Some new info on the root cause of the QMX smoke. The user had the case off and just installed it. On power up it smoked. Looking at the lower case half I can see a smoke/burn mark where it must have contacted the Pwr Hold pin coming from/to the PS#1 board. Wonder what else besides the PS1 board and the ADC has been damaged.?
toggle quoted message
Show quoted text
Hmmm... I've never figured out how to access the QMX usb on my Mac. I read it can be done, but it's so easy to do on the Pi I've never messed with it on the Mac.
Cliff,
You should put the radio into IQ mode and let Audacity process the incoming I and Q channels off of USB.
From what you said about your sweeps, you may already have the answer, though ;-(
The PCM1804 is quite likely damaged.
JZ John,
RF Sweeps and AF sweeps are pure garbage.?
No scope here.?
Using Raspberry Pi for testing.?
Regarding using Audacity just run audio output from the QMX into Audacity?
I thinking I'll just go ahead and order the IC. Looks very likely it's bad.
RF Sweep
AF Sweep
Roger that.
That check can be done but it is a bit complicated and only works if you can? actually hear signals.
First off, bad IF sweeps are a big clue. If one ADC channel has failed then IF sweeps will show no suppression of the unwanted sideband
If the sweeps are just noisy garbage, perhaps both ADC channels have failed.
Going deeper:
Tune in a constant carrier to produce an audio beat note near 700 Hz.
Using the oscilloscope, verify that clean sine waves exist at pin 1 of IC405 and IC406. The signals should be near 12KHz in frequency and 90 degrees out of phase. If so,? the entire RX chain ahead of the ADC is good.
Via the terminal program,? put the QMX in IQ mode and observe the I and Q channels via the PC Audacity audio app. You should see two channels of approximately equal signal amplitude. Otherwise, something is wrong.
JZ
Ok, no harm in trying.?
Before I order the part for the PS Board, is there a way to check if the PCM1804 is good. Save on shipping if I order both at the same time.
It can be done, Cliff. I successfully did one recently. Of course Jeff Moore is there to help!
JZ Ouch! That one looks like it's beyond me to change out. If it proves bad guess I'll give it a go, but ......
Cliff, yeah, I am also concerned that this is not the last layer of the onion that needs to be peeled down.
Overvoltage on the 5V supply rail tends to take out the PCM1804 ADC chip. Its analog front end runs on 5V, the digital portion on 3,3V. It is the weak link of all the devices that run on the 5V rail. That may be the source of the unstable readings you see with the "good" SMPS module.
Keeping fingers crossed...JZ Ok, thanks. I like Mouser an they ship for $5 and I get it next day as they are close to me. I'll look at Digikey.
If this fixes the PS board then I'm puzzled about what else is not right. Fluctuating 5V rail and won't shut down. Hopefully that is an easier fix. Q108 is pretty tiny. Fun to change it.
Ouch! Cliff, I think you found the culprit.? Digikey has the AO3415.
JZ Yes, Q108 is bad. Measurements 20 - 50 ohms.
Next question, looking at Mouser AO3415A doesn't come up. I get a PJ3415 to come up. Is there a better part number?
Agreed, Cliff. That looks OK. Try the same test on Q108. Let's see what we have there.
JZ John,
Drain-Source - Starts out at 10K then charges up to 22K+ (same regardless of +/- orientation, Gate-Drain 110K, Gate - Source 99K. Seems good to me. ?
Cliff , Remove the module from the main board. Locate Q106. Using an ohmmeter, check the resistance at Q106: gate-source, gate-drain, drain-source. All readings should be above 10K ohms. Any reading much less than that suggests a problem.
John,
How would I do that? Checking in circuit and comparing with the good and bad boards they seem very similar in all combinations of resistance measurements.
Duty cycle =0! That means the MCU is doing all it can to throttle back the 5V SMPS and is finding it has no control at all.
Possibly one or more damaged transistors on the 5V SMPS module.
I would check Q106 for shorts.
