开云体育

Hi,

New QMX here, Rev3 2024 received in Late February.? It powered up and everything has gone reasonably well.? However, I can't receive anything under S9.? The noise on the input is a constant solid noise.? I've gone through the operation manual, and done the diagnostics for RF Filter, image, etc, and the results are... bad.? 20db, no image rejection between the two sides, the rf filter sweeps don't start at 0 and go negative, they go between 0 and go +20.? The SWR and LPF sweeps turn out just fine as expected, though the 3rd harmonics seem high, I'm guessing due to whatever is going on.

The operation manual says "check the trifiliar core T401" etc, which I've done.? I have replaced both LM4562's, both 3253's, and here's the big diagnosing issue:? When I short pins 3 or 5 of U405, the audio output goes silent, and the bias voltage (from R401/R402) on T401 gets collapsed to .2V or whatever (The 3253's can handle the <1mA current no problem, so no damage here).? This shows that the 3253 is working.? In addition, the bias dissapears on the other side of the capacitor that is selected by IC402.? Also, a scope shows proper I/Q clock signals from the 5351 going into the mixer/switch.?

When I short pins 3 or 5 of U406, however, there is no change in audio.? The bias voltage all the way to T401 collapses as well.? I thought at first maybe I had a bad opamp or something, so I had replaced them both, but no change.? I measured the resistance between 1&2 and 6&7 on both IC405/406, both showing ~530 ohms (as expected) under no power condition.? Bias voltages are all around 2.45-2.5V as expected, getting passed through L401 to IC407.

The only thing I have left is to maybe swap the PCM1804 codec IC, as this is the only thing I can guess might be failing, but I don't have any of those spare until I make my next digikey order.?

Any suggestions?

Oh I forgot to mention, yes, I checked, rechecked, quadruple checked, more often than I would care to say that T401 trifiliar has the correct orientation/wiring, no shorts, etc.

Hi Martin,

Have you checked Q508?

If you have access to an oscilloscope, you could verify that clock0 and clock1 are 90 degrees out of phase. The same is true for the outputs of IC405 and IC406.

If that checks out, then I suspect the codec (IC407) has an issue.

Hans has a troubleshooting page on his website:



73
Evan
AC9TU

Yes, the radio was in CW mode.? Also, Q508 is good too, I can use the diagnostics to put it into TX mode temporarily by doing the SWR/LPF tests and watch the node at the drain go to 12V and back.

Also, yes, the clock signals are out of phase 90 degrees on my scope, look like nice perfect square waves with little ringing.?

I went through the QMX related posts I could see, have not seen anything similar to this problem in any of the list archives, and the troubleshooting guide doesn't show anything like this.? If the PCM1804 ends up being the issue, I'll take some screen shots of the diagnostics and post a summary at least, assuming I find out what the issue is.?

There aren't really any test points, otherwise I would run the i2s bus through the decoder on a different scope I have access to at work to see if there's a significant difference in the Left/Right channels of the PCM1804/IC407, but I don't see a good way to access all the i2c pins (since only some are shared with IC401 which is an SOIC-8 that's a lot easier to handle).

Hmm, would have to steal a signal generator from work, but can probably also use the TX of my KX3 as well [note to self: need a signal generator for the home lab some day].? I did some testing with the KX3 at zero power output into a dummy load which still produced more than enough signal (Said dummy load is also an attenuator and I've got an SMA output on it), but I never probed the I/Q signals at the PCM1804 (But it gets picked up by the audio of the QMX quite well) as I was at first thinking it was the mixer IC's or something further up the chain.? I'll see if I can get a good signal into the RX_IN node at the BPF and go from there up to ~500mV at the PCM1804, those levels should easily be handled by the entire RX chain.

So here's what I've got so far.? I'm going to see if I can attach a bunch of pictures for all the diagnostics I've done, but it's looking more and more like the PCM1804 is bad.? Here's the kicker picture:

Yellow and blue are the inputs to the PCM1804.? These are pins 1 and 1 of IC405/IC406, VINL+ and VINR+.? The input is generated by running 0.1W from the KX3 into a dummy load and tuning to about the same frequency.? Dummy load is 2x 100W 100ohm resistors (|| to make 50ohm), a 1kOhm resistor, and a 50ohm resistor forming the pi network.? It doesn't matter too much which trace is which, but from my understanding of i2s, logic probe channel 1 is the Left/Right clock [word select].? When LR is high, and the blue trace is at a maximum, I see basically all 1's.? When blue is low, and LR is high, it's all zeros.? I don't see similar behavior with the yellow input and the LR clock being low, it's always a bunch of 1's and zeros, probably random-ish noise I'm guessing.?

I'm also attaching the diagnostics screenshots at the end of this for reference of what I was seeing in the diagnostics.? I only included 40m, they're all basically the same.

I need to add a few other things to my digikey order before hitting send.?

If there are any other suggestions on spare parts for this system, please let me know, if I'm going to build a QMX of Theseus, might as well do it right...

Martin,

I have read one post where a lead was not soldered correctly on one of the large ICs.? It might be a good idea to check the pins of IC407 for soldering mistakes from the factory.?

73
Evan
AC9TU

Evan,

I had already checked, but I doubled checked just now the PCM1804 connections under the inspection microscope.? Each pin got a fine point soldering iron touch to get a good fillet, and a poke with a chisel-tip scalpel to confirm solid connection.? Sadly guessing it's the PCM1804 itself...

Martin,

I am not the expert, but I agree that the PCM1804 is the most likely problem.

73
Evan
AC9TU

Ps. If you have a VNA /antenna analyser that can be set to continuous mode you’ve got a low power signal generator already?

Thanks :)? Maybe I'll customize it with a stencil of a old style ship or something in the end, we'll see!? Digikey parts should arrive tomorrow according to Fedex, in theory, so I'll hopefully know soon.

Alrighty, I got the replacement PCM1804's in, and the replacement went off without a hitch.? I cleaned up the pads, put down a touch of paste, hit each pin with the knife edge soldering iron, gave it a good IPA clean, and it worked.? The only victim of all my debugging was 20m, first initial try and it was getting no receive on that band, which I traced back to having left the BPF split toroid (L401) unsoldered at the mid tap.? Putting that back in and it RX's just fine.? I can't say I enjoy the user interface as much as my KX3, but hey.?

I'll have to start using this bugger and get some time on it now that it's up and running.? Thanks for the suggestions folks.? I'll ask another question in another thread soon likely.

Cheers,

Martin