expressLRS

How do you check the id in the RX? I got the 900MHz version. The RX doesn't put any parameters in the CRSF menu.

The Happymodel EP RX has WiFi and a web interface where I can set the ID.

In the coming ELRS release, betaflight configurator is going to be able to work through this WiFi connection. Awesome.

Looks like the RX menu should show that, but Rx menu on my 2.0.1 Rx wasn't showing any menu data either on deviationTX:

www.expresslrs.org/2.0/software/model-co...g-model-match-number

Reading this it looks like we are supposed to:
1. Connect Rx to tx
2. Set model number in deviationTX model
3. Toggle on model match in eLRS tx settings
4. Rx is automatically programmed with the model number

hexfet wrote: How do you check the id in the RX? I got the 900MHz version. The RX doesn't put any parameters in the CRSF menu.

Thanks for the link. I really need to read the docs first

Model match does work, but an extra step is needed in belrik's list:
2.5 Click Re-Init (select the Re-Init box and press ENT)

This is needed to tell Deviation to send the new fixed ID to the tx module. I'll look at changing the code so the update happens automatically.

I haven't seen Deviation lose connection to the TX, but I'll do more testing. Please let me know if you have some procedure that can make it happen repeatably.

Apparently the RX device in CRSF-device page should show a message about incorrect model number if you try to access it from a model with the wrong number. Otherwise there's no other data shown. Seems odd to me that the RX shows no data but that's the behaviour so far.

The device info reported by the receiver says it has zero parameters, so that's consistent. The model mismatch error comes in the ELRS info message that Deviation is not doing anything with yet. It's on the list.

New test build is available - a29d310
Spoke too soon when I said model match was working, at least fully. The model id was being sent too early on power-up.

I also changed things so no need to press Re-Init any more. If the fixed id is changed, then the new model id is automatically sent to the tx module. Deviation displays a 0 fixed id as "None" but it still should work. 1-63 should be plenty anyway A fixed id of 64 or higher is sent to the tx module as 255, the "Off" value.

The "stuttering" when loading device parameters is related to the higher data rate. I need to add some buffering to fix it.

I tested a29d310 model match on the following scenarios described by ELRS page. www.expresslrs.org/2.0/software/model-config-match/#model-match

Scenario#	TX ModelMatch	TX Receiver ID	RX Model ID	Result
1	Off	Any	Off	Connects / Communicates as usual
2	Off	Any	A	Connects but does not communicate to FC
3	On	Any	Off	Connects but does not communicate to FC
4	On	A	A	Connects / Communicates as usual
5	On	B	A	Connects but does not communicate to FC

1. Correct
2. Correct
3. Wrong. The TX changes the RX ID and Connects / Communicates as usual.
4. Correct
5. Wrong. The TX changes the RX ID and Connects / Communicates as usual.

Sometimes the menu disappears while fiddling around with the ID's.

I think using the re-init key to set the RX-ID is more logical for Deviation than what is happening now.

Even more logical maybe would be to set the RX-ID in the ELRS-RX menu in Deviation? Why is this menu empty?

I also have a Betaflight menu option when I am connected. This menu is also empty. Can this menu be filled with pid profiles etc, etc?

Same here, I also had a CRSF BetaFlight menu option when I was using Tracer. I was not able to do anything with it - just an empty shell.

In the CRSF parameter system the menu is driven by the device. All the labels and options come from the device.

The ES900RX responds to a device info query with the name, which is why it appears in the device list. But that same response says it has zero parameters, so there's nothing to display after clicking on the device name.

I'll look at the model match soon. Working on the parameter loading issue.

The model match id is a parameter of the RX so maybe ELRS should implement a menu for it.

I'm sorry to say I think frsky and opentx is a messy pile of junk. I'm very happy I never had to use any of their freaky abominations.

I agree, this is not a deviationTX item, I will suggest it on the eLRS discord and see what people think.

mooiweertje wrote: The model match id is a parameter of the RX so maybe ELRS should implement a menu for it.

I'm sorry to say I think frsky and opentx is a messy pile of junk. I'm very happy I never had to use any of their freaky abominations.

