The AM Forum
December 13, 2024, 07:09:59 PM *
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   Go Down
  Print  
Author Topic: Hopefully back at it  (Read 25117 times)
0 Members and 1 Guest are viewing this topic.
N4LTA
Contributing
Member
*
Offline Offline

Posts: 1070


« on: November 02, 2020, 02:30:37 PM »

Am a few weeks away from having my shop back but I laid out a PC board today for Rodney's rig (also the one Wayne built). It is close to what I was working on when the fire happened.
Since Wayne had such good luck with it, I thought it might be a good start.

It should be here by the weekend and I can start building it. Maybe by the time it is complete , I will have some shop space available.

I have all the parts ready to go.


* board for classd d-e.JPG (66.34 KB, 566x723 - viewed 670 times.)
Logged
vk3alk
Member

Offline Offline

Posts: 264


« Reply #1 on: November 02, 2020, 08:11:17 PM »

Very good Pat.....

Should make things easier for others to build.....

I did change the turns ratio of the output transformers to enable a 7R load......
With the turns specified the load was 3R a little too low for me...

Just make sure your phasing is correct as well ..... I did have them out of phase and no output  Shocked

I still have the module and can connect it back up again quite easily if needed.....

Have taken other photos like the one uploaded for reference etc:

Also too I made up Steves Duty Cycle circuit .... the one on his WEB site to adjust things.....


Wayne



* CREE Class DE 4 FET Gate.jpg (113.62 KB, 800x480 - viewed 663 times.)
Logged
N4LTA
Contributing
Member
*
Offline Offline

Posts: 1070


« Reply #2 on: November 03, 2020, 09:58:30 AM »

Thanks Wayne,

I need to build Steve's circuit myself. Seem like I started a board a while back. I need to look for it. I'd also be interested in your discrete PWM generator.

Thanks

Pat
Logged
KQ6F
Member

Offline Offline

Posts: 124


« Reply #3 on: November 03, 2020, 10:08:46 AM »

Pat -

When making the output transformers it's important to wind the two primaries on OPPOSITE sides of the cores.  This eliminates most of the unwanted capacitance between them that otherwise would distort the half-sinusoidal waveforms at the FET drains. The secondary winding should be evenly distributed around the cores.

Rod KQ6F
Logged
N4LTA
Contributing
Member
*
Offline Offline

Posts: 1070


« Reply #4 on: November 03, 2020, 12:08:59 PM »

Thanks Rod

I had wondered about the secondary windings. Look forward to getting this built. Any info you can add would be  appreciated.

73

Pat
N4LTA
Logged
KQ6F
Member

Offline Offline

Posts: 124


« Reply #5 on: November 03, 2020, 03:05:11 PM »

Pat -

As Wayne suggested you might consider changing the transformer ratios to increase the load resistance.  It depends upon your power supply capability.  In general increasing the load resistance is a good thing since it means the modulator doesn't have to work so hard.  In my case the turns ratio was dictated by the limitations of my power supply.  If I had a more robust supply with higher output voltage I would increase the load R.  Of course there is a limit as to how high you can make the supply voltage, keeping in mind that if the flyback caps on the FET drains are chosen correctly you will get half-sinusoidal peak waveforms on the drains of 4 times the supply voltage.  In my case I used the CREE 1200 volt SiC devices which gives me tons of margin but if I were building the rig today I'd choose the 900 volt devices which are about 40% cheaper.

The CREE SiC FETs have very fast switching characteristics and I had no need to tailor the phase input waveforms to 40% on/60% off as others have done.  I just run a flip-flop at twice the carrier freq into the FET drivers directly through gates that turn off BOTH phases at Tx shutdown.  It's VERY important to shutdown both phases since going directly from the flip-flop outputs would leave one phase high at shutdown which unless your supply has current limiting would cause the drain currents in that phase to rise precipitously and blow the FETs.

One more thing.  If you change the transformer ratios you will likely need to change the values of the flyback caps to get the 4x peaks.  I am attaching a scope picture of the two phases in my rig.


* classd_e_002.jpg (206.04 KB, 1000x750 - viewed 636 times.)
Logged
vk3alk
Member

Offline Offline

Posts: 264


« Reply #6 on: November 03, 2020, 05:05:50 PM »

Good to hear from you Rod and all comments noted.....

Yes Pat would wind the output transformers as Rod suggested as those waveforms look extremely good.....
I wound mine interleaved and although look fine are not as good as Rod's....

