ERSKY9X Coding

[ozphoenix]

Hi Mike,
I just acquired a nice Freewing 1600mm 4-engine B17 - uses two (totally separate) 2200mAh 3S batteries (2 ESCs and 2 motors on each battery). I've installed an S8R in it -- all of that is working well.
I want to use 2 MiniLipo sensors (MLVSS) through the S.Port. I assume I'd have to change the S.Port ID of one(?). I've had a bit of a fiddle with the telemetry menus to see what I can pick up for inputs, but I cannot see how I would differentiate between the two sets of sensor inputs.
Any tips on how to go about this?
Regards,
ozphoenix

[MikeB]

The firmware should detect a second lipo sensor that is on a non-standard physical ID, and put the cell values into the cells 7 through 12 positions in the telemetry list. I think you will get a single value for the total pack voltage that is the total of all 6 cells however.

Mike

[ozphoenix]

Hi Mike,
Ok, thanks for getting back during the holidays - is there a 'map' of already used IDs for each type of S.Port sensor? If yes, where can I find it? If not, what range of IDs should be used for MLVSS-type products?
I'm out of time today, but I'm happy to do some experimentation in the coming days and give you feedback.
If the Ctot is for both batteries, are there any already-available programming tricks to give a total of each battery separately? Like, addition in a Scaler or Global, then announcement over audio?
Regards,
ozphoenix

[MikeB]

Try 0x0B for now (11 decimal), many others are allocated. I have a document from FrSky (under NDA) that specifies the physical IDs.
0x0C, 0x0D, 0x0E and 0x0F are also unused at present.

Recent test versions allow addition of two values in a Scaler. Try setting one scaler with source CELL1 and "ex Source" to CELL2. Then a second Scaler to Source CELL3 and "ex Source" to the first Scaler, both times with function "Add". The result of the second Scaler should be the pack voltage.

Mike

[ozphoenix]

Ok, understand NDAs (lots in business). Thought that might be the case, hence the way I asked But, thanks for the pointer(s).
As I said, out of time today - getting late now (here) and will be at the flying field by or before 6am tomorrow, so up really early.
Will now do some testing over the coming days and get back to you - it will be a week or so before the big bird gets up in the air - I'm waiting for a BEC to take over the supply for the electronics, instead of the little UBECs in the 4x ESCs, before I maiden it so I have time to play a bit.
Thanks, again, for the prompt info.
Regards,
ozphoenix

[Kilrah]

That info is given in the user manual of some sensors.

[ozphoenix]

Ok, thanks- will also watch for that.
regards,
ozphoenix

That info is given in the user manual of some sensors.

[ozphoenix]

The MLVSS documentation says that its default ID is '2'. It mentions that it can be changed, but does not give any available ranges for suggested changes. For now, I will go with Mike's info.
regards,
ozphoenix

Ok, thanks- will also watch for that.
regards,
ozphoenix

That info is given in the user manual of some sensors.

[ozphoenix]

@Mike,
Please educate me further on the PB14 Heartbeat and the Double Rate options (9XRPro with XJT external module):
1. Once selected/de-selected, is it necessary to re-bind to the Rx for the options to be fully implemented?
2. If you fail to re-bind, what do you expect the outcome to be?

Once I have gathered these answers and further information, I can explain more background to my enquiry.
Thanks in advance,
ozphoenix

[MikeB]

A default of 2 is interesting, the FLVSS is 1, the FasV sensor is also 2. You could use 1 and 2 therefore.
Re PB14 and Double rate. No rebind is necessary.

Mike

[ozphoenix]

