开云体育

WARNING!!!

?

Running the SSB calibration in automatic mode has the potential to burn out the finals.

?

I've read at least 1 other thread where running the SSB calibration resulted in failures in the output stage.

?

I now have a non-working QMX+ because of this.

?

I replaced the blown final, but there is still another problem in the transmit circuit.? When I attempt to transmit from the diagnostics screen, I get 0 output power.? If I put a scope on the gates of Q503-Q506, I see what I assume to be a good signal, it's at least at the right frequency.? Q507 heats up when attempting to transmit.? I also note that the duty cycle for the 5V SMPS jumps up to around 75%, I assume because it's trying to push out so much current.

?

I'm kind of stuck here, I have the tools (meter, scope, rework station), but I'm sorely lacking in troubleshooting skills for this type of problem.? This was my second QMX+ build, and I was ecstatic when it worked flawlessly on first power up.? I made some contacts on it, and decided to get it ready to try SSB.? Kicking myself for not reading more before diving in to the calibration or I might have not used the automatic option.

?

I'm going to go read what I can find regarding an issue like this.? If anyone can offer any troubleshooting tips they would be greatly appreciated.? I'm looking for what to check, and what a reasonable result would be.? It's built for 12 volts.? Currently (pun unintended), the supply voltage drops to about 7 volts during transmit when I have the power supply current limit set to 700 mA.? I checked and the unit seems to be able to sink as much current as I dare feed it.?

?

My first unit is on its way to Jeff.? I'm going to work through this one myself, I hope.

?

Thanks and 73.

?

Andy

WB9YXA?

?

?

Andy,

?

please read QMX QMX_plus PA problems, broken BS170.

Maybe the driver IC is not ok. It's also connected to 5V.

It's not enough.

?

73 Ludwig

?

On Mon, Apr 28, 2025 at 02:50 PM, WB9YXA wrote:

Running the SSB calibration in automatic mode has the potential to burn out the finals

Andy, to collect some data regarding this, could you share the following info:

?- What voltage was you power supply at (measured with multimeter)

?- Is the QMX+ a 12V build or 9V?

?- Do you remember what voltage the QMX+ displayed?

?- I presume you had a good dummy load attached,? please check to make sure any adapters/cables used to attach it are not flaky (at least one guy who reported a similar failure found that his bnc adapter was intermittent, and I had a similar issue)

?- What output power was your QMX+ giving on the highest power band?

?

Hans made some changes after the 001 build and again before the 007 build to minimize the chance of this happening, but it's obviously not bulletproof.? Very few people have had trouble,? so it would be good to understand the environment of those who do.

?

Stan KC7XE?

?

Stan

Stan, to explain my BS170's failure environment. On JS8 I can have long QSO's at 5W never an issue. The case gets a little warm to be expected at 5W output.? My BS170's failed each time on SSB callibration ver 004? and the latest 007.

?

Use the QRPLabs dummyload good connection with a short coax cable.

?

The first SSB calibration I used a small linear supply which I always use for operating at my desk.? The 9V wired QMX (ver 4, 80m - 15m) was at 9V to start with then collapsed to 6V when it reached the 15m band using the all automatic 'R' key.? Overheating BS170 failure.

?

The second time with the latest 007 beta software I used 2 LIFPO batteries in parallel with a 0.6V diode as I know it can deliver more current and the power lead is thick and short.? The voltage sat at 9.1V to start with then drops to the high 8V when at full output power using the QMX volt meter. Used the 'O' hey to complete the calibration frame by frame to allow for more cooling time. All calibration frames completed. The output had dropped to 2W testing afterwards from 5W, I did not observe output during the calibration.? Did a factory reset then reloaded 007 repeated the whole calibration again in case there had been a software issue.? Curves looked better than ever and I have them all saved as .jpg which I can email. Another factory reset loaded the pre-beta software no output increase still 2W.? Took the QMX apart, one BS170 had failed (strange failure, Fairchild brand), replaced 4 (now from Hans' stock), power back to 5W.? I always measure the BS170's to have the simalar voltage so this is not a factor. Running on pre-beta 027 software running well like before.? I might upload the SSB software and use it with VarAC without calibration, calibration should not be needed hopefully down the line?? These things take time and effort!

?

My first BS170 failure was with sending through a heavily boosted FT8 signal off a tablet using FT8CN software which every one wrote on this forum should not be possible.? Never tried that again.? So I have been lucky replacing the BS170's 3 times all tracks still intact.

Hans said calibration is for the brave which is true, but 2 BS170 calibration failures I am weary to try soon as I like my functional QMX; don't fix what aint broke.? Then again never say never..? ? ?

It's a 12 volt build.

I was getting 3-5 watts depending on the band.

QMX+ voltage display was a couple of 10ths less than the set supply voltage, which I have verified as being as accurate as my meter.