Have no doubt will connect my module back up again in time so will recheck the drain waveforms then....



Wayne
Logged
N4LTA
Contributing
Member
*
Offline Offline

Posts: 1070


« Reply #7 on: November 04, 2020, 12:51:14 PM »

My bench supply is a heavy duty linear lab supply rated at 0-120 volts at 3 amperes. Any ballpark ideas on how the flyback capacitors may need to change as the resistance rises? I am using the 1200 volt Cree FETS so I can run at a higher voltage and have an ample power supply. I have S Mica caps and some Johanson MLCC rated at 1000pF at 1 KV that are surface mount types.

I designed a PC board for the variable duty cycle for general use rather than this rig. I should get some prototype boards in a week or so.


* steves driver and duty cycle circuit.JPG (107.38 KB, 732x503 - viewed 674 times.)
Logged
KQ6F
Member

Offline Offline

Posts: 124


« Reply #8 on: November 04, 2020, 02:08:01 PM »

As the resistance goes up the flyback cap values go down.  I ran a SPICE model that shows that increasing the R to 7.7 ohms (by changing the transformer turns ratios) requires 510pf caps in order to maintain the same 4x peak half-sinusoids.  I have not verified this in actual practice.  Wayne may have further info.

Rod
Logged
N4LTA
Contributing
Member
*
Offline Offline

Posts: 1070


« Reply #9 on: November 05, 2020, 03:20:28 PM »

Thanks. I got the board in today and need to do some cuttout work for the fets.  I hope to get the FETs in from DigiKey tomorrow and will start putting things together.

Don't have my test bench up yet so there is no rush. 

Pat
N4LTA
Logged
N4LTA
Contributing
Member
*
Offline Offline

Posts: 1070


« Reply #10 on: November 06, 2020, 05:54:20 PM »

Another simple question. Do the gate drivers need to be heatsinked to the large heatsink or can they get by being mounted upright with bolt on heatsinks. I can go either way, but I'd rather not cut an opening in the circuit board to lay it down to mount to the main heat sink. I have some large T0220 heat sinks.

Pat
N4LTA
Logged
KQ6F
Member

Offline Offline

Posts: 124


« Reply #11 on: November 06, 2020, 06:01:17 PM »

Pat -

I have my drivers mounted vertically on TO-220 heat sinks and they run nice and cool.

Rod
Logged
N4LTA
Contributing
Member
*
Offline Offline

Posts: 1070


« Reply #12 on: November 06, 2020, 08:38:23 PM »

Thanks Rod. That's how I will build mine.

73
Pat
Logged
vk3alk
Member

Offline Offline

Posts: 264


« Reply #13 on: November 06, 2020, 09:50:37 PM »

Actually what you can do with the drivers is when testing use 12 volts as the supply level.....
After completion and are happy with efficiency etc: etc: I then reduce the voltage to the IXDDs and when the output power starts to drop a little then nip it back up a bit....
You might find that the voltage is only say 10 volts.... and the current is lower as well reducing the heat quite a bit....

Hopefully you get what I mean......  Grin

PS:  I do this more on 40M .....
With the H Bridge modules since there is only 1 driver .... with the IXDDs at 12 volts they can get hot .....
So after experimenting can reduce the level down to 8 volts dropping the heat noticeably ....



Wayne
Logged
N4LTA
Contributing
Member
*
Offline Offline

Posts: 1070


« Reply #14 on: November 07, 2020, 09:39:31 AM »

Thanks Wayne

I'll do that when I get around to testing. On 75 meters for the fist one, I doubt the driv ers will have to work hard with the SiC FETs.

I have to cut the holes for the lay down mounted fets this weekend. With no shop, I am looking at a hand drill, a hand nibbler and a file - Ugg!

My fingers already are itching

Pat
Logged
N4LTA
Contributing
Member
*
Offline Offline

Posts: 1070


« Reply #15 on: November 07, 2020, 12:50:42 PM »

Was easier than I thought. Still itching between my fingers. The heat sink is not here. I need it to mount the board and the FETs. It is due Monday. I may have one somewhere in boxes.



* IMG_5618.jpg (66.77 KB, 640x480 - viewed 642 times.)
Logged
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2692


Making AM GREAT Again!


« Reply #16 on: November 07, 2020, 08:25:10 PM »

Wow.  Looking good, Pat.

On another note, the Icom RX is dojng great.  Smells like a forest fire, and the DC input jack failed after about a week (solder joint)....  Put some heat to it and touched it up...  Still going great!

