ERSKY9X Coding

Post by **jhsa** » Sun Dec 20, 2015 8:54 pm

Mike, you corrected that on the version you sent me on the 13th Dec while you were fixing the BT.. I believe it is wrong on the version of 9th Dec. I think that was when I mentioned it to you. That is why I asked kaos to change Ad8 to AD10 on Pot4.

Kaos as Mike said, you must have something wrong either with the hardware or the programming. My pots and switches are all working correctly, and they were already with the same version you have installed now. The only problem was those 2 inputs were swapped as far as I remember.

João

willhac · Post by **willhac** » Sun Dec 20, 2015 9:16 pm

I just tested the 9Xtreme test version for the new multi protocol format.
I flashed the multi board with the file posted today on RCgroups.
Tested OK with a Flysky and a Frsky receiver (not 16 channels...)
But no telemetry with Frsky, not implemented yet ?

Post by **MikeB** » Sun Dec 20, 2015 9:30 pm

The telemetry does (apparantly) work on a 9XR-PRO. To where are you connecting the telementry signal from the MULTI module?
The MULTI signal is non-inverted, so currently will only work if connected to COM2 as the COM1 input is inverting on the 9Xtreme.

I may be able to get COM1 working, but that will need a software UART that operates with even parity to be working. I have a software UART working, but not using even parity!

Mike.

willhac · Post by **willhac** » Sun Dec 20, 2015 9:39 pm

It's connected on COM2.
Telemetry was working fine with the test version that fixed the reboot bug with MULTI/COM2 on the 9Xtreme.

kaos · Post by **kaos** » Sun Dec 20, 2015 10:07 pm

I am using V.3205, 09.12.2015, r218. I will check the analog/diagnostics and report back. It is a high impedance pot, but it worked when connecting to the P1 or P2.

edit: Mike, I tested with p4 at ADC10.
the diagnostic: P3: -100 to 100, A7:0000 - 07FF A9: 0000-0FFF

the pics taken when P3 and rocker pot both held at one end and the other, sorry my hand was shaky a little trying to hold cell phone and push button at the same time. The A9 0FFx, the last digits actually fluctuate some what I saw 8,6 C....

kaos · Post by **kaos** » Sun Dec 20, 2015 10:45 pm

MikeB wrote: What you are reporting indicates the 'new' pot connection is either not connecting, or is a high impedance.
In the analog scanning order, due to the physical connections, the P3 pot is read, then AD(C)8, then the throttle, then AD(C)10.
What this means is if the analog input to AD8 is missing or weak, then AD8 follows P3. Similarly, if AD10 is missing or weak, it follows the throttle stick.

yes, I think this is the problem, with what you said the AD8 and AD10 reversed. the AD10 follow the P3 and AD8 follow throttle as of now. it seems still reads AD10 but overridden/influenced by P3. I can see servo moving with limited movement when P3 is fixed.
so P3, AD10, throttle, AD8 are all read in sequence together but P1, P2 are independent?

Post by **MikeB** » Sun Dec 20, 2015 11:34 pm

willhac wrote:It's connected on COM2.
Telemetry was working fine with the test version that fixed the reboot bug with MULTI/COM2 on the 9Xtreme.

I may need to do a special test version to find out what's not right, as I currently have no way of testing this.
Just one test you could do first though. On the TELEMETRY menu, the first item is "Usr Proto". Please try setting this to "FrHub", the last item.

Mike.

willhac · Post by **willhac** » Sun Dec 20, 2015 11:49 pm

It's the same with "FrHub", no values.
For the info I didn't change the previous settings : COM2 func. = telemetry and COM2 baudrate = 115200, I suppose it's overriden ?

Post by **MikeB** » Mon Dec 21, 2015 12:14 am

Yes, the baudrate is forced to the required rate when the protocol is MULTI.
I'll try to put some debug information in tomorrow.

Mike.

Post by **MikeB** » Mon Dec 21, 2015 12:19 am

kaos: All the analog inputs are read in sequence. I just tested ADC10 on my AR9X using a 5K pot. It all worked fine.
I'm currently running a version of firmware from the 10th Dec. I'll look out the version you are running and try that tomorrow.
Do you have a lower value pot you could try as a test?

