Topics

Two QCX kits (80m and 40) - both with shorts on Si5351A - fixed


Nick Austen
 

Summary: check for shorts between legs of the surface mounted devices before you start (additional cautionary note to add to manual?)

Background: I first bought and built one 40m QCX - all worked first try - then accidentally blew it up with a stray remnant resistor leg shorting 12v to 5v - now fixed after replacing the processor, and other digital chips as well as voltage dropping diodes (first replacement Si5351A replaces twice as first replacement did not oscillate at crystal frequency). All perfect now.

Then bought two more (one to give as a gift). I built them in parallel. Neither worked (initially).

After searching and checking, I looked at the quadrature outputs of the Si5351A’s. One unit had an odd looking three-level signal on the oscilloscope- which I eventually realised was the sum of the two quadrature outputs - they were shorted.
After careful magnified viewing I saw a short right on the pins of this synthesizer (not put there by me).
Solder wick fixed that quickly.

The second unit had only one of the quad outputs oscillating- the other was at 0 vdc. Found another short - this time from the second quad  output to the ground pin. This was also fixed by quick application of solder wick.

All working now - phew.

Conclusion - don’t let it put you off - just check the legs of the Si5351A on the bare board before you start - it may save you some disappointment:

Still highly recommend - the best kit I have seen.

Thanks again,
Nick
(VK6FABE)


geoff M0ORE
 

I think this has been mentioned before but be careful when checking for shorts around IC's. Some multimeters feed out up to 15 Volts across the test probes which can blow a sensitive IC before you even start to assemble your kit.

I am amazed that a stray resistor leg shorted the 12 Volt line to the 5 Volt line. Does the manual not mention that the surplus length is cut off.

I use a bleeper designed by W4ZCB ( now sadly a SK) to check for problems which has a very low voltage across the test probes on an open circuit and only a few micro-amps on a short circuit.

Hans has mentioned in the past that the number of QCX's sold is around 4000 but given the number of builders who have purchased several units, the true number of purchasers must be only a few hundred!

On 11/12/2019 14:20, Nick Austen wrote:
Summary: check for shorts between legs of the surface mounted devices before you start (additional cautionary note to add to manual?)

Background: I first bought and built one 40m QCX - all worked first try - then accidentally blew it up with a stray remnant resistor leg shorting 12v to 5v - now fixed after replacing the processor, and other digital chips as well as voltage dropping diodes (first replacement Si5351A replaces twice as first replacement did not oscillate at crystal frequency). All perfect now.

Then bought two more (one to give as a gift). I built them in parallel. Neither worked (initially).

After searching and checking, I looked at the quadrature outputs of the Si5351A’s. One unit had an odd looking three-level signal on the oscilloscope- which I eventually realised was the sum of the two quadrature outputs - they were shorted.
After careful magnified viewing I saw a short right on the pins of this synthesizer (not put there by me).
Solder wick fixed that quickly.

The second unit had only one of the quad outputs oscillating- the other was at 0 vdc. Found another short - this time from the second quad  output to the ground pin. This was also fixed by quick application of solder wick.

All working now - phew.

Conclusion - don’t let it put you off - just check the legs of the Si5351A on the bare board before you start - it may save you some disappointment:

Still highly recommend - the best kit I have seen.

Thanks again,
Nick
(VK6FABE)


SkipF, NT1G <skip.flem@...>
 

PLEASE LIST these multimeters that put 12V on the testleads.
 Thankyouverymuch!
 de SkipF, NT1G


geoff M0ORE
 

Too many to list but just get another voltmeter and measure across the leads yourself.

On 11/12/2019 19:23, SkipF, NT1G wrote:
PLEASE LIST these multimeters that put 12V on the testleads.
 Thankyouverymuch!
 de SkipF, NT1G


SkipF, NT1G <skip.flem@...>
 

My Flukes 87s... show .655V  at 11MegOhms...
 And 3.55V when testing diodes.
 Both show the same.
And what do yours show?
Thanks!


ajparent1/KB1GMX
 

Simpson-260 for one.
Triplite 630
Triplite 230

Three that come immediately to mind as I have them.

However with modern logic that should be less than 1.5V!  Why the max voltage for SI5351 is 3.3V as a starting point.

