[SOLVED] Help - stuck in HI rate + LOW EEPROM alert

er9x is the best known firmware. It has a superb range of features and is well supported by the community. Well worth trying out.
Post Reply
nvd07
Posts: 63
Joined: Sat Aug 15, 2020 8:45 am
Country: France

[SOLVED] Help - stuck in HI rate + LOW EEPROM alert

Post by nvd07 »

Hi , I am running Triple Rates and have had trouble in flight, where all the sudden during a flight my ailerons got stuck in HI rates for the remainder of the flight. (This is on a EDGE QQ280 with large control surfaces and that made for hairy landings even though I have a lot of expo.)

My model is setup with triple rates this way: I assigned THR toggle switch so that all my LO rates become MID rates at the same time; however the HI rates come from the three dedicated D/R switches.

Granted , I might be asking for trouble, because I am flying with the 'EEPROM low mem' ALERT ... I have no idea if this alert could cause malfunctions / unexpected behaviour ? I am still on the m64 have not moved to the m128 yet.


So last time I was stuck in HI rates was 2 weeks ago ; this morning it happened again. After a less than perfect landing, this time I went straight into the menus to investigate:


It turns out that the MID/LO had a malfunction that did not affect all channels the same way:
-The Ailerons got stuck in HI regardless what I did with its D/R switch (and regardless what I did with the THR switch).
-However the Rudder and Elevator could still change between HI and MID with their D/R switches, the minor problem they had is they did not respond to MID/LO changes commanded by THR switch.

Next I powered OFF the Tx. Then powered it ON again. Went back into the menu... the triple rates are working perfect, LO/MID/HI on the three channels !!!
I would say this last ON/OFF test rules out a bad wiring / solder / toggle switch , as I only touched the POWER switch...

==> I would appreciate any help/advice on what could be causing this 'stuck in HI rate' ... especially as to whether it could be a consequence of 'low mem' or if it could not be related to it.

Thanks

EDIT : During two flights today I also experienced a maybe 1 second cut in the RPM of the motor during level flight, for no particular reason; but it could be anything: the ESC, the AR6335 receiver, the wiring ... it has never happened with other models.


Note I have videotaped the tests I described above, just in case, but even in low quality it is a 216 megs video and I am not sure I coud upload it to the forum and not sure it would bring any added value anyway.
Last edited by nvd07 on Sun Jun 20, 2021 11:23 am, edited 1 time in total.

User avatar
MikeB
9x Developer
Posts: 17990
Joined: Tue Dec 27, 2011 1:24 pm
Country: -
Location: Poole, Dorset, UK

Re: Help - stuck in HI rate + LOW EEPROM alert

Post by MikeB »

I've not heard of this before. Did you possibly change any trims while flying? A changed trim gets written to the EEPROM, and if that is low then there may be some unexpected problem.

Mike
erskyTx/er9x developer
The difficult we do immediately,
The impossible takes a little longer!
nvd07
Posts: 63
Joined: Sat Aug 15, 2020 8:45 am
Country: France

Re: Help - stuck in HI rate + LOW EEPROM alert

Post by nvd07 »

Yes I am positive I had to put in several clicks of elevator downtrim during that flying session, so this would match what you are explaining... ;)

Today, out of curiosity I tried to replicate the problem at home, by playing several times with the trim.
I could not replicate the aileron triple rates problem, at all, today.

However , instead, on three different occasions today the main MENU showed up unexpectedly (=without me pressing the menu key on the right of the screen), then after a few seconds it went away.
I think this has happened to me in flight before but I had dismissed the issue as 'probably me accidentally pressing a key'...

So the EEPROM low memory alert is one possibility to explain erratic behavior; and I should backup and free up some memory space to address this ;

now , is there also a posibility that I could have a ' bad processor / bad main board ' , causing unexpected stuff like this ?

