开云体育

Greetings all,

Last fall, I experimented with creating a QDX that could operate on 40, 20, and 15m. I wound L12 normally, only I stopped at the 40m tap, as I used a tap at 16 (I think?) turns and connected it to the PCB pad for the 80m tap for my 15m tap. I then assembled the 40m and 30/20m LPF networks as per the manual, but with the 80/60m LPF network being one designed for 15m (see discussion threads last fall). Due to a mistake in installing T1 (I did not get enough of the enamel removed, so it wasn’t passing voltage properly) and some cascading issues, I was never able to get it on the air. I was able to get good audio, image, and RF sweeps for 40/30/20m, but 15m, in addition to showing a lot of loss, never produced acceptable results.

In the process of troubleshooting (I was getting good sweeps on 40/20, but hardly any output into a dummy load), I ended up replacing the finals and IC5 (IC5 was the actual culprit). Afterwards, when probing the finals, I made a Fatal Mistake - my probe touched 2 pins and magic smoke appeared :-\ Replacing the finals and IC5 again, I found that while the microprocessor still worked and I could run all of the diagnostic sweeps, the QDX would not transmit and the sweeps looked horrible (see attached pics). For various reasons, I chose to just order a new QDX, set aside the original one, and revisit it at a later date.

As now is that later date, my first step is figuring out what is going on with the sweeps. Last fall, at the advice of one of the fellow forum members, I removed Q6 (the T/R switch) and put a 0.1uF capacitor across the Source and Drain pins (see attached pic), thus putting the QDX in a receive-only mode, and removed the finals and IC5. The attached sweeps were taken afterwards.

My immediate question is, given the sweep results that were produced, did I also fry IC3 or (maybe) IC4? Looking at the schematic, tracing from T1, I can see how Q6 could have been fried, which could have also taken out the TXCO, and then IC3. How can I check to see if IC3 (and IC4, if applicable) is operational? What about the TXCO (assuming it could have been damaged)? I don’t mind rewinding L12, winding it for the original spec, if that will help.

I appreciate the help!

Charles Johnson
KF4AYT

**

Q6 Bypass




Band Config

Note: This was after a factory reset when I was trying to determine why it wasn't transmitting.




Audio Sweeps





RF Sweeps

Thoughts from anyone? :-)

Thanks!

Charles Johnson
KF4AYT

Hi Charles,

I would suggest that IC4 is the culprit.? It is a Tayloe detector that provides the input to the I/Q sampling codex.? It could also be the op amps after (IC7 and IC9).? All three are run from the 5 volts line that got jolted with 12 volts when IC5 failed.? IC3 is a static selector for the BPF.? If this device failed, you would have significantly lower values for the magnitudes of the sweeps.? That does not mean it, too, is fried.? I would not start with that device.

Assuming the receiving current is under 160ma, you can test IC3 by measuring the voltages on the BPF capacitors.? Only one should have 2.5 volts.? The others should be close to 0 volts.? Check capacitors C39, C40,? C41, and C43.? They should all have 2.5 volts.? If not, then IC4 is most likely bad.

There is a troubleshooting page on the QRP-Labs webpage:


Do not transmit with the Q6 bypass.? That will cause issues with IC3, IC4, and maybe IC7 and IC9.

Above are my thoughts, as requested.? Please verify for yourself.
73
Evan
AC9TU

Evan,

Assuming the receiving current is under 160ma, you can test IC3 by measuring the voltages on the BPF capacitors.? Only one should have 2.5 volts.? The others should be close to 0 volts.? Check capacitors C39, C40,? C41, and C43.? They should all have 2.5 volts.? If not, then IC4 is most likely bad.

Having put the QDX-in-question on the bench (and in the Pana-Vise ;-) ), here are the observations I made:

?- Using a DROK buck converter that displays current draw to 2 decimal places (though exactly how accurate it is, I don’t know), with the supply voltage set to 9.00V, the receive current draw was fluctuating between 220 mA and 230 mA.

?- While the 5V regulator was outputting 5.00V (according to my Klein DMM) and the 3.3V regulator was receiving 5V and outputting 3.3V, the 5V regulator was overheating - very uncomfortable to the touch (as in ‘finger tip immediately jerks back’).

?- Measuring the BPF capacitors, one had 2.45V (measured on the side nearest IC3) while the others measured ~0V (I’m wanting to say maybe 50 mV?). As I did not have a computer connected, so as to change bands, I can’t definitively say that IC3 is working, but it would seem that way.

?- Measuring the capacitors connected to IC4 (on the side nearest IC4), C39 and C40 both metered 2.45V, while C41 and C43 metered only 0.9V. Based upon your above statement, that leads me to believe that IC4 is bad.

FYI, the PCB is a Rev 3A with the C41 via short cut effected and C71-C74 shorted with solder blobs.

The last time I recall dealing with an overheating 5V regulator was the first time that IC5 went out, so if IC4 is bad, A, would that explain the high current draw in receive mode, and, B, where else should I check for possible culprits? I ask b/c if I decide to revive this board and replace IC4 (at least), I’d like to make only one Digikey order :-)

-------

John,

I felt IC10 and it -may- have been a tad warm, but it was definitely not overheating.