--
Please reply on list so we can share.
Off list email goes to trash, I had to due to scrapers.


SkipF, NT1G <skip.flem@...>
 

And WHAT voltage(s) do you SEE?
 The word MODERN is interesting.
I'm seeing a 117K series shunted by a 17K, with a 9V
source. But I don't have a 260 to 'vouch' for.
Owner's manual is available on line.


On Wed, Dec 11, 2019, 2:55 PM ajparent1/KB1GMX <kb1gmx@...> wrote:
Simpson-260 for one.
Triplite 630
Triplite 230

Three that come immediately to mind as I have them.

However with modern logic that should be less than 1.5V!  Why the max voltage for SI5351 is 3.3V as a starting point.

--
Please reply on list so we can share.
Off list email goes to trash, I had to due to scrapers.


Hans Summers
 

Hi Geoff , Nick

Hans has mentioned in the past that the number of QCX's sold is around 4000 but given the number of builders who have purchased several units, the true number of purchasers must be only a few hundred!

You are far far out of date OM... current quantity is 8,885... 
Conclusion - don’t let it put you off - just check the legs of the Si5351A on the bare board before you start - it may save you some disappointment:

Still highly recommend - the best kit I have seen.
Thanks Nick... You should play the lottery OM. But avoid thunderstorms, they say lightening doesn't strike twice but... hi hi. 

Here we have assembled a large number of QCX kits for the shop http://shop.qrp-labs.com/assembled/qcx40 at a rate of 3-4 per week over the last 18 months or more... and I do not recall having seen a short on Si5351A pins once in all that time. I know it does happen occasionally, but I think it is pretty rare. Anyway I am glad you solved it and still enjoy the kits!

This does illustrate a problem with kits... it isn't easy to do a full Quality Control check before you ship it out. It isn't practical to examine the whole thing for every type of failure under a microscope, and count every resistor and capacitor, etc., for every kit shipped. When you have a completely manufactured product (not a kit), it is possible to power it up and do a Quality Control check which it either passes or it doesn't. Producing kits is a very tough business :-)

73 Hans G0UPL


George Korper
 

It is a kit builder's requirement to do incoming quality control not the suppliers or kit manufacturer. 
It is a very tough business, and considering the state of the bands QCX is an amazing success.
I can wait forever for the QSX; after 60 years of listening to SSB hams repeating themselves, 
I love being limited to CW! And getting an amazing education in the process. What a great group.
BTW it now has CAT a GPS and  new firmware. K3GK

On Wed, Dec 11, 2019 at 4:11 PM Hans Summers <hans.summers@...> wrote:
Hi Geoff , Nick

Hans has mentioned in the past that the number of QCX's sold is around 4000 but given the number of builders who have purchased several units, the true number of purchasers must be only a few hundred!

You are far far out of date OM... current quantity is 8,885... 
Conclusion - don’t let it put you off - just check the legs of the Si5351A on the bare board before you start - it may save you some disappointment:

Still highly recommend - the best kit I have seen.
Thanks Nick... You should play the lottery OM. But avoid thunderstorms, they say lightening doesn't strike twice but... hi hi. 

Here we have assembled a large number of QCX kits for the shop http://shop.qrp-labs.com/assembled/qcx40 at a rate of 3-4 per week over the last 18 months or more... and I do not recall having seen a short on Si5351A pins once in all that time. I know it does happen occasionally, but I think it is pretty rare. Anyway I am glad you solved it and still enjoy the kits!

This does illustrate a problem with kits... it isn't easy to do a full Quality Control check before you ship it out. It isn't practical to examine the whole thing for every type of failure under a microscope, and count every resistor and capacitor, etc., for every kit shipped. When you have a completely manufactured product (not a kit), it is possible to power it up and do a Quality Control check which it either passes or it doesn't. Producing kits is a very tough business :-)

73 Hans G0UPL


Nick Austen
 

Hi Hans
All understood and appreciated. I was obviously unlucky and hope that I am the only one :)
Just thought it could help someone else - as, although I was disheartened (still excited like a young boy when the kit is completed), the solution was quick and simple.

On the previous kit, I was initially daunted at the prospect of trying to replace the the match-head size 10-lead surface mount device - but did it twice and it was relatively easy.

