I could use some help in troubleshooting the amp.

But I have a confession to make. I had just finished building the 50W amp kit. I got through the adjustments okay (or so I think) but when I went to measure output I had none. Then I noticed the rookiest of all rookie mistakes. I had connected the amp coaxes to the QCX+ backwards. In other words, I transmitted into the output of the amp and had the dummy load on the input.  I guess because it was upside down on my bench, duh. Or there may be some other issue. Not sure what happens when you do what I did. I had keyed the radio a couple times for just a few seconds each and noticed no current draw on the bench supply. There was no smoke or explosions. then noticed the connection issue. Since I suspected no output I connected the QRP Labs 50 Ohm dummyload and checked the RF DCV from the connections provided for that. Nothing coming out at all. 0 Volts. The QCX seems to have survived with over 21VDC measured. 

Can someone help me save this amp?

Bryan, N0LUF

May have jumped the gun. While looking around I discovered some troubleshooting videos on youtube. Ill start there and come back after some discovery.
Bryan, n0luf

Update: 
I started with NA5Y's youtube video "QCX Amplifier Troubleshooting". 
I dont have a scope so couldn't see rf on the input but double checked QCX+ was giving 21VDC on the little diode detector on the dummy load. With  the daiwa meter it shows 5ish watts into t he qrp labs dummy load. So Im going to assume I didnt whack the radio too hard. I could be wrong here becasue when I connect an antenna my power only shows 1 watt on  the Daiwa. Maybe the load is causing an effect or failure in the QCX+ 20m after all.
Moving on for now...I started checking voltages...
The first check was at the left side of R10 - 12.7 volts (as supplied by bench supply)
The second voltage check was at the top pin of the variable pot which is also the the 5v regulator output, IC1. - 0 Volts

Before I take this out and replace it, is there anything else I might need to check first. A) I want to be sure the component is bad, and B) a Mouser order is $8 shipping, so I certainly want to be sure as I can be. So if I can check anything else I might have fried hooking up backwards, that I can preemptively test, please chime in.

Thank you so much for your help.
Bryan, n0luf

Bryan,

It is not clear what tests you have completed from the video that passed.  I would also confirm that you are testing first with the QCX in "practice" mode so that no RF is feeding the amp, then going back and trying with RF supplied by the QCX.

As to IC1, did you measure the input pin to verify that it was getting voltage from Q4?  

I would also confirm that you have power on the drains of the MOSFETs Q1 and Q2.  

I would check the gate to ground resistance with the power off and verify that it changes as you adjust R5.

Even with IC failed open, you should be able to do most of the tests in the video that do not require an oscilloscope.  You would not have much power amplification without the bias to the MOSFET gates.

The above are just suggestions. 
73
Evan
AC9TU

Update:
IC1 was okay. I didnt notice he was transmitting for that to measure correctly.
I went through all his other voltage measurements (ones that didnt require a scope) and all is well except for one place.
Where he measures the voltage from the doubler, between D3 and D5. running the same rail voltage of 12.7v he got about 70VDC between the two diodes. I got about 8VDC. Hopefully, that might be a clue to help someone smarter than I toward the next steps.

Meanwhile, I will disassemble and start looking closely at T1 and T2.

Bryan, n0luf

On Tue, Mar 29, 2022 at 08:57 PM, <bc3910@...> wrote:

Where he measures the voltage from the doubler, between D3 and D5. running the same rail voltage of 12.7v he got about 70VDC between the two diodes. I got about 8VDC. Hopefully, that might be a clue to help someone smarter than I toward the next steps.

Bryan,

The voltage at that point comes from rectifying and doubling the output of the amp.  Since the amp is not amplifying RF, there is not signal to double.  The value you are measuring seems reasonable if there is no RF being supplied to the amp.

In place of the oscilloscope, you can in some cases use an RF Probe.  I have one built from a kit that I tested for our Build Club:
https://www.qrpkits.com/rfprobe.html
Or if you have parts already:
https://www.nonstopsystems.com/radio/pdf-ant/antenna-dummy-article-6.pdf

That would at least let you know if you are getting RF from the QCX through to the IRF510s. 