-------

Regards,

Charles Johnson
KF4AYT

PS For the record, I had a 50-ohm BNC terminator connected to the antenna connector during the above tests.

Charles,

Based on your measurements, it could be that IC4 is bad.? It could also be IC5, IC7, IC9, or any of the ICs on the 5-volt line.??

I would lower the voltage to 7.5 volts to minimize the power dissipation on the 5-volt regulator.? It cannot go below 7 volts, or the regulator will not function properly.

Check pins 7 and 9 of IC4.? Both should be at 2.5 volts.? Same for IC3.

Check the gate voltages on the BS170s while receiving.? If any are not zero, then pull the BS170s and test again.? The BS170s are more accessible to replace than IC5.

Do a message search for IC7 and IC9 voltages.? That should bring up one of the prior posts where the pin voltages were measured.? Here is the start of one that I found:
/g/QRPLabs/message/95581

73
Evan
AC9TU

Evan,

Could Charles' clock generator have failed? That might be another way
that the Tayloe detector is not passing 2.5 volts out to all four
caps. Stuck on one decode.

JZ

John,

A failure of one of the two orthogonal outputs of the Si5351 could explain the symptoms.? He measured 2.5 volts on two of the capacitors.? A total failure would have 2.5 volts on only one of the capacitors.

He could measure the voltages on pins 2 and 14 of IC4 to see if they are 1.65 volts (half of the 3.3 volts peak squarewave output of the Si5351 chip).? Verification with an oscilloscope would be better.

73
Evan
AC9TU

Evan,

He could measure the voltages on pins 2 and 14 of IC4 to see if they are 1.65 volts (half of the 3.3 volts peak squarewave output of the Si5351 chip).

With the supply voltage set to 7.5V (the buck converter was reading ~7.42V output under load) and a current draw of ~210 mA, I metered Pins 2 and 14 of IC4, though I metered on the side of R40/41 that is away from the IC - I was afraid my DMM probe might slip and short something out (particularly if I tried to measure Pin 14 directly). Given that they are only 47 Ohm resistors, for these measurements, I figured that the effect of the resistors would be negligible.

As for my results, using a Klein DMM set to DC, I measured 3.170V on R40 (Pin 2 for IC4) and 3.146V for R41 (Pin 14). I haven't tried to do the math for the voltage drop across a 47-ohm resistor, so perhaps what I measured was not what IC4 is seeing, but that's what I measured :-)

Re: checking the gate voltages for the BS170s, there are currently no BS170s or IC5 installed. As for checking Pins 7 and 9 of IC3 & IC4, I forgot to do that, but will when I'm back in my workshop.

Regards,

Charles Johnson
KF4AYT

Evan et al,

Setting the QDX up on the workbench and getting out my DMM, with the QDX set to the 40m band (the default freq of 7.074, I believe), here are the requested measurements for IC4 and, just for good measure, ICs 7 & 9.

?

IC4

----

?

Pin 7 - 2.492V

Pin 9 - 2.492V

Pins 2 & 14 (Clk 0 & 1) - 3.170V (measured on the IC pin)

?

IC7

----

?

Pin 1 - 2.492V

Pin 2 - 2.492V

Pin 3 - 2.492V

Pin 4 - 0V

Pin 5 - 2.492V

Pin 6 - 2.492V

Pin 7 - 2.492V

Pin 8 - 5V

?

IC9

----

?

Pin 1 - 0.824V

Pin 2 - 0.920V

Pin 3 - 0.915V

Pin 4 - 0V

Pin 5 - 0.916V

Pin 6 - 0.921V

Pin 7 - 0.837V

Pin 8 - 5V

?

----

?

While I do not really understand what is going on in IC4, as I understand it, it is producing 2 audio signals which are then fed through IC7 and IC9 (audio op-amps) before reaching the ADC. The fact that the measured voltage levels for ICs 7 & 9 are pretty much 'what comes in is what goes out' tells me that something is wrong, for with the positive and negative inputs both the same, should they not be cancelling each other out and producing 0V on the outputs?

?

Hans, in a support message, suggested putting an o'scope on the Clk0/1 lines (among others), but until I can replace my non-working 465B (which should be soon!), all I can do is to theorize that since the measured voltage is 3.170V, which is not -quite- Vdd, perhaps a clock signal -is- being output, but rather than a nice square wave with the signal at 3.3V and 0V for equal halves of the time period, perhaps the signal is at 3.3V -most- of the time, but does go to 0V for short periods of time? It wouldn't be acceptable for a clock, to be certain, but it would help to explain my measurements :-)

?

I appreciate the help and insight!

?

Charles Johnson

KF4AYT

To all,

My apologies for the large type in my previous post. I wrote it in the Notes app on my iPad and when I copied and pasted it, despite my best efforts, I couldn't make it smaller :-/

Regards,

Charles Johnson
KF4AYT