Thread Tools
Jul 29, 2018, 01:08 AM
Registered User
Thread OP
Discussion

F-port on R-XSR via bi-directional inverter not working.(Solved)


Edit: Solved.See post #10

Hi friends, Please help me to solve this weird problem.

I built this bidirectional inverter and tested it on X4R receiver(flashed with Fport firmware) connected to SpracingF4evo @UART4 and its working great.I got telemetry and Rx input working good.

Then i wanted to do the same thing on my R-XSR .SO i flashed it with latest Fport firmware and swaped it with X4R.But it refused to work??I only got Receiver input signals but no Telemetry.I am running latest Inav firmware on my F4 board.So i know that my inav,F4 board and my DIY built inverter works(tested on X4R Rx).I spent lots of time flashing,reflashing my R-XSR but telemetry did not work.

Then just for test , i hooked up R-XSR to my F3 board without inverter and it works??So now i know there is no problem with R-XSR receiver.So now i am stumped.

Then i did another test.I soldered a wire to the uninverted pad 'P' and connected it to the Tx of UART (after the inverter towards flight controller board and while inverter is in place)and it worked.Then both receiver and telemetry worked.Please help me to find the problem.I have a access to the oscilloscope if anything i can do with it to diagnose the problem.So here is the situation:
Fport on :

X4R---bidirectional inverter-----F4 with inav-------works
R-XSR---bidirectional inverter-----F4 with inav-------Not working(only receiver works but no telemetry)
R-XSR------F3 board(without inverter)---works
R-XSR---bidirectional inverter-----------------F4 with inav-----works
|---->wire from uninverted 'P' pad to Tx ---^|


Thanks in advance.
Last edited by ninja_zx11; Aug 02, 2018 at 01:07 PM.
Sign up now
to remove ads between posts
Jul 29, 2018, 06:39 AM
Carbo (-nator) ;-)
I can confirm the issue, eventually it is a level issue with the inverter. But why not simply connect P pad with UART TX without additional inverter and switch "set serialrx_halfduplex = ON"?

Inverter is only useful imo, if there is no inverted SPort/FPort pad available, like with R9 mini (where this inverter works fortunately).
Last edited by Allerhopp; Jul 29, 2018 at 06:56 AM.
Jul 29, 2018, 07:58 AM
Registered User
Thread OP
Quote:
Originally Posted by Allerhopp
I can confirm the issue, eventually it is a level issue with the inverter. But why not simply connect P pad with UART TX without additional inverter and switch "set serialrx_halfduplex = ON"?

Inverter is only useful imo, if there is no inverted SPort/FPort pad available, like with R9 mini (where this inverter works fortunately).
Thanks a lot Allerhopp.

The reason I am trying to use inverter because I read somewhere here on Rcgroups that there was some reason why futaba designed it with inverters and bypassing inverters can compromise the signal. I am no expert and have no idea about it.Was just trying to play safe.
If I won't be able to solve the problem,then I will use uninverted signal from pad P.

Is there anyway we can trace and correct the level issue?Did you mean the voltage level issue.
I have access to oscilloscope and I need guidance to trace the issue.Thanks again.
Jul 29, 2018, 08:29 AM
Carbo (-nator) ;-)
If you want to find the reason, you can compare FPort Hi/Lo signal levels between the two RX when connected to the bidirectional inverter.

I do not think there is a good reason for inverted UART levels/logic for such a short connection. Positive said: UART level/logic will work for 5". I would always prefer to reduce complexiity, if possible.
Jul 29, 2018, 01:17 PM
Registered User
Thread OP
Quote:
Originally Posted by Allerhopp
If you want to find the reason, you can compare FPort Hi/Lo signal levels between the two RX when connected to the bidirectional inverter.

I do not think there is a good reason for inverted UART levels/logic for such a short connection. Positive said: UART level/logic will work for 5". I would always prefer to reduce complexiity, if possible.
Both signal levels are found same.

R-XSR --No telemetry,only Rx inputs working
X4R--All working

-------RXSR--------------X4R
Fport-3.3v---------------3.3v
Rx-----4.8v---------------4.8v
Tx-----3.4v---------------3.4v All voltages are measured with oscilloscope.