Glad to hear your back building.  Always liked reading about your projects!

--Shane
KD6VXI
Logged
N4LTA
Contributing
Member
*
Offline Offline

Posts: 1070


« Reply #17 on: November 09, 2020, 02:34:36 PM »

Did some PC Board on the Modulator. Put Nigel's entire modulator minus the output filter on a single board. Checked it a few times and then sent for two prototype boards.

Using a couple of T0 220 FETs that Nigel recommended over the older FETs he first used. I tested them previously before the fire. Should make a nice compact high powered modulator.

Pat
N4LTA


* nigels complete.JPG (105.98 KB, 751x511 - viewed 653 times.)
Logged
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2692


Making AM GREAT Again!


« Reply #18 on: November 09, 2020, 02:58:28 PM »

Did some PC Board on the Modulator. Put Nigel's entire modulator minus the output filter on a single board. Checked it a few times and then sent for two prototype boards.

Using a couple of T0 220 FETs that Nigel recommended over the older FETs he first used. I tested them previously before the fire. Should make a nice compact high powered modulator.

Pat
N4LTA

I've had good luck with the IRF640 as the switches.  Used them a few times.

--Shane
KD6VXI
Logged
N4LTA
Contributing
Member
*
Offline Offline

Posts: 1070


« Reply #19 on: November 15, 2020, 09:34:10 PM »

Wayne

When you changed the transformer ratio to move to a higher modulating resistance, do you remember what your turns ratio was? I assume you changed it to to double to Rods resistance and am guessing x 1.414 or close.

I'd like to get the supply voltage up somewhere near 60-70 volts

Thanks

Pat
N4LTA
Logged
vk3alk
Member

Offline Offline

Posts: 264


« Reply #20 on: November 16, 2020, 04:57:52 AM »

Hi Pat...

Sounds like your really moving ahead with the TX....

Have attached a photo of the cores that are wound interleaved ....
I added an extra 2 turns either side of the center tap so 14 turns in all.....
If you wanted too ....  maybe wind an extra 2 turns overall making a total of 16T....
Test at low voltage just to check what load you are getting and reduce until you get what you want...that's what I did....

With 50 volts carrier voltage with a 7R load thats quite a bit of power .....
60-70 volts would probably test things a bit  Huh

You have to remember as well the voltage of the FETs used in your PWM...
If your using the IRFB4227 there only rated at 200 volts.....
Using a Mark to Space ratio of 60:40 to obtain a carrier voltage of 60-70 voltages would put them very close to their max....
Probably a bit too close  Grin

I have done this before and is quite exciting .....
Operating near their voltage limit they can seem Ok for a while and then unexpectedly .... pufffff


Wayne


* P1030384.JPG (471.56 KB, 1600x1200 - viewed 670 times.)
Logged
vk3alk
Member

Offline Offline

Posts: 264


« Reply #21 on: November 16, 2020, 05:09:39 AM »

Sorry my eyes saw double and got confused....

I misquoted the turns.....

The input is the same that is 6T either side of the center tap....
The output turns are reduced from 11 to 7T...

Sorry about that....


Wayne


Logged
KQ6F
Member

Offline Offline

Posts: 124


« Reply #22 on: November 16, 2020, 10:47:51 AM »

Pat -

The 60-70 volt voltage you mentioned, is this TOTAL supply voltage or voltage at carrier?

If it's TOTAL then you will be the same as what I have and the transformer turns are optimized for maximum power output (and about 3.7 ohms modulation load resistance).

Rod
Logged
N4LTA
Contributing
Member
*
Offline Offline

Posts: 1070


« Reply #23 on: November 16, 2020, 12:33:52 PM »

Rod

it would be the supply voltage, not the modulated voltage.

Thanks

Pat
Logged
N4LTA
Contributing
Member
*
Offline Offline

Posts: 1070


« Reply #24 on: November 16, 2020, 12:54:23 PM »

Thanks Wayne

I was using Nigels PWM circuit and he was running it at over 100 volts, so I didn't pay attention.

Hoping to get good power but at a lower current is the reason for wanting a higher resistance. Have a nice linear power supply 0-120 volts at 3.5 amps max.

Thanks for the info and drawings . I hope to get some testing done in the next week but we hope to move into some more areas in the house this week if things go well.
That may well slow me down. Boy do I miss my small drill press.

Pat
Logged
Pages: [1] 2   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 0.178 seconds with 18 queries.