@Mike,
Today, I typed a very long and detailed post regarding a bind failure and subsequent testing, reproduction and verification of that failure.
However, I didn't make a copy before hitting 'post' and now that material seems to be lost. I now need to go out for the day and will not be back at the desk until tomorrow. I intend to provide you with all of my testing configuration and results if you need them, but cannot re-type my post until tomorrow, which I intend to do.
However, if you have nothing better to do and have the particular equipment and time and inclination, please look at the combination of ErSky9x R222c9 on 9XRPRo with XJT external module and S6R and/or S8R Rxs and new options (either or both, it doesn't matter -- the error still occurs but time before failure varies) of PB14 Heartbeat and/or Double Rate enabled (and, subsequently disabled). This error does NOT occur with QX7 internal XJT module or with QX7 with external XJT module (moved from my 9XRPro) using the same option settings and ErSky9x version.
With this 9XRPro/Ersky9x/Options/SxR Rx configurations and combinations of PB14 Heartbeat and/or Double Rate options enabled, I am losing bind/radio link consistently and reproducibly -- typically between 2-4 seconds and 5-10 seconds cycle on/cycle off (mostly, Failsafe then engages and, subsequently, disengages), but variable and the time seems to depend on the hardware combination (and particular SxR Rx in use) and software settings chosen (enabled or not enabled). If BOTH (must be both) PB14 Heartbeat and Double Rate are disabled, the loss of link does not occur. It occurs faster if both options are selected. It occurs if only one option is enabled, but not as often. And, I found one set of S6R configuration options changed (AUX1 and AUX2 disabled, when they were originally enabled and working) in one plane during this current circumstance.
My testing (four different planes/Rx, two different radios but same version of software, one external XJT module) came about because yesterday one of those planes engaged Failsafe and flew away even while close in to our field, only to smash into the side of a tree that was still within our normal flying distance. We found it by using the RSSI signal and meter on the 9XRPro, but that signal was disappearing/reappearing during our entire search.
Apologies for brevity of this post, but the last post attempt took an hour or so to create and I do not have that time remaining today - will do it tomorrow, if you still need it.

[ozphoenix]

Hi Mike,
Trying to change ID on one of the MLVSS.
Using the same (two-wire) cable that I successfully use for SxR updates and configurations, connected to back of 9XRPro with the XJT module removed and MLVSS powered separately. The MLVSS LED pulses fast at first, then slower when I connect the radio, as it should.
The SCANNING message stays for a long time and shows one result:
Id 1 AppID 300

Then, nothing. No other device found. No offer to change anything, no ID 2 (MLVSS) found.

I've tried powering the device both before and after I choose Change SPort ID option on the menu, but the SCANNING message never ends (until I hit EXIT).
Tried two different (new) MLVSS units.
Tried a new S8R which had just now been re-flashed using the cable - it shows no ID, either.
Ideas?
regards,
ozphoenix

Try 0x0B for now (11 decimal), many others are allocated. I have a document from FrSky (under NDA) that specifies the physical IDs.
0x0C, 0x0D, 0x0E and 0x0F are also unused at present.

Recent test versions allow addition of two values in a Scaler. Try setting one scaler with source CELL1 and "ex Source" to CELL2. Then a second Scaler to Source CELL3 and "ex Source" to the first Scaler, both times with function "Add". The result of the second Scaler should be the pack voltage.

Mike

[MikeB]

I'll look into the SxR problem you have. You should only enable the PB14 option if you have done a hardware mod to connect the heartbeat signal to the PB14 connection. I need to check and document that option for the 'PRO.
Changing the ID, when you have the scanning showing Id 1 AppID 300, press MENU LONG and you should get the option to select a new ID. Change the ID to the required value and press MENU to set it.

ID 1 is the default for the FLVSS, so is what I expected for the MLVSS.

Mike

[ozphoenix]

Hi Mike,
Ok, was suspecting that you might say something about a required hardware mod for PB14, though I don't think that the need for a hardware mod (in advance) has been evident anywhere in the thread. It was implied that the QX7 has this ability already included as part of the build, but no comment on the hardware mod required in the XJT/9XRPro variant. However, the Double Rate option causing the issue is also still a quandary.

As for ID, that message of AppID only appeared once - all other times, the message of SCANNING stayed on the screen and the AppID message did not appear.

Not sure why that is, but will keep trying.

Regards,
ozphoenix

I'll look into the SxR problem you have. You should only enable the PB14 option if you have done a hardware mod to connect the heartbeat signal to the PB14 connection. I need to check and document that option for the 'PRO.
Changing the ID, when you have the scanning showing Id 1 AppID 300, press MENU LONG and you should get the option to select a new ID. Change the ID to the required value and press MENU to set it.

ID 1 is the default for the FLVSS, so is what I expected for the MLVSS.

Mike

[ozphoenix]

Ok, just tried the Change ID again on one MLVSS and think I might have got it - got the AppID 1 message, LONG pressed Menu, got the 'new ID' message, changed it to 11(dec), got the 'set Idle', 'Set Active', etc messages and done. Will try it again in the B17 with 2x batteries, etc tomorrow and report back.
Thanks and regards,
ozphoenix

Hi Mike,
Ok, was suspecting that you might say something about a required hardware mod for PB14, though I don't think that the need for a hardware mod (in advance) has been evident anywhere in the thread. It was implied that the QX7 has this ability already included as part of the build, but no comment on the hardware mod required in the XJT/9XRPro variant. However, the Double Rate option causing the issue is also still a quandary.

