MMDVM Clone No TX/RX - TXCO Issue?

MMDVM_HS Hat hardware
Post Reply
W3MFP
Posts: 4
Joined: Wed Jan 12, 2022 1:07 am

MMDVM Clone No TX/RX - TXCO Issue?

Post by W3MFP »

I have two MMDVM_hs_hat clones that have the same issue and is suspect, suggesting a PCB issue?

Symptom: No TX/RX. No orange LED. Use mmdvmcal and CW shows up way off freq, set it fro 433.0 or 446.0 and spec an shows spike at 484.1 MHz

I have upgraded mmdvm FW and downgrade multible times with no issue, so no issue with communications between the STM32 and RPi0.

I think the issue is with the PCB traces related to the ADF7021 and/or the TXCO, since the FW and STM32 look good, mmdvmcal talks to the ADF7021 (turns CW on/off, RSSI displays and changes,...) but freq on spec an is always 484.1 (also 483.4 ish). Does this 484 MHz look familiar?

I did try the 12MHz FW and it did not change symptoms, so I went back to 14MHz FW. MMDVM shipped with 1.4.17, upgraded to 1.5.2, same symptoms.

The makings on the TXCO do not look familiar.

Maybe G8SEZ could clarify:
G8SEZ wrote: ↑Tue Feb 18, 2020 9:51 am
If the TCXO that clocks the AD7021 RF chip is not running you get very similar symptoms.
viewtopic.php?f=30&t=2438&p=12003&hilit=g8sez#p12003

Any help or thoughts is appreciated.

73s,
W3MFP
W3MFP
Posts: 4
Joined: Wed Jan 12, 2022 1:07 am

Re: MMDVM Clone No TX/RX - TXCO Issue?

Post by W3MFP »

First off, thanks much for the interest and insights!
Exactly what boards? Either find the silkscreen name and board revision number or post a decent photo of the boards or maybe the auction/site where you got them.
https://www.ebay.com/itm/264973895527
Photos later

Funny thing is I bought one from him earlier and it worked fine, so I thought I has safe buying again. The other one with the same symptoms was bought from a different seller. All Chinese clones but shipped from the US (much easier to return).

Board 1: bought from "harhap" - works
Board 2: bought from someone else - does not work
Board 3: bought from "harhap" - does not work

Obviously, Board 2 & 3, from "different sellers" were made by the same morons. Funny Board 1 works at all (very reliable). A different run? All three look identical, with NO manuafature markings, but have pin and LED markings.

A good lesson in Chinese clones ;-(
I take it you mean red LED but exactly which LED are you talking about? the Power LED, the SRV LED, the PTT LED, the DMR LED??
Sorry, Orange LED is DMR mode. Actually somethimes comes on for 30 seconds after two minute boot, then never comes on again. Green power and red "Activity" are normal slow and fast flashing. PTT never comers on.
W3MFP wrote: Fri Jan 21, 2022 3:44 am Use mmdvmcal and CW shows up way off freq, set it fro 433.0 or 446.0 and spec an shows spike at 484.1 MHz
Now that seems super fishy,....
Yes, I agree.
You are hitting 'm' in MMDVMCal, and then pressing the spacebar right?
Yes once, but that test is far too complicated for this issue. I did do it earlier, but with no DMR and PTT LEDs it would never had worked.

I use CW to see if freq changes and it does not, at all, but it does enable TX and kill TX of the spurious signal.

MMDVMcal: I press spacebar to turn on/off CW carrier and that raises 484 MHZ or kills it, so comms to ADF7021 is working. Also if I raise or lower TX freq (433 vs 446 MHz) spurious 484 Mhz does not change. F or f also does not change spurious freq. Using any modulation mode also does not change spurious TX. Almost like random oscillation.
Actually MMDVMCal talks to the STM32, and the STM32 talks to the ADF7021, and then the ADF7021 talks back to the STM32 and back to the software on the RPi.

What happens when you go to different frequencies, like what happens when you go to a completely different band like 2m?
I have not tried 2M, but will soon. I agree swapped parts ...

I think the it is very telling I can enable TX and get RSSI readings, so the STM32 is talking to the ADF7021, but it can not change freq or modulate.

Lack of DMR and PTT LEDs I think shows STM is not getting confirmation from the ADF.
You can't mix and match stuff like that, whatever TCXO the board has the firmware HAS to be configured to match.

Also are you sure you've installed the correctly configured firmware that matches the board you have?
Yes, I have triple checked the FW and am using right one. Since I could not read the part number on the TXCO, I tried the 12MHz version of FW to see if there was any difference on the symptoms, and now change at all. I am back on correct FW. I even tried the FW versions by Dxxx that have GPIO init pins. No change.
Markings don't really matter, the question is what frequency is it oscillating at? Maybe you can get your hands on a scope.
Yes, I have been wanting to do that, but it is under the OLED board and I thought my 12 MHz test proved it was not that. I think it might not be oscillating at all.
Yeah, I dunno where he got that from, since that thread was basically about a serial communications problem between the RPi and the STM32. The TCXO doesn't have any connections to the STM32, so a faulty TCXO isn't going to effect serial coms. It however could be an issue in your specific problem but not in that other persons "failure to init/serial port" problem.
Yes, I think his point is that if the STM32 is working, and the TXCO is not, then you can flash it all you want, but it will no TX.

