Page 1 of 3 123 LastLast
Results 1 to 10 of 25

Thread: Tube amp SMPS

  1. #1

    Tube amp SMPS

    Hello all, first post.


    I'm not an engineer, but trained in electronics 16 years in the field. Built many tube amps before, but no switching supplies. I'm trying to put a tube amp in my car, so I need a switching supply. Most of the circuit is already built with remote power relay/driver transistor, filaments, and the PWM chip all wired up and producing sawtooth waveforms at the Ct/Rt pin. All that's left is the transformer, here's the simulation I hope I did it correctly.


    calc1.JPG

    Note that I'm not using push pull topology, I'll be using a pair of IRFZ46N in parallel to switch the primary. Not sure what that's called, in tube land single ended. lol I know the primary wire size isn't right, I'll probably have to use 3/4 pieces of .012" wire twisted together to handle the amps.

    Anyhow the turns ratios look about right for the B+ I'm after, I'm just not sure about RDSon value. Planning to pull around 10A peak from the primary, I need around 250ma on the secondary to run a stereo push pull vacuum tube output stage and a few spare milliamps for the voltage amp/phase splitter stages.

    Thanks in advance.

  2. #2
    .... Silvio's Avatar
    Join Date
    Apr 2015
    Location
    Malta
    Posts
    659
    Blog Entries
    9
    Quote Originally Posted by miniman82 View Post
    Hello all, first post.


    I'm not an engineer, but trained in electronics 16 years in the field. Built many tube amps before, but no switching supplies. I'm trying to put a tube amp in my car, so I need a switching supply. Most of the circuit is already built with remote power relay/driver transistor, filaments, and the PWM chip all wired up and producing sawtooth waveforms at the Ct/Rt pin. All that's left is the transformer, here's the simulation I hope I did it correctly.


    calc1.JPG

    Note that I'm not using push pull topology, I'll be using a pair of IRFZ46N in parallel to switch the primary. Not sure what that's called, in tube land single ended. lol I know the primary wire size isn't right, I'll probably have to use 3/4 pieces of .012" wire twisted together to handle the amps.

    Anyhow the turns ratios look about right for the B+ I'm after, I'm just not sure about RDSon value. Planning to pull around 10A peak from the primary, I need around 250ma on the secondary to run a stereo push pull vacuum tube output stage and a few spare milliamps for the voltage amp/phase splitter stages.

    Thanks in advance.
    Hi I have seen your simulation on Exellent IT and there are a few corrections that you have to make.
    1) RDS on is taken from the datasheet of the switching mosfets and for IRFZ46N is 16.5 milli ohms (0.0165 ohms)
    2) Using push pull topology is better for low voltages thus you need 2 primary windings 12v-0-12v
    3) if you are using a TL494 I am not sure if it can drive your fets directly
    4) if stabilization is not so important you can use IR2153 chip which is easier to use and can drive your fets directly
    5) your minimal output current is too low and as you can see from below the output choke is rather large
    6) remember to use fast switching diodes for the output such as UF4007 or similar
    7) If you can post a schematic of your smps so that we can see what are your intentions.
    8) Your oscillator frequency is double the switching frequency so TL494 must oscillate at 94Khz for 47Khz switching frequency

    regards

    Silvio

  3. #3
    Thanks Silvio,


    I'm using UC3844 for the PWM chip, not optimal but I'm cheap and using what's on hand. Had to build a 555 based charge pump to get it above 15v so it would start running, but it seems to stay running till 10v so that must be under voltage dropout. Still lots of fiddling to do, but here's the updated EIT screenshot.

    calc1.JPG

    I seem to be having some trouble finding an ultra fast bridge rectifier for this, any suggestions? Guess I can always make one out of diodes, but would be nice to have that in one package to cut down on board space.

    Also probably going to wind my own choke for it.
    Last edited by miniman82; 01-05-2017 at 09:21 PM.

  4. #4
    .... Silvio's Avatar
    Join Date
    Apr 2015
    Location
    Malta
    Posts
    659
    Blog Entries
    9
    I seem to be having some trouble finding an ultra fast bridge rectifier for this, any suggestions?

    I don't really know of ultra fast bridge packages. I guess 4 1n4148 dont take much space

    Parameters in chart seem ok now


    Good luck and regards

    Silvio

  5. #5
    Update: have the SMPS running now, but after a few moments the 400v rectifier shorts. I chose RFU5TF6S for both the 250v winding and 400v winding rectifiers, thinking the 600v rating would be enough but apparently not? Any special reason a 600v rectifier is not high enough for the application, or am I making a mistake testing it with no load?

