The AM Forum
March 28, 2024, 07:15:58 AM *
Welcome, Guest. Please login or register.

Login with username, password and session length
 
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: [1] 2 3   Go Down
  Print  
Author Topic: 40 meter push pull RF deck  (Read 65329 times)
0 Members and 2 Guests are viewing this topic.
ka1tdq
Member

Offline Offline

Posts: 1509


Red part turned in for a refund.


« on: March 13, 2016, 09:10:00 PM »

I would like to build a 2 FET, push-pull RF deck for 40 meters.  My plan is to use the frequency generator in the picture at 2x the frequency.  Use a 74F74 after that to get two phases 180 degrees out (it works with my current CW transmitter).  Then copy Steve's duty cycle adjuster, and run that into the IXDD614's. 

That's the easy cookie-cutter part...

I'm trying to fit this onto a 14" x 14" aluminum sheet and I don't want to re-engineer halfway through the project.  My question is what do I use for an output transformer for the FETs?  My current CW transmitter uses 8 FB-43-1020 cores to make a 1:1 transformer (4 stacks of 2).  I know that'll work, but is it best?  If something smaller will work, that'd be great because I want to fit it onto a smaller footprint. 

The series capacitor in the picture goes from 33 to 275pf (should be plenty for 40 meters).  I'll pick up a broadcast style air variable at this coming hamfest for the parallel cap. 

Jon
KA1TDQ


* IMG_2002.JPG (1911.3 KB, 3264x2448 - viewed 1239 times.)
Logged

It’s not just values, it’s business.
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2648


Making AM GREAT Again!


« Reply #1 on: March 15, 2016, 05:22:31 PM »

Research a TLT.   Mini coax and twisted pair on a torroid.  Commonly known as a Transmission Line Transformer.

WA1GFZ,  Frank,  has done extensive work on these I believe.

Work MUCH better than a binocular core.

--Shane
KD6VXI
Logged
ka1tdq
Member

Offline Offline

Posts: 1509


Red part turned in for a refund.


« Reply #2 on: March 27, 2016, 06:34:53 PM »

Thanks Shane.  Good talking to you on 40 by the way a little while ago. 

I also got a response from a guy in Australia, and he used brass tubing in the cores for excellent primary to secondary coupling.  The secondary wire goes through the tubing.  It looks very neat!

I'm focusing on the driver right now.  I've ordered the parts to build Steve's VFO from the /2 flip flop to the output IXDD's.  The only difference is that my flip flop will be a 74F74.  I figured I'd build the driver separately.  That way if I get a small RF deck to work and want to go bigger, I can just unplug the driver and have that part all done.

I'm going all fancy too... SMA chassis connectors for the output.

Jon


* IMG_0007.JPG (1914.24 KB, 4032x3024 - viewed 1289 times.)
Logged

It’s not just values, it’s business.
ka1tdq
Member

Offline Offline

Posts: 1509


Red part turned in for a refund.


« Reply #3 on: April 24, 2016, 10:02:58 PM »

Ok, I finally got around to building the duty cycle adjuster.  I'm getting 13.8 volt peaks (supply voltage to the IXDD's), but the duty cycle doesn't adjust at all when I move the 2k pots from one extreme to another.  It's the same with the signal driving the IXDD's, only it's a 5 volt peak signal.  And, from the picture, it looks to be >50% duty cycle, which is not good.

I'm pretty darn sure I cook-booked it correctly from the class E website.  I'm using a 5 volt pp signal from a DDS VFO for 20 meters.  That then goes to a 74F74 flip flop to divide the frequency by 2 and get a 40 meter, dual phase square wave.  The rest is from the class E website VFO schematic (from the 33pf capacitor on to the IXDD's).

Jon


* IMG_0116.JPG (2213.97 KB, 4032x3024 - viewed 1137 times.)

* IMG_0117.JPG (1877.07 KB, 4032x3024 - viewed 1105 times.)
Logged

It’s not just values, it’s business.
K6IC
Contributing
Member
*
Offline Offline

Posts: 745


« Reply #4 on: April 25, 2016, 12:37:26 PM »

Hi Jon,

Do you have a Link to the Duty-Cycle Adjuster diagram?

Have searched the E Forum,  and Class E Ratio sites,  but have only found the 2-Band VFO schematic,  and believe that that may not be it,   as it dates from about Feb '09 ...

Thanks,  Vic
Logged
KF1Z
Member

Offline Offline

Posts: 1796


Are FETs supposed to glow like that?