73
Evan
AC9TU
 

Evan,
Thank you. I will look into the probe
 I may have a few parts around. 

But for the other comment, I do have the radio connected just as he showed and rf should be present in the amp. If the voltage measurement he makes between d3 and d5 is due to no or little RF then I think the irf510s are blown (most likely since i did hookup backwards) or a winding is shorted or some problem in that row of  diodes arrangement perhaps?

Bryan,

Can you make the bias adjustment as described in the manual?  That is done with the QCX in practice mode (no RF output) and an amp meter connected in line with the power supply.  If that works, I would think the IRF510s are OK and look to the transformers as the possible issue.

Most of the time, it is poor connections with the leads of the inductors and transformers in these types of kits.  It can be hard to burn or scrape off the enamel insulation.  I scrape the wires and crank up my temperature-controlled iron when soldering them. 

The RF probe would at least tell you if you are getting the signal to the gates of the IRF510s.

If you plan to build more kits, I would suggest that you look into purchasing an oscilloscope, at least 20MHz capable.  I like the Siglent series of scopes.  I have read good reviews of the Hantek and Rigol series as well.  Hantek is the lowest cost of the three, with Rigol offering lower-priced models with lesser specifications than the Siglent that I have.

73
Evan
AC9TU

Evan,
Yes  I can, and did,  adjust the bias current with QCX in practice mode. 
I found an RF sampler I made years ago for the Oscope I used to have, which was designed to go in line on the RF output of a radio. I used to use it for setting up my PSK31 settings. I have converted it into a makeshift  RF probe and will attempt the remainder of the checks mentioned in the video It should at least tell me that RF is present and increasing in amplitude as I go through the signal path later today.

I checked all the windings on the toroids and transformers. Things seem okay however I can't tell much about confirming T1 whether it is wired into the board correctly. I did double and triple check it before soldering but with it now in the full circuit, it shows continuity on all 6 pads. I 'm guessing that this is normal.

Bryan,

The potential issue with T1 is getting the phase for the secondary wrong.  If that is the case, instead of push-pull, the two IRF510s will both conduct simultaneously and cancel each other.  That is when the oscilloscope is needed; to verify the phase.  If this is the case, you should see RF on both gates and almost none on T2.

Since all of the other checks were OK, this is the last thing that I can think of that could be causing the issue.  At least one of the IRF510s is conducting as the bias adjustment confirms.  Even with the other one failed open, you should see some RF out.

I checked my 40meter 50watt PA with my RF probe and did see first a slight drop (there is a 6DB attenuator on the input), the same voltage on the gates, then a significant increase on the drains of the IRF510s, and finally another increase in voltage on the output of T3.

At Input BNC     R3/R4 Junction     IRF510 Gates       T3 Primary     T3 Secondary     Output BNC
    10.4                     6.43                       9.7                        20                    43                       37.4

I measured 39watts out.  Verifies that the RF probe is working OK.  NOT lab grade.

The above measurements are made with 13.8 volts to both the QCX+ and the 50watt Amp.

I hope the above helps.
73
Evan
AC9TU

Resolved: (sorta)
I managed to finally get a little RF probe  to work, primarily as just an indicator since I didnt have the exact values for the diode and resistor.

Unbelievable, not having any rg6 patch cable lying around I had scabbed the big stuff from my HF rig and bulkhead. That meant using adapters to get it all connected. I found two bad adapters. One just had the inner conductor not making contact due to corosion but there was a a pl259 to BNC that was dead short. Long story short, no RF into the amp. The amp actually works, but exhibits lower power than I expect.

I recalibrated the bias voltage /current and managed to get 30 watts at 20V. at 2-3 amps. I still need to beef up my power cord to make sure that it is not limiting current but is there anything else I should look into for low output?

