Hello everyone!

After assembling my QMX had a bug that caused an over voltage of VDD of 4.2V: /g/QRPLabs/topic/101385726

After fixing this, receiving works fine. However, the QMX behaves strangely when transmitting. Shortly after switching on, transmission is possible. With dummy load in CW I can get a signal in the station transceiver, it sounds clean. According to PSKreporter a FT8 transmission on 40m on magnetic loop is received all over Europe. After a short time (1-2 minutes) it becomes less clear. The signals in CW and FT8 sound unclean and rough, according to PKSreporter I am no longer received. After an even longer time nothing seems to be transmitted at all. Only after switching off the QMX and waiting a while does it work again for a short time.

After switching on, the sweeps on all bands look almost identical to those in the manual. This will remain the case until transmission is started. If, after some time, the signal sounds dirty, the RF sweep will look like this:

The Audio sweep:

These are examples from the 40m band, but this is the case on every band.

To be on the safe side, everything has been done with an input voltage of 7V. I could not detect any abnormal heating of any component. However, the symptoms point to a thermal problem. I hope that the high VDD has not destroyed anything. Does anyone have any idea how I can find the cause of the problem described?

73 - J?rg, DG0JDE

开云体育

Is your host computer a Dell? The Dell audio driver does some funny things with the audio, that might make your power go away...

Good Luck!

73, Paul -- AI7JR

Paul,

Interesting guess about the cause. But no, I don't have a DELL computer. The problem also occurs without a computer, i.e. in CW mode only with QMX, key and antenna.

73 - J?rg, DG0JDE

Perhaps unrelated but on my first QCX+5W it started a WSPR transmission cleanly by degenerated to barely receivable with lots of noise wideband after minute or so?

This was totally stand-alone. ? I traced the transmit problem to the MS 5351A and a tiny puff of duster gas on that chip would put it right for a few moments. ?

RX had also ?a problem and I found the demodulator chip was also faulty at the same time so I think it was a thunderstorm beforehand that was the root cause of both.?

I measured again with the oscilloscope. Also at CLK2 directly at the IC204 the frequency is about double the set frequency. Is there any other explanation than a defect in the IC204?

73 - J?rg, DG0JDE

Hi Hans,

Sorry, I didn't describe it clearly enough: I measured with the oscilloscope while pressing the T key in the diagnostic screen. That is, during normal transmission. It doesn't matter which band I use. I always measure the double frequency.

73 - J?rg, DG0JDE

On Wed, Sep 27, 2023 at 10:48 AM, J?rg / DG0JDE wrote:

I used an oscilloscope to look at the signals at the gate and drain of the Q503-Q506. At the gate everything was fine, even though the RF output became dirty after a while. At the drain, the signals also appeared normal at first. From the moment the RF output became unclean, a twitching could be seen. I can hardly describe this with my limited knowledge of English. I suspect that the output signal is temporarily interrupted.

I did not measure the drain waveforms, but I always measure the output power and the spectrum. I did see some strange phenomenon similar to what you describe, without any previous fault of the power supply units.

I did not have time to do followup investigation but there is something mysterious about the drain side of the final amplifier and the modulator.

I built mine with the WSTS (1:1 transformer for 9V supply) but based on my analysis this transformer can exert some mild but sustained strain on the final transistors and the ferrite core. When I measured the transformer temperature, it gets quite hot.

Also, when the final transistors are dead, the driver can put out on the order of 0.1W and depending on the FET's fault mode enough power can come through the transformer and LPF. So be careful when you test for the presence of the output. The signal coming from the driver has a lot more harmonics than the one from the properly functioning final amp.

Ryuji,

Nothing has gotten hot on my QMX yet because I have only run it on 7V as a precaution. I built it for 12V. Also, I have only run the QMX with a dummy load. Except for a short FT8 test on a magnetic loop that had been precisely tuned by VNA. Therefore, and based on the test results so far, I don't believe that any of the final stages are broken.

But until the double frequency problem is solved, I will not be able to investigate the first transmission problem. There is almost nothing at the output, the LPFs work fine here.

73 - J?rg, DG0JDE

I did it: my QMX is now fixed. I solved the double output frequency problem by replacing the IC204 with a new SI5351A. This also solved the problem with the unclean signal during transmission. I have now called CQ into the dummy load via WSJT-X for 1 hour and the signal remained clean all the time. This weekend I will try some real QSOs.

In summary, Q107, Q109 and IC204 were broken. After replacing them without any other changes, the QMX now works.

73 - J?rg, DG0JDE