We seek to understand and document all radio transmissions, legal and otherwise, as part of the radio listening hobby. We do not encourage any radio operations contrary to regulations. Always consult with the appropriate authorities if you have questions concerning what is permissible in your locale.

Author Topic: Class D output match  (Read 6129 times)

Offline OgreVorbis

  • Full Member
  • ***
  • Posts: 141
    • View Profile
    • DosaidSoft
    • Email
Class D output match
« on: September 02, 2019, 2050 UTC »
Hello Stretchy and Redhat, I'm back.

I am going to change the output match on my transmitter to T200-2 or T200A-2 because the binocular thing is not working very well.
I am not sure how to do the windings. How many turns? Maybe either of you have a photo of how you wrapped it.
If I am doing class D, I shouldn't need any other components for matching, just the balun?

Is this how it's done?


Thanks  :)
Radio and Programming Blog: http://dosaidsoft.com/wp/

Offline Stretchyman

  • Hero Member
  • *****
  • Posts: 655
    • View Profile
Re: Class D output match
« Reply #1 on: September 03, 2019, 0541 UTC »
I didn't like the binoculars either and used T200-2s, one either side.

So you want to match the output?

Couple of ways....

Use a 1:1 transformer and then conjugate match the O/P (series L,  parallel C). Turns wise, not sure on either but use some scrap wire and experiment, start with 5T on the tran and series L with 1nF C.

Other method is simpler but not as 'adjustable'.

Simply work out your O/P Z, R=V/I, this will be lower than 50R, divide one by the other and take square root to give turns ratio. Standard stuff.

So, say, amp takes 10A@50V=5R.

50/5=10. Square root of 10 is 3 and a bit so use 3:1.

Ok

Str.
'It's better to give than receive' so why Rx when you can Tx!

                                              ;)

Offline OgreVorbis

  • Full Member
  • ***
  • Posts: 141
    • View Profile
    • DosaidSoft
    • Email
Re: Class D output match
« Reply #2 on: September 03, 2019, 1429 UTC »
I didn't like the binoculars either and used T200-2s, one either side.

So you want to match the output?

Couple of ways....

Use a 1:1 transformer and then conjugate match the O/P (series L,  parallel C). Turns wise, not sure on either but use some scrap wire and experiment, start with 5T on the tran and series L with 1nF C.

Other method is simpler but not as 'adjustable'.

Simply work out your O/P Z, R=V/I, this will be lower than 50R, divide one by the other and take square root to give turns ratio. Standard stuff.

So, say, amp takes 10A@50V=5R.

50/5=10. Square root of 10 is 3 and a bit so use 3:1.

Ok

Str.

OK, thanks! The thing I am confused about is physically how do I wrap the wire on the core? This may seem obvious, but it is not clear to me. Should it be like how I drew in the image or is it incorrect? I don't have space to use two separate cores, but I will use T200A which is a thicker model of the one you mention.

I tried to understand based on this image, but it is not clear:
Radio and Programming Blog: http://dosaidsoft.com/wp/

Offline redhat

  • DX Legend
  • ******
  • Posts: 1585
  • USA
  • Music is my drug.
    • View Profile
    • Email
Re: Class D output match
« Reply #3 on: September 03, 2019, 1622 UTC »
First off, what turns ratio are you needing? Everything I build is 1:1 ratio to keep modulator current levels in sane territory.  I also recently changed the output topology in my transmitter to a coaxial wound binocular type core.  It does get warm at 750 watts carrier, but not hot.  I needed an isolated output in the PA stages to facilitate line derived B+ in the following generations of the transmitter.

FWIW, binocular type output transformers are the standard in solid state PA's from DC to 200 MHz.

+-RH
Somewhere under the stars...
Airspy HF+, MLA-30/Mini-whip/Chi-Town Loop
Please send QSL's and reception reports to xfmshortwave [at] proton [d0t] me

Offline Stretchyman

  • Hero Member
  • *****
  • Posts: 655
    • View Profile
Re: Class D output match
« Reply #4 on: September 03, 2019, 1810 UTC »
Nigel uses a single core with turns wound in the opposite direction (most important or it won't work) just use seperate cores then it won't matter.

I use 1:1 with 5 Turns each side and conjugate match the o/p.

You need to 'experiment'!

Str.
'It's better to give than receive' so why Rx when you can Tx!