JZ Ludwig,
The QMX will not boot at 7V. 9V will let it boot so the screen shot if with 9V.
As it sits there the 5V varies up and dow a little. If I go to 10V in it varies up 5.8 or so. On May 29, 2024, at 15:59, DH8WN via ??< DH8WN@...> wrote:
Cliff,
it looks like a problem in the control loop.
How and where did you measure VCC?
If really 6 V and more D108 on board #1 must be broken or disconnected. But this isn't the reason for the high VCC. It's only a lack of protection.
Please show a screen shot of Terminal - Hardware tests - Diagnostics. But use only around 7 V input to protect the QMX. And what is VCC in this situation.
73, Ludwig
|
According to the picture in the manual, it looks like you start the right side by winding at the bottom and moving up, and the left side by starting at the top and moving down.? In other words they are both wound anti-clockwise as looking at them from the same end.? I'll need to figure this out soon, too!? :-)
|
Hmmm... I've never figured out how to access the QMX usb on my Mac. I read it can be done, but it's so easy to do on the Pi I've never messed with it on the Mac.
toggle quoted message
Show quoted text
Cliff,
You should put the radio into IQ mode and let Audacity process the incoming I and Q channels off of USB.
From what you said about your sweeps, you may already have the answer, though ;-(
The PCM1804 is quite likely damaged.
JZ John,
RF Sweeps and AF sweeps are pure garbage.?
No scope here.?
Using Raspberry Pi for testing.?
Regarding using Audacity just run audio output from the QMX into Audacity?
I thinking I'll just go ahead and order the IC. Looks very likely it's bad.
RF Sweep
AF Sweep
Roger that.
That check can be done but it is a bit complicated and only works if you can? actually hear signals.
First off, bad IF sweeps are a big clue. If one ADC channel has failed then IF sweeps will show no suppression of the unwanted sideband
If the sweeps are just noisy garbage, perhaps both ADC channels have failed.
Going deeper:
Tune in a constant carrier to produce an audio beat note near 700 Hz.
Using the oscilloscope, verify that clean sine waves exist at pin 1 of IC405 and IC406. The signals should be near 12KHz in frequency and 90 degrees out of phase. If so,? the entire RX chain ahead of the ADC is good.
Via the terminal program,? put the QMX in IQ mode and observe the I and Q channels via the PC Audacity audio app. You should see two channels of approximately equal signal amplitude. Otherwise, something is wrong.
JZ
Ok, no harm in trying.?
Before I order the part for the PS Board, is there a way to check if the PCM1804 is good. Save on shipping if I order both at the same time.
It can be done, Cliff. I successfully did one recently. Of course Jeff Moore is there to help!
JZ Ouch! That one looks like it's beyond me to change out. If it proves bad guess I'll give it a go, but ......
Cliff, yeah, I am also concerned that this is not the last layer of the onion that needs to be peeled down.
Overvoltage on the 5V supply rail tends to take out the PCM1804 ADC chip. Its analog front end runs on 5V, the digital portion on 3,3V. It is the weak link of all the devices that run on the 5V rail. That may be the source of the unstable readings you see with the "good" SMPS module.
Keeping fingers crossed...JZ Ok, thanks. I like Mouser an they ship for $5 and I get it next day as they are close to me. I'll look at Digikey.
If this fixes the PS board then I'm puzzled about what else is not right. Fluctuating 5V rail and won't shut down. Hopefully that is an easier fix. Q108 is pretty tiny. Fun to change it.
Ouch! Cliff, I think you found the culprit.? Digikey has the AO3415.
JZ Yes, Q108 is bad. Measurements 20 - 50 ohms.
Next question, looking at Mouser AO3415A doesn't come up. I get a PJ3415 to come up. Is there a better part number?
Agreed, Cliff. That looks OK. Try the same test on Q108. Let's see what we have there.
JZ John,
Drain-Source - Starts out at 10K then charges up to 22K+ (same regardless of +/- orientation, Gate-Drain 110K, Gate - Source 99K. Seems good to me. ?
Cliff , Remove the module from the main board. Locate Q106. Using an ohmmeter, check the resistance at Q106: gate-source, gate-drain, drain-source. All readings should be above 10K ohms. Any reading much less than that suggests a problem.
John,
How would I do that? Checking in circuit and comparing with the good and bad boards they seem very similar in all combinations of resistance measurements.
Duty cycle =0! That means the MCU is doing all it can to throttle back the 5V SMPS and is finding it has no control at all.
Possibly one or more damaged transistors on the 5V SMPS module.
I would check Q106 for shorts.
JZ Ludwig,
The QMX will not boot at 7V. 9V will let it boot so the screen shot if with 9V.
As it sits there the 5V varies up and dow a little. If I go to 10V in it varies up 5.8 or so. On May 29, 2024, at 15:59, DH8WN via ??< DH8WN@...> wrote:
Cliff,
it looks like a problem in the control loop.
How and where did you measure VCC?
If really 6 V and more D108 on board #1 must be broken or disconnected. But this isn't the reason for the high VCC. It's only a lack of protection.
Please show a screen shot of Terminal - Hardware tests - Diagnostics. But use only around 7 V input to protect the QMX. And what is VCC in this situation.
73, Ludwig
|
Thanks JZ. Finally happy to solder T507 in place.
--
Phill EA5JHA
|
Is there any where that matching jumpers (JPxxx) and respective pins are listed for the various kits? It is often difficult to match them up on the schematic pages. Maybe there is some logical way that they are all connected, but it has escaped me.?
--
73, Dan? NM3A
|
Was it built form a kit did it work then failed later.
Was it ready made working then failed later. Or has it never worked.
I can look at it for you?
I am in the UK.?
Philip de G7JUR
|
Re: Reasonably priced CW paddles
All of my paddles are hand-me-downs, except for my Nye Viking paddle/keyer that I bought for work. My favorite paddles right now are my CT-73MX from UR5CDX.
I've been using my W5JH Black Widow paddles for POTA, but I'd like to find something better, as I'm hoping to do a SOTA peak while I still am able...
Russ, va3rr
toggle quoted message
Show quoted text
On Fri, May 31, 2024 at 07:39 AM, KY4LV wrote: I appreciate the laugh I got from seeing the expensive paddle folks swing by
this post.
|
Re: Reasonably priced CW paddles
I guess the concept of "reasonably priced" varies quite a bit. For me, I consider BamaTech and Begali to be on the higher end of paddle pricing, with the TP-III and the Simplex clocking in over $200US shipped here. Certainly not the elite pricing tier, but a lot for CW paddles. I can appreciate the desire to have nice things, though, but it doesn't seem like what the OP was asking about.
On the reasonable end (aka cheap), I like the CW Morse paddles a lot. The Ali Express metal paddles are pretty good as well. Either will get the job done if that's what you are looking for.
Brian
n1bs
|
Re: Reasonably priced CW paddles
I rarely, if ever, use the electronic keyer, but I use a sideswiper that could be used with it if I wanted to. You might consider making your own key. It's just a switch after all.
--
73, Dan - W2DLC
|
Re: Reasonably priced CW paddles
I'll echo the votes for CW Morse and VK3IL paddles.
Also N6ARA sells kits and complete tiny paddles for about $20. They're great performers.
I got some AliExpress paddles and had to order two sets, make a complete working unit out of the two, and send one back. They're only a good value if you get a working one.
I appreciate the laugh I got from seeing the expensive paddle folks swing by this post.
|
Re: Reasonably priced CW paddles
Tom, et al,
I have the Begali Simplex and I love it. However, it's not exactly what I want to haul out to a part or summit. I've tried a few, but hands down, the best one by far for portable ops is the BamaTech TP-III.?
--
73, Dan? NM3A
|
Re: Low Supply of 0.6mm Copper Wire
On Fri, May 31, 2024 at 09:40 AM, Hans Summers wrote:
The cause is coupling between the transformer windings. You can improve it by trying to push the 0.6mm wire single-turn further away from the 10-turn 0.33mm wire turns. Having a piece of insulated hookup wire would tend to improve that separation automatically because the insulation takes?up a little space so enforces at least a small separation.?