Thanks !
User avatar
MikeB
9x Developer
Posts: 17990
Joined: Tue Dec 27, 2011 1:24 pm
Country: -
Location: Poole, Dorset, UK

Re: Help - stuck in HI rate + LOW EEPROM alert

Post by MikeB »

The low EEPROM memory alert indicates there are less than 200 bytes of free EEPROM. On the model select page, beside each model is a number, this is the number of EEPROM bytes needed to store the model. How big is the model you are using?
The main code that runs continuously calls a routine that checks for the current model to be written to the EEPROM (which will be needed if you have changed a trim). Since writing to the EEPROM is slow (several milliseconds per byte), this could cause the radio to hang while the model is written, particularly for large model files. To avoid this, while waiting for an EEPROM byte to be written, the main code is called (recursively). EEPROM accesses include a lock to prevent this happening for more than one level of recursion, but even this one level will use up a lot more stack than normal operation.
Possibly the stack is overflowing (into the data space), maybe only with low EEPROM.
Clearly, as you say, backing up a model and freeing some EEPROM space is a useful test (If you have a serial voce module, you may back up models to the SD card on the voice module!).
I think I may be able to remove the recursion by moving the EEPROM updating check outside the main code. This might help.
Since I'm not sure that even a stack overflow would cause the problems you have seen, it may be that you processor has developed a problem that occurs while writing to the EEPROM.
Testing normal operation while NOT using any trims is another useful test (no EEPROM writes unless you modify a model or change models).

Mike
erskyTx/er9x developer
The difficult we do immediately,
The impossible takes a little longer!
nvd07
Posts: 63
Joined: Sat Aug 15, 2020 8:45 am
Country: France

Re: Help - stuck in HI rate + LOW EEPROM alert

Post by nvd07 »

MikeB wrote: Sat Nov 07, 2020 10:42 pm How big is the model you are using?
So my free space is 135 bytes, and most of my models are bigger than this:

- the EdgeQQ (with the triple rate probems) is currently 177 bytes,
- another model I use often is 299 bytes, a heli is 155, but I have not seen any problem on these (they have no triple rates though).

MikeB wrote: Sat Nov 07, 2020 10:42 pm The main code that runs continuously calls a routine that checks for the current model to be written to the EEPROM (which will be needed if you have changed a trim). Since writing to the EEPROM is slow (several milliseconds per byte), this could cause the radio to hang while the model is written, particularly for large model files. To avoid this, while waiting for an EEPROM byte to be written, the main code is called (recursively). EEPROM accesses include a lock to prevent this happening for more than one level of recursion, but even this one level will use up a lot more stack than normal operation.
Ok so I get a notion it is tricky to keep the whole system operating smoothly while doing these 'slow write' tasks, these are interesting details of the internals of the system. I noticed that there is a horizontal cursor blinking in the upper left corner during these 'EEPROM write' times.

MikeB wrote: Sat Nov 07, 2020 10:42 pm Possibly the stack is overflowing (into the data space), maybe only with low EEPROM.
Clearly, as you say, backing up a model and freeing some EEPROM space is a useful test (If you have a serial voce module, you may back up models to the SD card on the voice module!).
Ok / no voice module so will backup onto the PC. I also have plans to take care of my additionnal TH9X with the m128 processor in it ( right now it is still stock / untouched , but as bad weather is coming in I will do the hardware mod + er9x reflash to it ). By the way could you offer an estimate of how much additionnal free space I would get with this m128 Tx set , say if I was to transfer all my current models from my m64 into the m128, how much additionnal free space would I get compared to my current 135 bytes (!) ?

MikeB wrote: Sat Nov 07, 2020 10:42 pm I think I may be able to remove the recursion by moving the EEPROM updating check outside the main code. This might help.
If you see any benefit for the rest of the er9x community, I mean other than just myself ; otherwise if I am the only person, I do not mean to create additionnal work to you. it is probably not worth it and I am sure you have more exciting stuff / new functionnalities to implement ! and I am aware I am still hanging onto my m64 for the sake of it...

MikeB wrote: Sat Nov 07, 2020 10:42 pm Since I'm not sure that even a stack overflow would cause the problems you have seen, it may be that you processor has developed a problem that occurs while writing to the EEPROM.
Testing normal operation while NOT using any trims is another useful test (no EEPROM writes unless you modify a model or change models).
Ok I will keep testing and monitoring it in the coming days, both attended / operating the sticks and switches, and unattended / stiiting still as well ; if there is a hardware problem it might show up again over hours of operation. If anything new happens , I will report back here.

Thanks very much for the help and the detailed answers as always.

User avatar
MikeB
9x Developer
Posts: 17990
Joined: Tue Dec 27, 2011 1:24 pm
Country: -
Location: Poole, Dorset, UK

Re: Help - stuck in HI rate + LOW EEPROM alert

Post by MikeB »

I've just posted the change I thought of, it was not difficult. I tested all was still working OK on my '2561 radio.
If the stack overflowed it is possible it ended over the top of the telemetry receive buffer, that is filled during an interrupt. This could cause the problems you have reported.
This change was well worth me doing, I never really liked the recursion I had before!