                                              ;)

Offline OgreVorbis

  • Full Member
  • ***
  • Posts: 141
    • View Profile
    • DosaidSoft
    • Email
Re: Class D output match
« Reply #5 on: September 03, 2019, 2242 UTC »
First off, what turns ratio are you needing? Everything I build is 1:1 ratio to keep modulator current levels in sane territory.  I also recently changed the output topology in my transmitter to a coaxial wound binocular type core.  It does get warm at 750 watts carrier, but not hot.  I needed an isolated output in the PA stages to facilitate line derived B+ in the following generations of the transmitter.

FWIW, binocular type output transformers are the standard in solid state PA's from DC to 200 MHz.

+-RH

I think 1:1 is probably fine. It looks like you have the output balun as your only matching. I don't know if that's true or not. I'm not an electrical engineer so I don't understand the "isolated output line derived B+" part.

I see some amplifiers that do the matching with one balun as you do, but I've seen others (especially VHF amps) that use three baluns. One for each side of the amp and then one final on the output. Is there an advantage to this?

Really all I am trying to do is get my efficiency up above 5 MHz. Everything on my PCB has been optimized (no long traces, plenty of caps, good groundplane, good waveforms). I've narrowed the problem to the output match and I think it is the fact that I'm using wire on my output balun instead of coax. I'll try what you did first, redhat, with the coax because that shouldn't be hard. Then Ill try a T200 if it still doesn't work.
« Last Edit: September 03, 2019, 2315 UTC by OgreVorbis »
Radio and Programming Blog: http://dosaidsoft.com/wp/

Offline OgreVorbis

  • Full Member
  • ***
  • Posts: 141
    • View Profile
    • DosaidSoft
    • Email
Re: Class D output match
« Reply #6 on: September 04, 2019, 0303 UTC »
See my last post.

I decided to do a little more research. Given my current PCB and my love for simplicity, I am trying to find a way to use only a single balun for matching - nothing else. I am able to get very high efficiencies up to 5.5 MHz, so it just needs a little extra push. I looked at your video more closely, redhat, and now I see that you have a class D tank circuit followed by a row of caps and then your 1:1 balun. My PCB as it is doesn't have space for a row of caps like this. Maybe I could just do the 1:1 and then follow it with a single DC blocking cap?

So I collected some images of different types of balun options.

With my lack of space, I don't think the first one will be very easy and it will require mods to the PCB. The second one isn't bad, but I have doubts that it will work for the frequencies I want. It looks like the last one is the one I should try. I should be able to do it with my current binoculars and without mods to the PCB. What do you guys think?



Radio and Programming Blog: http://dosaidsoft.com/wp/

Offline Stretchyman

  • Hero Member
  • *****
  • Posts: 655
    • View Profile
Re: Class D output match
« Reply #7 on: September 15, 2019, 1413 UTC »
Pictures don't explain the method!

However I know the first 2 well but not the 3rd.

Just use the one Nigel uses with opposed windings...

Str.
'It's better to give than receive' so why Rx when you can Tx!

                                              ;)

Offline redhat

  • DX Legend
  • ******
  • Posts: 1585
  • USA
  • Music is my drug.
    • View Profile
    • Email
Re: Class D output match
« Reply #8 on: September 16, 2019, 1438 UTC »
I think 1:1 is probably fine. It looks like you have the output balun as your only matching. I don't know if that's true or not. I'm not an electrical engineer so I don't understand the "isolated output line derived B+" part.

I see some amplifiers that do the matching with one balun as you do, but I've seen others (especially VHF amps) that use three baluns. One for each side of the amp and then one final on the output. Is there an advantage to this?

Really all I am trying to do is get my efficiency up above 5 MHz. Everything on my PCB has been optimized (no long traces, plenty of caps, good groundplane, good waveforms). I've narrowed the problem to the output match and I think it is the fact that I'm using wire on my output balun instead of coax. I'll try what you did first, redhat, with the coax because that shouldn't be hard. Then Ill try a T200 if it still doesn't work.

Sorry for the late reply.  My PA is designed to look into a 50 ohm balanced load.  The output transformer's only job is to convert the 50 ohm unbalanced load to something the PA can use.  electrically, it is configured just like the last picture you posted showing a BLF188XR or similar pallet with a 4:1 output transformer.  One of the benefits of CMCD, like class E is the voltage multiplication of the PA; peak output volts 3-3.5x supply volts.  This makes a 1:1 output transformation practical with the high voltage devices we have now.

I would guess the reason behind the split 4:1 output balun topology used on all the CMCD guy's stuff is the ease of construction.  I think it would have practical benfits too when dealing with higher currents encountered at power with large impedance transformations.  The more ferrite in the system, the less likely it is to saturate and cause problems.  FWIW, I did notice the output transformer in my transmitter running warmer when I converted it to a standard transformer output topology.  This is most likely because there is now twice as much voltage across the windings as before, since I reconfigured the transformer to voltage mode from current mode.