Moral of the story, assume nothing. Check everything.
Bryan, n0luf

I would follow Evan's advice.  I know the videos you refer to and they are very good. There are 5 (I think) on building the amp and a further "Troubleshooting" video.  You need to follow all the steps in the troubleshooting video as I mention below.

Did you start at the PTT section and make sure that is working O.K. You then need to check what is happening at L9 when switching between Rx and Tx.  That then leads onto Q4... is it being turned on.  Is there a positive voltage on the input of IC1 ? 

If you have checked as above then, if there is no 5v output IC1 is not looking good. We don't know if you did the checks outlined above as you stated you checked the input to R10 (I think it's the input you measured) which would show that there is 12 volts there but doesn't tell us if the transistor is turning on and passing that voltage to IC1.

Just to be sure and narrow down what is happening might I suggest....  go through the troubleshooting video again and do the testing at each stage as shown in it. You do NOT need to hook up the QCX.... indeed it would be better if you don't at this stage. Just supply 5 volts to the amp's PTT switch and it would be easier if you could just switch this off and on easily....  I use a press button switch for this. Now you can supply 12v to the amp and by means of the PTT 5v switch you can change between Tx and Rx.  Having got the 5v PTT line set up you can now go through the testing steps shown in the series of videos. You will be switching between TX and Rx as you test voltages at all the places shown in the video. You do NOT want to connect the QCX as you do not need a signal through the amp at this stage. Once you have verified the voltages for Tx and Rx at each stage you will then probably have identified the problem. If everything checks out as it should you may then need an oscilloscope, but you do need to check the DC voltages before you get to that stage.

Make a chart of the voltages as you test each step shown in the videos, have two columns one for Rx and the other for Tx voltages. You then have something to go on and can compare with the same test points on the videos.

Please attach a dummy load while testing just to be on the safe side. Probably don't need it as you are not putting a signal through the amp and amplifying it but it won't hurt to attach one anyway.  Once you have followed the steps in the video about troubleshooting then you should know a good bit more about the amp than before.

In a nutshell...   provide 12v or so to power amp. Arrange 5v to the PTT to switch amp between Tx and Rx. Follow the troubleshooting video. Do NOT connect up the QCX.

Hope that helps.

Reg                                G4NFR

Bryan,

Point F should be close to the supplied voltage when keyed without any RF input to the amp.  This is because there is no RF pumping to get the voltage up using D2, D9, D10, D4, and C10 as a voltage doubler.  9.3 volts seems low.  It should be closer to 12.75-1.4=11.35volts though not too low.  I would verify that the diodes are installed correctly.  Also, check the D4/D8 junction to confirm that there is supply voltage there when the PTT is enabled, and no RF is being fed to the input BNC.  Finally, check on Q3 Drain voltages.  It should be the same as point F.

What happens to point F when you supply RF to the input?  It should be over 2x the peak RF voltage to ensure that the diodes remain off.

You may have stated what band it is built for, but could you please repeat it as I can not seem to find it.  Higher frequencies will act differently than lower frequencies.  I am comparing it to my 40meter QCX and Amp.

73
Evan
AC9TU

Evan,
Per our private conversation I verified that the QCX was putting 5.3Watts to the qrplabs dummy load connected to the other end of the patch coax, so its verified passing RF correctly. 
Practice mode ON:
Diodes4, 9, 10, 2, 8, and 6 in with all stripes pointing toward T3.
Near V+ voltage at junction of D4 and D8. 11.8v.
Q3 gate RX=6.92v ; TX .046v.
Q2 Drain rx= .049 ;TX=-9.37V

Practice mode OFF:
Point F with RF. RX=.067; TX 31.7V

20 Meter amp.
I was actually hoping that I had a bad patch cable but thats not the case. The amp is producing about 30watts at 20V and about 15W at 12.7v supply voltage.

Thank you,
Bryan, n0luf

Bryan,

Here are the power curves for each band:



Based on your measurements, I believe it is working as designed, though maybe a little low at 20volts (36watts vs 30watts).  I would not do anything else at this time other than having fun using it.

73
Evan
AC9TU

--
regards,
Bryan, N0LUF