I was hoping that he would elaborate on TXCO and ADF7021 issues.
User avatar
G8SEZ
Posts: 553
Joined: Fri Apr 13, 2018 8:26 pm

Re: MMDVM Clone No TX/RX - TXCO Issue?

Post by G8SEZ »

KE7FNS wrote: Fri Jan 21, 2022 6:04 pm
W3MFP wrote: Fri Jan 21, 2022 4:47 pm I was hoping that he would elaborate on TXCO and ADF7021 issues.
I guess you'll have to wait for him to see the thread.
My comment was aimed at the point that the ADF7021 serial interface to the F103 CPU on the HAT is clocked by the internal clock supplied by the TCXO.

If something is wrong with the TCXO, i.e. a soldering problem or it has failed, then the HAT can communicate with the RPi but mmdvmhost cannot tell that the ADF7021 isn't working. The CPU has its own 8MHz crystal, if that's not running then the RPi can't see the modem at all.

Hope that's clear enough. I had one of my HATs fail to work in this way a few years ago, I fixed it by fitting a new TCXO, the fitted part was DOA.
--

Brian G8SEZ
W3MFP
Posts: 4
Joined: Wed Jan 12, 2022 1:07 am

Re: MMDVM Clone No TX/RX - TXCO Issue?

Post by W3MFP »

Here are the photos, but unfortunately, the important ADF7021 portion is blocked by the OLED.

https://imgur.com/a/M7Lnkjs

Purchased on eBay at: https://www.ebay.com/itm/264973895527
First one bought in Nov. 2021 works fine. Second ordered from same vendor does not work. Third from another vendor also does not work, but likely from the same clone factory.

To be clear:

Green LED is PWR - On at all times

Red LED is SVC - Flashes slow until STM32 boot is complete, then flashes fast.

Orange LED is DMR - Sometimes lights up about 3 minutes after pistar and mmdvm boot up, but then goes off.

COS, PTT, ... NEVER light up.

I have tried most of the mmdvmcal commands, and the work as the below:

SPACEBAR - turns carrier on and off, but only carrier on spec an is about 484 Mhz, regardless of frequency setting of E command.

D command does nothing, except brings up CW 484 MHz carrier.

M & B commands do nothing, except CW 484 MHz carrier. Useless, as NO modulation works.

C command is useful, as when I use it, then F or f commands, no freq. change occurs.

V command work as one expects. Also verifies my upload of different version.
BTW, I have been using “sudo pistar-mmdvmhshatflash hs_hat” to upgrade FW. I

R command is very interesting. It does display a changing RSSI value, and it does go up or down, if I key up DMR radio, but not very good, just like if t was off frequency, which I suspect it is, since the C command CW carrier tests always are off freq. But I think this does verify serial comms from the ADF7021 to the STM32 do work.


Symptoms Summary:
- Only PWR & SVC LEDs are normal, DMR LED rarely lights, COS & PTT LEDs never light.
- Never have TX or RX on the frequency programmed
- Random freq. 484 MHz will show up on spec an, but never with modulation. SPACEBAR does disable/enable the 484 MHZ carrier.

- FW upgrades and downgrade work fine, implying the STM32 and FW is OK. Delivered with 1.4.17 and upgrade to 1.5.2 fine. Also tried 12 MHz FW, No change. I am back to v1.5.2 14MHz Simplex with: “sudo pistar-mmdvmhshatflash hs_hat”

-R RSSI command does result in a display with changing values, that implies the STM32 can communicate with the ADF7021.

- Two similar boards with no manufacture markings have the exact same symptoms.

The constant 484 MHz spurious carrier sure sounds like there is a wrong component or other issue with the PLL or TXCO section of the ADF7021 circuitry. I am setting up RF test to verify 14MHz TXCO is working or not.

Sorry, maybe I should have called this thread "Possible ADF7021 Design Issue of Clone MMDVM_HS_hat".

Thanks!
W3MFP
Posts: 4
Joined: Wed Jan 12, 2022 1:07 am

Re: MMDVM Clone No TX/RX - TXCO Issue?

Post by W3MFP »

UPDATE of Symptoms:

NEVER say Never ;-). I have since found that the "bad" unit DOES show a orange DMR LED, but only when an active TG is talking, but obviously only when on wifi and thus the DMR BM server has traffic to send to it. But the mmdvm does not TX DMR correctly (details below). Had been testing at work with a spec an and then I had no wifi access, so just doing mmdvmcal RF tests, and thus no DMR LED. I have never seen this unit RX from my MD380 no TX in DMR correctly. I still do not think it is modulating nor setting the frequency correctly. What had been confusing me is that I had been testing the mmdvm board by TX on the MD380 and RX on the mmdvm and it never showed the DMR light and it is trying to RX on the wrong frequency.

Regardless of the UHF freq. set in pistar, 443.0 or 446.0, the unit transmits a carrier only on 484.1 (roughly).

When I set pistar to 144.500 I finally found it on 157.150 MHz. When I set pistar to 144.600 I found the spurious carrier on 157.250 MHz. I will do the MMDVMcal carrier and other tests later, but my attempts to get the MD380 to RX the 157.150 or 157.250 on DMR, neither worked.

The mmdvm does correctly enable and disable the 157.150 correctly, per the COS LED, per traffic on the BM TGs it is monitoring.
Post Reply