|
I've been having lots of trouble with my QCX and have been directed to look at T1, just like everyone else. I've been over it a zillion times and can't find a wiring error. Anyway, since I've now officially written off the build and am waiting for QCX 1200 to arrive, I've been poking at it.
The attached scope capture shows Clock 1 and Clock 2, which page 103 says should be in quadrature. I can understand the weird waveform because the outputs of the Si5351A are not unloaded. Anyway, the two waveforms are clearly not in quadrature. The mode is set to normal, and the freq is 7020 kHz, which is confirmed on the scope.
Looking at the schematic, it seems to me that if these two are not in quadrature, nothing downstream will work in quadrature either.
Suggestions?
--
Dave Casler
KE?OG
Colorado, USA
www.YouTube.com/davidcasler
|
Also note the large voltage discrepancy between channels. That's because I mis-set the probe for 1x instead of 10x. The shapes look identical to the photo above when set for 10x.
--
Dave Casler
KE?OG
Colorado, USA
www.YouTube.com/davidcasler
|
Hi Dave, CLK2 is used to drive the transmit side of things.? CLK0, CLK2 go off to the receiver.
CLK0, CLK1 are turned off during transmit. There may a short on the two lines going to IC4 or IC1 may be faulty. Have a look around C21, R36, C22, Q7 as the two signal lines from IC1 also go this area.? There may be a short, cap around the wrong way, solder splash, etc. Good luck with the fault finding. Keep us updated!
PS I enjoy your Youtube channel ;-)
73 David VK2JDR
|
Whops, I was reading the schematic in a hurry sideways. . .
The X marks are the test points so disregard what I said about the signals from iC1 going to op amp area. I'll get back to work now :-( David
|
Hi Dave
Clk0 and Clk1 are supposed to be in quadrature during Receive. Your scope probes should attach to the Clk0 and Clk1 test points. The first thing to do is test that all is well with your measurement setup. Please take another 'scope screenshot, with all the x1 x10 etc sorted out to remove any possibility of error. Then swap the probes on Clk0 and Clk1 and take another screenshot. This will verify that you have the 'scope correct etc.?
Another check: make sure your menu "3.9 Band" is 40m.?
73 Hans G0UPL
On Fri, Nov 17, 2017 at 7:43 AM, David Rushworth <david.1.rushworth@...> wrote: Whops, I was reading the schematic in a hurry sideways. . .
The X marks are the test points so disregard what I said about the signals from iC1 going to op amp area. I'll get back to work now :-( David
|
Hi Dave Cassier -? KE?OG
My eyes see roughly 90° phase shift between traces. Did you expect
180°?
Hans has writ the wisdom so I will sit back and await more nice
captures.
John, VK6JY
On 17/11/2017 11:14 AM, Dave KE?OG
wrote:
I've been having
lots of trouble with my QCX and have been directed to look at T1,
just like everyone else. I've been over it a zillion times and
can't find a wiring error. Anyway, since I've now officially
written off the build and am waiting for QCX 1200 to arrive, I've
been poking at it.
The attached scope capture shows Clock 1 and Clock 2, which page
103 says should be in quadrature. I can understand the weird
waveform because the outputs of the Si5351A are not unloaded.
Anyway, the two waveforms are clearly not in quadrature. The mode
is set to normal, and the freq is 7020 kHz, which is confirmed on
the scope.
Looking at the schematic, it seems to me that if these two are not
in quadrature, nothing downstream will work in quadrature either.
Suggestions?
--
Dave Casler
KE?OG
Colorado, USA
|
Hi Dave Casler -? KE?OG - afterthought.
My eyes see roughly 90° phase shift between traces. Did you
expect 180°
Hans has writ the wisdom so I will sit back and await more nice
captures.
Your RIGOL can probably display vertical "cursor" lines which you
can move
across so one is at a key point on the yellow trace and another on
the blue
trace at the corresponding point of its waveform.
Bingo!?? It may read out the difference in time or phase with good
precision.
