RE: Problem with KPA100 power output; high current

[Courtney Krehbiel KD6X]

Hi All,
 
I have K2 S/N 5788 with KPA100 which I recently completed.  Everything
seemed to be working perfectly, including the KPA100.  Over the weekend,
I removed the top cover KPA100 to install the KDSP2 module.  I was very
gentle with the KPA100.  The receiver worked fine after assembly, so I
moved the unit across the room from the workbench to desk and rerouted
the antenna coax.  The first time I tried to light up the KPA100, the
power supply folded back due to high current.
 
I've spent the last few days dusting off my troubleshooting skills, and
have come up with the following observations.  I've also tried to review
and follow troubleshooting hints from this reflector.  My test equipment
is limited to a DVM and an old external analog SWR bridge, so I'm
somewhat handicapped.  I'm hoping someone has some ideas or similar
experience.
 
1. K2/KPA100 was working originally as far as I could determine during
construction and alignment steps.  Tested ok into dummy load and
KAT100/long wire on the air.
2. Problem is not band specific.  I'm using a good 50 ohm dummy load for
all tests.
3. K2 works normally into a dummy load antenna with KPA100 removed.
4. C224 .047 uFat AUX RF output of K2 had an internal short.  Replaced
this with new part.  Not sure if this could have been cause or effect.
No impact on observed problem after replacing part.
5. Have repeatedly seen HI CUR on K2.  Moved current limit to 425 mA
from original setting of 350 since I thought the DSP module might have
pushed things over the top.  This makes it better with the top open.
Putting the KPA100 back on the K2 causes the supply to fold back.  (I
expect this is due to a better ground connection between the KPA100 and
K2.)  All measurements made with KPA100 removed and sitting to the right
side of K2.
6. Power output seems to be at or near max both <10 and >10watt setting.
Relays click at the right time, and power output into a dummy load on
the analog SWR bridge takes a significant jump when high power kicks in.
(I used the K2 solo to "calibrate" the external SWR bridge.)  The amp
seems to be working.
7. Power control doesn't seem to affect power setting or power output.
Control is very jumpy at the far CCW end of the range, but that's the
extent of the adjustment ability.
8. I've carefully looked at the built in SWR bridge and components
around U5 and U6.  I used a jewelers loupe to check each connection
during assembly.  It's possible I missed something, but I was very
careful and methodical.  
9. Both R26 and R27 seem are set near middle of range.  I read 26.8Kohms
to ground at both U5 pins 3 and 5.
10. VRFDET seems to sit between .8v and 1.4v depending on band and
whether high or low power.  There is no movement in VRFDET when I adjust
the Power control.  (Just like no change in output power level on
external analog SWR.)
11. DVM measured frequency seems to correlate to tuned frequency.  I
doubt the circuit is oscillating.
 
I would be very appreciative of any ideas, recommendations, or
suggestions.
 
Thanks very much, and 73,
 
  -- Courtney Krehbiel
     KD6X
     K2 S/N 5788
 
 
_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
 http://mailman.qth.net/mailman/listinfo/elecraft   

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com

[Courtney Krehbiel KD6X]

Hi Tom,

Thanks for your help and questions.  I've answered them in-line below.
Also thanks to several others who alerted me to the issue with the DSP
connection pins being a bit too long on late models and potentially
causing a short between the DSP board and the KPA heatsink and/or
shield.  I'll trim those for sure.  However, since the shield and
heatsink are ground, a DSP to ground short would be bad for the DSP, but
shouldn't have impacted the KPA100.  (Right??)  For what it's worth, the
KDSP2 seems to be working just fine.

  -- Courtney Krehbiel
     KD6X

-----Original Message-----
From: Tom Hammond [mailto:[hidden email]]
Sent: Friday, March 02, 2007 9:11 AM
To: Courtney Krehbiel
Subject: Re: [Elecraft] RE: Problem with KPA100 power output; high
current

Courtney:

>I have K2 S/N 5788 with KPA100 which I recently completed.  Everything
>seemed to be working perfectly, including the KPA100.  Over the
>weekend, I removed the top cover KPA100 to install the KDSP2 module.  I

>was very gentle with the KPA100.  The receiver worked fine after
>assembly, so I moved the unit across the room from the workbench to
>desk and rerouted the antenna coax.  The first time I tried to light up

>the KPA100, the power supply folded back due to high current.

When you say 'light up'... do you mean the first time you APPLIED POWER?
Or the first time you APPLIED RF to the KPA100?

[Courtney] The initial failure was noted the first time I applied RF to
the KPA100 after the DSP module was installed.  The supply immediately
folded back and had to be turned off and back on to reset and restore
DC.

Did the power supply STAY in an overcurrent protect mode? Or did it just
'cycle' back to on-line immediately, placing the K2 back in power-up
mode (as if you'd just turned it on)?

[Courtney] The power supply effectively shut down.  I had to restart it
to restore DC.

>I've spent the last few days dusting off my troubleshooting skills, and

>have come up with the following observations.  I've also tried to
>review and follow troubleshooting hints from this reflector.  My test
>equipment is limited to a DVM and an old external analog SWR bridge, so

>I'm somewhat handicapped.  I'm hoping someone has some ideas or similar

>experience.
>
>1. K2/KPA100 was working originally as far as I could determine during
>construction and alignment steps.  Tested ok into dummy load and
>KAT100/long wire on the air.
>
>2. Problem is not band specific.  I'm using a good 50 ohm dummy load
>for all tests.
>
>3. K2 works normally into a dummy load antenna with KPA100 removed.
>
>4. C224 .047 uF at AUX RF output of K2 had an internal short.  Replaced

>this with new part.  Not sure if this could have been cause or effect.
>No impact on observed problem after replacing part.

I think... the only effect this might have had would have been IF (which
it was shorted) you accidently MIS-plugged the +12VDC line into the AUX
RF connector. Doing this WILL FRY one of the relays in the K2 and CAN
cause some strange operation.
[Courtney] I have never plugged the +12VDC line into the RF.  The length
of cable wouldn't let me do that.  I also have them very clearly color
coded per the suggestion in the manual.  However, I did have the RF and
ground reversed on the connector briefly when initially testing the
KPA100.  That's probably what happened to C224.  

>5. Have repeatedly seen HI CUR on K2.  Moved current limit to 425 mA
>from original setting of 350 since I thought the DSP module might have
>pushed things over the top.  This makes it better with the top open.
>Putting the KPA100 back on the K2 causes the supply to fold back.  (I
>expect this is due to a better ground connection between the KPA100 and
>K2.)  All measurements made with KPA100 removed and sitting to the
>right side of K2.

Ummm... 'normal' CAL CUR values should be around 3.50A, not anything in
mA.

[Courtney]  My error.  Right numbers, wrong decimal point.  I was
writing this from memory.  It is 3.50A to 4.25A.  Maybe I'm sloppier
than I think I am.  :-)

I'm beginning to suspect a possible parasitic in the amp, but this
should ONLY occur if you attempt to TRANSMIT WITH the amp, and then
(usually) only on 40M.

[Courtney] Most of my testing has been in the 80 and 20M bands into a
dummy load.

>6. Power output seems to be at or near max both <10 and >10watt
setting.
>Relays click at the right time, and power output into a dummy load on
>the analog SWR bridge takes a significant jump when high power kicks
in.
>(I used the K2 solo to "calibrate" the external SWR bridge.)  The amp
>seems to be working.

You mean that the analog SWR bridge 'jumps' when you go from 10W TO 11W,
even without KEYING anything? I think I'm lost now...

[Courtney] Sorry, I'm putting the K2 into Tune mode and then increasing
the power across the 10W to 11W boundary.  The reading on the external
SWR meter takes a big jump and goes from about 20% of scale to full
scale.  I "calibrated" the 20% point to be about 10 watts using the K2
barefoot without the KPA100.

>7. Power control doesn't seem to affect power setting or power output.
>Control is very jumpy at the far CCW end of the range, but that's the
>extent of the adjustment ability.

This is at ALL power level settings? Including <10W?

[Courtney] Correct.  The K2 power adjustment works just fine with the
KPA100 removed.  With the KPA100 attached and the RF going through it's
RF jack, the adjustment doesn't not work for <10W or >10 watt settings.
The power output level takes a large jump as the K2 Power adjustment
moves past that boundary.

