Eachine CG023 protocols - what are they?

victzh
Offline

12 May 2015 20:45 #32468 by victzh

Replied by victzh on topic Eachine CG023 protocols - what are they?

It works fine until you get a model with the same protocol but different interpretation of trims - either ratio of trim to stick, or even the sign - one model can add trim value, another subtract!

Please Log in or Create an account to join the conversation.

goebish
Offline
I Void Warranties

12 May 2015 20:51 - 12 May 2015 20:52 #32469 by goebish

Replied by goebish on topic Eachine CG023 protocols - what are they?

Right, but if it works it could be added as a specific protocol option for this model.

Last edit: 12 May 2015 20:52 by goebish.

Please Log in or Create an account to join the conversation.

hexfet
Offline

13 May 2015 00:38 #32476 by hexfet

Replied by hexfet on topic Eachine CG023 protocols - what are they?

Agreed some models could redefine the original use, but a protocol option is a decent solution. The extra agility is a significant difference for some of the small quads. The yaw rates on mine are anemic without it.

Please Log in or Create an account to join the conversation.

goebish
Offline
I Void Warranties

13 May 2015 01:02 - 13 May 2015 01:04 #32477 by goebish

Replied by goebish on topic Eachine CG023 protocols - what are they?

My CX10 (blue pcb

) is yaw anemic too (compared to my v272)...
As I know my RX aircraft ID I can implement it in my TX without waiting for the emulation layer RX code, thanks for the tip, I'll try it

Last edit: 13 May 2015 01:04 by goebish.

Please Log in or Create an account to join the conversation.

goebish
Offline
I Void Warranties

14 May 2015 18:22 - 14 May 2015 19:13 #32567 by goebish

Replied by goebish on topic Eachine CG023 protocols - what are they?

I think I completed the protocol reversal, eg, it should be possible now to fly more than one machine at the same time with Deviation

vasiliy_gr, as you have 2 Devo radios, a EAchine X4 and a CG023, can you check if it works ? (make sure FIXED ID is set to different values on the 2 transmitters)
I can't test myself.

Updated builds available at same url
www.deviationtx.com/downloads-new/catego...-cg023-nrf24l01-test

Last edit: 14 May 2015 19:13 by goebish. Reason: bad url

Please Log in or Create an account to join the conversation.

victzh
Offline

14 May 2015 18:32 #32568 by victzh

Replied by victzh on topic Eachine CG023 protocols - what are they?

I will move the XN297 emulation code to nRF24 source/header as PhracturedBlue suggested so that the development of unidirectional protocols will go on. I will handle reception later - it does not influence the transmission too much.

Please Log in or Create an account to join the conversation.

vasiliy_gr
Offline

14 May 2015 18:33 #32569 by vasiliy_gr

Replied by vasiliy_gr on topic Eachine CG023 protocols - what are they?

goebish wrote: I think I completed the protocol reversal, eg, it should be possible now to fly more than one machine at the same time with Deviation

vasiliy_gr, as you have 2 Devo radios, a EAchine X4 and a CG023, can you check if it works ? (make sure FIXED ID is set to different values on the 2 transmitters)
I can't test myself.

Updated builds available at same url
www.deviationtx.com/downloads...-nrf24l01-test

Certainly I'll do test. But please at first clarify to me about ID-s. I do understand your idea that fixed ID-s should be different. But if I will use non-fixed ID-s - should it also be working simultaneously without collision? Some fixed ID and non-fixed ID?
I am making a test plan...

Please Log in or Create an account to join the conversation.

goebish
Offline
I Void Warranties

14 May 2015 19:06 #32574 by goebish

Replied by goebish on topic Eachine CG023 protocols - what are they?

non fixed ids should work too.

Please Log in or Create an account to join the conversation.

goebish
Offline
I Void Warranties

14 May 2015 19:08 #32576 by goebish

Replied by goebish on topic Eachine CG023 protocols - what are they?

victzh wrote: I will move the XN297 emulation code to nRF24 source/header as PhracturedBlue suggested so that the development of unidirectional protocols will go on. I will handle reception later - it does not influence the transmission too much.

Thanks, I'll wait for that to be in the default branch before pushing anything.

Please Log in or Create an account to join the conversation.

vasiliy_gr
Offline

14 May 2015 21:00 #32584 by vasiliy_gr

Replied by vasiliy_gr on topic Eachine CG023 protocols - what are they?

Ok, I tested those new builds. In general - it is working. Both with non-fixed id-s and different fixed id-s. Two quads worked without interference.

If I started both quads first - they worked with the first TX. If I started quad-TX-quad-TX they worked separately. If I set equal ID-s both quads worked with both TX-s but with half responce (looks funny). So this seems to be correct.

Also mentioned some strange things. But I really do not know, if it is somehow related to the topic we discuss. But I will list them also.

If I used non-fixed ID-s and after initial connect turned off and on TX (quad intact) there was different behaviour on 10 and 7e. With 10 - auto-reconnect, with 7e - no reconnect (quad continue blinking). With fixed ID-s reconnect with both TX-s. It seems to be related to 7e's loadable modules and its memory, I guess.

If I connected two TX-s with different fixed ID-s to two quads and then turned off one of TX-s - their light stayed solid. They started blinking only when I turned off second TX. With non-fixed ID-s every quad started blinking when turning off its personal TX no matter of the second one. May be my fixed ID-s were to close to each other (11111 and 11112)?

Really I never before tried to run two quads simultaneously. So it may be normal. But I mention this as it attracted my attention.