As for ID, that message of AppID only appeared once - all other times, the message of SCANNING stayed on the screen and the AppID message did not appear.

Not sure why that is, but will keep trying.

Regards,
ozphoenix

I'll look into the SxR problem you have. You should only enable the PB14 option if you have done a hardware mod to connect the heartbeat signal to the PB14 connection. I need to check and document that option for the 'PRO.
Changing the ID, when you have the scanning showing Id 1 AppID 300, press MENU LONG and you should get the option to select a new ID. Change the ID to the required value and press MENU to set it.

ID 1 is the default for the FLVSS, so is what I expected for the MLVSS.

Mike

[ozphoenix]

Mike,
BTW - just a curiosity question - after changing to ID 11, if I exit the page which advise that ID was changed and then shows ID 11 AppId 0300, if I re-enter the Change SPort ID option again, the device is found but it shows ID 1 AppId 0300 (not 11). I can change it again and it tells me that it is changed to 11. Do you know if it really WAS changed to 11 in the first effort? If so, why does it then show ID 1 again? Haven't had time to test the two devices (ID 1 and ID 11(?)) in a plane yet -- will do that later today -- but thought I'd ask, since I have a few moments to type right now.
BTW #2 - Happy New Year!

[MikeB]

I'm running "d2" on my 'PRO. I just tested changing the ID using a Vario (I don't have a MLVSS), and that changed from 0 to 3 correctly.

Mike

[ozphoenix]

I updated my Pro to 'd2' to be consistent with you.
I tried an FrSky SPort inline 40A current sensor - it started as '0', changed to '2' properly and was still '2' when I came back to it, then changed it to '0' again, properly.
I tried two Varios - they both changed properly, as for you.
I tried a GPS Sensor - it changed from '3' to '5' properly and back again to '3' - saved properly each time.
I tried 5 different MLVSS (all recent production, one has already been returned to FrSky for warranty repair - SPort not responding at all, repaired and returned to me working ok, it seems). Of the 5 MLVSS units, 4 will not hold the changed ID - they consistently report as changed after the 'State' messages, but show the original ID when run again.
However, one MLVSS (not the one already repaired by FrSky, but another of the new ones) will OCCASSIONALLY change and hold - for IDs 0,1,2,3,5 that I tried but not 11, so far. And, it is very hit-and-miss -- every 5-6 tries, it will save and hold, the rest of the time, it reports as changed but has not changed when re-run. So far, I have not been able to determine any consistent sequence that will cause a firm 'set-and-hold' for a new ID.
I will keep trying to figure it out, but maybe there is something slightly different in the timing of the SPort of the MLVSS when compared to other SPort devices?

I'm running "d2" on my 'PRO. I just tested changing the ID using a Vario (I don't have a MLVSS), and that changed from 0 to 3 correctly.

Mike

[ozphoenix]

Got '7, 8, 9, 10, 11' to 'stick' on the sample of one that works, but not every try - still hit-and-miss.

I updated my Pro to 'd2' to be consistent with you.
I tried an FrSky SPort inline 40A current sensor - it started as '0', changed to '2' properly and was still '2' when I came back to it, then changed it to '0' again, properly.
I tried two Varios - they both changed properly, as for you.
I tried a GPS Sensor - it changed from '3' to '5' properly and back again to '3' - saved properly each time.
I tried 5 different MLVSS (all recent production, one has already been returned to FrSky for warranty repair - SPort not responding at all, repaired and returned to me working ok, it seems). Of the 5 MLVSS units, 4 will not hold the changed ID - they consistently report as changed after the 'State' messages, but show the original ID when run again.
However, one MLVSS (not the one already repaired by FrSky, but another of the new ones) will OCCASSIONALLY change and hold - for IDs 0,1,2,3,5 that I tried but not 11, so far. And, it is very hit-and-miss -- every 5-6 tries, it will save and hold, the rest of the time, it reports as changed but has not changed when re-run. So far, I have not been able to determine any consistent sequence that will cause a firm 'set-and-hold' for a new ID.
I will keep trying to figure it out, but maybe there is something slightly different in the timing of the SPort of the MLVSS when compared to other SPort devices?

I'm running "d2" on my 'PRO. I just tested changing the ID using a Vario (I don't have a MLVSS), and that changed from 0 to 3 correctly.

Mike

[ozphoenix]

Just went back to a couple of the 'non-changing' units - they still don't change.