Mike.

kaos · Post by **kaos** » Mon Dec 21, 2015 12:20 am

Mike:
I just ran a test. I have some 5K pots and I hook one to ADc10 (measured at 5.7K), now the P3 completely take over the reading. there is no movement of the 5K pot any more. In calibration the P3 bar and p4(hooked to ADC10) moves perfectly with P3 (before there was a slight difference), and the 5K pot is not moving the bar at all. And after calibration, the 5K pot has no effect on servo at all either.
the pic shows the analog data of P3 and 5K pot at each end:

edit: wow you are reading my mind or I am reading yours

Post by **jhsa** » Mon Dec 21, 2015 12:48 am

Please post a couple pictures from your hardware menu.. Switches and pots..

João

kaos · Post by **kaos** » Mon Dec 21, 2015 1:09 am

here are they:

kaos · Post by **kaos** » Mon Dec 21, 2015 1:23 am

jhsa: do you have revA board or revB? Mine is revB. There could be a difference. (I don't know for a fact.)

Post by **jhsa** » Mon Dec 21, 2015 1:41 am

Rev B

Post by **jhsa** » Mon Dec 21, 2015 1:50 am

Kaos, I've sent you a PM..

João

kaos · Post by **kaos** » Mon Dec 21, 2015 2:23 am

Jhsa: I just loaded your test version to Tx.

Yep, that is it. You are using a different version. And there is no problem with your version. Something has changed from your version to the published version. Both pots works normally now, the 5K or my Rocker.

Post by **jhsa** » Mon Dec 21, 2015 2:40 am

On the version you had before, you would have to connect the pot to ADC8 but select AD10.. That was corrected on this version I have sent you..
So, all working now?

João

Post by **jhsa** » Mon Dec 21, 2015 2:42 am

kaos wrote:Jhsa: I just loaded your test version to Tx.

Something has changed from your version to the published version. Both pots works normally now, the 5K or my Rocker.

The version I sent you is actually newer.. So, something changed from the published version to the version I sent you

João

kaos · Post by **kaos** » Mon Dec 21, 2015 3:17 am

revert my statement. Everything is OK only AD10.

I just tested AD8. Both 5K and my rocker has some issue.
My rocker is slightly affected by p3 but operating on its own no issue, only if I don't use rocker and move p3 there is slight movement on the ch8 servo. With a 5k pot. just in hardware set up menu when moving from line to line the ch8 servo will jump. once calibrated, turning the 5k pot will end up with a shaky servo at certain angle, the servo just vibrates. but p3 has no effect on servo/ch8.

Jhsa: May be that is why, your mod with 3 position switch and wiper has no issue. Beause your version the AD10 is used for pot which has no issue, but AD8 when used with pot, there is a problem. Can you try your set up and pot on ad8 to see what happens. and try just a regular toggel switch for ELE and put one pot on AD8.

edit:
I have to revert my revert.

actually ad10 still have some effect from the throttle. not much but a little which means when p4 is not touched by moving throttle, servo will move a little. and when changing menu item to item you can see the servo moves a little.

ReSt · Post by **ReSt** » Mon Dec 21, 2015 9:03 am

@Daedalus66

As I said, just my personal preferences

Reinhard

Post by **jhsa** » Mon Dec 21, 2015 11:06 am

Kaos, my setup is:
ELE - AD10
Taranis Slider (Pot) - AD8

No interaction with throttle whatsoever as far as I can see. But will check again..

Post by **MikeB** » Mon Dec 21, 2015 12:01 pm

You described using only a single wire connection to the ADC8/10 input and using common power and ground connections to several pots and switches. This is not a good arrangement. My testing is connecting a 5K pot directly to the three connections for ADC8/10.
Please try this, if possible, using ADC8.
Because ADC10 may be used to convert a 2-pos switch to a 3-pos switch, there is some code in place, only used if ADC10 is not a pot input, that 'debounces' a 3-pos switch input. I may not have this disabled correctly on some versions of the firmware depending on the order of AD8 and AD10 in the hardware menu for setting up P4 and P5.

Mike.