>8. I've carefully looked at the built in SWR bridge and components
>around U5 and U6.  I used a jewelers loupe to check each connection
>during assembly.  It's possible I missed something, but I was very
>careful and methodical.

 From personal experience, I (almost) always find that my errors are
those I've looked at sometimes several times, and mentally say, "It
CAN'T be that, I KNOW I did it correctly!".

[Courtney] I absolutely agree.  We're the worst at critiquing our own
work.  If you know any way to counter that effect, I'm interested.
Again, the KPA100 was working for several weeks while I built the KAT100
and KDSP2 module, so I doubt it's a construction error.  It's always
possible a stray piece of wire or small solder ball gets lodged in the
wrong place when I removed the cover.  That's the sort of thing I've
been looking for.  Obviously, I try to keep the work area super clean,
but things like that are always possible.  I have a background in
electronics from my early career years, although the skills are pretty
rusty these days.

>9. Both R26 and R27 seem are set near middle of range.  I read
>26.8Kohms to ground at both U5 pins 3 and 5.
>
>10. VRFDET seems to sit between .8v and 1.4v depending on band and
>whether high or low power.  There is no movement in VRFDET when I
>adjust the Power control.  (Just like no change in output power level
>on external analog SWR.)
>
>11. DVM measured frequency seems to correlate to tuned frequency.  I
>doubt the circuit is oscillating.

You DVM includes a freq. counter? Up to what frequency?
[Courtney] Per the DVM specs, up to 10MHz.  It's not supposed to be very
accurate, but it seems to work.  I measured output frequency in both 80
and 20 meter bands, and it correlated with the tuning dial pretty well
even at 14 MHz.

Good luck,

Tom   N0SS

[Courtney] Thanks for your penetrating questions!  :-)

_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
 http://mailman.qth.net/mailman/listinfo/elecraft   

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com

[Courtney Krehbiel KD6X]

Hi All,
 
I found my problem, and thought I repost my findings back to the
reflector to help others who may have similar issues.  There are a few
red herrings in my troubleshooting observations, and I am deeply
grateful to Tom Hammond, N0SS and Ron, ZL1TW for their time and advice
to help me zero in on the problem.
 
First of all, the problem with C224 was most likely caused during
construction test when I errantly reversed the RF leads going to the
KPA100.  I'd conveniently forgotten about this momentary assembly error.
 
The installation of the DSP module was also a red herring.  Thanks to
several who suggested I trim the leads on the newer version DSP module
to make it sit closer to the control board and less likely to short to
the KPA100 heatsink/shield when the cover was on.  Done.  It fits much
better now.
 
The real cause of my trouble was dead 1N5711 diodes in the internal SWR
bridge circuit. The diodes in both the KPA100 and KAT100 were fried and
seemed to look like 2 to 4 Kohm resistors.  I'd been comparing
resistance readings between the two circuits thinking it wasn't likely
that both were dead.  Wrong!  The problem wasn't obvious to me because
of the effect of the surrounding circuit, but closer testing with the
diode feature on the DVM got me very suspicious.  Once out of the
circuit, they were obviously bad.  With the RF detector circuit dead and
measuring no output power, the K2 feedback control process just kept
trying to drive the unit harder until it reached full bore and
high-current would ensue.
 
As a helpful hint, I was able to remove and replace the diodes from the
top side on both the KPA100 and KAT100 with minimal disassembly.  I
looped a very fine piece of steel wire under one diode leg, and pulled
up gently while applying heat. With one end now free, it was easy to
heat the other lead and remove the diode. There was just enough room to
get the tip of a solder sucker in to clean out the hole, and I dropped
the replacement with leads pre-trimmed and resoldered from the top. It
was about a 30 minute operation for both the KPA100 and KAT100. I
figured this approach was less traumatic to the hardware than the
disassembly necessary to reach these diodes from the bottom.

Thanks again to all who helped!

