N4LTA
Contributing
Member
Offline
Posts: 1070
|
|
« on: August 25, 2019, 12:29:37 PM » |
|
Made some progress this weekend on my Class D system. I completed the low voltage poeer supply, a second PWM generator and a large capacitor board for the main power supply.
The cap board has 14000uF at 250 volts worth of filtering. I plan on regulating the output at 150 volts.
Here are a few pictures:
The PWM board based on Steve's circuit
The low voltage power supply board and power transformer. Toroidal power with regulated +12, -12 and +5 volt output plus raw +20 and -20 volts
Test board for four SiC MosFETS
|
|
|
Logged
|
|
|
|
N4LTA
Contributing
Member
Offline
Posts: 1070
|
|
« Reply #1 on: August 25, 2019, 10:35:08 PM » |
|
I got the drivers, the bypass caps and the four MOSFET mounted tonight. Anyone see anything wrong that jumps out?
The drivers are tedious to mount but they all seem to be installed without shorts as tested with a ohmeter.
Also tested the PWM generator and it is working well.
Next I need to build the audio Pre PWM filter. I have a board and hope to build it tomorrow.
I ordered a Hammond 7 inch rack panel and enclosure to house the complete PWM system. It will probably need some steel structure addition to handle the 35 pound toroidal power transformer. It is overkill for sure at 1400 VA.
Thanks for looking and comments are appreciated.
Pat
|
|
|
Logged
|
|
|
|
N4LTA
Contributing
Member
Offline
Posts: 1070
|
|
« Reply #2 on: August 27, 2019, 03:28:45 PM » |
|
Tested the UCC 2501 Modulator board last night and it is working fine. Frequency was 151 KHz
Layed out an audio processor board using Steve's design as a guide with phase select and a negative peak limiter and a 7.5 KHz filter
Also made another bipolar power supply board. Plotted the boards this morning and just finished etching them. Will drill and stuff them tonight.
I plan on using a larger Hnmmond rack enclosure for the big power supply and PWM generator and driver and then using a smaller 3" rack enclosure for the compressor and audio chain.
Thanks for looking
Boards about half etched in H2O2 and hydrochloric acid (Muriatic acid)
Pat N4LTA
|
|
|
Logged
|
|
|
|
vk3alk
Member
Offline
Posts: 263
|
|
« Reply #3 on: August 27, 2019, 07:46:45 PM » |
|
All looks good and things moving along nicely For the PWM are you using Steves output design ( Class D ) or a half bridge like Nigels..... Have you thought about the LPF yet and the material to use.... Wayne
|
|
|
Logged
|
|
|
|
N4LTA
Contributing
Member
Offline
Posts: 1070
|
|
« Reply #4 on: August 27, 2019, 08:19:02 PM » |
|
Wayne,
I plan to use a half bridge. I haven't designed the filter yet, but I have quite a few cores from my past 500 Khz work. My voltage will be higher so it will require a different filter. I will be closer to Nigel's output stage because my RF stages will be with SiC and GaN devices run at higher voltage and less current.
I am shooting for 300 watts carrier on 80 meters with four SiC Mosfets. Not having done PWM AM modulation before - I am experimenting. Never having built any PWM system I am learning and having a good time. I'll let the smoke out of some Mosfets for sure.
BTW - TI has a nice 500 KHz discrete PWM generator made for class D audio. will be easy to drop it to 150 KHz or so and it should adapt well for AM modulation.
The offer full design notes and full double sided PC Board Gerber files. I ordered ten boards from Seeed Studio for $4.17 pus $13 shipping from China. I have used them before and they do a good job. If anyone wants a board, let me know and I'll send you one or two. We will see how they turn out, but the last ones they made were nice boards.
Pat
|
|
|
Logged
|
|
|
|
KD6VXI
Contributing
Member
Offline
Posts: 2683
Making AM GREAT Again!
|
|
« Reply #5 on: August 27, 2019, 09:29:34 PM » |
|
What's the chip number? I'm interested.