Post by **MikeB** » Mon Dec 21, 2015 12:14 pm

willhac wrote:It's the same with "FrHub", no values.
For the info I didn't change the previous settings : COM2 func. = telemetry and COM2 baudrate = 115200, I suppose it's overriden ?

Just checking (probably correct from before), you do have "FrSky COM Port" set to 2?

Mike.

kaos · Post by **kaos** » Mon Dec 21, 2015 2:49 pm

The 5k pot i used to test was all by itself with 3 wires v,gnd,AD8 or10.
Actually the 5k pot at both port had more influence by P3 or throttle than my rocker.
With either pot, AD8 had more interference than AD10.
I will use the FW jhsa sent to me to make a video to show you. That version is much better than the posted test version.

Post by **jhsa** » Mon Dec 21, 2015 3:05 pm

Mike, I PM'ed Kaos a version from 13th Dec you sent me while you were fixing the BT.
This version had already the sources named better than the one from 9th Dec..

João

willhac · Post by **willhac** » Mon Dec 21, 2015 3:55 pm

MikeB wrote: Just checking (probably correct from before), you do have "FrSky COM Port" set to 2?

Mike.

Yes it's set correctly to 2. (I noticed that when UsrProto is changed, "FrSky COM Port" is forced back to 1)
I also reflashed the 9Xtreme with dec.15th version to check, and the multi board with older multiprotocol_18-11.hex, telemetry works either with "UsrProto"=Frsky or FrHub.

Post by **MikeB** » Mon Dec 21, 2015 4:04 pm

kaos: I've just posted a specific test version (here: viewtopic.php?f=7&t=4676&p=65894#p65894).
This makes the selection order "NONE, AD10, AD8". This is necessary because of the "debounce" code, on AD10, I mentioned above.
In the DIAGANA screen, an input on ADC8 is displayed as A8, and one on ADC10 is displayed as A9.

Please try this with a 5K pot directly connected to ADC8 and/or ADC10. Check the hardware setting as it may now be different.
I've just tested this and all worked correctly.

Mike.

kaos · Post by **kaos** » Mon Dec 21, 2015 4:34 pm

will do. I will check it and report back. I am at the office. Will do this when I get home.
Also, it comes to mind, one thing I did not test is my own servo on the HD camera. Whether that shakiness at certain angle is my servo or the tx (using jhsa sent version). But that thing has been sitting there for >1 1/2 yr who knows. Last time I used it, it was fine.

The version jhsa gave me is a lot better for sure. The Dec 9 version has nothing to do whether the servo is slightly off or not, the signal is just "bled though" from throttle or p3 to AD8 or AD10.

Post by **MikeB** » Mon Dec 21, 2015 5:03 pm

willhac: Special debug version for MULTI telemetry here: viewtopic.php?f=126&t=7751&p=103194#p103194. I may also have added a fix!
If MULTI telemetry still doesn't work, press UP long, then RIGHT a few times to get to the DEBUG display.
Then let me know what the numbers on the first 2 lines are.
The bottom 3 lines are the start of the FIFO buffer for the COM2 input when used for telemetry.

Mike.

openrcforums

ERSKY9X Coding

Re: ERSKY9X Coding

Re: ERSKY9X Coding

Re: ERSKY9X Coding

Re: ERSKY9X Coding

Re: ERSKY9X Coding

Re: ERSKY9X Coding

Re: ERSKY9X Coding

Re: ERSKY9X Coding

Re: ERSKY9X Coding

Re: ERSKY9X Coding

Re: ERSKY9X Coding

Re: ERSKY9X Coding

Re: ERSKY9X Coding

Re: ERSKY9X Coding

Re: ERSKY9X Coding

Re: ERSKY9X Coding

Re: ERSKY9X Coding

Re: ERSKY9X Coding

Re: ERSKY9X Coding

Re: ERSKY9X Coding

Re: ERSKY9X Coding

Re: ERSKY9X Coding

Re: ERSKY9X Coding

Re: ERSKY9X Coding

Re: ERSKY9X Coding

Re: ERSKY9X Coding

Re: ERSKY9X Coding

Re: ERSKY9X Coding

Re: ERSKY9X Coding

Re: ERSKY9X Coding