? ?? ???????? 73, JohnY
|
Hi Dave Casler -? KE?OG - after my afterthought.
Refocused my eyes and thinking - the traces are IN PHASE. Sorry - JohnY
|
Hello Dave, Friday, November 17, 2017 I've been having lots of trouble with my QCX and have been directed
to look at T1, just like everyone else. I've been over it a zillion
times and can't find a wiring error. Anyway, since I've now
officially written off the build and am waiting for QCX 1200 to arrive, I've been poking at it. The attached scope capture shows Clock 1 and Clock 2, which page
103 says should be in quadrature. I can understand the weird
waveform because the outputs of the Si5351A are not unloaded.
Anyway, the two waveforms are clearly not in quadrature. The mode is
set to normal, and the freq is 7020 kHz, which is confirmed on the scope. Looking at the schematic, it seems to me that if these two are not
in quadrature, nothing downstream will work in quadrature either. Suggestions? Thyose look in phase, but as you say, of different amplitude and a x1 and X10 probe setting will load it differently, possibly the cause of the different shape waveforms? Someone more knowledgeable will confirm? -- 2E0ILY Best regards, Chris mailto:chris@... -- Best regards, Chris Wilson (2E0ILY)
|
Hello JohnY,
Opps, replied before seeing this, which is orpahened off the thread
when read in my mail reader. I would be interested seeing the
waveforms with both probes at X10 though.
Friday, November 17, 2017
toggle quoted message
Show quoted text
Hi Dave Casler -? KE?OG - after my afterthought. Refocused my eyes and thinking - the traces are IN PHASE. Sorry - JohnY -- 2E0ILY Best regards, Chris mailto:chris@...
--
Best regards, Chris Wilson (2E0ILY)
|
Hi Guys,
I double-checked that 3.9 is set for 40 meters, as it should be.
I took extra care to have the two channels identical. 10x setting on both the scope and probes, calibrated each probe, DC coupling on both probes. Per Hans, I attached the two probes to Clk0 and Clk1 for this screen, and then reversed for the next. Note that I moved the second channel up so it's obvious that the two test points are in phase.
And now swap probes between the two clock outputs.
--
Dave Casler
KE?OG
Colorado, USA
www.YouTube.com/davidcasler
|
Hi Dave. I assume your scope triggering is set up so that you trigger off
one channel and use this one trigger to trigger both channels and not trigger
off each individual waveform?? This is sometimes overlooked and can make
the two channels appear to be in phase when they are not.
Steve K1RF
toggle quoted message
Show quoted text
Sent: Friday, November 17, 2017 12:09 PM
Subject: Re: [QRPLabs] QCX: Aren't Clk0 and Clk1 supposed to be in
quadrature?
?
Hi
Guys,
I double-checked that 3.9 is set for 40 meters, as it should
be.
I took extra care to have the two channels identical. 10x setting on
both the scope and probes, calibrated each probe, DC coupling on both probes.
Per Hans, I attached the two probes to Clk0 and Clk1 for this screen, and then
reversed for the next. Note that I moved the second channel up so it's obvious
that the two test points are in phase.
And now swap
probes between the two clock outputs.
--
Dave
Casler
KE?OG
Colorado, USA
www.YouTube.com/davidcasler
|
Hi Dave
Have you checked with a DVM to see if there is a short circuit between Clk0 and Clk1? It's as though they are shorted together, fighting each other. 1/4 of the cycle is flat at the top, perhaps when both Clk1 and clk0 are "high"; 1/4 of the cycle is flat at the bottom, when both Clk0 and Clk1 are low; in the other two 1/4 cycles when Clk0 and Clk1 are opposite each other, they are somewhere fighting in the mid-voltage area.?
So to me that would be something to check. Just check for continuity with a DVM. If there is continuity then perhaps there is a solder splash at the Si5351A pins 9 and 10, or a solder bridge.
73 Hans G0UPL? ?
On Nov 17, 2017 8:09 PM, "Dave KE?OG" < casler28@...> wrote: Hi Guys,
I double-checked that 3.9 is set for 40 meters, as it should be.