Mike
erskyTx/er9x developer
The difficult we do immediately,
The impossible takes a little longer!
User avatar
jhsa
Posts: 19480
Joined: Tue Dec 27, 2011 5:13 pm
Country: Germany

Re: Help - stuck in HI rate + LOW EEPROM alert

Post by jhsa »

Mike, what about the possibility of adding an External Eeprom module? I use a few on some of my projects, but they are i2c.
I believe there are also SPI Eeprom chips.. They shouldn't need many components also. Quite easy to connect, I would say.

João
My er9x/Ersky9x/eepskye Video Tutorials
https://www.youtube.com/playlist?list=PL5uJhoD7sAKidZmkhMpYpp_qcuIqJXhb9

Donate to Er9x/Ersky9x:
https://www.paypal.com/cgi-bin/webscr?cmd=_s-xclick&hosted_button_id=YHX43JR3J7XGW
nvd07
Posts: 63
Joined: Sat Aug 15, 2020 8:45 am
Country: France

Re: Help - stuck in HI rate + LOW EEPROM alert

Post by nvd07 »

MikeB wrote: Sun Nov 08, 2020 9:02 pm I've just posted the change I thought of, it was not difficult.
Ok thank you for this improved version!
MikeB wrote: Sun Nov 08, 2020 9:02 pm If the stack overflowed it is possible it ended over the top of the telemetry receive buffer, that is filled during an interrupt. This could cause the problems you have reported.
Over the past 3 days I tried to put myself in the correct conditions to replicate at home all this, but no success... Therefore I cannot blame 'low eeprom+firmware' and should not rule out a ' main board / outside temperature ' problem yet.

I will move to the new firmware version before next flying session. I might stay in 'low eeprom' conditions and see what happens...
The flying sessions seem to be the best way to reveal problems... :mrgreen:
nvd07
Posts: 63
Joined: Sat Aug 15, 2020 8:45 am
Country: France

Re: Help - stuck in HI rate + LOW EEPROM alert

Post by nvd07 »

Ok , before moving to er9xProv822v , something new to report with 822u, somehow related:

After the flying session tonight I edited my biggest model (299 bytes, the one that did not have any problems) to tweak the throttle response through a curve. The model is still 299 bytes because I already had the curve in it; it was just not being used in my Mix. While enabling it back and editing the curve points I did get 'EEPROM overflow Alert - press any key' screens. :idea: This is the first time I am seeing this overflow screen (but if it happens in flight you probably leave the flight displays unobstructed...).
The menus are handling this fine: next I can get back to editing the other points of the curve.

Note, in the course of switching between models, the free memory is now 225 instead of 135 bytes, without me deleting anything.
The warning at startup has gone because I am over 200 bytes free mem. Although I do not quite get what is going on, this 'temporary bump' over 200 free bytes has happened before.

Anyway I guess this new overflow alert is the connexion to what Mike was describing.

And probably it would have been healthier if I had kept as much free mem as the size of my biggest model... :roll:
User avatar
MikeB
9x Developer
Posts: 17990
Joined: Tue Dec 27, 2011 1:24 pm
Country: -
Location: Poole, Dorset, UK

Re: Help - stuck in HI rate + LOW EEPROM alert

Post by MikeB »

There will be some confusion over the size of the free space. When you write a model to the EEPROM it writes it to a temporary file, then swaps the temporary file for the original file, but doesn't delete the original file, so the actual "free" space is the size of the undeleted file and the unsued space added together. When saving the model the undeleted file is deleted first, then the saving process is started.
The size of the undeleted file is therefore the size of the last model stored, which causes the variation in the reported free space.
You are, however, correct that it is best to make sure the reported free space is larger than your largest model.

To answer an earlier question, a '128 (or '2561) processor has 4K of EEPROM instead of 2K, so moving your models to that would add 2K to your free space.

Mike
erskyTx/er9x developer
The difficult we do immediately,
The impossible takes a little longer!
nvd07
Posts: 63
Joined: Sat Aug 15, 2020 8:45 am
Country: France

Re: Help - stuck in HI rate + LOW EEPROM alert

Post by nvd07 »

MikeB wrote: Tue Nov 10, 2020 10:48 pm The size of the undeleted file is therefore the size of the last model stored, which causes the variation in the reported free space.
Ok thanks for these explanations.
MikeB wrote: Tue Nov 10, 2020 10:48 pm You are, however, correct that it is best to make sure the reported free space is larger than your largest model.
So today I backed up my models onto my PC, then flashed the new 822v version;
I let things reformat the eeprom at startup, then had a look, free space was 1725 with just the default model01 at 56 bytes.
Then I put all my models back into the Tx , noticed I had at best, 315 free. Did a quick test, I went into the curve point editing, managed to trigger the eeprom overflow; next I deleted some models, in order to clear up some additional free space, taking advantage of my recent backup, until I was able to confirm there are no longer any overflow messages when I edit the curves.

So with this 'more sensible' configuration, I went to fly with the EDGE QQ which has triple rates...

And surprisingly I got locked into HI ailerons rate again, during the first flight... (it was uneventful this time because earlier I had tamed down the HI rates 'just in case' )

So all this makes me think that even though during the previous days I was definitely having EEPROM overflows without knowing about them, we see now that my 'ailerons stuck in HI rate' events seem to be another independent problem.

So for the rest of the flying session today, I duplicated the EDGE QQ model and got rid of the Triple Rates, freed as much additionnal space as needed, then it flew OK on Dual Rates, for the three remaining battery packs.
I did get a few throttle cuts but I now blame them on the receiver going into failsafe.
MikeB wrote: Tue Nov 10, 2020 10:48 pm To answer an earlier question, a '128 (or '2561) processor has 4K of EEPROM instead of 2K, so moving your models to that would add 2K to your free space.
Sounds good, that means I will be able to put about 6 additional models around 300 bytes each onto the m128.
I will probably try again the triple rates on the m128 too... :mrgreen:

Thanks again for all the help.
nvd07
Posts: 63
Joined: Sat Aug 15, 2020 8:45 am
Country: France

Re: Help - stuck in HI rate + LOW EEPROM alert

Post by nvd07 »

So here is the report of the Triple Rates experiments I made this morning with the newly converted FLSKY '128 Tx:
(errors edited out on friday nov 20th in red after viewing a nov 6th video of the '64 - sorry for the confusion)

With my '128 Tx, near the end of the first flight with the EDGE QQ, the Ailerons also became stuck in HI rates, as they were doing on my '64 Tx.
However, some things differred from as with my '64, the Rudder and Elevator rates did not fully function properly on my '128; only MID and HI remained available...

I am attaching vids of the '128 below:

Ailerons: stuck in HI / cannot get to MID through D/R switch (& cannot get to LO by THR switch, because to do so, MID should be engaged first)
VID_20201119_128_AILERONS_stuck_in_HI.mp4
(14.63 MiB) Downloaded 304 times
Elevator and Rudder: can switch OK between HI/MID through the D/R switches ; but once in MID , cannot move to LO by the THR switch (and this is unlike on my '64)
VID_20201119_128_ELEVATOR_lost_LO_rate.mp4
(21.81 MiB) Downloaded 264 times
VID_20201119_128_RUDDER_lost_LO_rate.mp4
(23.75 MiB) Downloaded 252 times

On 2 different areas:
- I noticed that using the same eePee model for the EDGEQQ, and same Ytech 4-in-1 MPM module, it does indeed display telemetry data on my '128 screen, whereas on my '64 Tx it does not display it (however my '64 is able to show FRSKY D16 or FLSKY AFDHS2A telemetry; and from this I had incorrectly concluded the AR6335 DSMX receiver of the EDGE QQ was unable to send any telemetry).
IMG_20201119_64_DSMX_noTelemetry.jpg
IMG_20201119_128_DSMX_telemetry.jpg
- Next, on the 'throttle cuts' problems I was having with my '64, I found out they did not come from EDGEQQ's FAILSAFE becoming engaged ; I recently tested FAILSAFE operation and found out that all my channels, including the throttle channel (!) were set to 'Hold'. So the throttle cuts on the previous flights ('64) did not come from Failsafe. And on the four flights with the '128 i did not experiment any throttle cut - but it is probably too early to make a conclusion.



As the EDGEQQ is the only one in my fleet with a DSMX receiver, this protocol may be related to the problems I am experiencing - although technically I have no idea why and how this could happen...

