开云体育

ctrl + shift + ? for shortcuts
© 2025 开云体育
  1. QRPLabs
  2. Topics

Stupid! Stupid! Stupid! I killed my QCX receiver!

 

OK, folks.? This is going to be the most stupid thing you've heard all day.? Maybe all week.? Or maybe in your lifetime.? Bear with me.? I have some 'splaining to do.? This will be a long read.? And I need some advice.?

I found a great little inexpensive plastic enclosure that the QCX fits nicely in.? That will be the subject of a future post, however.?

Once I had the QCX CCA mounted in the enclosure, I found that the tab of the 7805 regulator was too tall and the enclosure lid would not fit.? I removed the regulator from the CCA and looked at the leads.? They are thicker near the body of the regulator, necking down into a smaller diameter about 3/16 inch away from the body.? I took an emery board and carefully filed the leads down so that they were a uniform thickness all the way along their length.? This allowed the regulator to slip into place flush with the CCA, which gave me enough headroom to close the lid on my enclosure.? Neato!? Just what I was hoping for!? So I soldered the regulator into place on the CCA...? BACKWARDS.? Yep.? I actually did this.? And I didn't notice it.? Now, I'm no noob.? This ain't my first rodeo.? I'm a Senior RF/Microwave Test Engineer with 38 years of experience at a major aerospace company and I've been a ham for over 40 years.? I was a bit distracted due to non-radio related matters.? Still, this kind of thing just doesn't happen to me.? I'm usually very, very thorough!? Well, as I said, I didn't notice the mistake.? I powered up the QCX.? The display lit up for an instant, then the 1 A fuse in my power lead failed (note to everyone -- ALWAYS use a fuse in your power lead because you never know when something unexpected will occur).? I started looking for a cause, and I almost immediately noticed the incorrectly installed regulator.? DUH!!!? I removed it again.? Unfortunately, the plated-through vias pulled out of the board.? Damn!? I looked over the board and it looked like it should still work, so I went ahead and soldered the regulator back into place, correctly this time, and powered up the QCX again.? The display lit up.? I had audio in the headset.? I hit the key, and I had full transmit power.? The VFO was right on frequency.? The CW note sounded clean in my K3's receiver.? I was overjoyed!? That is, until I realized that I didn't hear any signals coming in on the QCX.? I switched the antenna over to the K3, and the pan scope showed strong signals all across the band.? The QCX wasn't hearing any of them.? I switched K3 into my dummy load, dialed the K3's transmitter down to 0.1 Watts, and hit the key.? The QCX did not hear the K3's signal.? CRAP!? I set up the QCX to allow me to run through the alignment checks again.? It ain't hearing itself.? No 700 Hz tone.? No received signal at all.? I get good audio, and it has the characteristic sound of the 200 Hz filter.? Everything seems to be functioning correctly.? But the receiver is not hearing a bloody thing.? What the devil did I kill?? I can't believe this!? I was sooooo close to being all finished with this project, and now it's broken!?

Anyone have a suggestion as to where to look for the failure?? This one has me stumped.?

73/72,
Dean K5DH? --? very depressed!
20m QCX msn 612
 

Hello Dean,

If you think thats stupid. Try checking the frequency of CLK0 with the 12 volt pin on the three pin header. Instead of the freq. pin next to the clock.

I would start with the mixer. And work your way down the signal chain towards the unity amp.

-Tomasz (N6SRN)
 

Hi Dean

It sounds to me that you probably fried IC4, the FST3253, which is the mixer (Quadrature Sampling Detector, a.k.a. Tayloe Detector). If that is blown, the op-amps would be just amplifying noise, then filtering it with the 200Hz filter... it could explain the lack of signals. You could use the onboard DVM to take a DC voltage reading at pins 2, 3, 5 and 6 of IC5, these should be around 2.5V (approximately midrail) - if they are not, then again it points to a fried FST3253.?