I took extra care to have the two channels identical. 10x setting on both the scope and probes, calibrated each probe, DC coupling on both probes. Per Hans, I attached the two probes to Clk0 and Clk1 for this screen, and then reversed for the next. Note that I moved the second channel up so it's obvious that the two test points are in phase.
And now swap probes between the two clock outputs.
--
Dave Casler
KE?OG
Colorado, USA
|
On Fri, Nov 17, 2017 at 09:09 am, Dave KE?OG wrote:
Hi Guys,
I double-checked that 3.9 is set for 40 meters, as it should be.
I took extra care to have the two channels identical. 10x setting on both the scope and probes, calibrated each probe, DC coupling on both probes. Per Hans, I attached the two probes to Clk0 and Clk1 for this screen, and then reversed for the next. Note that I moved the second channel up so it's obvious that the two test points are in phase.
And now swap probes between the two clock outputs.
--
Dave Casler
KE?OG
Colorado, USA
Hi Dave, It looks like maybe CLK0 and CLK1 are shorted to each other since there appear to be three logic levels instead of just High and Low. One level when both CLK0 and CLK1 are? High, a second level when one of the clocks is Low and the other clock is High, and a third level when both clocks are low. Check for a short. 73, Warren K2GQT
|
It looks to me like the Clk0 and Clk1 outputs are shorted to each other.
?
-Lee-
?
WA3FIY
?
------ Original Message ------
Sent: 11/17/2017 12:09:00 PM
Subject: Re: [QRPLabs] QCX: Aren't Clk0 and Clk1 supposed to be in quadrature?
?
Hi Guys,
I double-checked that 3.9 is set for 40 meters, as it should be.
I took extra care to have the two channels identical. 10x setting on both the scope and probes, calibrated each probe, DC coupling on both probes. Per Hans, I attached the two probes to Clk0 and Clk1 for this screen, and then reversed for the next. Note that I moved the second channel up so it's obvious that the two test points are in phase.
And now swap probes between the two clock outputs.
--
Dave Casler
KE?OG
Colorado, USA
|
Hold the fort! It turned out to be a stray subatomic solder ball!
Here's the new scope screen grab: Woohoo! This is quadrature!
I have a microscope for use with surface mount devices, given to me by viewer Michael Lloyd, and it revealed a miniscule solder blob that shorted Clk0 and Clk1 right across the top of the Si5351 leads.
Here's a view of the left side of the Si5351 (upside down--the Celestron microscope camera inverts). Looks okay.?
Here's the right side of the Si5351. Does not look okay. Lots of stray solder. I used a dental pick and a tiny screwdriver to pop the solder ball out and clean up a bit between the leads.
So, although I still need to work on getting C1 to behave, the rest of the alignment runs well. The receiver is definitely single-signal now and much better!
Many thanks for the suggestion to look for stray bits of this and that, and again thanks to Michael Lloyd for the microscope and camera!
Now to get T1 working properly with C1!
--
Dave Casler
KE?OG
Colorado, USA
www.YouTube.com/davidcasler
|
Good news on the rogue solder ball find.? Sounds like you are on the way now.
I believe this type of issue will be causing a lot of the problems that are being posted.? There is not much room for error with tracks very close together.
73 David? VK2JDR?
|
Woo Hoo!!! Success! I'm glad that the microscope and camera were helpful! I had a "stray subatomic solder ball" on an OHR100 once and it was under a thru hole component. It took two weeks to find it and I found it with the very microscope that you found yours with. They'll drive a person nuts. In fact, before I found it, I asked my psychiatrist about the voices in my head and she said that I didn't have a psychiatrist. Figures.
Finding and fixing problems entitles one to claim the Lone Watie award. Here's where it comes from:
:o) Every time I have a problem with a kit I think "Endeavor to persevere" and declare war on the "solder balls"
73
NE5U
Mike
|
Dave KE?OG