New test build is available - 6659b95 . Just a small change that reduces the uart errors but doesn't eliminate completely. Still have a couple more things to try.

I wasn't able to duplicate the model match behavior. Both cases 3 and 5 worked as expected, though I don't have an FC hooked up to the check the control connection. The LED on the RX was flashing three times fast which indicates a model mismatch. The RX does not show up on the config page. Telemetry does work though.

Maybe something's off with my test. The procedure I used was:
1. Start with both rx and t8 turned off. They are working with model match on with id 16.
2. Turn on rx, wait for wifi connection.
3. In browser change model id from 16 to 255. Power off receiver.
4. Power on t8.
5. Power on receiver. After a few seconds it starts the model mismatch LED sequence.
6. Power off rx and t8. Power on rx, wait for wifi, verify model id is still 255.
Then repeat except set rx model id to 10 instead of 255.

I was thinking from the perspective that the hardware has no WiFi.
I have several models and want to add a new one. To do so I have to turn on model match on the TX. To be able to do so, I first have to turn off model match and turn on model match again to send model id to the receiver. This is an abomination.

I think I want to turn on both TX and RX and send the ID to the RX with the re-init key. This is how it has always worked with Deviation. Sending the ID to the RX from the RX menu can also make sense.

The model match only blocks communication between RX and FC. It has no effect on communication between T8 and RX.

I just flew 3 lipo's on a race quad using 6659b95. All good. ELRS and Deviation are fabulous.
Anything specific needs testing on the 6659b95? I have never noticed any fluctuations on the rates during flight. But I am still flabbergasted/overwhelmed by the short response time when I fly. Maybe when I am used to the speed I will notice something..

Thanks for the testing! .No, nothing specific to test.

Been spending a lot of time on the serial data errors. One part of the problem is a small bug in ELRS. An ELRS build with this PR reduces the errors. With that change in place I think I see the final boss. At the higher data rate ELRS takes the opportunity to send more than one message in a single interval. Need to think on how best to handle that in Deviation. It may not even happen at 500Hz packet rate - ELRS adjusts what it sends based on how much time is available.

On the model match subject, one thing to understand is that Deviation does not send the model ID to the RX. It sends it to the TX module. The module will try to use it to connect to the RX. If the device menu on the TX is used to change model match from Off to On, the module then will program the RX to the configured model ID.

Also the ELRS module keeps some parameters separate based on the model ID even if no receiver is matched to the model ID. So if model match is on with fixed id 16, if the fixed id is changed to 17 and 17 has never been bound to a receiver, then in the CRSF device menu the ModelMatch will be Off.

So considering the case of adding a new model (receiver) there are four cases: TX model match on or off, and RX model ID at 255 or 0-63.
1. TX model match off, RX with model id 255. This would be the case with a brand-new RX and in Deviation setting the fixed ID to a value that hasn't been bound to any receiver before, or going to the device menu and turning model match off. In this case when both T8 and RX are powered on they will connect. Go to the TX device page and change ModelMatch to On. The RX will be programmed with the selected fixed ID.

2. TX model match off with fixed ID 20, RX with model id 16. Turn on T8 and RX. Go to device page and set ModelMatch to On. RX is bound to ID 20. This takes about 30 seconds with the ES900RX. LEDs go through several states but eventually solid on.

3. TX model match on with ID 16, RX model id 255. Power on T8 and RX. Go to device page and set ModelMatch to Off. Then set ModelMatch On. RX is bound with ID 16.

4. TX model match on with ID 16, RX model id 10. Power on T8. Go to device page and set ModelMatch to Off. Turn on RX and it will connect. On device page set ModelMatch On. The RX will be programmed with model ID 16.


All this seems fairly reasonable to me, but I'm not a UI guy One thing I don't want to do is send CRSF device parameter settings outside of the menu presented by the device. For one thing in Deviation the parameter data only exists while in the CRSF device page. Also it just seems like opportunity for confusion if parameters could be changed in multiple places.

Test build is updated - faea251 . With this build and the ELRS update in the previous post the link runs 1.87Mbps with no errors.