Only difference i found when RXSR working after connecting uninverted wire from pad P to Tx of FC,is that the min volt jumps up around 1/2 v.You can see the difference in the traces i attached.
Last edited by ninja_zx11; Jul 29, 2018 at 01:23 PM.
Jul 29, 2018, 02:32 PM
Carbo (-nator) ;-)
In fullduplex mode/halfduplex off, we have this procedure:

RX FPort sends control data, it gets inverted and routed to FC RX port. FC TX port sends then telemetry data, it gets inverted and appears on FPort, where RX listens to it.

OK, we know, the only difference in setup is RX. If you watch the telemetry signal part, which appears simultaneously on FC TX and FPort, you can see FPort telemetry data level. And this level is OK for X4R, but not for R-XSR.

In the end you found out, a R-XSR needs a higher signal level on FPort/SPort than a X4R.

If you want to dig deeper, you should flash SPort firmware again and watch the SPort level of a original FrSky sensor. Btw. FrSky SPort passthrough telemetry is also not working with R-XSR and this inverter. With a X4R it works also ....

But again, it is not worth the effort imo, better use the P-pad. I would apprecuiate a link to the "bypassing inverter issue", to find out the arguments.

Edit: Eventually this is also a explanation for some inexplicable sensor lost messages, i have experienced occasionally. I will start digging ...
Last edited by Allerhopp; Jul 29, 2018 at 02:56 PM.
Jul 29, 2018, 07:33 PM
Registered User
Thread OP
Thanks for the explanation!
Do you think it would be a good idea to leave the inverter in place and connected and just jump the P pad of receiver straight to the Tx port of FC.In this way the receiver signals will pass through the inverter and will be inverted properly and who cares if there is a occasion glitch (only if bypassing inverter creates any noise or something) in the telemetry signal travelling through the P pad jumper.But again i have no knowledge of it and have seen so many people flying with uninverted hacked receivers.

Quote:
Originally Posted by Allerhopp
I would apprecuiate a link to the "bypassing inverter issue", to find out the arguments.
Did you mean to creating a new thread?Sorry i didn't get it.
Jul 29, 2018, 11:52 PM
Carbo (-nator) ;-)
Quote:
Originally Posted by ninja_zx11
Did you mean to creating a new thread?Sorry i didn't get it.
My wording in english is sometimes strange, i know, sorry. I meant your remark here:

Quote:
Originally Posted by ninja_zx11
The reason I am trying to use inverter because I read somewhere here on Rcgroups that there was some reason why futaba designed it with inverters and bypassing inverters can compromise the signal.
I would like to read the arguments. I do not understand, why it is better for the signal, to pass two inverters (RX and diy), instead of using the original signal from P-pad. P-pad gives access to the signal before it is inverted by the R-XSR internal inverter. IIRC there was a reason for inverted logic in longer transmission routes, but within a few inches it does not make a difference imo.
Jul 30, 2018, 11:26 AM
Registered User
Thread OP
Quote:
Originally Posted by Allerhopp
My wording in english is sometimes strange, i know, sorry. I meant your remark here:



I would like to read the arguments. I do not understand, why it is better for the signal, to pass two inverters (RX and diy), instead of using the original signal from P-pad. P-pad gives access to the signal before it is inverted by the R-XSR internal inverter. IIRC there was a reason for inverted logic in longer transmission routes, but within a few inches it does not make a difference imo.
Yes sure i will start a new thread.But i found this one with some discussions.

Just wondering if you had same problem of using inverter with R-XSR receiver.Thanks
Jul 31, 2018, 08:32 PM
Registered User
Thread OP
Thanks a lot to 'teralift' aka 'jflyper' for giving me suggestion of replacing the resistors on the Tx side of the inverter.Its working now!!
On TX side, he suggested me to
substitute:
4.7K ---> 330R
47K ---> 4.7K

@Allerhopp I am sorry i didn't get proper trace at first.I was not capturing the complete signal.I should have zoomed out more.You were right with logic level issue!!Thanks
Last edited by ninja_zx11; Aug 01, 2018 at 12:12 AM.
Jul 31, 2018, 11:28 PM
Carbo (-nator) ;-)
Good catch! I am still wondering, what may be the advantage of this signal path
R-XSR GPIO <--> R-XSR inverter <--> DIY inverter splitting signal in RX and TX <--> RX and TX FC GPIO
over this signal path:
R-XSR GPIO <--> FC GPIO

