reverse' problem, something happened (a new bug) which made the connection / error correction less robust. As you mentioned, the quality/reliability of the connection for these chips is really related to how well the code is written for them. It's possible that nothing changed with the quality of the nRF24L01+ daughterboards (in this latest batch of boards/remotes), but an inadvertent side-effect of this software update is the dropouts some of us are experiencing.

 

Hey - anybody with a perfectly working GT wanna donate their remote and motor controller for some testing? For the good of the community? If you live in cold weather, it's wintertime anyway

 

The problem seems to be related much more to the latest batch of remotes/motor controllers. Owners of both CGTs and Bamboo GTs have had the problem, and it seems to occur more frequently with the newer batch of remotes where the left button is now 'power level down' rather than 'instant reverse'.

This would support the theory that recent production boards might have a problem with the transceiver daughterboards on both the motor controller and the remote, with Julian identified.

 

yea EvolveinLA - the simplest explanation is often the right one - and I know we are speculating but the correct answer / solution will be found - the main change recently was in the remote - so likely either a new bug in the code or a specific problem with the nRF transceiver remote side or board side or for some potentially both

don't get me wrong - I still like the board and want Evolve to do well - but we do need more reliable control

 

Isn't there a V1 to a V2 (protocol?) setting which can be set in the remote? Was the old remote V1 and the new one V2? Was there a change in the protocol they're using?

 

I got my Carbon GT from Evolve UK on Friday morning, the day after they got it true the Spanish Customs from Australia.
And right away I see that the battery and and signal indicator on the LCD screen drops in and out a lot.......
This made me scared to ride, but I gave it a go and started of in eco, now 2 days and 1000 dropout later I still haven't lost control of the board 1 time.....It works perfect. Only the data on the screen drops out.
So the signal to the LCD Screen drops out, BUT IM STILL IN CONTROLL OF THE BOARD.

And Yesterday I found a great post on Reddit, posted by an Engineer, and after reading that, and my own experience whit the dropouts Im sure hes 100% correct.

The Controls receive info via Bluetooth, like speed and battery %, and its constantly receiving new data.

But the Board itself have a RC CONTROL. And this dont brake connection, , even if the bluetooth does.


Read this post from reddit:


Would be great if someone who works at Evolve would confirm this, or tell us exactly what sort of control the boards have.

Anyway, IM SOOOOO HAPPY WHIT MY C-GT. Its way better then expected. I could not have been happier right now. .


Last a BIG thanks to Evolve UK, And especially Mark, for awesome customer service.

Ride safe guys.


Cheers from Spain.

 

Kenneth - I think he's 100% wrong - lol - he admits that in the bottom of his reddit post with his edit that he no longer thinks its bluetooth - see this thread (all the pictures we have of the remote and board show a radio that is not bluetooth or rc):
Evolve - radio is using nRF24L01

enjoy your board

 

I bet it's simply the channel selection / the pairing process that is a little different between V1 and V2 - can we assume the V2 remote is backwardly compatible with V1 (yea it may just be in V2 they are now using channels in addition to the original 13 from that first FCC filing - there could be an addendum for more channels)

 

The Boston vlogs guy, who has been doing some nice reporting on his experience with the GT, just crashed and fractured his arm. Not due to disconnection, but by inadvertently hitting the reverse button on his 1st gen remote (don't know which is actually a more serious problem).

 

Yeah just watched it on Youtube. This is why the new remote have the gear down system whit the left Botton now....Can`t slam it in reverse when your in Eco-Fast or GT mode.

Quite funny that he always talk about safety gear....Yet the first thing you see when watching the video is him riding whit no helmet. Kind of stupid Imo.


I actually like the remote....only complain I have is that the trigger seems way to fragile, Had it for 4-5 rides and its already super-loose


Cheers all