You say TX is working, and that implies the Si5351A is Ok. But TX only tests Clk2 of the Si5351A. So you should check Clk0 and Clk1. You can easily do this using the onboard test equipment of the QCX itself. Use "9.6 Signal gen." and set the signal generator to, say, 4MHz. Then go to "9.5 Frequency" which is the frequency counter. You should be able to hold a wire from the "Freq" test input pin, to Clk0 and Clk1 and measure 4MHz, shown on the display. Then you will know that the Si5351A is Ok.?

All the op-amps run with the full supply voltage anyway, not 5V. The only components running on 5V are the LCD, IC2 (Processor), IC1 (Si5351A) via two series diodes, IC4 (FST3253) and IC3 (74ACT00).?

If you find out that the FST3253 is the fried item, it should be possible to replace it with a new one. They are easily sourced. I could send you one (free of charge) but that would take time to reach you... up to you!

Good luck...

73 Hans G0UPL

On Thu, Dec 7, 2017 at 4:09 AM, K5DH <k5dh@...> wrote:
OK, folks.? This is going to be the most stupid thing you've heard all day.? Maybe all week.? Or maybe in your lifetime.? Bear with me.? I have some 'splaining to do.? This will be a long read.? And I need some advice.?

I found a great little inexpensive plastic enclosure that the QCX fits nicely in.? That will be the subject of a future post, however.?

Once I had the QCX CCA mounted in the enclosure, I found that the tab of the 7805 regulator was too tall and the enclosure lid would not fit.? I removed the regulator from the CCA and looked at the leads.? They are thicker near the body of the regulator, necking down into a smaller diameter about 3/16 inch away from the body.? I took an emery board and carefully filed the leads down so that they were a uniform thickness all the way along their length.? This allowed the regulator to slip into place flush with the CCA, which gave me enough headroom to close the lid on my enclosure.? Neato!? Just what I was hoping for!? So I soldered the regulator into place on the CCA...? BACKWARDS.? Yep.? I actually did this.? And I didn't notice it.? Now, I'm no noob.? This ain't my first rodeo.? I'm a Senior RF/Microwave Test Engineer with 38 years of experience at a major aerospace company and I've been a ham for over 40 years.? I was a bit distracted due to non-radio related matters.? Still, this kind of thing just doesn't happen to me.? I'm usually very, very thorough!? Well, as I said, I didn't notice the mistake.? I powered up the QCX.? The display lit up for an instant, then the 1 A fuse in my power lead failed (note to everyone -- ALWAYS use a fuse in your power lead because you never know when something unexpected will occur).? I started looking for a cause, and I almost immediately noticed the incorrectly installed regulator.? DUH!!!? I removed it again.? Unfortunately, the plated-through vias pulled out of the board.? Damn!? I looked over the board and it looked like it should still work, so I went ahead and soldered the regulator back into place, correctly this time, and powered up the QCX again.? The display lit up.? I had audio in the headset.? I hit the key, and I had full transmit power.? The VFO was right on frequency.? The CW note sounded clean in my K3's receiver.? I was overjoyed!? That is, until I realized that I didn't hear any signals coming in on the QCX.? I switched the antenna over to the K3, and the pan scope showed strong signals all across the band.? The QCX wasn't hearing any of them.? I switched K3 into my dummy load, dialed the K3's transmitter down to 0.1 Watts, and hit the key.? The QCX did not hear the K3's signal.? CRAP!? I set up the QCX to allow me to run through the alignment checks again.? It ain't hearing itself.? No 700 Hz tone.? No received signal at all.? I get good audio, and it has the characteristic sound of the 200 Hz filter.? Everything seems to be functioning correctly.? But the receiver is not hearing a bloody thing.? What the devil did I kill?? I can't believe this!? I was sooooo close to being all finished with this project, and now it's broken!?

Anyone have a suggestion as to where to look for the failure?? This one has me stumped.?