--Shane KD6VXI
|
|
|
Logged
|
|
|
|
N4LTA
Contributing
Member
Offline
Posts: 1070
|
|
« Reply #6 on: August 27, 2019, 09:55:00 PM » |
|
Look Here http://www.ti.com/tool/TIPD108Uses a SMT op/amp and a rail to rail comparator and a voltage reference source Samples are available for the op/amp and possibly the comparator. Pat N4LTA
|
|
|
Logged
|
|
|
|
N4LTA
Contributing
Member
Offline
Posts: 1070
|
|
« Reply #7 on: September 06, 2019, 01:15:35 PM » |
|
The analog discrete PWM boards came today. I don't see how they do it. Extremely high quality boards. Cost about $4 and chang3 for ten boards. Shipping was a little over $12
Took a little over a week - As I said - If anyone want to experiment - I'll send you one.
|
|
|
Logged
|
|
|
|
N4LTA
Contributing
Member
Offline
Posts: 1070
|
|
« Reply #8 on: September 11, 2019, 09:46:12 AM » |
|
Finished the PWM generator for the PWM power supply. This is the final circuit board. I added a couple of voltage regulators and RCA type connectors to make hookup easier. The board can run off 9-20 volts now.
|
|
|
Logged
|
|
|
|
N4LTA
Contributing
Member
Offline
Posts: 1070
|
|
« Reply #9 on: September 15, 2019, 12:01:27 PM » |
|
Got the enclosure assembled and am ready to punch the front panel for the controls and three meters. The filter cap board has ten 1800 uF 250 volt capacitors. The PWM boar, and the low voltage power supply board are ready. I may or may not rework the power driver board for T0220 FETs to use a more modern MOSFET with higher currrent rating.. I also need to pick a heatsink. With a newer 60 amp set of MOSFETS, it probably won't need to be large. If I redo the board, I can put a RCA jack on the board to make it plugable to allow changing to a different PWM board for experimenting.
The transformer is reduced to a 800 VA one to reduce the weight. Should still be plenty to do more than the legal limit. Output votage from the high voltage supply will be 150 volts regulated.
Hope to do the meter punching this afternoon - ugh!
|
|
|
Logged
|
|
|
|
vk3alk
Member
Offline
Posts: 263
|
|
« Reply #10 on: September 24, 2019, 11:22:54 PM » |
|
Hi Pat... Wonder what stage your up to now....? Have been playing around with a CMCD 75M Transmitter and seems to be more or less in the ball park.... If you cannot get Nigel's TX to work this might be worth a try.... Made it only using 2 FETs but 4 or more shouldn't be a problem.... The photos show the testing only at low power levels but run it up to 150Watts for appox 30 minutes into a dummy load .... the FETs temperature got to 38 degrees C ( not sure what that is in F ).... The efficiency is just over 90% post filter but probably with a little more jiggling could go a little higher but as its only an experiment will leave it at that... Might wind the Transformer using coax just to see if I can squeeze a little more though... The Class E looking waveforms are at the Drains of the FETs...... Wayne
|
|
|
Logged
|
|
|
|
N4LTA
Contributing
Member
Offline
Posts: 1070
|
|
« Reply #11 on: September 26, 2019, 10:17:11 AM » |
|
Looks good Wayne. I am interested in the design.
I am putting the PWM boards in a Rack Mount enclosure and have spent the last week waiting on a punch for the meters. I got it yesterday so I will spend some of the weekend metalworking.
Meters are always fun to mount! Getting the hole is one thing and matching up the 4 mounting studs is even more fun. I also have to punch or some other way cut the rectangular power cord entry on the back. File work most likely. Waiting for a illuminated toggle switch from Mouser also. Should be here today.
I backed down a bit on the power supply and am using a 600VA transformer rather than the HEAVY 1500 VA. It fits the enclosure better and should have no problem at 350-400 Watt output. It is good for 15-20% more per the specs. I doubt I ever push more than 400 Watts.
I finished one of the TI boards with the 06 type SMT resistor and caps. Went to the eye doctor yesterday to see if I needed new glasses. That is definately as small as I want to go. I had a special pair of reading glasses made. My eyes are so weird that I can use the drugstore ones.
I have a good friend who bought a 3CX1200 and socket as a initial special for hams by Eimac. He had it for years and I got it. So that may be another project. I am barefoot on HF. I have a nice 3600 volt power supply that I used with a 4-1000 160 meter amp and a 4CX1500 6 meter amp years ago.
I'll post some pictures later this weekend.
Pat
|
|
|
Logged
|
|
|
|
N4LTA
Contributing
Member
Offline
Posts: 1070
|
|
« Reply #12 on: September 26, 2019, 05:48:08 PM » |
|
Decided to go ahead and punch the three meter holes this afternoon. First my 3/4" punch got a slug of 1/8" aluminum stuck in it. The meter punches themselves require a 3/4" hole for the drive stud. Not getting it loose, I used a 7/8" punch to finish.
Them my really stupid mechanical stunt for the day. I had a 2 1/8 in punch that was just a bit two small. I bought a 2 1/4 greenlee and it arrived today.
After the 3/4 and the one 7/8" holes were done I popped in the big punch and quickly had three large meter holes. I grabbed a meter to test fit and Uh O I used the wrong punch!