The dual inverter solution is equivalent for the signals, only more complex imo. If you still have contact, can you ask?
Aug 01, 2018, 12:10 AM
Registered User
Thread OP
Quote:
Originally Posted by Allerhopp
Good catch! I am still wondering, what may be the advantage of this signal path
R-XSR GPIO <--> R-XSR inverter <--> DIY inverter splitting signal in RX and TX <--> RX and TX FC GPIO
over this signal path:
R-XSR GPIO <--> FC GPIO

The dual inverter solution is equivalent for the signals, only more complex imo. If you still have contact, can you ask?
Yes i will ask.

Too bad frsky didn't release any info about inverted or non inverted signal integrity.Yes you are right,more components between Rx and Fc,more chances of failures.
Aug 02, 2018, 04:35 AM
AKA jflyper

How to derive proper resistor values


Glad to hear the inverter working.

Following is the trick I've used when I derived the values for my X4R-SB.

Take a look at this drawing.

Name: Document_pdf(1ページ).png
Views: 160
Size: 264.4 KB
Description:

- Top schematic is a DC-equivalent circuit of the inverter's tx side, connected to a receiver's s.port signal pad.
  • Small dotted circle labelled Hi-Z are circuit connected with very large resistance (impedance), so we can ignore them.
  • Large dotted square is the receiver. It basically has a series resistor to protect the rest of the circuit and a pull-down resistor. Since rest of the receiver is Hi-Z, these two resistors connected in series to form Rs against ground.
  • Resistor R1 comes into account only when the switch is closed (TxD = Hi).

- Lower left is a direct rewrite of the top schematic.

- Lower right has two schematics, with switch open (TxD = Lo) and closed (TxD = Hi).

By simple synthesis of series and parallel resistors, we can define VL and VH as in the drawing.

Now, we are ready to derive R1 and R2.
  1. By actually measuring VL with some R2, we can compute (or estimate) the value of Rs (Solve VL equation with VL = measured voltage for Rs).
  2. Then, using the value of Rs, desired value of R2 can be computed by solving VL equation with VL = target high voltage.
  3. Lastly, using the value of Rs and the R2, desired value of R1 can be computed by solving VH equation with VH = target low voltage.
Done.
Last edited by teralift; Aug 02, 2018 at 09:55 AM.
Aug 02, 2018, 06:17 AM
Carbo (-nator) ;-)
Very impressive!

In the end it means, R-XSR has another input circuit than X4R. But still one question is not answered: If you have access to the signal on RX pcb, before it is inverted, it there nevertheless an advantage, to use the dual inverter signal path? I do not know one argument.

TIA
Aug 02, 2018, 06:47 AM
AKA jflyper
Quote:
Originally Posted by Allerhopp
Very impressive!

In the end it means, R-XSR has another input circuit than X4R. But still one question is not answered: If you have access to the signal on RX pcb, before it is inverted, it there nevertheless an advantage, to use the dual inverter signal path? I do not know one argument.

TIA
If an access to non-inverted signals are available, there is no reason to use the inverter.


Quick Reply
Message:

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion R-XSR to CL Racing F4S FC F.Port Question Malito77 Multirotor Drone Electronics 1 May 28, 2018 09:42 PM
Discussion SP3 - BF 3.3.0 - R-XSR F Port lyvfd8142 Multirotor Drone Electronics 3 Mar 04, 2018 11:17 PM
Help! Kakute F4 AIO V2 not working with Frsky R- XSR Bugs You 2 Multirotor Drone Electronics 0 Feb 03, 2018 01:18 PM
Discussion FRUSTRATED...Switches will not work on Vector via Tanaris STUCARIUS Eagle Tree Systems 7 Feb 29, 2016 08:51 PM
Discussion Tarot 2D Gimbal motors not working. can't connect to motor port via usb JoopDrum Multirotor Drone Talk 3 Sep 15, 2015 11:32 PM