Please Log in or Create an account to join the conversation.

goebish
Offline
I Void Warranties

14 May 2015 21:18 - 14 May 2015 23:20 #32589 by goebish

Replied by goebish on topic Eachine CG023 protocols - what are they?

vasiliy_gr wrote: Ok, I tested those new builds. In general - it is working. Both with non-fixed id-s and different fixed id-s. Two quads worked without interference.

Great.

If I started both quads first - they worked with the first TX.

That's normal (one way protocol).

If I started quad-TX-quad-TX they worked separately. If I set equal ID-s both quads worked with both TX-s but with half response (looks funny). So this seems to be correct.

It is.

If I connected two TX-s with different fixed ID-s to two quads and then turned off one of TX-s - their light stayed solid. They started blinking only when I turned off second TX. With non-fixed ID-s every quad started blinking when turning off its personal TX no matter of the second one. May be my fixed ID-s were to close to each other (11111 and 11112)?

I understand why using ids close to each other can be an issue, I'll fix that, but there'll always be a risk of collision, even with stock TXs

.... looks like there are only 63 possible combinations, that sucks if you buy 2 TXs and they both use the same RF channel ...

The protocol is weak, there's no channel hopping, it only use a fixed frequency for bind stage (chan 0x2D) then it only uses 1 frequency for packets that's computed from txid. I thought txid was 2 bytes long but your tests show it's only 1 (well, 6 bits actually)...

Last edit: 14 May 2015 23:20 by goebish.

Please Log in or Create an account to join the conversation.

goebish
Offline
I Void Warranties

14 May 2015 23:13 - 14 May 2015 23:22 #32590 by goebish

Replied by goebish on topic Eachine CG023 protocols - what are they?

vasiliy_gr wrote: If I used non-fixed ID-s and after initial connect turned off and on TX (quad intact) there was different behaviour on 10 and 7e. With 10 - auto-reconnect, with 7e - no reconnect (quad continue blinking). With fixed ID-s reconnect with both TX-s. It seems to be related to 7e's loadable modules and its memory, I guess.

I think you're right:

txid = (Crc(&Model, sizeof(Model)) + Crc(&Transmitter, sizeof(Transmitter))) ;

This code returns a different value depending of variables address in RAM, which may be different at each startup on the 7e, but stay the same on other Devos until you switch protocols.
This is borrowed from flysky protocol ... maybe this has to be fixed, thanks for noticing

Last edit: 14 May 2015 23:22 by goebish.

Please Log in or Create an account to join the conversation.

PhracturedBlue
Offline

15 May 2015 00:00 #32594 by PhracturedBlue

Replied by PhracturedBlue on topic Eachine CG023 protocols - what are they?

This is a design choice. Most protocols will not re-bind on a Tx power cycle since we tend to use rand32 to seed the lfsr (unless you use a fixed-id). cmpang consistently tells me that we should fix it, but I do not really think it is a bug. If you want your binding to survive a tx reboot, use a fixed-id.

Please Log in or Create an account to join the conversation.

goebish
Offline
I Void Warranties

15 May 2015 00:01 #32595 by goebish

Replied by goebish on topic Eachine CG023 protocols - what are they?

Thanks for clarifying.

Please Log in or Create an account to join the conversation.

goebish
Offline
I Void Warranties

16 May 2015 16:58 #32650 by goebish

Replied by goebish on topic Eachine CG023 protocols - what are they?

victzh wrote: It works fine until you get a model with the same protocol but different interpretation of trims - either ratio of trim to stick, or even the sign - one model can add trim value, another subtract!

Nevermind, I tried and it has no effect on the 3D X4.
I only tried on yaw, but it's not faster with "dynamic trims".

Please Log in or Create an account to join the conversation.

Durete
Offline

16 May 2015 22:42 #32665 by Durete

Replied by Durete on topic Eachine CG023 protocols - what are they?

goebish wrote:
victzh wrote: It works fine until you get a model with the same protocol but different interpretation of trims - either ratio of trim to stick, or even the sign - one model can add trim value, another subtract!

Nevermind, I tried and it has no effect on the 3D X4.
I only tried on yaw, but it's not faster with "dynamic trims".

Some models benefit from "dynamic trims", some others not.
For example, the tamed Syma X11 is a blast flying with DeviationTX. His new brother Syma X13 don't benefit from this trick and fly with the same pitch/roll/yaw rates as with stock TX.

Please Log in or Create an account to join the conversation.

goebish
Offline
I Void Warranties

16 May 2015 22:51 #32666 by goebish

Replied by goebish on topic Eachine CG023 protocols - what are they?

Ok, I'll leave it as a protocol option, disabled by default.

Please Log in or Create an account to join the conversation.

goebish
Offline
I Void Warranties

16 May 2015 23:06 #32668 by goebish

Replied by goebish on topic Eachine CG023 protocols - what are they?

What would be a good name for this protocol ?

Please Log in or Create an account to join the conversation.

Durete
Offline

16 May 2015 23:11 - 16 May 2015 23:17 #32669 by Durete

Replied by Durete on topic Eachine CG023 protocols - what are they?

goebish wrote: Ok, I'll leave it as a protocol option, disabled by default.

I guess is a good solution, because I know at least one more model using this same protocol, the Attop YD829 (in the slow boat to my home after you cracked the frequency hopping pattern

).
Maybe others model benefit from "Dynamic trims"

About the name for the protocol, I have no idea . You are the father of the creature

Last edit: 16 May 2015 23:17 by Durete.

Please Log in or Create an account to join the conversation.

czajunia
Offline

17 May 2015 00:01 #32673 by czajunia

Replied by czajunia on topic Eachine CG023 protocols - what are they?

goebish wrote: What would be a good name for this protocol ?

XN297? Sorry for not being more creative.

Please Log in or Create an account to join the conversation.

Time to create page: 0.076 seconds

Force desktop display