  6. #6
    .... Silvio's Avatar
    Join Date
    Apr 2015
    Location
    Malta
    Posts
    659
    Blog Entries
    9
    Show your schematic and some wave forms so we will try to understand what is causing the diodes to blow. It could be you need to snub them and apart from that if there are any HV spike somewhere it will surely exceed the voltage limit of your diodes. Also there should be some kind of small load at least to settle out any stray voltages on the output

  7. #7
    Apologies, but I have no software for that. I could draw it out, but I have very little time having a family and all. It's simple enough to visualize though, I built it from scratch using info from the PWM chip data sheet and ideas from my head. lol

    UC3844 driving a pair of IRFZ46N in parallel. Transformer as in the model above, frequency is around 50khz. Gates have 22ohm resistors with ferrite beads, mosfet current sense resistor value is .5 ohm, voltage feedback taken from a resistive divider off the 250v line (for now), etc.

    I actually may have hit on something today. It was very unstable for some reason, drawing more current than it should have (nearly 2A) and the current sense resistor was getting very warm. Well I had a .1uf cap on the current sense line to ground (discovered it wouldn't even run without it), but it turns out to be much more stable with only a .0047uf there instead. It settled down immediately with the .0047uf, now only drawing .75A from the 13v power supply and I can actually get some shots of various waveforms. These pictures are with the 400v section disconnected since it was killing rectifiers, and a 30ma load on the 250v section.

    Ignore the busy perfboard, I realize it's not what an SMPS should be built on but this is only the first try at this.

    (Apologies also for the rotated shots, this is a phone camera)

    IMG_0964.jpg

    This is the gate waveform, looks like some ringing there. Probably excessive capacitance on the gate circuit, the leads between the UC3844 and MOSFETS is longer than it ought to be. Scope set on .5/div with 10x probe, timebase set to 5us.

    IMG_0965.jpg

    This is what I see on the drain pin of the MOSFETS, not really sure what it should look like yet as there's no snubbing or anything else attached yet.

    IMG_0966.jpg

    As you can see though, it's happily putting out a very stable 250vdc. Secondary has no noise to speak of, so the CLC filters seem to be doing their jobs.

    IMG_0967.jpg


    Thank you very much for the help so far, this is a fun project and I haven't destroyed anything yet so I must be doing something right!

  8. #8
    .... Silvio's Avatar
    Join Date
    Apr 2015
    Location
    Malta
    Posts
    659
    Blog Entries
    9
    Miniman82, I can see a lot of ringing in your waves. Try snubbing the primary windings. Find a small cap (ceramic or film EX 1-10nf that can withstand the voltage peaks in your case 100v I think is enough) that will produce 3 times the ringing frequency when you fix it across drain and source of each fet. Your series resistor will be calculated as (R=1 divided by 2pie X C X F) where C is the capacitor placed in farads and F is the new ringing frequency in Hertz. Try out the calculated value of the resistor to put in series and find the nearest value available. I would suggest that you put 2 of these one on each mosfet between drain and source. Check out the result and see the dissipation of the series resistors with the caps and also the load they are imposing on your circuit. However many combinations exist and you can try several of these. Trial and error will get your best result.

    Note put a current limiter in your supply rail just in case there is some form of disturbance to the switching frequency. If all seems ok then you can proceed.

    Hope that helps

    Silvio

  9. #9
    Silvio,

    Yes there is some ringing present, I assumed as much since I hadn't gotten to the point of adding snubbers yet. Thank you for the suggestions, I will attempt to damp the oscillations as you describe. If that looks promising I will reconnect the 400v rectifier, since any damaging pulses should be cleared up.

  10. #10
    .... Silvio's Avatar
    Join Date
    Apr 2015
    Location
    Malta
    Posts
    659
    Blog Entries
    9
    @ miniman82 One other thing I forgot to mention. Try to filter out your pulses going to the gates otherwise these will be amplified in the switching transistors. Shorting the path as much as possible from IC to gates. Try to add some ferrite beads to your wires going to the gates, a small cap between gate and source in the pf range say 10pf or so. just try and see what happens, I am trying to figure out where that ringing is coming from. Also decouple your supply to the pwm chip as close as possible to it. try soldering a 100nf ceramic cap directly to its feet. Those long wires crisscrossing are all antennas

Similar Threads

  1. Vacuum Tube SMPS
    By printer2 in forum diysmps
    Replies: 21
    Last Post: 05-13-2016, 03:29 AM
  2. Tube Amplifier SMPS with Heater and Bias Supply
    By ballpencil in forum diysmps
    Replies: 3
    Last Post: 09-29-2014, 03:32 PM
  3. Tube SWPS
    By sdman in forum Switching Power Supplies Informations
    Replies: 0
    Last Post: 03-24-2014, 05:42 PM
  4. Power Supply for X-Ray Tube
    By S0syk in forum diysmps
    Replies: 0
    Last Post: 07-30-2013, 12:45 PM
  5. My Tube amp by L6L
    By cakum in forum Tubes & Valves
    Replies: 5
    Last Post: 04-06-2013, 11:55 AM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •