Topics

QCX-20 ... low output power (~1.5 W) #qcx #20m


Rob French (KC4UPR)
 

Hello!

So, I recently pulled out my 20-meter QCX and plugged it into my dummy load via my new watt meter.  This is the first time I've measured its power, and I was curious to see how it did.

Powered via my 13.8 VDC bench supply, I got an output power of ~1.3 watts!

I switched to a fully charged 14.8 VDC LiPo (~16 V) and got 1.8 watts.  SWR in both cases was 1.00 (using the QRP Labs dummy load).

My first fear was maybe one of the finals transistors being bad.  I inspected with a magnifying glass and saw no visible damage.  Then I looked at this FAQ:  https://www.qrp-labs.com/qcx/qcxfaq.html#lowpower

I didn't observe the behavior where the output power would change after a second or two (the bad capacitor).  Based on reviewing that page, my guess is that I need to mess with the turn spacing and/or number of turns on L1-L3... does that sound right?  Will those change the output power that much?

Thanks,
-Rob KC4UPR


Gregg Myers
 

Hi Rob,

I'm not sure a visual inspection of these TO-92 transistors would reveal much. On a IRF510 I have seen some visual indication on the metal part when I smoked one, but the plastic BS170 could conceivably die less dramatically with no visual evidence. However, I think the finals in this case would tend to fail all at once so if one goes, they all go. And normally you would replace all of them at the same time if needed.

So assuming it is not the final transistors, the LP filter is a good place to look. Most people find unwinding a turn on the inductors might bring up the power to more expected levels. This can be done without removing the whole inductor. Just unsoldered one lead, unwrap a turn, and try it. It is not too uncommon that people accidentally wind an extra turn on their inductor in my experience. Only count a turn as the wire passing through the center of the toroid. If that doesn't work, then you might start investigating the LP filter capacitors.

Someone also just posted an issue with Q6 that resulted in low power. But in that case, that was a transistor substitution.

I might also add that even though you feel you are under-powered, a 4x increase in power is only 1 S unit or so. Not that dramatic a difference in actual performance! But I understand the effort to get it right.

Also, are you sure your new wattmeter is accurate? Can you verify it on another source?

Good luck & 73,
Gregg w7grm

Let us know what you find. 

On Thu, Mar 19, 2020 at 10:05 PM Rob French (KC4UPR) <kc4upr@...> wrote:
Hello!

So, I recently pulled out my 20-meter QCX and plugged it into my dummy load via my new watt meter.  This is the first time I've measured its power, and I was curious to see how it did.

Powered via my 13.8 VDC bench supply, I got an output power of ~1.3 watts!

I switched to a fully charged 14.8 VDC LiPo (~16 V) and got 1.8 watts.  SWR in both cases was 1.00 (using the QRP Labs dummy load).

My first fear was maybe one of the finals transistors being bad.  I inspected with a magnifying glass and saw no visible damage.  Then I looked at this FAQ:  https://www.qrp-labs.com/qcx/qcxfaq.html#lowpower

I didn't observe the behavior where the output power would change after a second or two (the bad capacitor).  Based on reviewing that page, my guess is that I need to mess with the turn spacing and/or number of turns on L1-L3... does that sound right?  Will those change the output power that much?

Thanks,
-Rob KC4UPR


Rob French (KC4UPR)
 

Greg,

Thanks for the response.

I can't promise the new watt meter is super accurate, but... I have an RS-HFIQ, which consistently registers right around 5W like it's supposed to.  I also have a uBITX that puts out 4-12W depending on the band.  So this all seems reasonable.

I will check Q6 just to be sure that it is the expected transistor.

When I get some time to mess around with it again, I will try taking off a turn... I'll probably just try the steps in the FAQ, i.e. squeezing/stretching windings, progressing to removing and/or adding turns as is described in the FAQ. 

Thanks,
-Rob KC4UPR


Paul AI4EE
 

I would replace the LPF capacitors first. Make sure they are NP0. if that doesn't help, then modify the turns.


On 3/20/2020 12:05 AM, Rob French (KC4UPR) wrote:
Hello!

So, I recently pulled out my 20-meter QCX and plugged it into my dummy load via my new watt meter.  This is the first time I've measured its power, and I was curious to see how it did.

Powered via my 13.8 VDC bench supply, I got an output power of ~1.3 watts!