Just fyi - not sure what is meant to happen here, but I daisy-chained the '11' unit and a '1' unit to the radio cable and only the '11' unit shows up on your SCANNING, regardless of where in the daisy-chain they are located. Just fyi.

Will now go to the hobby shed and see if these two units ('11' and '1') will show up in the Telemetry of the radio when installed in a plane.

Got '7, 8, 9, 10, 11' to 'stick' on the sample of one that works, but not every try - still hit-and-miss.

[ozphoenix]

Ok, some feedback - some good news, some interesting, some bad
Configuration: two 2200S 3Cell batteries each at about 50% or 11.6V, 2x MLVSS (let's call it ID 1 for battery 1 and ID 11 battery 2)
1. With two MLVSS installed and connected, the Ctot value shown on my Telemetry operational screen shows 23.2V for about 4-5 seconds, then reverts to 11.6V and doesn't change (something not right there -- based on what you previously mentioned, I expected only 11.6V always).
2. On the Telemetry Cell Scaling page, battery 1 shows up as Cells 1,2,3 (with about 3.88 to 3.89v each), battery 2 shows up as Cells 4,5,6 (with about 3.88 to 3.89v each) -- not 7,8 and 9 but good, so far.
3. If I disconnect either bat2 balance lead or MLVSS ID11 SPort lead, Cell4 goes to 0.0V, but Cells 5,6 stay at the last reading (say, 3.89V). The only way to clear these values back to 0.0V seems to be to power cycle the radio - exiting to any other screen and re-entering the Cell Scaling screen doesn't do it. Reconnecting bat1/MLVSS SPort re-establishes the correct reading on Cells 4,5,6 -- so, a bit of value clearing needed here, I think.
4. To create a usable read-out of the new battery values, I will use the Scaler addition method you suggested. Assume that I have a Scaler already set up in SC1 (I do). It doesn't matter if I do or not -- this next bit also will happen in SC1 and it is 100% reproducible. Start a new Scaler (SC2) and scroll to Cel4 in the Source field. Go to the 'ex Source' field and scroll towards Cel5, but stop at SC1. If you step one more step (to SC2), the screen flashes blank for a moment, then the word 'MESSAGE' in large letters appears at the top of the screen for just an instant, the radio beeps and is returned to the last operational screen you were on. If you leave SC2 blank and use SC3 for this test, then you can advance to SC2 in the 'ex Source' field and the radio does its reset when you try to step to SC3. Worryingly, with the Scaler crash, you also lose bind to the Rx. If you wait for half a minute or so, it seems to come back, but not always -- and, power cycling the radio doesn't seem to re-bind every time - I had to do a re-bind to the Rx on one such event (maybe I didn't wait long enough). But, on one similar such radio crash, the four engines of the plane fired up for about 1/2 second, the control surfaces cycled and bind was lost -- ouch! if my props were on and near my body or the plane wasn't locked down ) Seems like there's some coding fix required here.

EDIT: BTW, I checked on the QX7 and this effect (4) is the same - crashes the page.

EDIT NOTE: BTW, on either radio, I've just now found that if you MENU LONG in the 'ex Source' field to bring up all the available options, then choose the Cel4,5,6 from that full list, the crash does not happen. So, it is something to do with scrolling passed the currently selected SCx when using the arrow keys while on the 'ex Source' field itself.

So, a couple of steps forward, but a couple back - the second MLVSS appears, there are a few tidy-ups required.

I'll try to move forward with a couple of other parts of this project by using the MENU LONG option selection, for now.

Thanks in advance,
ozphoenix

[ozphoenix]

Mike,
I've been staring at this project too long ) I was trying to set up the Scalers using the Add method and noticed that the voltages reported for the ID1 MLVSS seemed to be too similar to those reported by the ID11 MLVSS - too coincidental.
I looked at the Telemetry Cell Scaling page and still saw voltages reported for Cells 1-6, but when I looked at the Telemetry Logging page, there were ticks (data being received) beside Cells 1,2,3, and 7,8,9 (but not 4,5,6).
So, now I am confused as to which values are being reported for which cell positions and which MLVSS is actually being reported correctly.
I have to take a step back from it and wait until you've had a look at what I've found and reported so far - your wider and deeper vision will surely clarify some things for me.
Thanks in advance,
ozphoenix

[ozphoenix]

Mike,
Just FYI - tried some of these ID changes using the external module bay pins of my QX7 - still same result - one MLVSS reports ID '11' but will now not change back to any other number, the other MLVSS units will not change from '1'. In other words, consistent with the 9XRPro.