Believe it or nor I was able to enlarge the hole with the new punch reasonable accurately - not as well as I had planned though. Meter placement is all that good.
|
|
|
Logged
|
|
|
|
Opcom
Patrick J. / KD5OEI
Contributing
Member
Offline
Posts: 8296
|
|
« Reply #13 on: September 26, 2019, 07:37:28 PM » |
|
If the hole that looks lower were elongated upward, would the bottom of the meter still cover the bottom of the hole? If so, can you beat flat the slug that came from that hole, enlargen the center hole in it, and put it in the panel hole? Then put the punch in, offset to it elongates the hole enough? Placing the panel on a 1" thick wood surface with a generous cutout for the punch and clamping it well. Then punch the elongation! Only will have one chance to do that. Otherwise a new panel. It's only 1/8".
|
|
|
Logged
|
Radio Candelstein
|
|
|
N4LTA
Contributing
Member
Offline
Posts: 1070
|
|
« Reply #14 on: September 26, 2019, 10:23:36 PM » |
|
I ended up using a file and got it OK looking. All that planning ruined by being stupid!
Not The first time!
Pat
|
|
|
Logged
|
|
|
|
N4LTA
Contributing
Member
Offline
Posts: 1070
|
|
« Reply #15 on: October 13, 2019, 03:45:34 PM » |
|
Getting some more metal work done. Have to add the keying relay. It is due this week. Need to test the voltage on the supply and at no load and design a series MOSFET regulator. I have the PC Board done. Also to wire the Cap board and the low voltage supply board. The bridge is mounted on a small heatsink. Probably didn't need it as the full supply current is about 5 amps.
|
|
|
Logged
|
|
|
|
N4LTA
Contributing
Member
Offline
Posts: 1070
|
|
« Reply #16 on: October 14, 2019, 05:39:52 PM » |
|
Got the enclosure pieces all back together and etched the voltage regulator board. Waiting for a 10 amp power switch and a couple of relays that should show up Wednesday.
Will build the voltage regulator board tonight.
|
|
|
Logged
|
|
|
|
vk3alk
Member
Offline
Posts: 263
|
|
« Reply #17 on: October 14, 2019, 10:21:23 PM » |
|
Hi Pat... Everything seems to be coming along well..... Is your TX going to be in another box ? Have you decided on what materials you will use in your PWM LPF yet....?. Have tried 3 different types and can detail those if you like.... Might be best to build the transmitter first and see what load it places on the modulator and then design the Filter for whatever it is like 6 ohms / 12 ohms etc: With the TX are you going to try Nigels ..... has he sent you any details at all as he seems to be quiet at the moment.... Hope all is OK with him....... Are currently building another 160M TX .... all the modules are built and tested and are building the case just like you so might upload some photos for interest sake.... Sounds like your regulator is just a regulator Can send you the circuit for the one I use which has variable output voltage and over current with SWR rollback which is a good safe guard really.... You could build that afterwards and install at a later date perhaps... Thats enough... Wayne
|
|
|
Logged
|
|
|
|
N4LTA
Contributing
Member
Offline
Posts: 1070
|
|
« Reply #18 on: October 15, 2019, 10:52:44 AM » |
|
Thanks Wayne.
Right now I am trying for a simple regulator like Nigel describe on one of his projects. I hope to regulate a few volts above the PS voltage at full load, so that the power disappation on the pass Mosfet is reasonable. Thinking something like 3 volts with 4 amps or 12 Watts. When the transmitter is not running, current will drop and voltage will soar but the current will be low. Regulation is not a good word for the circuit, it is really an anti voltage soaring circuit. That hopefully will get rid of the relays in older circuits. May not work and I may have to build a regulator with an error amp. The transmitter will go into another enclosure. I do have a board etched that I can do bench testing with.
The filters will probably be on larger toroid forms but I will have to see what the transmitter draws. I have wondered if the really nice looking air core speaker crossover coils would work well. Some are available with AWG 12 and 14 wire.
I would be very interested in your regulator and your experiences with PWM filter materials. Also in your 160 meter rig.
Pat
|
|
|
Logged
|
|
|
|
N4LTA
Contributing
Member
Offline
Posts: 1070
|
|
« Reply #19 on: October 19, 2019, 03:48:57 PM » |
|
Spent some time refining and slightly revising the power supply on the driver board. Made a new board for T0220 MOSFETs and built and tested it this morning.
Hope to install it later today or tomorrow and maybe let some smoke out.
Got the MOSFETS mounted on the heatsink and the board attached to the heatsink. the board will mount above the chassis and the first inductor will mount from under the board as will the main power supply input .47 mylar capacitor.
Got the HV power supply running and burning in. Line voltage here is a little high and I am a few volts higher than expected. This is with the bleeder resistor drawing 32mA.
