W6PQL 150w? 1296 amp

35 views
Skip to first unread message

Ray Cannon - W7GLF

unread,
Apr 14, 2022, 6:38:16 PMApr 14
to PNW Microwave
I thought I would pass on my so far less than spectacular results of trying to get power out of my W6PQL 150 watt 1296 amplifier.

This was owned by KD7TS although it sure looks like the prebuilt version.  The soldering is very clean.  I have played with the bias and have no problem getting the bias current on each device up to as much as 3 amps. 

When I set the level on each device to 500ma I cannot seem to get more than 75 watts out of the amplifier with 8 watts in and running at 28 volts.  When it is up to like 3 amps per device I see about 100 watts out.  Still nowhere close to 150 watts.

I have tried snowflaking but I could not find anyplace that seems to make any difference.

I will spend some more time trying to figure out what is happening.

Mike Lewis

unread,
Apr 14, 2022, 6:45:26 PMApr 14
to Ray Cannon - W7GLF, PNW Microwave
Does that version have a trimmer on each input, to vround as i recall?  The values of the fixed cap on the input also chaged over time. Seems like ww tested a few values trouvleshooting Pauls rf deck, eventally got 170W out with 10 or 11W in.

Sent from my T-Mobile 4G LTE Device
Get Outlook for Android

From: pnw-mi...@googlegroups.com <pnw-mi...@googlegroups.com> on behalf of Ray Cannon - W7GLF <w7...@qsl.net>
Sent: Thursday, April 14, 2022 3:38:15 PM
To: PNW Microwave <PNW-Mi...@googlegroups.com>
Subject: [PNW Microwave] W6PQL 150w? 1296 amp
 
--
You received this message because you are subscribed to the Google Groups "PNW Microwave" group.
To unsubscribe from this group and stop receiving emails from it, send an email to PNW-Microwav...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/PNW-Microwave/84b4b9cb-563a-4cbb-8482-a206e5b25d58n%40googlegroups.com.

Ray Cannon - W7GLF

unread,
Apr 14, 2022, 7:01:12 PMApr 14
to PNW Microwave
Yes it has a 1-3 pF trimmer on each input.  I remember when Mike bought it and I am sure he got the pre-assembled board/heatsink.  I bought some from Mouser that are 2-6pF.  I also bought some 2 pF, 3 pF and 5 pF capacitors to play with.  Do you remember where you set the bias?  Jim mentioned in the build instructions 500 ma for device.  That is where I set it.  According to his graphs I should be able to get 170 watts out with 8 watts in but my amp is multiplying by 10.  It would be interesting to know just which parts you changed... 

Nick Krachmalnick

unread,
Apr 14, 2022, 9:02:10 PMApr 14
to Ray Cannon - W7GLF, PNW Microwave
Hi Ray,

When I asked W6PQL about setting the bias, he responded with the below:

"Bias voltage is also 28v (same as VDD); while watching the VDD current, connect VDD to a 28v current-limited power supply limited to 2 amps. Connect the bias to 28v and note the VDD current. Turn one of the controls to minimize the current. Set the other control for 500ma, then turn up the first one for 500ma more (1 amp total for the two devices)".
Jim

Initially using 28V as the bias voltage, I could not adjust the bias down to 500ma /device. Experimently we and found that by lowering the voltage to ~ 20V would provide adjustment of 500ma/device. (~ 4 V measured at each bias point on each device). Will test that tomorrow and see how much RF output is available.

73's,

Nick
N0CYT




--

Ray Cannon - W7GLF

unread,
Apr 16, 2022, 9:54:23 PMApr 16
to PNW Microwave
Interesting.  I have no problem getting bias below 500 ma.  I am using rev 7.11 of the board.  I think this is the correct schematic (hard to tell since version number is not on the schematic).  I note that it is using 6v zeners.  I assume you are doing the same...
2xkitschematic.pdf

Ray Cannon - W7GLF

unread,
Apr 16, 2022, 10:07:02 PMApr 16
to PNW Microwave
Putting message from W9BTV in this conversation to keep stuff together...

Hi Ray,

 

There were several different versions of the 1296 board that Jim produced.  I spoke with Jim on my board a while back when I was building mine and revised the later rev boards to have two trimmers on each input vs one fixed value and one trimmer.   Depending on which rev board you may obtain success by replacing the fixed caps with another trimmer of  around same value on the input side of the amp.   I was having the same challenge with my board, not reaching the rated RF output of 160 to 180 Watts with 10W drive .   Mike (K7MDL) and I tried various values on the input with no success.  Based on my ongoing discussion with Jim for my Rev board (Rev 7.3) from 2012,  I incorporated those changes  and after my amplifier was producing and easy 170 to 180 watts with 9 to 11 watts drive.