When deriving your PA voltage supply from the AC mains, there is no longer a safe signal ground.  All signals going into and out of the PA have to be isolated from ground.  This is the next step in my design evolution and adding isolation to the PA output is one of those steps.  It also means that the modulator and RF drive will also have to be transformer coupled or more likely, optically coupled through the use of photocouplers.

+-RH
« Last Edit: September 16, 2019, 1613 UTC by redhat »
Somewhere under the stars...
Airspy HF+, MLA-30/Mini-whip/Chi-Town Loop
Please send QSL's and reception reports to xfmshortwave [at] proton [d0t] me

Offline Stretchyman

  • Hero Member
  • *****
  • Posts: 655
    • View Profile
Re: Class D output match
« Reply #9 on: September 16, 2019, 1725 UTC »
Looks like you need to look at the Super Senior etc.

Str.
'It's better to give than receive' so why Rx when you can Tx!

                                              ;)

Offline redhat

  • DX Legend
  • ******
  • Posts: 1585
  • USA
  • Music is my drug.
    • View Profile
    • Email
Re: Class D output match
« Reply #10 on: September 16, 2019, 1938 UTC »
Until it can do cquam or other modes, no thanks.

+-RH
Somewhere under the stars...
Airspy HF+, MLA-30/Mini-whip/Chi-Town Loop
Please send QSL's and reception reports to xfmshortwave [at] proton [d0t] me

Offline Stretchyman

  • Hero Member
  • *****
  • Posts: 655
    • View Profile
Re: Class D output match
« Reply #11 on: September 16, 2019, 2010 UTC »
Soz, was meant for OgreV.

Definitely the simplest implementation of mains powered PWM.

Oh for Cwuam surely a DtoA and a linear...?

Not much use for anything digital here and doubt it will catch on commercially. Maybe it's MASSIVE in the US like most things over there??!!

Str.
'It's better to give than receive' so why Rx when you can Tx!

                                              ;)

Offline redhat

  • DX Legend
  • ******
  • Posts: 1585
  • USA
  • Music is my drug.
    • View Profile
    • Email
Re: Class D output match
« Reply #12 on: September 16, 2019, 2018 UTC »
Oh for Cwuam surely a DtoA and a linear...?

Nope, tradition high level PWM with phase modulated carrier drive.  All the pre-correction is done in the exciter.  The old transmitter was linear all the way from 0dBm to 400W carrier, at about 30% efficiency.  I'm trying to get away from that...

+-RH
Somewhere under the stars...
Airspy HF+, MLA-30/Mini-whip/Chi-Town Loop
Please send QSL's and reception reports to xfmshortwave [at] proton [d0t] me

Offline OgreVorbis

  • Full Member
  • ***
  • Posts: 141
    • View Profile
    • DosaidSoft
    • Email
Re: Class D output match
« Reply #13 on: October 06, 2019, 1841 UTC »
That's interesting. I didn't think you could do CQUAM on a PWM type system. If you don't mind, maybe you could explain it?
There is the Polar Explorer which is able to do class E with SSB.

On the original topic:
I replaced the balun with a coaxial wound one. I found a 1% decrease in efficiency (more or less the same).
I will not give up. I have now sent for revision 4 of my PCB. If this one doesn't work, then I have no clue. The waves were a little rounded on the current one, but this new one has only straight buses and the crystal is hooked up directly.
Radio and Programming Blog: http://dosaidsoft.com/wp/

Offline redhat

  • DX Legend
  • ******
  • Posts: 1585
  • USA
  • Music is my drug.
    • View Profile
    • Email
Re: Class D output match
« Reply #14 on: October 08, 2019, 0002 UTC »
If the efficiency stays the same, then most likely we are dealing with either output transformer saturation, or more realistically not enough dead time in the PA drive phases causing shoot-through (overlap).  The output miller capacity canl cause the drive waveforms to stretch on turnoff, causing overlap and loss of efficiency.

As far as AM stereo is concerned, the exciter provides phase modulated RF to the transmitter (or PA) and an envelope signal.  The transmitter combines the two to produce an C-QUAM signal.  This is the way AM stations have been doing it for 30 or more years.  You can also use this method to produce other signals, such as DRM, SSB, etc.

+-RH
Somewhere under the stars...
Airspy HF+, MLA-30/Mini-whip/Chi-Town Loop
Please send QSL's and reception reports to xfmshortwave [at] proton [d0t] me