Thanks again,
Nick
VK6FABE


Arv Evans
 

Hans

Unfortunately it is impossible to perform quality control on assembled kits at the purchaser's location.  Also probably impossible to perform knowledge tests on those assembling the kits.  Even Heathkit never found a solution to assembler introduced problems.  

With over 8000 workinking kits out there it would seem that the design is solid.  Quality of shipped kits must be relatively good or the units would not be this popular.

Arv


On Wed, Dec 11, 2019, 2:11 PM Hans Summers <hans.summers@...> wrote:
Hi Geoff , Nick

Hans has mentioned in the past that the number of QCX's sold is around 4000 but given the number of builders who have purchased several units, the true number of purchasers must be only a few hundred!

You are far far out of date OM... current quantity is 8,885... 
Conclusion - don’t let it put you off - just check the legs of the Si5351A on the bare board before you start - it may save you some disappointment:

Still highly recommend - the best kit I have seen.
Thanks Nick... You should play the lottery OM. But avoid thunderstorms, they say lightening doesn't strike twice but... hi hi. 

Here we have assembled a large number of QCX kits for the shop http://shop.qrp-labs.com/assembled/qcx40 at a rate of 3-4 per week over the last 18 months or more... and I do not recall having seen a short on Si5351A pins once in all that time. I know it does happen occasionally, but I think it is pretty rare. Anyway I am glad you solved it and still enjoy the kits!

This does illustrate a problem with kits... it isn't easy to do a full Quality Control check before you ship it out. It isn't practical to examine the whole thing for every type of failure under a microscope, and count every resistor and capacitor, etc., for every kit shipped. When you have a completely manufactured product (not a kit), it is possible to power it up and do a Quality Control check which it either passes or it doesn't. Producing kits is a very tough business :-)

73 Hans G0UPL


Nick Austen
 

Hi Geoff

The stray resistor leg was one still floating around after clean-up - I have been building electronics things on and off for about 45 years :)
I had enjoyed the kit for about 24 hrs - and, through lack of self discipline, was listening to 40m on the QCX while cleaning up :) - a discarded lead was accidentally projected into the air and down onto the kit.

Thanks for reiterating the issue of meter voltages and their potential to cause damage. I typically use a CRO first and should take more case with the meter.
Interesting to note that for my cheap "Micron" (no-name) meter, I measure the following voltages across the probes for the selected resistance range settings:
  • Diode-check => 2.88 V
  • 200 ohms => 2.88 V
  • all other ranges => 560 mV 
Not 15 V but still worth keeping in mind.
Thanks.
Nick
VK6FABE


ajparent1/KB1GMX
 

Us old farts that have been in the industry for 30 or 50 years have seen it.
Its common to hand soldered items.   You have to be looking for it.... Solder
with internal flux can often spatter.  It happens more so with some fluxes
and a hotter [than needed] irons as the flux heats up and goes from solid
to gas and the solder with it.  That sputter can end up everywhere and
anywhere.  That with modern tight lead spacing and traces close to each
other often shows as hairline shorts.   Little heat or solvent often breaks
them.

This is less common with wave soldering, radiant heating, and IR heated SMT
soldering systems as there is less solder and the flux is not inside it.  That
added to the thermal ramp to heating and cooling tends to avoid this.

Allison
--
Please reply on list so we can share.
Off list email goes to trash, I had to due to scrapers.


George Korper
 

Allison, 
I built my last QCX with  poor quality solder and had excessive sputtering and was not satisfied with it.  Please recommend a solder that you have had good performance with. Preferably available on Amazon.
George


On Wed, Dec 11, 2019, 9:29 PM ajparent1/KB1GMX <kb1gmx@...> wrote:
Us old farts that have been in the industry for 30 or 50 years have seen it.
Its common to hand soldered items.   You have to be looking for it.... Solder
with internal flux can often spatter.  It happens more so with some fluxes
and a hotter [than needed] irons as the flux heats up and goes from solid
to gas and the solder with it.  That sputter can end up everywhere and
anywhere.  That with modern tight lead spacing and traces close to each
other often shows as hairline shorts.   Little heat or solvent often breaks
them.

This is less common with wave soldering, radiant heating, and IR heated SMT
soldering systems as there is less solder and the flux is not inside it.  That
added to the thermal ramp to heating and cooling tends to avoid this.