So my Triple Rate issue is rather obscure and puzzling...
One thing I could do is setup one of my FRSKY D16 airplanes with triple rates as well, and then see if my triple rates work fine with it.
Any another suggestion appreciated, in terms of what I could investigate next; I could also show / post here an eePee file of my model, in case I would have programmed triple rates incorrectly. :?:
Also, if some other er9x users are satisfactorily running the combination of Triple Rates + DSMX protocol, please chime in and let me know - so far , no idea whether it is widely being used or not !

As the weather was flyable I carried on with my flying session by switching memory to the other 'fallback' EDGEQQ model where there are only dual rates and no triple rates. I enjoyed telemetry data :D and I played with gyro stabilization gains - it looks like this AR6335 AS3X is able to straighten my lines quite impressively through the wind gusts ! 8-)
Last edited by nvd07 on Fri Nov 20, 2020 8:27 am, edited 1 time in total.
User avatar
Kilrah
Posts: 11108
Joined: Sat Feb 18, 2012 6:56 pm
Country: Switzerland

Re: Help - stuck in HI rate + LOW EEPROM alert

Post by Kilrah »

You should really post your eepe file for people to look at.
nvd07
Posts: 63
Joined: Sat Aug 15, 2020 8:45 am
Country: France

Re: Help - stuck in HI rate + LOW EEPROM alert

Post by nvd07 »

Ok here is the eePe file of the troubled model ... :?
Attachments
2020-11-19_TroublesomeTripleRates.eepe
(1.75 KiB) Downloaded 271 times
User avatar
MikeB
9x Developer
Posts: 17990
Joined: Tue Dec 27, 2011 1:24 pm
Country: -
Location: Poole, Dorset, UK

Re: Help - stuck in HI rate + LOW EEPROM alert

Post by MikeB »

If/when the radio has the problem, please go to the main screen that shows the switch positions and report whether the switches are responding on that screen.

Mike
erskyTx/er9x developer
The difficult we do immediately,
The impossible takes a little longer!
nvd07
Posts: 63
Joined: Sat Aug 15, 2020 8:45 am
Country: France

Re: Help - stuck in HI rate + LOW EEPROM alert

Post by nvd07 »

MikeB wrote: Thu Nov 19, 2020 8:40 pm If/when the radio has the problem, please go to the main screen that shows the switch positions and report whether the switches are responding on that screen.
OK understood - next flyable weather should be sunday; i should be able to check it out then
( all attempts at home / while not flying, have been unsuccessful ... )
User avatar
MikeB
9x Developer
Posts: 17990
Joined: Tue Dec 27, 2011 1:24 pm
Country: -
Location: Poole, Dorset, UK

Re: Help - stuck in HI rate + LOW EEPROM alert

Post by MikeB »

I've just posted a test version for the '128 that includes stack space checking. The STAT screen shows how much unused stack space there is, the DEBUG screen also shows some of the stack contents.
I don't have a '128 radio available to test this, although it has worked previously.
It would be useful to know how much spare stack space you have in normal operation (including receiving telemetry), and also how much space is reported if/when the triple rates stop working.

Mike
erskyTx/er9x developer
The difficult we do immediately,
The impossible takes a little longer!
User avatar
MikeB
9x Developer
Posts: 17990
Joined: Tue Dec 27, 2011 1:24 pm
Country: -
Location: Poole, Dorset, UK

Re: Help - stuck in HI rate + LOW EEPROM alert

Post by MikeB »

I just realised the two switches that appear to stop working (AIL and THR) are the two that are re-routed as part of the telemetry mod.
How have you done that mod?

Mike
erskyTx/er9x developer
The difficult we do immediately,
The impossible takes a little longer!
nvd07
Posts: 63
Joined: Sat Aug 15, 2020 8:45 am
Country: France

Re: Help - stuck in HI rate + LOW EEPROM alert

Post by nvd07 »

Yes the same had occurred to me when I tried to investigate this on my own: these are the re-routed switches...

I wired both my radios according to the 'Gruvin9x' variant:
https://github.com/MikeBland/er9x/blob/ ... etails.pdf
ie I only added wires and cut the track, I re-used the existing SMD resistors of the stock board.

Because the problem occurs on the 2 re routed switches, I also considered it would be possible I could have done some ESD damage to the pins of the ATMEGA when soldering onto them... :cry:
However I would be kind of expecting to also get trouble from AIL and THR on the other aircraft as well, as I have been using these switches often (however, with THR assigned to other things than triple rate); so I set this theory on the backburner for a while.