Just went back to a couple of the 'non-changing' units - they still don't change.

Just fyi - not sure what is meant to happen here, but I daisy-chained the '11' unit and a '1' unit to the radio cable and only the '11' unit shows up on your SCANNING, regardless of where in the daisy-chain they are located. Just fyi.

Will now go to the hobby shed and see if these two units ('11' and '1') will show up in the Telemetry of the radio when installed in a plane.

Got '7, 8, 9, 10, 11' to 'stick' on the sample of one that works, but not every try - still hit-and-miss.

[MikeB]

I've just posted a specific test version for you to try.
This should fix the Scaler ex_source editing bug.
The first MLVSS sensor should populate cells 1-6, the second should populate cells 7-12. Ctot should be the total of ALL cells.
I tested using a single FLVSS, changing its ID. I did find that the ChangeID scanning operation didn't find the FLVSS unless it had a LiPo connected to the measurement pins.
Please test this version carefully as I thought I saw a watchdog reboot happen on mine.

Mike

[ozphoenix]

Ok, many thanks, Mike - will only get a chance to test it in about 4-5 hours from now - will report back later today (our time).
Regards,
ozphoenix

I've just posted a specific test version for you to try.
This should fix the Scaler ex_source editing bug.
The first MLVSS sensor should populate cells 1-6, the second should populate cells 7-12. Ctot should be the total of ALL cells.
I tested using a single FLVSS, changing its ID. I did find that the ChangeID scanning operation didn't find the FLVSS unless it had a LiPo connected to the measurement pins.
Please test this version carefully as I thought I saw a watchdog reboot happen on mine.

Mike

[ozphoenix]

Mike,
I haven't tested just yet because of a small problem - I managed, over time, to end up with 3 of the MLVSS 'stuck' at ID 4 so wanted to change at least one to another number and now cannot - they will not change, using the Pro and the back module pins. Then, I noticed an interesting issue -- if I scan with the MLVSS connected, get the current ID (say, 4), then disconnect the MLVSS but go ahead and choose to change the ID number (to, say 11), the Pro then proceeds to say 'Set Idle State', 'Set Active State', 'ID Changed' and it then shows Id 11 AppID 0300!! But, the MLVSS device has not been connected since I first scanned and got the original ID number (4). As if it is not even communicating with the MLVSS during the 'change ID' part of the process.
For now, I can grab another MLVSS that is still set to '0 or '1' and also use one of the '4' units, but maybe the Change SPort ID code needs a re-visit? I know you have had it working with FLVSS and I have it working with other devices, as noted earlier, but these MLVSS units confound me
Regards,
ozphoenix

[MikeB]

I'll add two extra telemetry items Cls1 and Cls2 (For Cells1 and Cells2 unless I come up with better, 4-character, names). These will be the total cell voltage for each of the two MLVSS (or FLVSS) sensors. Ctot will continue to be the total of all the cells from both sensors.

Did you try changing the ID with a LiPo connected?
The change operation doesn't need or use any response from the sensor!

Do you have any of the FrSky devices that allow connecting a PC to a SPort device (e.g. STK, FDDlite)? If so, there is a FrSky PC (Windows) program that allows the ID to be changed.

Mike

[ozphoenix]

Hi Mike,
Ok, thanks for these added bits - will try them when available.
Yes, I have tried with Lipo connected - no different, but will keep trying.
Yes, I have an operational STK and I've been looking for the Windows program to change ID but haven't been able to locate one - I've tried the FrSky-rc web site, but their software only configures or updates - I cannot see an option to change ID anywhere - what am I missing?

[ozphoenix]

I briefly tried the new Ersky9x you posted - I am using MLVSS with ID 1 and 4 (only choices I have at present). I now get no readings at all in Cells 4-6 nor Cells 7-9 I have to go out for about 1.5hours - will continue testing when I return, but think you might have retired from the battlefield by that time. I will keep posting what I uncover, anyway, for you to see when you have available time.
thanks
ozphoenix

[MikeB]

This is the program from FrSky (no longer on their web site) that changes sensor ID's (at least some of them). Try the FLVSS tab.

Please try, if you haven't already, just using the MLVSS with ID 4. I only have one FLVSS available (my other one is in a model in the loft so I could get it later). When I tried that on ID 11 I got cells 7-9 showing up (I wasn't using the 'PRO though, another radio with internal XJT).

Mike

Edit: It looks like the ID value shown is one more than the actual value, so my FLVSS shows ID 2 when it is actually ID 1.