73/72,
Dean K5DH? --? very depressed!
20m QCX msn 612


 

Well...? so far I have incorrect voltages on pins 5 and 6 of IC5 (both less than half of what they should read), and no output at all on CLK0 and CLK1 using the onboard counter function and my bench counter.? I will check the two CLK signals with my o'scope when I have time, but I'm not feeling really good about my chances here.? If both ICs are blown, is it even worth the trouble to attempt repairs??
--

73/72,
Dean K5DH
20m QCX msn 612
 

I am feeling with you. You should really check clock 0 and clock 1 with a scope. If you have a two channel scope, you should even observe the 90° phase difference between the two clock outputs of the DDS chip which is required for the Tayloe detector. It is strange that clock 2 (for transmitting) seems to be working still, but otherwise .....? I'd be interested how to remove such SMD devices from such a densely populated PCB without frying other componentes thermically and how to get replacement devices back in without thermal problems and without creatings shorts between the narrow speced legs.

Frying semiconductors can be interesting, As I have found out (not really deliberately) BS170s generate a lot of smell and astonishingly much smoke for such a small transistor, howeever MPS751s (Q6) only crack up noiselessly, not much dramatic effect there. But both devices have only three leges and are much easier to remove ;-) than a SI5351A or FST3253.

73, Klaus DL2QB (WN2Z)



P.S.: Reading your vita and that you were working for a major aerospace company: You weren't involved in the Apollo 13 mission (the one where a oxygen tank blew up midway between Earth and Moon), were you?





This is the fried Q6
 

But both devices have only three leges and are much easier to remove ;-) than a SI5351A or FST3253.
Klaus,

I'm not so sure.
Possibly the best way I have for removing that sort of device is an embossing tool. (Cheap hot air gun.) Cut a hole in some tinfoil the size of the chip, heat it and lift off.
Other ways include lifting one pin at a time, various methods, needle or wire under pins but more risky.

Resoldering, no problem. If you bridge pins use solder wick.

OK, I admit cutting off 3 legs and soldering to the stumps is easier but not if you insist on cleaning out the holes:-)

73 Alan G4ZFQ
 

If the Si5351 were to be replaced, I would just cut off the legs flush with the chip using a scalpel or similar. Make sure to only press down until the leg is free from the chip, not all the way down to the board. The FST3253 legs could be cut off with a small sidecutter. Collect the pins with the soldering iron and clean up the area with flux and solder wick/braid. Drag solder to fix the new one, lots of flux.

Virus-free.

 

If the Si5351 were to be replaced, I would just cut off the legs flush with the chip using a scalpel or similar. Make sure to only press down until the leg is free from the chip, not all the way down to the board. The FST3253 legs could be cut off with a small sidecutter. Collect the pins with the soldering iron and clean up the area with flux and solder wick/braid. Drag solder to fix the new one, lots of flux.
Daniel

I've never tried the scalpel method. It seems to me it needs to be done with a lot of care and patience.
Cutting legs with a sidecutter seems really risky. Many sidecutters will displace the leg as it cuts. The movement may damage the pad or track.

I advise anyone who has not done it before to get a scrap board, maybe from an old computer or satellite receiver etc.. and try first.

73 Alan G4ZFQ
 

Guys,

There is a product called quickchip which I have successfully used in the past to remove SMD parts from closely packed boards. You can find it on Amazon?

Regards,

Steve G0XAR
 

You're thinking of ChipQuik, not Quickchip; that was the original low
temperature desoldering product. The Fast Chip product is a similar
one from a competing maker. The stuff isn't cheap but it's very
effective; it can be a real board saver if you have to get parts off a
dense surface mount board, especially if you do not have a hot air
rework station.

Here's a link to the ChipQuik kit:

The original version containing lead is also still available:


On Mon, Dec 18, 2017 at 10:27 AM, Stephen Farthing G0XAR JO92ON97
<squirrox@...> wrote:
Guys,

There is a product called quickchip which I have successfully used in the
past to remove SMD parts from closely packed boards. You can find it on
Amazon


Regards,

Steve G0XAR

1 - 10 of 10