Also, at some point the '64 radio fell because the lanyard broke. so I had to replace 3 broken toggle switches, and the above were among them. As I was quite mad it had happened, I next suspected that on top of this , I had done an incorrect repair - however the 'diag switch' screen looked OK. And this would not explain anthying on the '128 as i did not repair it.

Tomorrow after looking at the main screen as you suggested, I could also do another experiment where I would re-affect things :
1) quit using THR in my triple rates, and replace it with GEAR , and see what happens
2) swap AIL and RUD within my triple rates, and see what happens
User avatar
MikeB
9x Developer
Posts: 17990
Joined: Tue Dec 27, 2011 1:24 pm
Country: -
Location: Poole, Dorset, UK

Re: Help - stuck in HI rate + LOW EEPROM alert

Post by MikeB »

I've just posted a new test version for the '128. I'm guessing here, but I think there may be a problem with (bad/corrupted?) telemetry data. The firmware includes auto detection of an addon called TelemetreZ (from SmartieParts). This normally makes it very easy to do the telemetry mod (no soldering), but it reads the THR and AIL switches and passes their values to the main board over the telemetry serial connections, along with real telemetry data. When this is detected, the operation on PC6 and PC7 (the alternate THR and AIL pins) is changed to output and their voltage level set by the data from the TelemetreZ. It is possible for incorrect telemetry data to be detected as TelemetreZ data, thus upsetting the operation of the THR and AIL inputs on PC6 and PC7.
This test version skips looking for TelemetreZ data if the module protocol is set to MULTI.

Mike
erskyTx/er9x developer
The difficult we do immediately,
The impossible takes a little longer!
nvd07
Posts: 63
Joined: Sat Aug 15, 2020 8:45 am
Country: France

Re: Help - stuck in HI rate + LOW EEPROM alert

Post by nvd07 »

MikeB wrote: Sat Nov 21, 2020 10:07 pm ... It is possible for incorrect telemetry data to be detected as TelemetreZ data, thus upsetting the operation of the THR and AIL inputs on PC6 and PC7.
Well it is a very convincing explanation to everything I have seen so far!

When I looked into the AR6335 DSMX receiver, it is a 'short-range-telemetry-intended-for-indoor-people-who-should-not-ask-too-many-questions-about-telemetry-anyway'. I bought it for AS3X+weight reasons, and the EDGEQQ my only aircraft with this.

So with your explanation I did another ground test with the regular firmware this morning: I locked the plane inside the house, and went out with the Tx. Soon the TSSI dropped to zero. From there I played with AIL and THR. Bingo! An unwanted menu popped up by itself and also the AIL + THR switches became frozen in the main screen ! This is the first time I am able to replicate things on the ground.
When I went back in the house, normal operation did not restore itself even when I got close to the receiver. So the Tx probably had switched to 'Smartieparts' operation you described, and had 'latched' onto it.

Even during some other tests when staying close to the receiver there are some fishy things going on with DSMX TSSI sometimes remaining stuck at 0 - but my telemetry settings may be incorrect ( before all the above, I unticked the 'DSM A as RSSI' setting I had been using so far but did not sound quite right to me ).

And by cycling the power of the RX on-off-on, i discovered that the menu of the TX showed up ... with my hands off the TX at all times. :lol:
MikeB wrote: Sat Nov 21, 2020 10:07 pm This test version skips looking for TelemetreZ data if the module protocol is set to MULTI.

Mike
Thanks very much for this newer 'test version'. I will now move to this newer test version and will let you know how it goes. I am pretty sure you guessed everything right! If there is anything else you would like me to test with the previous version let me know I can always go back to it.
:D
nvd07
Posts: 63
Joined: Sat Aug 15, 2020 8:45 am
Country: France

Re: Help - stuck in HI rate + LOW EEPROM alert

Post by nvd07 »

MikeB wrote: Sat Nov 21, 2020 10:07 pm I've just posted a new test version for the '128.
...
This test version skips looking for TelemetreZ data if the module protocol is set to MULTI.

Mike
So this afternoon I flashed the er9x-128StackTez.hex firmware and I am happy to report that I did not get any AIL/THR trouble.
I first repeated the tests on the ground with the EDGEQQ / AR6335 DSMX , got far away, and did not experience any frozen AIL+THR and associated Triple Rate malfunction; no unwanted menu either.

Then I went for a flying session. The triple rates went great, so at the end of first flight I cranked up the percentage of the HI rate of the ailerons back to 100% as I have no fear anymore of getting locked into HI rates. Flew again and again, another aircraft for a while, then the EDGE again, I did a total of 5 battery packs with the EDGE - maybe 50 minutes of air time, no issues with these triple rates!

