开云体育

What did I do wrong?

I built my 50W PA for 20M - This is my third QRPLabs kit, having successfully built two QMX first time 'round (high and low band versions) so I get the idea of how the manuals and instructions tend to be set up, and I think I got the PA completely right, except that the input SWR is a mess, hovering around 3.0 at 14 Mhz. This of course isn't going to work.? I don't want to smoke my QMX. I'm crazy about these little radios. They both work so brilliantly.?

(Note, I was manually keying the PA to test before the latest 1_00_016 firmware came out, thanks for that, PA key-up is working as expected with the new FW)

R5 is adjusted as per the manual, plus tried a few percentage points +/- either way - no improvement. My antenna is a meticulously trimmed-for-20 dipole with excellent performance. I've also tried a range of different voltages from 12-20, and there's no meaningful difference in the SWR sweep when I change that variable. Also, the nano-VNA more or less agrees with the SWR sweep that shows high SWR on 20M. And of course, I'm doing these sweeps with the power amp keyed.?

I'm not an expert in PA, or really analog anything design, so I'm at a loss here. Nothing really sticks out on the schematic at this point in my education.? I re-wound the trifilar T1 a second time just for fun with a brand new set of wires, but the outcome is identical. High input SWR.? What if I removed a turn from the T1 transformer? Would that shift this nice low SWR dip I see at 10 mhz? Or added MORE turns?
?
If I wanted to run at 30M, then of course it looks like I'd be in excellent shape here, but I need 20.? As I understand it, having a SWR of 2 into a PA isn't even unusual, but 3 is a whole other story.?

I would appreciate if the group could share their knowledge and help me better understand what's going on and suggest avenues I could pursue to correct this. And please be kind, remember I'm new to this and trying to get into the hobby.?

Thanks!
Andy KA1PLU

Hi,
what you can always do is to add a 5W 3dB attenuator to the PA input to lower the worst VSWR to 2:1.
Or you can change the internal 3dB to a 6dB attenuator.
In both cases you will need a double power at the?PA?input to get the same PA output power as now.
In the first case you will also loose 3dB at the receive.
--
73 Bojan S53DZ

This does sound easy enough - to make the attenuator attenuate more, although it looks to me that there's a very happy frequency that this amp wants to operate, which implies that somewhere, this frequency of 10 Mhz is being set by a certain combination of components - capacitors, inductors, etc?

Could it be C2 or C3? And I don't know about changing the number of turns on T1, but maybe? I could experiment with a few things I guess.? Any suggestions for starting points from the crowd? I've attached the schematic for convenience.?


Andy KA1PLU

On Wed, Feb 14, 2024 at 12:04 AM, <kontraptioneering@...> wrote:

it looks to me that there's a very happy frequency that this amp wants to operate, which implies that somewhere, this frequency of 10 Mhz is being set by a certain combination of components - capacitors, inductors, etc?

Could it be C2 or C3? And I don't know about changing the number of turns on T1, but maybe? I could experiment with a few things I guess.? Any suggestions for starting points from the crowd? I've attached the schematic for convenience.?

This isn't the first time this has come up here...I'm curious as well.?

I don't know much, but it is interesting to me/ I wonder why 1 uf caps are used on the input here. On the 10W linear PA kit, it goes straight into BS170s for the first stage, and then some 0.1 uf caps before the IRF510s....... And the input transformer is different.? ?
The WA2EBY also uses 0.1 uf on the input.?

-Nate N8BTR

likely causes....

D1 bad or backward.
R2/3/4 in the wrong places or wrong values.
or T1 has a shorted turn(s).

I'll toss in soldering problems.

Note: the 10W linear is not even close to the 50w in general design.

--
Allison
------------------
Post online only,?
direct email will go to a bit bucket.

Hi Andy, Ron,

Ron, good point to check the T/R switching, but check all switching diodes: the RX and TX.

My 50W (QCX) PA is also built with the 20m LPF.
I have just tried?quickly driving it with the QMX and the VSWR measured with the QMX was between 1.7:1 to 2.0:1 for 40m to 20m band. The driving RF power was around 4W.
However, when the PA is unkeyed, it means in RX position, than you should have the thru connection to the output so to the dummy load. The VSWR in this case is not 1:1 but it is not unbearable either. This is the R.L. what the receiver feels.

In addition the internal input 3dB attenuator is built as PI attenuator (330 | 22 | 330 Ohm) for 60 Ohm to 60 Ohm, so there are also some parts of a dB of inherent R.L. difference.
--
73 Bojan S53DZ

Wonderful suggestions - I will check these this evening. T1 has been checked and double checked (and wound twice just for fun) so I don't suspect it. But the other parts deserve a triple check, and it's even possible that D1 is in fact bad or backwards.

I should add that RX performance seems fine with the PA inline (and it must be powered with the green light on only, of course) and power consumption when keyed also increases to a value that seemed inline with expectations with a modest input power, but perhaps I made an error. I will borrow the de-soldering gun from work this evening and report back to the group with my findings and any cautionary tales.?

I really appreciate everyone helping me with my amp. Thank you!

The amp is (more ore less) working. Heap down your abuse upon me for my malignant deeds, of which I will speak of presently....




Center pin of q4 got disconnected from one of the resistors.?

Works now, but SWR is high.
On a short piece of coax from TX to PA, the SWR is 2.5 and QMS shows 5.4W. Uncomfortable.
Long coax shows 1.9 SWR and 2.4W, PA draws 45W at 18V, heat is good, definitely getting out.?

The amp is (more ore less) working. Heap down your abuse upon me for my malignant deeds, of which I will speak of presently....




Center pin of q4 got disconnected from one of the resistors.?

Works now, but SWR is high.
On a short piece of coax from TX to PA, the SWR is 2.5 and QMS shows 5.4W. Uncomfortable.
Long coax shows 1.9 SWR and 2.4W, PA draws 45W at 18V, heat is good, definitely getting out.?






Getting out though!
Weirdness with the cable has me head scratching, and I prefer to run the long (5 meter) cable from the TX to PA for the apparent attenuation. Did Amazon ship me spark plug wire with BNC on the end? LOL wouldn't surprise me.??

Thanks for all the help. I have other data, but this is the 3rd time I've tried to post these images and the forum keeps losing the post. Still feels like a hack, so wondering what others think?
Andy KA1PLU