Testing was done with the QRP-LABS 20 watt dummy load.

?

I read the document that Ludwig referred to.? When transmitting, the signals coming from IC503 are 180 degrees out of phase, but are not symmetrical.

?

Currently, I have a strange symptom.? When I turn the radio on, it draws about 600 mA, and Q507 heats up.? The strange part is that the current drops to around 100 mA when I enter the diagnostic screen.? Upon further investigation, I find that the current output on my power supply shows 652 mA when transmitting, and 652 mA when I'm not sitting in the diagnostics screen.? I find that to be suspiciously coincidental.

?

Once I get to the diagnostic screen, I see the transmitter voltage sitting at 0.4 V.? The wiki document Ludwig referred to indicated that to be problematic.? That voltage jumps to about 0.8 V when I attempt to transmit.? When I transmit, the 5 V SMPS duty cycle jumps to around 75%.

?

I plan on replacing IC503 and even though 3 of the 4 finals tested good, I'm going to replace them as well.? However, I don't see how that's going to remedy the current draw problem when not in the diagnostic screen.

?

Sure would like to understand the mechanism that causes the current to drop when I enter the diagnostics screen.? I don't think I fully understand what number 6 in the wiki doc is trying to say.? It indicates a non-zero transmitter voltage as undesirable, but doesn't really discuss where to look for the problem or how to fix it.

?

Thanks.

?

Andy

?

?

?

Well, I'm stuck.

?

Replaced the finals and IC503, still have the same symptom, high current draw until I go in to the diagnostic screen.

?

I'll ponder this some more,? but I'm stumped as to what changes when I go in to the diagnostics screen.

?

Help?

Andy,

?

When in the diagnostics screen, Q507 is shut down, so close to 0 volts on the drains of the BS170.

When receiving, Q507 turns on, the drains are near the power supply voltage, this drastically reduces the capacitance?

from drain to source across the BS170's and keeps that capacitance from shorting out any incoming signal.

?

Do the BS170's get hot?? If not, then perhaps the drains are shorted to ground somehow,

or perhaps one end or the other of L502.? A quick check from one end of L502 to ground

with an ohmmeter will verify a short (with power off).

?

If the BS170's get hot then is either set of BS170 gates high?

They should not be high when receiving.

?

When receiving, the gates of all BS170's should be around 0 volts, I suspect yours are around 5 volts.

When receiving, the TX line into pins 2,5,10,13 of IC503 should be near 0 volts, those are AND gates so the outputs to?

the gates should also be low (each gate has a low output unless both inputs are up around 3 volts, in which case?

the output goes high to 5 volts).??

?

You can verify all of this with a voltmeter.? Using the voltmeter, you will probably find an error around IC503?

when verifying the second paragraph above.? When holding the board such that the print on IC503 is right side up,

then pin 1 is in the lower left, pin 7 is lower right, pin 8 is upper right, pin 14 is upper left.

https://www.mouser.com/datasheet/2/308/74ACT08-D-1801606.pdf

?

Jerry, KE7ER

?

?

On Mon, Apr 28, 2025 at 08:09 PM, WB9YXA wrote:

On Tue, Apr 29, 2025 at 03:55 AM, WB9YXA wrote:

the transmitter voltage sitting at 0.4 V.? The wiki document Ludwig referred to indicated that to be problematic.?

Andy, 0.4 V in Diagnostics and without transmitting is fine. "0.x V" means all between 0 V and 0.999... V.

?

73 Ludwig

?

OK, so mine failed in exactly this situation. I have yet to replace the finals (the transistors have arrived so I should be able to replace them tomorrow.)

?

QMX built for 12 V supply.

Supply of 10 V from a bench PSU.?

QMX displayed 10 V.

Good Bird dummy load rated for 25 W, adaptors all good.

My QMX is a high-band unit and was giving around 4 W on 10 m.

?

Two of the removed transistors had failed and the other two were fine, I believe both the failed ones were one of the push-pull pairs.

?

I hope that helps!

?

Once repaired I shall avoid doing an SSB calibration!

?

Piggly

I measured the voltages on IC503, and here are the results:

?

?

Is IC501 stuck on?? Its inputs are Vdd and CLK2.? As I look at IC501 and IC502, it would seem that 1 of the two chips would always be 'on' based on the way they are wired.? I tried putting a scope on pin1 of IC501 to see the CLK2 signal, but it just looks like noise on the scope.? I'm still learning how to use my digital oscilloscope, triggering seems a bit more finicky than my old analog scopes.

?

The finals do get warm, it just takes longer for me to feel the heat because of the heat sink.

?

There don't appear to be any measurable shorts to ground at L502.

?

Now what?? There isn't much else to the circuit to cause an issue.

?

Andy

?

?