I am convinced my problem of triple rates is now solved ! :P
i will even increase the endpoints of the ailerons beyond 100 to 125% to allow to increase my HI throws even more - no fear anymore !

Thank you very much Mike for the great help as always, and for the modified firmware as well. :D
nvd07
Posts: 63
Joined: Sat Aug 15, 2020 8:45 am
Country: France

Re: Help - stuck in HI rate + LOW EEPROM alert

Post by nvd07 »

Hi , I am back onto this thread; as mentionned in the older posts I am a happy camper running my 128 with the following firmware from the thread named 'Er9x test versions':
MikeB wrote: Mon Nov 25, 2013 11:50 am ...
er9x-128StackTez.zip
21-Nov-2020 21:59
Special test version for '128 with stack trace included
Also disable TelemetreZ processing when multiprotocol selected
...
I would like to request , if possible at all :?: , a m64 version with the same 'disable TelemetreZ' ;
I have absolutely no idea if it would technically be possible on the m64 and how much work it would require from you Mike...

I should have asked first - maybe the 'disable Tez' condition has become standard in the newer binaries such as the version 822w below - if so, the m64 firmware in the ZIP would do what I need ...
MikeB wrote: Mon Nov 25, 2013 11:50 am ...
er9xProv822w.zip
03-Jan-2021 11:31
Fix (possibly!) PC4(L-WR) not working as extra digital input
...
:?: :?:
Thanks
User avatar
MikeB
9x Developer
Posts: 17990
Joined: Tue Dec 27, 2011 1:24 pm
Country: -
Location: Poole, Dorset, UK

Re: Help - stuck in HI rate + LOW EEPROM alert

Post by MikeB »

er9xProv822w.zip should have the TelemetreZ disable in it. It will now always be in.

Mike
erskyTx/er9x developer
The difficult we do immediately,
The impossible takes a little longer!
nvd07
Posts: 63
Joined: Sat Aug 15, 2020 8:45 am
Country: France

Re: Help - stuck in HI rate + LOW EEPROM alert

Post by nvd07 »

MikeB wrote: Sun Feb 07, 2021 7:13 pm er9xProv822w.zip should have the TelemetreZ disable in it.
Ok I downloaded it in order to update my m64 with it
MikeB wrote: Sun Feb 07, 2021 7:13 pm It will now always be in.
Great ! I will be able to use future versions then; my hardware will not be tied forever to ' v / Stack Tez ' ... Thanks !
nvd07
Posts: 63
Joined: Sat Aug 15, 2020 8:45 am
Country: France

Re: Help - stuck in HI rate + LOW EEPROM alert

Post by nvd07 »

Sorry to revive this topic - I think that this issue :
MikeB wrote: Sat Nov 21, 2020 10:07 pm ... The firmware includes auto detection of an addon called TelemetreZ (from SmartieParts). This normally makes it very easy to do the telemetry mod (no soldering), but it reads the THR and AIL switches and passes their values to the main board over the telemetry serial connections, along with real telemetry data. .... It is possible for incorrect telemetry data to be detected as TelemetreZ data, thus upsetting the operation of the THR and AIL inputs on PC6 and PC7. ....
could be back in another form, on the m64 only, as I never did get to properly test the 822w m64 mentionned in previous post.

Tonight I decided to give my m64 and my 128 a good round of firmware updates:
On the m128 , going from er9x-128StackTez from november 2020, to 822y from may 3rd 2021, everything worked fine, before and after.

On the m64 though, going from 822v i think, to 822y , it went from a known condition where it would get 'stuck in HI rate when the telemetry fades away (for the THR/AIL reasons explained by Mike in this thread)' , to a different condition, where it is now 'permanently stuck ( same THR and AIL switches frozen )' right from startup (it even does this with a nearby S6R Rx always powered on, ie before I attempt to turn on the Tx).

-This is the first time I am testing on my m64 the cure for the AIL/THR/TelemetreZ/smartieParts unwanted detection/mixup.
-My understanding is that these frozen AIL/THR I am seeing on this 822y should affect everyone with a m64 with the AIL / THR reroute for telemetry mod...
-However Magilla who was the requestor of this 822y - stated that he had a m64 ,
magilla wrote: Tue Apr 20, 2021 8:02 am ...
I am using turnigy 9x with m64 processor. ...

and I did not see him complain about 822y in his subsequent posts... so my head starts spinning !

