Hello again Bert, I have in my possession an old "a/f stage amp" that tops out @ 780 watts per channel ( both channeis driven!). The outputs are Mosfet (2sk176 + 2sj56) and are obviously in push pull. I've been told these mosfets will happily go all the way up to 10 Meg. (it also can be configured as a 1600 watt mono block!)
My questions are, can I use it to power up a Tesla Coil by giving it Positive feedback to make it oscilate, (It would be easy to tune to the dominant Ferq. of the secondary that way.)or does the feedback coil do that for me?
Can I split the value of the feedback "tank capacitor" and use them to tie the f/b coil to the gates of the Mosfets or do I tie the f/b coil straight to the amp input? I also imagine that the Feedback coil, would have to be centre tapped to tie in with Class B operation.
Can I split the value of the feedback "tank capacitor" and use them to tie the f/b coil to the gates of the Mosfets or do I tie the f/b coil straight to the amp input?
Also, would I need to use protection Diodes across all of the individual mosfets or just across the P channel and N channel? (4+ and 4- mosfets per ch. ie. push pull parallel )
(So many questions, so little time!)lol!
Cheers again, Mick. ;-))
Re: yet another question
Posted by Bert Hickman on 8/22/2008, 6:17 am, in reply to "yet another question"
--Previous Message-- : Hello again Bert, : I have in my possession an old "a/f stage amp" : that tops out @ 780 watts per channel ( both channeis driven!). : The outputs are Mosfet (2sk176 + 2sj56) and are obviously in push pull. : I've been told these mosfets will happily go all the way up to 10 Meg. : (it also can be configured as a 1600 watt mono block!) : : My questions are, : can I use it to power up a Tesla Coil by giving it Positive : feedback to make it oscilate, (It would be easy to tune to the dominant : Ferq. of the secondary that way.)or does the feedback coil do that for me? : : Can I split the value of the feedback "tank capacitor" and use : them to tie the f/b coil to the gates of the Mosfets or do I tie the f/b : coil straight to the amp input? : I also imagine that the Feedback coil, would have to be centre tapped to : tie in with Class B operation. : : Can I split the value of the feedback "tank capacitor" and use : them to tie the f/b coil to the gates of the Mosfets or do I tie the f/b : coil straight to the amp input? : : Also, would I need to use protection Diodes across all of the individual : mosfets or just across the P channel and N channel? (4+ and 4- mosfets per : ch. ie. push pull parallel ) : : (So many questions, so little time!)lol! : : Cheers again, : Mick. ;-)) :
Hi Mick,
Those are some excellent questions. I don't know the answers, but here are some thought and suggestions. Even though your amplifiers are designed to operate in linear mode, I'd recommend using soft switching and driving them hard into saturation (rail to rail) to reduce output switching losses and drive the system in bursts to get longest sparks. This also permits you to use digital control for switching, burst length, etc. - otherwise you'll have a CW system that make hot arc-like flaming corona but short sparks.
Free-wheeling diodes are also essential across each group of MOSFETS. You will likely need to substantially beef up your power supply capacitors to handle high current high frequency energy that will be circulating between the amplifier and load. This means using both large DC electrolytics AND a polypropylene "DC Link" capacitor in parallel.
There are a couple of ways to drive your system. One easy way is to simply connect the output of your amp directly into the base of your secondary, either with a direct connection or through a fairly large valued polypropylene coupling capacitor. This method is simplest, since it requires no tank circuit and feedback control/switching is simpler.
You can also drive a series combination of primary inductor and tank cap off one audio channel, with the other end of the tank circuit grounded, forming a "half bridge" circuit. You might even be able to parallel amplifier outputs to double output current capability as long as you are switching them softly and from rail/rail.
Another other option is to drive the inputs from opposite phase (but otherwise in sync) to form a "full bridge" output to double toe voltage swing "seen" by the primary tank circuit. In any event, you will need to look closely at the circuitry for the amplifiers to make sure their transient responses are identical and there are no surprises (like significant internal phase shifts or excessive time delay from input to output) that will cause you grief when when you try to drive these at the operating frequency of your system.