Allison
--
Please reply on list so we can share.
Off list email goes to trash, I had to due to scrapers.


ajparent1/KB1GMX
 

What did you use?

I do not advise using any of the no-lead solders.  They do not rework well and
generally need more heat.  More pain that gain.

The standard here in my lab is the same for the last 40+ years Kester 44, 37/63 SN/PB alloy
in .031" diameter.  I also use the .020 and .040 diameter of the same at any given time
I have all three sizes in 1pound spools on the shelf.  Its not cheap.

KESTER part number  24-6337-0027 

An alternate some use is the 245 no clean in the .031 63/37 also kester.

https://www.amazon.com/Kester-24-6337-0027-Solder-Alloy-Diameter/dp/B0149K4JTY/ref=pd_cp_328_1/130-0155267-6249125?_encoding=UTF8&pd_rd_i=B0149K4JTY&pd_rd_r=1601493a-5f26-41c2-9b3c-65ccfb1d8bc2&pd_rd_w=iS80j&pd_rd_wg=98Cqk&pf_rd_p=0e5324e1-c848-4872-bbd5-5be6baedf80e&pf_rd_r=6MGM1H1D328D9YW4MACZ&psc=1&refRID=6MGM1H1D328D9YW4MACZ


--
Please reply on list so we can share.
Off list email goes to trash, I had to due to scrapers.


ajparent1/KB1GMX
 

one last thing... soldering iron.

The average one I see in use by people is a reasonable quality temperature controlled
unit turned too far up with a cruddy tip with rough edges.  Replace the tip turn it down
and try to not use the tip as a chisel.   

Hint put it back in the holder without cleaning it after making a joint.  Clean it on the
way out and make sure its shiny clean before touching the joint.  If it gets dark and
grungy too fast its likely way to hot.


Or an el cheapo that run at 900 degrees or hotter, with a cruddy black tip.
Get a soldering iron!

Both users tend to have terrible soldering and complain they can't get the solder to melt
and they burn up boards in the process.  The tip has to be clean to transfer heat.  If you
beat it up or use poor solder and high heat (trying to get it to work) that tends to destroy
tips amazingly fast.  I've see a few kill a tip in hours.
 
I tend to use tips for literally thousands of hours (years) of build time.  Quality solder,
good care, and moderate temps.  

I bought the current Weller WES51 some 20 years ago and the current tip is more than
3 years old.  The only thing was to repair the cord to the handle.  My other iron is a
really old TCP60 (curie point temp control) from about 1979ish, the .25" wide tip on
that is maybe 10 or more years old.  I use that for SMT parts as its wide!

-------
Please reply on list so we can share.
Off list email goes to trash, I had to due to scrapers.


George Korper
 

I love that iron and had mine for years and just replaced with the same.
Now thanks to you and Amazon, I will have a good solder. Appreciate the 
help. Thank you.

On Thu, Dec 12, 2019 at 4:27 AM ajparent1/KB1GMX <kb1gmx@...> wrote:
one last thing... soldering iron.

The average one I see in use by people is a reasonable quality temperature controlled
unit turned too far up with a cruddy tip with rough edges.  Replace the tip turn it down
and try to not use the tip as a chisel.   

Hint put it back in the holder without cleaning it after making a joint.  Clean it on the
way out and make sure its shiny clean before touching the joint.  If it gets dark and
grungy too fast its likely way to hot.


Or an el cheapo that run at 900 degrees or hotter, with a cruddy black tip.
Get a soldering iron!

Both users tend to have terrible soldering and complain they can't get the solder to melt
and they burn up boards in the process.  The tip has to be clean to transfer heat.  If you
beat it up or use poor solder and high heat (trying to get it to work) that tends to destroy
tips amazingly fast.  I've see a few kill a tip in hours.
 
I tend to use tips for literally thousands of hours (years) of build time.  Quality solder,
good care, and moderate temps.  

I bought the current Weller WES51 some 20 years ago and the current tip is more than
3 years old.  The only thing was to repair the cord to the handle.  My other iron is a
really old TCP60 (curie point temp control) from about 1979ish, the .25" wide tip on
that is maybe 10 or more years old.  I use that for SMT parts as its wide!

-------
Please reply on list so we can share.
Off list email goes to trash, I had to due to scrapers.