:arrow: Is anyone with a m64 and the AIL/THR telemetry reroute seeing the AIL/THR switches frozen with the new er9x firmwares ?
IMG_20210613_202346.jpg
IMG_20210613_202346.jpg (106.03 KiB) Viewed 12145 times
IMG_20210613_202356.jpg
IMG_20210613_202356.jpg (108.77 KiB) Viewed 12145 times
nvd07
Posts: 63
Joined: Sat Aug 15, 2020 8:45 am
Country: France

Re: Help - stuck in HI rate + LOW EEPROM alert

Post by nvd07 »

As I was unsure about what state of hardware modifications I had left my m64 with back then,
tonight i did more investigations and I rolled back to the Firmware version shown below.

My m64 has the NPN inverter in, and it seems to work as expected - with the MultiProtocolModule in, i need to go into the 'radio menu' and enable the command 'send inverted telemetry command to MPM' for the communication to go well between the MPM and the atmega. AIL/THR switches unfrozen at startup. Works fine with the XJT module in too (or with no module at all), never any frozen AIL/THR at startup.

For now, I do not see anything I would want to change on the hardware of this m64, so I will stick to the older firmware as I cannot get the HW to 'like' the latest 822y FW.

the firmware I rolled back to:
m64_822v_rollback_date.jpg
m64_822v_rollback_date.jpg (86.15 KiB) Viewed 12085 times
screenshots with MPM in:
m64_822v_withMPM_protocol_MPM_FWver_shown_20210615_212016.jpg
m64_822v_withMPM_protocol_MPM_FWver_shown_20210615_212016.jpg (70.18 KiB) Viewed 12085 times
m64_822v_withMPM_unfrozen_AIL_THR_20210615_212054.jpg
m64_822v_withMPM_unfrozen_AIL_THR_20210615_212054.jpg (80.41 KiB) Viewed 12085 times
screenshots with XJT in:
m64_822v_withXJT_protocol_20210615_211227.jpg
m64_822v_withXJT_protocol_20210615_211227.jpg (81.91 KiB) Viewed 12085 times
m64_822v_withXJT_telem_menu_20210615_211200.jpg
m64_822v_withXJT_telem_menu_20210615_211200.jpg (88.75 KiB) Viewed 12085 times
m64_822v_withXJT_unfrozen_AIL_THR_20210615_211133.jpg
m64_822v_withXJT_unfrozen_AIL_THR_20210615_211133.jpg (82.86 KiB) Viewed 12085 times

On the other hand , with the atmega128 (not shown in this post), i flew sucessfully tonight with Firmware 822y,
the MPM and a volantex cub (FHSS) - everything went fine - my m128 'likes' 822y. :D
nvd07
Posts: 63
Joined: Sat Aug 15, 2020 8:45 am
Country: France

Re: Help - stuck in HI rate + LOW EEPROM alert

Post by nvd07 »

Ok I found the source of the newer problem I was having with my m64 (in the two previous posts : AIL+THR frozen at startup).

it turns out that from several months of only using the m128 and updating just the m128 , I had forgotten what I once knew about flashing the m64.

First hint was when I realized that the new 'hardware menu' in the m64 looked different from the one I got in the 'rollback/restore'... In fact the new firmware image I was using was not the correct one for my hardware.

In retrospect, the VERSION shown on the screenshot on my June 13 post, did not show the required FRSKY . The incorrect flashing could be deteced from that screen.

Now with the correct: er9x-frsky.hex rather than the incorrect er9x.hex , my AIL/THR switches are unfrozen

IMG_20210619_175801.jpg
IMG_20210619_175801.jpg (53.28 KiB) Viewed 12024 times
IMG_20210619_175741.jpg
IMG_20210619_175741.jpg (49.99 KiB) Viewed 12024 times
Sorry about the confusion I created here. :?

The lesson is that we have to always keep in mind that:

-For the m128, the user must declare the AIL+THR / telemetry mod/reroute inside the hidden Hardware menu, from a single firmware image which remains the correct one , whether or not the hardware mod is present.

-For the m64, even though there is also a hidden Hardware menu, it is smaller and this is not where the user could declare any AIL+THR / telemetry mod/reroute; rather, before all things, the user must pick the specific firmware image which has FRSKY in the filename, and make sure to use it when flashing the atmega 64.

:mrgreen:

Post Reply

Return to “er9x”