-- Courtney Krehbiel  KD6X


________________________________

From: Courtney Krehbiel
Sent: Friday, March 02, 2007 8:32 AM
To: [hidden email]
Cc: Courtney Krehbiel
Subject: RE: Problem with KPA100 power output; high current


Hi All,
 
I have K2 S/N 5788 with KPA100 which I recently completed.  Everything
seemed to be working perfectly, including the KPA100.  Over the weekend,
I removed the top cover KPA100 to install the KDSP2 module.  I was very
gentle with the KPA100.  The receiver worked fine after assembly, so I
moved the unit across the room from the workbench to desk and rerouted
the antenna coax.  The first time I tried to light up the KPA100, the
power supply folded back due to high current.
 
I've spent the last few days dusting off my troubleshooting skills, and
have come up with the following observations.  I've also tried to review
and follow troubleshooting hints from this reflector.  My test equipment
is limited to a DVM and an old external analog SWR bridge, so I'm
somewhat handicapped.  I'm hoping someone has some ideas or similar
experience.
 
1. K2/KPA100 was working originally as far as I could determine during
construction and alignment steps.  Tested ok into dummy load and
KAT100/long wire on the air.
2. Problem is not band specific.  I'm using a good 50 ohm dummy load for
all tests.
3. K2 works normally into a dummy load antenna with KPA100 removed.
4. C224 .047 uFat AUX RF output of K2 had an internal short.  Replaced
this with new part.  Not sure if this could have been cause or effect.
No impact on observed problem after replacing part.
5. Have repeatedly seen HI CUR on K2.  Moved current limit to 425 mA
from original setting of 350 since I thought the DSP module might have
pushed things over the top.  This makes it better with the top open.
Putting the KPA100 back on the K2 causes the supply to fold back.  (I
expect this is due to a better ground connection between the KPA100 and
K2.)  All measurements made with KPA100 removed and sitting to the right
side of K2.
6. Power output seems to be at or near max both <10 and >10watt setting.
Relays click at the right time, and power output into a dummy load on
the analog SWR bridge takes a significant jump when high power kicks in.
(I used the K2 solo to "calibrate" the external SWR bridge.)  The amp
seems to be working.
7. Power control doesn't seem to affect power setting or power output.
Control is very jumpy at the far CCW end of the range, but that's the
extent of the adjustment ability.
8. I've carefully looked at the built in SWR bridge and components
around U5 and U6.  I used a jewelers loupe to check each connection
during assembly.  It's possible I missed something, but I was very
careful and methodical.  
9. Both R26 and R27 seem are set near middle of range.  I read 26.8Kohms
to ground at both U5 pins 3 and 5.
10. VRFDET seems to sit between .8v and 1.4v depending on band and
whether high or low power.  There is no movement in VRFDET when I adjust
the Power control.  (Just like no change in output power level on
external analog SWR.)
11. DVM measured frequency seems to correlate to tuned frequency.  I
doubt the circuit is oscillating.
 
I would be very appreciative of any ideas, recommendations, or
suggestions.
 
Thanks very much, and 73,
 
  -- Courtney Krehbiel
     KD6X
     K2 S/N 5788
 
 
_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
 http://mailman.qth.net/mailman/listinfo/elecraft   

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com

[Mike Markowski]

Courtney and all,

Courtney Krehbiel wrote:
> ...The real cause of my trouble was dead 1N5711 diodes in the internal SWR
> bridge circuit. The diodes in both the KPA100 and KAT100 were fried...
 > With the RF detector circuit dead and
> measuring no output power, the K2 feedback control process just kept
> trying to drive the unit harder until it reached full bore and
> high-current would ensue...

I had exactly this problem a few weeks ago.  In my case, though, it
seems a near lightning strike (in winter!) or some other ESD source
zapped just one of the 5711s in the bridge.  But after that 3 diodes in
the t/r switch shorted, followed by scorching a few resistors in the PA
area as well as one in the t/r switch.  Ouch.

With Gary's patient help as well as a few component replenishments from
Scott (parts@elecraft) I was back in business again.  My lesson learned
is disconnect the antenna even in winter time.  It still feels good to
have the rig working again!

73,
Mike ab3ap
_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
 http://mailman.qth.net/mailman/listinfo/elecraft   

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com