 

  1. Remove the two fixed capacitors (believed to be 2.2pf) on the input side of the devices and replace with a 2-6PF trimmers.  So in all… each side ended up with  two 2-6pf caps I ordered from Digi key.
  2. The output side where the output coax is soldered to the board was revised:  a  connection is taken from the middle of the trace…there is a break in this trace at the end which was intended to be shorted across so the output could be end-launched as an option.  If yours is bridged it is essentially adding a small capacitance at the end by including that separated piece of trace. To resolve this, remove the solder bridge and just let that piece ‘float’ by itself.  I believe this made the biggest difference in the output.
  3. I also reflowed the output tabs on the transistors to ensure a good connection and solder flow.  It’s a bit challenging after the fact because they are now connected to the copper heat spreader so Flux and a very high wattage iron will do it.

 

If  you would like to talk about this off line or I can share anything else of my experience I had with mine getting it finally working, I be happy to share.

 

73,

Paul

WA9BTV

My reply to is:

Thanks for the info Paul.  I am not crazy about replacing input caps with trimmers but I certainly could try replacing them with different values of fixed caps.  Using trimmers adds more inductance and it seems to me will make them trickier to adjust having 2 variable caps per stage which will interact with each other.   I don't see any of his boards with extra trimmers - i looked through multiple versions.

Your point #2 is very interesting.   I suspect your board is different than mine which uses branchline couplers on input and output - not sure where you are referring to as the output coax on mine is connected directly to the branchline coupler port.  I don't see any breaks or bridges.  I have attached a photo of mine.  My board is rev 7.11.  

I tried reflowing tabs both input and output with no change.

I tried snowflaking with no change (maybe I did not try hard enough - see below).  I double checked my Bird with another way of measuring power but no change.  I checked input/output SWR and that looks OK.

I checked the gate bias voltage and it is 4.05 volts (data sheet say 3 to 5 so I am good here) and the idling current is 500ma for each side.  I did notice that if I increased the idling current the power out increased but I don't think that is a good idea.  I am nervous sure about setting idling current to like 2 Amps per device.

One thing I noticed on the datasheet is the part is spec'ed at 2 GHz for 10 dB gain.  Everyone seems to be talking about seeing like 17x gain (12.3 dB) when I am seeing 10 dB of gain.  I guess the gain must be higher at 1296.

I just read the paper by HB9BBD from 2010 EME conference (see http://www.ntms.org/eme/presentations.php)  and think I need to try snowflaking again taking some hints from what he did..  He was using version 7.2 of the board and had a lot of snowflakes on the board.
On Thursday, April 14, 2022 at 6:02:10 PM UTC-7 nick...@gmail.com wrote:
20201230_195745_resized.jpg

Mike Lewis

unread,
Apr 16, 2022, 11:57:01 PMApr 16
to Ray Cannon - W7GLF, PNW Microwave

Here is a 7.1 version board.  Only 1 cap on the input.

20200829_002159.jpg

Mike Lewis

unread,
Apr 16, 2022, 11:59:20 PMApr 16
to Mike Lewis, Ray Cannon - W7GLF, PNW Microwave

Correction, only 1 fixed cap on the input, the other is a trimmer to ground.  I do not know the values.

Nick Krachmalnick

unread,
Apr 17, 2022, 11:13:19 AMApr 17
to Mike Lewis, Ray Cannon - W7GLF, PNW Microwave
Good morning all,

Yes Ray, same schematic and using same 6V Zeniers. Also measured the resistance from the bias voltage input point to the device and all devices totalled up correctly. Will look at the board to see if the version # survived the board creation process.

We downloaded the stencils and used the positive version to create the boards. Got sidetracked yesterday so did not measure the output. I did order additional 2-6 pf variables to add on the inputs.

73's

Nick



--

Fred W

unread,
Apr 17, 2022, 3:20:14 PMApr 17
to Nick Krachmalnick, Mike Lewis, Ray Cannon - W7GLF, PNW Microwave
Gentlemen,

I would like to thank you for this thread. I can understand your frustration, but in the meantime my knowledge of mosfet amps is increasing greatly. Exactly why building equipment is so valuable.

Please continue with this fine dialog.

73's
Fred - N7FW

Ray Cannon - W7GLF

unread,
Apr 17, 2022, 4:11:20 PMApr 17
to PNW Microwave
I double checked and my board matches the schematic.  I just tested and I can get bias down to 220 ma.  Don't know what is different.  At 500 ma I see gate voltage (Vgs) at 4.15.  Data sheet says Vgs should be between 3 and 5 volts.  Perhaps our devices are just different.  Anyway  am beginning to do snowflaking and will take a picture when I am done.  Already found one place on input coupler in same place as HB9BBD.   I figure it will take a while to find positions.

Nick Krachmalnick

unread,
Apr 17, 2022, 5:37:25 PMApr 17
to Ray Cannon - W7GLF, PNW Microwave
The only thing I can see on the boards is below: 

W6PQL
2 2009

Fred did your boards come out better and see the rest of the words or is that all there is (as in Feb - 2009 maybe)?