« Reply #5 on: April 26, 2016, 01:29:32 PM »

the 2 band vfo schematic has duty cycle adjustment.

http://www.classeradio.com/vfo_2_band.pdf
Logged

K6IC
Contributing
Member
*
Offline Offline

Posts: 745


« Reply #6 on: April 27, 2016, 01:43:18 PM »

Hi Bruce,

Did look at that schematic again yesterday,  and it DOES have 2 K Duty Cycle pots,  but had felt that there was another diagram floating around that used Comparators,   so was not certain just which diagram that Jon was referring to ... probably just defective memory cells.

Thanks,   Vic
Logged
ka1tdq
Member

Offline Offline

Posts: 1509


Red part turned in for a refund.


« Reply #7 on: April 28, 2016, 08:49:53 AM »

I'm switching gears on this project.  I previously built a 2-FET push-pull CW rig for 40 that puts out 140 watts, and now I'm going to build a 4-FET push-pull.  I'll use an IXDD on each FET since it's harder to drive on 40 meters.  I found a nice chassis at a thrift shop and the whole thing should look pretty cool. 

Jon
Logged

It’s not just values, it’s business.
ka1tdq
Member

Offline Offline

Posts: 1509


Red part turned in for a refund.


« Reply #8 on: May 04, 2016, 11:48:40 AM »

I've gathered all the parts to build this schematic. I've verified with a scope that my DDS VFO will do single phase square waves with variable duty cycle at 5 volts + peaks.

I'm going to build it. Does it look right (for 40 meters)?

Jon


* image.jpeg (1748.77 KB, 4032x3024 - viewed 1432 times.)
Logged

It’s not just values, it’s business.
steve_qix
Contributing
Member
*
Offline Offline

Posts: 2599


Bap!


WWW
« Reply #9 on: May 04, 2016, 01:21:09 PM »

Everything looks quite reasonable !

Regards,  Steve
Logged

High Power, Broadcast Audio and Low Cost?  Check out the class E web site at: http://www.classeradio.org
ka1tdq
Member

Offline Offline

Posts: 1509


Red part turned in for a refund.


« Reply #10 on: May 04, 2016, 01:33:32 PM »

Cool, thanks! I'll start putting pieces together on the chassis.

This will be a relatively low power rig, but I can just use it for local talking. It'll be nice just to be able to turn the rig on and talk. I'll fire up the three 3-500's for nighttime battle conditions and use my other transmitter.

Jon
Logged

It’s not just values, it’s business.
ka1tdq
Member

Offline Offline

Posts: 1509


Red part turned in for a refund.


« Reply #11 on: May 04, 2016, 08:41:19 PM »

This will be the general layout.  A 1/2" block of teflon will go beneath the series capacitor to isolate it from ground.  And the coil will be cut down to proper size.

Jon


* IMG_0155.JPG (1994.25 KB, 4032x3024 - viewed 1295 times.)
Logged

It’s not just values, it’s business.
ka1tdq
Member

Offline Offline

Posts: 1509


Red part turned in for a refund.


« Reply #12 on: May 07, 2016, 05:45:54 PM »

90% of the drilling is done.  I'm just waiting on a vernier dial to come in for the loading cap, and then I'll break out the soldering iron. 

Jon


* IMG_0157.JPG (2225.47 KB, 4032x3024 - viewed 1287 times.)
Logged

It’s not just values, it’s business.
W3GMS
Contributing
Member
*
Offline Offline

Posts: 3063



« Reply #13 on: May 07, 2016, 06:25:00 PM »

With those components you could certainly run more than one lonely FET!

Joe-GMS
Logged

Simplicity is the Elegance of Design---W3GMS
ka1tdq
Member

Offline Offline

Posts: 1509


Red part turned in for a refund.


« Reply #14 on: May 07, 2016, 06:38:39 PM »

Yeah, it's a learning curve.  I need to get the concept with one FET, and then I can expand out.  I've never built one before, so I'm still thinking, "How in the hell is this thing going to work?" 

But true... it's a lot of big stuff for <100 watts carrier output, but I do want to get a working class E RF deck.  I'll worry about the modulator later.

Jon
Logged

It’s not just values, it’s business.
ka1tdq
Member

Offline Offline

Posts: 1509


Red part turned in for a refund.


