Hi,

Got a QMX on the workbench, and pulled up the terminal emulator window to run diagnostics. I'm currently seeing a great yellow trace on the ADC/IQ sweep that looks normal, but a very very erratic red trace. Definitely not a sine wave and absolutely not 90 degrees out of phase. Is this looking synonymous with a PCM1804 failure, or T407? (the trifilar one) being wound wrong?

--

Daniel ZL2DTL/ZL4YOTA 

 

 

Could be either one.  You can measure which one with a signal generator and oscilloscope.  Use a signal generator that can put a sine wave into the antenna jack (around 3mVp-p to 10mVp-p), and an oscilloscope to measure at the inputs of the PCM1804.  For example, put a 14.100MHz 5mV sine wave in, tune your QMX to 14.100MHz in CW mode, and you should measure about a 50mV clean sine wave at each of the 4 differential inputs of the PCM1804 (there is a 10x gain through the op-amps just before the PCM1804).  If one of the differential pairs looks bad, then look back, likely to the trifilar transformer.  If they all look good, it is likely the PCM1804 that is bad.

 

You can take this measurement at the PCM1804, pins 4 & 5 (inphase) and the corresponding 4th & 5th pins down on the other side of the chip (quadrature).  Alternately if it is easier, you can measure at the outputs of the op amps IC405 pins 1 & 7 (inphase) and IC406 pins 1 & 7 (quadrature).  If you have a 2-channel scope, you can measure inphase and quadrature at the same time and verify that they are 90 degrees offset.  But if you only have one channel, it is ok, because if they look like good sine waves at somewhere near the expected amplitude, they are likely also offset correctly (and you can see the offset in the diagnostic IQ scan test).

 

Special note: Make sure that you unplug the paddle/key and/or put the QMX in practice mode before doing this measurement.  Because if you accidentally bump the key and transmit into your signal generator, it will likely destroy it.  Don't ask me how I know....

 

Stan KC7XE

I also have a problem with bad image rejection. I/Q signal sweep shows anequal amplitude of I and Q signals. I first checked and rewound the trifilar transformer T401, but without results.

Today I made measurements with an oscilloscope and a signal generator. Signals on pins 1&7 on IC405 and IC406 shows equal amplitude and correct phase shift. I checked the same signals on the PCM1804 pins.

Can I conclude that I also have a problem with the PCM1804 chip?

You can find attached my image and I/Q sweep and oscillograms from pins 1&7 on IC405 and IC406.

 

73 de Zrinko Ivan, 9A7AM

It would be helpful for me to see a 2-channel oscilloscope trace of pin1-IC405 with pin1-IC406, and pin7-IC405 with pin7-IC406, to see the 90-degree offset, and any relative voltage offset between I&Q.  You are showing the 180-degree offset of the pairs of differential pins.

From your I/Q Scan, it looks like the "I" (inphase) signal has attenuation on it compared to the "Q" signal.  This could indeed point to a problem with the PCM1804, but I have never seen reports of it failing in this manner, with just one input reduced in amplitude.   Such attenuation could also come from one poor connection on T401, or something else via the IC405 path into the PCM1804.  It doesn't look like such pre-1804 attenuation is happening using your current screenshots, but it is a bit hard to tell without actually doing the differential sum of the PCM1804 input pairs and comparing them.  But we may be able to see a bit better comparison with the traces I mentioned above.

Stan

 

I took the measurements as you suggested. It seems to me that everything is as it should be. This behavior of the PCM1804 is also unusual for me, an atypical failure. It might be a fake chip.
Oscillograms are attached.

 

73 de Zrinko Ivan, 9A7AM

 

Those certainly look good to me.  I guess the next step is to replace the PCM1804, and hopefully see the problem go away.

One thing to remember in these measurements is that we are pumping in a totally unrealistic RF signal level (S9 +100dB) to enable us to easily measure it, so there is some possibility that there is a small-signal issue that is masked by all of this, like a noisy resistor or capacitor or chip I/O.  But I have no idea how find such a thing, so let's hope it is the PCM1804.

Stan KC7XE

Thank you Stan! I hope changing the chip will fix the problem. I will let you know the result.

 

Zrinko

Hi Stan! I replaced the PCM1804 and now everything works great. I don't know if it is possible that the original TI chip was not installed, but a copy that did not work well? Thank you again for your help!

 

73 de Zrinko Ivan, 9A7AM

Excellent!  I hope you have lots of fun with this great radio.

I would like to make contact with you on the air sometime - I don't think I have ever made a contact with Croatia...

Stan KC7XE

I have no doubt that I will have a lot of fun with this radio. Although I am primarily a CW operator, I am eagerly awaiting SSB. With SSB it will be a complete and unique device.
It would be a pleasure to to me to meet you on the band!

 

73, Zrinko, 9A7AM

Thank you for this thread, both to the OP and for everyone who took time to respond. I had nearly the same symptoms and oscilloscope readings as well as success after replacing the PCM1804. I’m going to make a longer post but essentially I built a second qmx+ and a qmx that both had this same failure mode. I’ve replaced the ic on the qmx with success and will try the qmx+ tomorrow.