No other numbers or letters as in v... 
I tried to download the stencils we used, but I am not able to access Jim's website or the data. Maybe undergoing maintenance or an outage?

My voltages after setting the currents varied slightly from 4.12 V on # 4 XRF-286 to 4.19 V on all the rest of the XRF-286's. I wrote them down on the amp for future reference (with the bias supply set @19.6 V DC). 





--

Nick Krachmalnick

unread,
Apr 17, 2022, 7:53:04 PMApr 17
to Ray Cannon - W7GLF, PNW Microwave
I was able to get the positive stencil from W6PQL website. 

Looks like our etched boards are version 7.2   2/2009 per the stencil.

image.png
In the testing process, I found that I had a bad on/off switch on the aux plastic 23 cm Xvtr from SG Labs. Took some time to replace that. Will pick this up tomorrow again.


--

Ray Cannon - W7GLF

unread,
Apr 17, 2022, 11:05:07 PMApr 17
to PNW Microwave
I am not sure how Jim does his revision numbering.  Mine is rev 7.11 but does that mean July 2011 or does it mean 2007 November or perhaps neither of those - don't know.  In any case your board is definitely different than mine although it seems to have the same bias circuit.  Odd that the two bias circuits are not laid out identically but looking at then carefully assuming the parts are installed correctly they are electrically identical.  On my board the two bias circuits are identical.

I took a break from fooling around with amplifier today and will go back at it tomorrow.  So far I have the power up from 75 watts to about 95 watts out with 8 watts in with the single snowflake I added. Still more to do.  I wonder about adding piston caps (hard to find) since I don't see a sharp peak when using 1-3 pF trimmers.  Maybe there is one that I just don't see? 

I did just notice that KL7UW has a schematic that shows 10 pF input caps from the input hybrid coupler while mine shows 2.2 pF caps.  I might try changing those to see what happens.

Nick Krachmalnick

unread,
Apr 18, 2022, 6:39:20 PMApr 18
to Ray Cannon - W7GLF, PNW Microwave
Hi Ray,

I am getting at this point only about 10-15 watts output after adjusting the input trimmer caps only. I need to do some more work on this. I have additional trimmers 2- 6 pf coming should be here this weekend. I got 25 trimmers. Enough for Fred and I + 1 other potentially. I was only able to see the wattage using one of my birds with a 100 Watt - 0.95 to 1.26 GHz slug. 

Are you using a 23 cm slug or a UHF KW for tuning?

I did verified all interconnecting cable, voltages and connections. No shorts or opens.

My test setup:

FT-817 (2M) -->XVTR (23CM)-->SX-1100 Wattmeter (~10watts 1296 visible on TX)-->input of 1296 Amp. Output of 1296 Amp to Bird wattmeter (100J slug) --> 500 Watt 50 Ohm termination load. 

73's

Nick




--

Ray Cannon - W7GLF

unread,
Apr 18, 2022, 9:31:35 PMApr 18
to PNW Microwave
I an using 400-1000 250 watt UHF slug.  I do have a 1.1-1.8 Ghz slug but it is only 10 watts.
Currently I am using my ICOM 1271A which puts out like 8 watts to 1296 amp.  Output is going to Bird 250 W UHF slug and then to my 1296 antenna (no dummy load for more than 100 watts).  No one hears me on 1296 anyway...

Are you sure your bias circuits are wired correctly?  We should be able to figure out what voltage should be by modeling bias circuit.  We know what the Vgs is (on my amp 4.15 Volts) so it seems like we should be able to do a circuit analysis.  Lessee...

Vbias goes through 2*510 ohm resistors in series or 1020 ohms in series with 6v zener diode.  I assume you should measure 6.1 volts at the junction between the diode and r14 (I see 6.1V).   The resistors (100 ohm, 200 ohm pot and 300 ohm o ground) are cleverly chosen to add up to 600 ohms so each 100 ohms drops it from 6 volts by one volt.  So on the pot one end should be 5V and the other end should be 3V.  I verified this on my amp.  I am confused about you guys saying you are reducing the bias voltage to 20 volts - why would that matter if the zener diodes are installed correctly?  Is it possible the zeners are installed backwards?  The stripe (cathode) should be pointing away from ground.

I have done some snowflaking and now have the amp putting out 110 watts.  I suspect there is more snowflaking to be done - hard to find the correct places.  So far I have just sorta lucked into them by looking at what HB9BBD did and trying the same places.

Ray Cannon - W7GLF

unread,
Apr 18, 2022, 10:29:18 PMApr 18
to PNW Microwave
I guess it is not possible to have reversed the diodes - you would not be able to get any bias voltage but are the resistors all correct values?  The pot should adjust between 3 and 5 volts.  I did read somewhere about the devices with flanges not dissipating power as well as the XRF286S which do not have flanges but I wouldn't think that would affect bias current.
Reply all
Reply to author
Forward
0 new messages