« Reply #15 on: May 08, 2016, 10:14:55 PM »

It's all done except for adding the vernier dial on the loading cap.  Gravity pulls it out of position, so I need to add some resistance to movement.  Plus, this'll give me some fine adjustment capability.

The shunt cap turned out to be 150pf.  It's all I had. 

I'll use 13.8 vdc for the drain for testing before I put 48 volts to it.

Jon


* IMG_0195.JPG (2505.95 KB, 4032x3024 - viewed 1166 times.)

* IMG_0196-2.JPG (2377.98 KB, 4032x3024 - viewed 1169 times.)
Logged

It’s not just values, it’s business.
WD5JKO
Member

Offline Offline

Posts: 1996


WD5JKO


« Reply #16 on: May 09, 2016, 07:23:12 AM »

The shunt cap turned out to be 150pf.  It's all I had. 

   Jon,

   I marvel at the incremental improvements in your construction. Your latest invention is really beautiful. I do have a comment though. Look up that 540v transient zener diode, and look for capacitance versus voltage across the diode. You might be surprised at how much capacitance these devices have.

   With class E, the peak drain voltage at maximum output is ~ 3.8x the DC drain voltage. You could round that off to 4, and the Vdd to 50v, and get 200v peak drain voltage. Then from there how much headroom do you want to provide, 1.5x, 2x? I don't know, but you are over 2x now. Also zener diodes used like this have a delay time before they clamp; perhaps the lead inductance along with the device capacitance are the primary cause?

   One possibility is to use 2 or 3 lower voltage surface mount zeners in series. Using them in series will lower the capacitance by the number of devices in series, and increase the zener power available by the number of zeners. The copper in the ground plane under the PA FET could be carved out to make pads to install the zeners in a leadless (no inductance) arrangement.

   I just wrote this before my morning coffee, so consider that too as my points made might be off base.

Jim
Wd5JKO
Logged
ka1tdq
Member

Offline Offline

Posts: 1509


Red part turned in for a refund.


« Reply #17 on: May 09, 2016, 10:19:57 AM »

Thanks Jim for the compliment.  I'm beginning to like this open frame construction.  The only thing with this one is that no one can touch the variable capacitor while transmitting, or should touch the variable capacitor.

I used the 540 volt zener that was in Steve's schematic.  I'm only guessing, but I'm imagining that the zener will clamp before the voltage gets to the maximum drain voltage of the FET, being 900 volts. 

If the capacitance turns out to be a problem, I will go ahead and replace it with surface mount ones.  I just need to look at the gate waveforms to make sure everything is kosher. 

...and I can't turn it on until the guy ships my vernier dial.  I ordered it on Saturday and the guys hasn't shipped it yet.  Too bad I didn't order it from Amazon, and then their drones could quickly deliver it to my door.

Jon
Logged

It’s not just values, it’s business.
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2648


Making AM GREAT Again!


« Reply #18 on: May 09, 2016, 09:52:52 PM »

Can't wait to hear the new rig Jon.

We have a pipeline usually,  so should be FB hearing this one.

--Shane
KD6VXI
Logged
ka1tdq
Member

Offline Offline

Posts: 1509


Red part turned in for a refund.


« Reply #19 on: May 09, 2016, 10:36:24 PM »

Ok, it has a pulse but it's not there yet.  

I rigged up a 3-gang electrical box which makes testing easier.  I can turn on the drive, drain, and choose 12 or 48 volts for the drain with the flick of a switch.  Attached is a picture.

I'm getting about 54% efficiency no matter what I do.  I've been varying both capacitors and I get different power levels, but they all end up being about 54%-ish.  At the current setting, I'm getting 6.3 watts output with 11.7 watts input (Id = .85 amps and Vdrain = 13.8 volts).  

Also, it looks like there is more than one waveform going into this thing.  I'm using a dirty Chinese DDS VFO and I'm thinking that I might have to put a low pass filter on it, or something.  The waveforms are there, but they're blurry.  So much for my $17 scope from the thrift store.  

So, that's where things sit right now.  I'm still waiting for my vernier dial.  I'm just playing with the loading cap carefully.

Jon

**  Oh yeah, and DON'T accidentally touch the variable cap.  Youch!


* IMG_0215.JPG (667.53 KB, 2048x1536 - viewed 1129 times.)

* IMG_0216.JPG (589.67 KB, 2048x1536 - viewed 1071 times.)