I switched to a fully charged 14.8 VDC LiPo (~16 V) and got 1.8 watts.  SWR in both cases was 1.00 (using the QRP Labs dummy load).

My first fear was maybe one of the finals transistors being bad.  I inspected with a magnifying glass and saw no visible damage.  Then I looked at this FAQ:  https://www.qrp-labs.com/qcx/qcxfaq.html#lowpower

I didn't observe the behavior where the output power would change after a second or two (the bad capacitor).  Based on reviewing that page, my guess is that I need to mess with the turn spacing and/or number of turns on L1-L3... does that sound right?  Will those change the output power that much?

Thanks,
-Rob KC4UPR


Hans Summers
 

Hi Rob, Gregg

I tested an assembled QCX-20 here yesterday for a customer. We leave the reverse polarity protection diode in circuit, of course. With the power supply set to 12V the RF power output was 1.2W. I consider that way too low. I look for at least 3.0W at 12V supply before considering calling it OK. With the assembled QCXs we make here, 40m kits are usually easy and give good power output immediately. The 20m kits are a lot more sensitive and I commonly find that initially the power output is rather low, 2W or less (at 12V supply). So some playing with the number of turns on L1/2/3 is needed.

In my case yesterday, I took 1 turn off each toroid and 2 turns off L3. The power increased to 4.3W, at 12V supply. Plenty happy with that! 

Generally the number of turns is specified to provide the theoretically calculated inductance when the turns are spaced out evenly around the core, with a small gap at the ends. 

What you can do, to see if removing turns would help, is play around with squeezing the turns together and spacing them out. Spacing out the turns reduces inductance. Squeezing them together, increases inductance. If you squeeze the turns bunched up closely together and the power output goes way down, and if the power output is maximum when the turns are spaced out evenly - then this is an indication that the inductance is too high, and that you need to remove a turn. 

73 Hans G0UPL

On Fri, Mar 20, 2020 at 7:30 AM Gregg Myers <gregg.w7grm@...> wrote:
Hi Rob,

I'm not sure a visual inspection of these TO-92 transistors would reveal much. On a IRF510 I have seen some visual indication on the metal part when I smoked one, but the plastic BS170 could conceivably die less dramatically with no visual evidence. However, I think the finals in this case would tend to fail all at once so if one goes, they all go. And normally you would replace all of them at the same time if needed.

So assuming it is not the final transistors, the LP filter is a good place to look. Most people find unwinding a turn on the inductors might bring up the power to more expected levels. This can be done without removing the whole inductor. Just unsoldered one lead, unwrap a turn, and try it. It is not too uncommon that people accidentally wind an extra turn on their inductor in my experience. Only count a turn as the wire passing through the center of the toroid. If that doesn't work, then you might start investigating the LP filter capacitors.

Someone also just posted an issue with Q6 that resulted in low power. But in that case, that was a transistor substitution.

I might also add that even though you feel you are under-powered, a 4x increase in power is only 1 S unit or so. Not that dramatic a difference in actual performance! But I understand the effort to get it right.

Also, are you sure your new wattmeter is accurate? Can you verify it on another source?

Good luck & 73,
Gregg w7grm

Let us know what you find. 

On Thu, Mar 19, 2020 at 10:05 PM Rob French (KC4UPR) <kc4upr@...> wrote:
Hello!

So, I recently pulled out my 20-meter QCX and plugged it into my dummy load via my new watt meter.  This is the first time I've measured its power, and I was curious to see how it did.

Powered via my 13.8 VDC bench supply, I got an output power of ~1.3 watts!

I switched to a fully charged 14.8 VDC LiPo (~16 V) and got 1.8 watts.  SWR in both cases was 1.00 (using the QRP Labs dummy load).

My first fear was maybe one of the finals transistors being bad.  I inspected with a magnifying glass and saw no visible damage.  Then I looked at this FAQ:  https://www.qrp-labs.com/qcx/qcxfaq.html#lowpower

I didn't observe the behavior where the output power would change after a second or two (the bad capacitor).  Based on reviewing that page, my guess is that I need to mess with the turn spacing and/or number of turns on L1-L3... does that sound right?  Will those change the output power that much?

Thanks,
-Rob KC4UPR


K5DH
 

I bought an older (2017) used 20m QCX this week via a well-known amateur radio internet swap-shop venue.  Transmitter power output was barely 1.5 Watts.  I replaced C25 and C26 with mica caps, and adjusted the turn counts on L1, L2, and L3 exactly like Hans noted in the post above: 1 turn each off L1 and L2, and 2 turns off L3.  Power is now a little over 4 Watts! 

Thanks for the help, Hans! 

--
73/72,
Dean K5DH
20m and 40m QCX'es


jsctcooper@...
 

I'm at this point in my build on a 20m unit.  Would you recommend going with less turns on L1, L2, and L3 from the start? 


Rob French (KC4UPR)
 

So, I finally got around to working on this.  I appear to have solved my overall power output problem, but I do have a couple of questions:

(1) Did I take off TOO many windings?
(2) NOW what's going on?

Re: (1)... I proceeded by taking 1 winding at a time off from L1, L2, and L3, in sequence, checking power after removing 1 winding from 1 toroid.  Using a fully charged 14.4V LiPo battery (so ~16V to start), when I got to 2 windings removed each from L1 through L3, I had ~3.5W.  So I removed one more winding from L3, and am now at about 4.5W using the ~16V supply.  I figure I'll leave it at that, BUT, my question is, have I removed too many windings, especially from L3?  Am I going to end up generating a bunch of spurs/harmonics?  (I do not have a way to check for this.)

Re: (2)... The other thing that's happening is that it seems like about 30-50% of the time when I key the transmitter, it only shows ~300mW output!  Sometimes when this happens, it seems like if I immediately re-key it, it will pop back up to 4.5W.  It seems like the first time I key it, it is about 300mW, but it also happens later, and seems to be random... but almost always, if I see this happen, and I quickly let up on the key and then re-key it, it pops back up to 4.5W.

I have verified whenever this happens that I have good RF power measured across the drains of the BS170s.  After that it gets a little trickier.  Trying to use the built-in RF watt meter, I have touched various pads from the top of C29 (works fine, good power), all the way out to the output pin of the BNC jack, and I never seem to get any good power reading, even when I am getting good output on my external watt meter!  So I am assuming I'm having contact issues between the probe and the solder joints, probably flux.  This is obviously making troubleshooting difficult... any suggestions would be appreciated.

Regardless of my measurement issue, any ideas as to what could be causing this varying power issue?  I will say that it's pretty binary... power output is either ~4.5W, or it's ~300mW.  "Seems" like maybe a solder joint issue on the toroids, and maybe just the vibration of re-keying is causing a good connection to be made??? But when I soldered those coils back in, I held the soldering iron on the joint for quite a while too ensure the enamel burnt off, and I also have tried to reheat each joint on the coils to ensure a good connection.

Are there any capacitor issues that could be causing this?

What about anything with the keying circuit and/or L4?

Thanks!
-Rob KC4UPR


Mike Easterbrook
 

I had similar experience with intermittent low power after taking one too many turns off L3. 2 turns removed total current jumped from 450 to 900 mA on key down & finals too hot to touch! Put the turns back - slightly less power out (abt 4W) but no instability. Other thing noted was that key shaping was poor - see my earlier post.
Mike 9M2LXM 


Rob French (KC4UPR)
 

Mike, did you ever resolve the key shaping? Also,  did you ever adjust L4?


Ronald Taylor
 

Rob, I went back to some old notes and see that I did exactly what you have done in turns removal on a couple of QCX20s. 2T off of L1 and L2 and 3 turns off of L3 for a total of L1=14T, L2= 15T and L3=13T. In all cases, my spectrum analyzer showed that I was still within FCC guidelines for harmonic suppression. Certainly your results could vary, but that is what I saw on mine. Hope that helps.The shift in power from 300 mW to full output and back really sounds like a continuity issue in the lowpass filter somewhere. Or also possible is L4 continuity. All of these have the dreaded possibility of not enough enamel removed as you alluded to. Hope you find that… 73 … Ron


On Sat, May 30, 2020 at 7:19 AM Rob French (KC4UPR) <kc4upr@...> wrote:
So, I finally got around to working on this.  I appear to have solved my overall power output problem, but I do have a couple of questions:

(1) Did I take off TOO many windings?
(2) NOW what's going on?

Re: (1)... I proceeded by taking 1 winding at a time off from L1, L2, and L3, in sequence, checking power after removing 1 winding from 1 toroid.  Using a fully charged 14.4V LiPo battery (so ~16V to start), when I got to 2 windings removed each from L1 through L3, I had ~3.5W.  So I removed one more winding from L3, and am now at about 4.5W using the ~16V supply.  I figure I'll leave it at that, BUT, my question is, have I removed too many windings, especially from L3?  Am I going to end up generating a bunch of spurs/harmonics?  (I do not have a way to check for this.)

Re: (2)... The other thing that's happening is that it seems like about 30-50% of the time when I key the transmitter, it only shows ~300mW output!  Sometimes when this happens, it seems like if I immediately re-key it, it will pop back up to 4.5W.  It seems like the first time I key it, it is about 300mW, but it also happens later, and seems to be random... but almost always, if I see this happen, and I quickly let up on the key and then re-key it, it pops back up to 4.5W.

I have verified whenever this happens that I have good RF power measured across the drains of the BS170s.  After that it gets a little trickier.  Trying to use the built-in RF watt meter, I have touched various pads from the top of C29 (works fine, good power), all the way out to the output pin of the BNC jack, and I never seem to get any good power reading, even when I am getting good output on my external watt meter!  So I am assuming I'm having contact issues between the probe and the solder joints, probably flux.  This is obviously making troubleshooting difficult... any suggestions would be appreciated.

Regardless of my measurement issue, any ideas as to what could be causing this varying power issue?  I will say that it's pretty binary... power output is either ~4.5W, or it's ~300mW.  "Seems" like maybe a solder joint issue on the toroids, and maybe just the vibration of re-keying is causing a good connection to be made??? But when I soldered those coils back in, I held the soldering iron on the joint for quite a while too ensure the enamel burnt off, and I also have tried to reheat each joint on the coils to ensure a good connection.

Are there any capacitor issues that could be causing this?

What about anything with the keying circuit and/or L4?

Thanks!
-Rob KC4UPR


Mike Easterbrook
 

Hi Rob
No - have not done any more on key shaping. I was hoping for more feedback from the group as don't fully understand the circuit. I would have thought (wrongly?) that malfunction in Q4/6 would result in no keying at all.  Allison did make a comment on L4.  Small adjustment either way reduces RF out so did not look further. I'll take another look at this with respect to key envelope & revert.

50 years since I used CW & very slow progress getting up to speed so a few key clicks aren't going to upset too many folks yet. Got me wondering if this is an uncommon problem or just reflects few users with the means (or motivation) to look at it.
 73 Mike 9M2LXM


Rob French (KC4UPR)
 

One more update.  Hopefully this will help narrow things down.

This occurs about 10% of the time, where when I go key down, I get ~300 mW output power.  

An additional thing I noticed...

When I measure (using the built-in RF power meter) the level at the drains of the final amp transistors, I always get a solid "5.2" on the display.  However, when I measure it at the output BNC center pin, it varies between 3.8-4.5... or else 0.32 or so.  So even when I am getting "good" output power, it's still not the same as what I'm seeing at the output of the finals.  More to the point, when I measure it at the transistors, the reading is very steady... constant 5.2.  When I measure it at the LPF output, it wanders a little.  

So I suspect this all indicates more and more that it's a problem in the continuity of the LPF, although I'm still surprised that I never see it change while key down (except for the slight wandering during the high power out).  What is the chance that this could be a problem with the LPF capacitors, and/or the coupling capacitor (C29 I think)?  Should I just go replace these all as a matter of course?

Also, for the coils, should I just remove, rewind, and reinstall, possibly with manually removing the enamel?  I attempted to use the "burn of the enamel" technique... maybe that just didn't work for me. 

Regards,
Rob KC4UPR


Bill Cromwell
 

Hi Rob,

You may or may not have continuity problems in the low pass filter. Measuring power at the transistor drains is NOT valid. Measurements there will include elevated harmonic energy that is removed by the low pass filter. That is why it is called "low pass filter".

When you say you measure output power at the BNC output pin how do you do that? Four or five watts would blow out my RF probe if I tried to do that and if it didn't I would still not be measuring across 50 ohms. There are always things I don't know so I would appreciate if you could explain that to me. It would be more convenient than connecting dummy loads with cables, connectors, adapters, and measuring there.