I used a large 200 watt 50 ohm resistor to test the full load voltage. It dropped to 157 volt. This works out to about 492 watts. This is about 19 volts difference from no load to full load. Hopefully I can design a simple regulator to lower this difference substantially. I plane to do something similar to what Nigel has described. I appreciate all that Nigel had published that made this project possible.
One question that I am pondering is should the T/R relay switch the AC side of the power supply going to the bridge (easier to switch AC with a zero crossing with a relay) or to switch the DC input to the PWM board. Opinions would be appreciated.
|
|
|
Logged
|
|
|
|
vk3alk
Member
Offline
Posts: 263
|
|
« Reply #20 on: October 20, 2019, 05:32:03 PM » |
|
Hi Pat...
Noticed you were online so thought would reply quickly....
I leave the power supply permanently connect to the PWM FETs. The T/R switching is to enable / disable the IR2110 driver and have no problems doing this at all.... The pulse train is stopped ......
The TX pulse train is stopped by disabling power to the Flip Flop.....5 volts....
Will post details of the PWM LPF that have used soon....
Wayne
|
|
|
Logged
|
|
|
|
N4LTA
Contributing
Member
Offline
Posts: 1070
|
|
« Reply #21 on: October 20, 2019, 05:51:15 PM » |
|
Thank Wayne. That is what I will do.
Did more testing this afternoon. Now messing with a voltage regulator and letting some smoke out. A 175 volt supply with 18000 uF filter caps makes a nice pop when things go south.
The zener string feeding the gate of a power mosfet via a 47 ohm resistor has given me problems. The mosfet seems to saturate and I get the same output voltage at the source as I do at the drain input. Must be doing something wrong. I have attached an image of what I have tried. I have a 150 volt zener string that measures 150 volts as expected but the output is identical to the input at about 175 volts.
In searching the am technical forum archives I find this circuit for a screen regulator that may be helpful. With no load the power dissipated should be and as the supply droops under load, the power dissipated at full load with a few volts dropped on the MOSFET should also be doable with a decent heat sink.
|
|
|
Logged
|
|
|
|
vk3alk
Member
Offline
Posts: 263
|
|
« Reply #22 on: October 20, 2019, 11:01:02 PM » |
|
Hi Pat....
OK on your regulator..... There was a posting on this type of Regulator not so long ago on the Forum I think.....
Well to be honest ...... I would never use this circuit in this application. Its just my opinion though ... for whatever that's worth as Nigel uses something similar without any problems ......
What I would do .... is finish your TX without this regulator and assess its performance then.... At a later date you could use feed forward like what Steve uses on his WEB site....maybe..... You might find that it works well at the 300 Watt level but at 400 watts not so much.....for whatever reason. At the receiving end no one would know the difference between the power levels.... A nice sounding transmitter is the main thing running a higher power level could come later...
Wayne
|
|
|
Logged
|
|
|
|
vk3alk
Member
Offline
Posts: 263
|
|
« Reply #23 on: October 21, 2019, 04:20:50 AM » |
|
What a great response to your regulator post...... Might have some suitable zener diodes in my junk box......I hope..... If so will build and abuse it a bit....
|
|
|
Logged
|
|
|
|
N4LTA
Contributing
Member
Offline
Posts: 1070
|
|
« Reply #24 on: October 21, 2019, 06:29:31 PM » |
|
Wayne
I got some good advice but am still puzzled as to why my first try did not work at all. It was built like the schematic above missing some protective circuitry but still should have functioned.
The only real difference was I used a 10uF capacitor instead of the .01uF cap which is a big difference but i'm not sure it would cause the circuit to not work at all. I have etched a new circuit board with the added protection zener from gate to source and will test it as soon as I get it built but I am not confident. I am doing something wrong, but I can't find it. I am using the same Modfet that Nigel suggested and it should do the job rated at 60 amps and 250 volts. I am getting the same voltage at the source as at the drain and the diode string is 150 volts as it should be. I built two boards and tested the first one with clip leads and had the above result. A clip lead slipped and I had a short and the dropping resistor for the zener string blew up is flame and pretty much ruined the board - Using a multimeter - the Mosfet seemed OK, but I built another board using non electrolytic caps and a new FET with the same results.
Unless the .01uF 1 KV cap rather than the much larger cap changes things - then I am puzzled. I have been making the assumption that the FET pins out as a standard T0220 Mosfet - I better check that before I build the next version.
I am not worried about power dissipation since the supply droops substantially under load and the delta V of full current will be 5-6 volts at 3 amps or so.
Any suggestions or assistance is greatly appreciated.
Pat N4LTA
|
|
|
Logged
|
|
|
|
|