* IMG_0200.JPG (1590.86 KB, 4032x3024 - viewed 1061 times.)
Logged

It’s not just values, it’s business.
ka1tdq
Member

Offline Offline

Posts: 1509


Red part turned in for a refund.


« Reply #20 on: May 09, 2016, 11:11:55 PM »

Here's a little better waveform.  

Gate is on bottom, drain is on top.

I've played around with things, and I'm guessing that my loss could be coming from the drain transformer.  I've even removed the shunt capacitor, but efficiency actually went down when I did that.  I put it back. 

Jon


* IMG_0201.JPG (1800.85 KB, 4032x3024 - viewed 1170 times.)
Logged

It’s not just values, it’s business.
WD5JKO
Member

Offline Offline

Posts: 1996


WD5JKO


« Reply #21 on: May 10, 2016, 05:46:07 PM »

I'm using a dirty Chinese DDS VFO and I'm thinking that I might have to put a low pass filter on it, or something.  The waveforms are there, but they're blurry.  So much for my $17 scope from the thrift store.  

Jon,

  I attach the manual to that Chinese signal generator. I am using mine for 2 tone audio tests. The 2nd harmonic is 40 db down as long as you don't overdrive the output op amps.

  As to your scope traces, are you sure the scope probes are set to 10X and the compensation capacitor is set properly? Those scope probe ground clips are pretty long too for 40m class E. As to the efficiency issue, I think your at a wall until you can square up that drive signal which looks more like a sine wave than a square pulse.

Also review this link concerning the MHS-5200a (that is what you are using isn't it?):
https://hackaday.com/tag/mhs-5200a/
https://github.com/przemekklosowski/mhs52xx/raw/master/MHS5200.pdf

Jim
Wd5JKO

 

* MHS5200.pdf (596.52 KB - downloaded 6610 times.)
Logged
ka1tdq
Member

Offline Offline

Posts: 1509


Red part turned in for a refund.


« Reply #22 on: May 10, 2016, 06:26:27 PM »

That is the VFO I'm using.  I'm running it at 7.0 volts peak to get it to drive the IXDD's, which is probably too high for the VFO.  My plan to square things up is to run a sine wave out of the VFO at about 3 volts peak.  I'll run that into a hex buffer and adjust the VFO amplitude until I get about 40% duty cycle.  I can mount that chip directly onto the copper.  

I experimented around with lots of stuff last night, including the drain transformer.  I went down to 1:1 and things got worse.  I then went to a single core with a 1:2, but the core got hot.  I'm going to use 4 cores instead now with a 1:2 turns ratio (2x2 cores stacked).  Bigger is better.  

I'm now seeing how these transmitters are tuned.  Basically you tune for maximum power output on the loading cap, and then back off a little bit.  The series cap is for determining drain current and power output, up to a point.  At a certain point you can't reduce that cap's value any more.  

Jon
Logged

It’s not just values, it’s business.
ka1tdq
Member

Offline Offline

Posts: 1509


Red part turned in for a refund.


« Reply #23 on: May 15, 2016, 06:36:38 AM »

I'm up to 89% efficiency, but I had to go to 75 meters to do it.  The waveforms going into the FET look much more square at the lower frequency.  I'm getting 9.4 watts output with 13.8vdc drain voltage with .77 amps being consumed (10.58 watts total). 

I'd be just as happy to turn this into a 75 meter rig, since I don't already have one.  I didn't size the components though for the band.  My loading capacitor is all the way fully un-meshed.  The series capacitor is about half way (about 125 pf) and the coil is about 2.5uH. 

Where should I go from here to hit >95%?

Jon 


* IMG_0212.JPG (2115.59 KB, 4032x3024 - viewed 1146 times.)

* IMG_0213.JPG (2620.37 KB, 4032x3024 - viewed 1185 times.)
Logged

It’s not just values, it’s business.
ka1tdq
Member

Offline Offline

Posts: 1509


Red part turned in for a refund.


« Reply #24 on: May 15, 2016, 07:30:23 AM »

...and here's another waveform at 48 volts on the drain.  Efficiency drops to 85% at this drain voltage and I'm getting 82 watts output.  Not bad with only 15.5 watts being wasted. 

Jon


* IMG_0215-2.JPG (1900.58 KB, 4032x3024 - viewed 1088 times.)
Logged

It’s not just values, it’s business.
Pages: [1] 2 3   Go Up
  Print  
 
Jump to:  

AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands
 AMfone © 2001-2015
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
Page created in 0.066 seconds with 19 queries.