73,

Bill KU8H

On 5/31/20 9:43 AM, Rob French (KC4UPR) wrote:
One more update.  Hopefully this will help narrow things down.
This occurs about 10% of the time, where when I go key down, I get ~300 mW output power.
An additional thing I noticed...
When I measure (using the built-in RF power meter) the level at the drains of the final amp transistors, I always get a solid "5.2" on the display.  However, when I measure it at the output BNC center pin, it varies between 3.8-4.5... or else 0.32 or so.  So even when I am getting "good" output power, it's still not the same as what I'm seeing at the output of the finals.  More to the point, when I measure it at the transistors, the reading is very steady... constant 5.2.  When I measure it at the LPF output, it wanders a little.
So I suspect this all indicates more and more that it's a problem in the continuity of the LPF, although I'm still surprised that I never see it change while key down (except for the slight wandering during the high power out).  What is the chance that this could be a problem with the LPF capacitors, and/or the coupling capacitor (C29 I think)?  Should I just go replace these all as a matter of course?
Also, for the coils, should I just remove, rewind, and reinstall, possibly with manually removing the enamel?  I attempted to use the "burn of the enamel" technique... maybe that just didn't work for me.
Regards,
Rob KC4UPR
--
bark less - wag more


Rob French (KC4UPR)
 

Bill, I had a dummy load connected abs wad using the built in RF power meter to measure across the dummy load, at the center pin solder joint (on the back of the PCB).

That all makes sense about harmonics.  Maybe more salient is that the transistor output seems rock steady whereas the LPF output appears to wander a little (when it's not dropping to 300 mW).

Rob  KC4UPR


Roger Hill
 

Rob

Just a dumb question, as usual from me: are you sure the dummy load itself is good, no intermittent connections etc?


73

Roger

G3YTN

---
***************************
Roger Hill
***************************


On 2020-05-31 10:34, Rob French (KC4UPR) wrote:

Bill, I had a dummy load connected abs wad using the built in RF power meter to measure across the dummy load, at the center pin solder joint (on the back of the PCB).

That all makes sense about harmonics.  Maybe more salient is that the transistor output seems rock steady whereas the LPF output appears to wander a little (when it's not dropping to 300 mW).

Rob  KC4UPR


Rob French (KC4UPR)
 

Roger,  my dummy load and associated wattmeter are known "good enough" based on use with two other rigs. 


Roger Hill
 

Ok great. That eliminates the dummy load then.

73
Roger
G3YTN

On 31 May 2020, at 12:31, "Rob French (KC4UPR)" <kc4upr@...> wrote:
Roger,  my dummy load and associated wattmeter are known "good enough" based on use with two other rigs. 


Bill Cromwell
 

Hi Rob,

So you were measuring across a 50 ohm dummy load. Aome of the variation *could* be in the dummy load or whatever bits you are using to connect it. As you hold the key down things begin to warm up. Even if they don;t get overheated there are still physical changes. That could account for the bit of wobble you are seeing.

The drop to 300 mW being intermitt is no doubt a poor connection somewhere. You know what to do about that:) I have a radio here with a lot of miles on it that I got cheap as 'not working right'. It had a similar problem to yours but on receive. I discovered a fractured winding on an inductor and had to rewind it. Look for things like that. Use a plastic prod to poke parts while transmitting into the dummy load (for brief periods). That will help you find the 'flaky' part.

73,

Bill KU8H

On 5/31/20 10:34 AM, Rob French (KC4UPR) wrote:
Bill, I had a dummy load connected abs wad using the built in RF power meter to measure across the dummy load, at the center pin solder joint (on the back of the PCB).
That all makes sense about harmonics.  Maybe more salient is that the transistor output seems rock steady whereas the LPF output appears to wander a little (when it's not dropping to 300 mW).
Rob  KC4UPR
--
bark less - wag more


Mike Easterbrook
 

Hi Rob
This sounds very similar to my experience. Using a straight key l get 4W most times but occasionally >1W - in this condition total current if I dare keep key down long enough is 900mA versus 450mA for "normal" i.e. 4W output.  All these at 13.8 v at input connector. Compressing turns or adding one turn to L3 or reducing supply volts to 12 eliminates problem. If I change to paddle & key a series of dits I can can see about 1in20 dits on the power meter (analogue) where the output drops. This avoids trying to troubleshoot key-down with rapidly frying finals!

What total current are you drawing when output is low or high?

My best guess is that something in the PA or shaping is on the edge of self-oscillation.