But depentend on the isolation material the capacity between the windings could be significant higher than without.
|
Re: Z-match tuner in matching enclosure to QMX
#qmx
#z-match
Hi Michael
I am writing to you off-list.?
On Fri, May 31, 2024 at 11:48?AM OZ1MIC, Michael via <Michael= [email protected]> wrote: Hi Hans?
Thank you for your quick reply.?
The idea of using two bottoms would be doable, but I suppose I would need to purchase two QMX enclosures, on order not to leave you with two front parts?
The side panels, I can make myself.
73 de OZ1MIC, Michael?
|
Re: Z-match tuner in matching enclosure to QMX
#qmx
#z-match
Hi Hans?
Thank you for your quick reply.?
The idea of using two bottoms would be doable, but I suppose I would need to purchase two QMX enclosures, on order not to leave you with two front parts?
The side panels, I can make myself.
73 de OZ1MIC, Michael?
|
Re: Reasonably priced CW paddles
BaMaKey paddles are very high quality and reasonably priced. 73
|
Re: Z-match tuner in matching enclosure to QMX
#qmx
#z-match
Hello Michael
There is no source for a raw enclosure; they are custom-made for QRP Labs according to our specifications, by a factory in China. I think about the best you could do would be to use two bottom panels of a QMX, these are undrilled and unprinted. But I do not have end panels...?
On Fri, May 31, 2024 at 11:10?AM OZ1MIC, Michael via <Michael= [email protected]> wrote: Hi?
@Hans
I would like to make a Z-match antenna tuner for the QMX.
Is it possible to buy a "raw" matching enclosure (no print and no holes) or could you point to your source.
Thank you?
/Michael?
|
Phill,?
The two winding scenarios you describe produce identical end products. Set side by side, you could not tell the difference between them.? Overthinking? Umm, probably :-) JZ
Thank you, Gentlemen.?
I'm still not clear but that says a lot more about my failings than your explanations.
If I lay the binocular flat on the bench with the holes pointing towards me it seems there are two ways to wind it.
1. Start both windings at the end nearest me, pass through and turn both outward, away from each other, Back towards me and then back in at the bottom.
The right hand "tube" is therefore wound clockwise whilst the lefthand is wound anticlockwise.
2. Wind the righthand tube as above but start the lefthand tube from the other end. In this case both are wound clockwise.
I know I might be over thinking this but in the first case the magnetic fields will be opposite whilst in the second they are aligned.?
I appreciate Hans might have allowed for this on the pcb so my question, before I solder it in is still which way should I wind it??
Kind regards
PS I know, I should have done EE not Chemistry!!!
--
Phill EA5JHA
|
Hi?
@Hans
I would like to make a Z-match antenna tuner for the QMX.
Is it possible to buy a "raw" matching enclosure (no print and no holes) or could you point to your source.
Thank you?
/Michael?
|
Thank you, Gentlemen.?
I'm still not clear but that says a lot more about my failings than your explanations.
If I lay the binocular flat on the bench with the holes pointing towards me it seems there are two ways to wind it.
1. Start both windings at the end nearest me, pass through and turn both outward, away from each other, Back towards me and then back in at the bottom.
The right hand "tube" is therefore wound clockwise whilst the lefthand is wound anticlockwise.
2. Wind the righthand tube as above but start the lefthand tube from the other end. In this case both are wound clockwise.
I know I might be over thinking this but in the first case the magnetic fields will be opposite whilst in the second they are aligned.?
I appreciate Hans might have allowed for this on the pcb so my question, before I solder it in is still which way should I wind it??
Kind regards
PS I know, I should have done EE not Chemistry!!!
--
Phill EA5JHA
|
Is there anyone in the UK who can fix non functioning QMXs?
|
Dj Merrill
Phill
va3rr
Daniel Conklin
KY4LV