There are a variety of ways to get feedback, ranging from a "tickler" winding, E-field sensing from the secondary, primary current sensing, and/or secondary base current sensing. For a variety of reasons, I would recommend using current sensing via one or more current transformers (CT. If you direct drive your secondary, use base current sensing (via a simple current transformer). If you are driving a tank circuit, use a CT to detect primary current. In either case, you want to drive switch the output of your amp(s) at (or very near) the current zero for ZCS mode of operation.
Finally, you'll want to control burst width and burst-burst timing for various effects, and to prevent current into your primary tank circuit from climbing to excessive values. This is where digital control becomes essential, especially when using half or full drive to your tank circuit. You may need to add a current monitoring circuit that will automatically shut down the input drive signal once a predetermned amount of peak primary or secondary current has been reached. Some good ideas for current monitoring, burst control, etc. are on Steve Ward's site. BTW, IGBT's are actually better suited for this mode of operation than MOSFETS.
--Previous Message-- : --Previous Message-- : Hello again Bert, : I have in my possession an old "a/f stage amp" : that tops out @ 780 watts per channel ( both channeis driven!). : The outputs are Mosfet (2sk176 + 2sj56) and are obviously in push pull. : I've been told these mosfets will happily go all the way up to 10 Meg. : (it also can be configured as a 1600 watt mono block!) : : My questions are, : can I use it to power up a Tesla Coil by giving it Positive : feedback to make it oscilate, (It would be easy to tune to the dominant : Ferq. of the secondary that way.)or does the feedback coil do that for me? : : Can I split the value of the feedback "tank capacitor" and use : them to tie the f/b coil to the gates of the Mosfets or do I tie the f/b : coil straight to the amp input? : I also imagine that the Feedback coil, would have to be centre tapped to : tie in with Class B operation. : : Can I split the value of the feedback "tank capacitor" and use : them to tie the f/b coil to the gates of the Mosfets or do I tie the f/b : coil straight to the amp input? : : Also, would I need to use protection Diodes across all of the individual : mosfets or just across the P channel and N channel? (4+ and 4- mosfets per : ch. ie. push pull parallel ) : : (So many questions, so little time!)lol! : : Cheers again, : Mick. ;-)) : : : Hi Mick, : : Those are some excellent questions. I don't know the answers, but here are : some thought and suggestions. Even though your amplifiers are designed to : operate in linear mode, I'd recommend using soft switching and driving : them hard into saturation (rail to rail) to reduce output switching losses : and drive the system in bursts to get longest sparks. This also permits : you to use digital control for switching, burst length, etc. - otherwise : you'll have a CW system that make hot arc-like flaming corona but short : sparks. : : Free-wheeling diodes are also essential across each group of MOSFETS. You : will likely need to substantially beef up your power supply capacitors to : handle high current high frequency energy that will be circulating between : the amplifier and load. This means using both large DC electrolytics AND a : polypropylene "DC Link" capacitor in parallel. : : There are a couple of ways to drive your system. One easy way is to simply : connect the output of your amp directly into the base of your secondary, : either with a direct connection or through a fairly large valued : polypropylene coupling capacitor. This method is simplest, since it : requires no tank circuit and feedback control/switching is simpler. : : You can also drive a series combination of primary inductor and tank cap : off one audio channel, with the other end of the tank circuit grounded, : forming a "half bridge" circuit. You might even be able to : parallel amplifier outputs to double output current capability as long as : you are switching them softly and from rail/rail. : : Another other option is to drive the inputs from opposite phase (but : otherwise in sync) to form a "full bridge" output to double toe : voltage swing "seen" by the primary tank circuit. In any event, : you will need to look closely at the circuitry for the amplifiers to make : sure their transient responses are identical and there are no surprises : (like significant internal phase shifts or excessive time delay from input : to output) that will cause you grief when when you try to drive these at : the operating frequency of your system. : : There are a variety of ways to get feedback, ranging from a : "tickler" winding, E-field sensing from the secondary, primary : current sensing, and/or secondary base current sensing. For a variety of : reasons, I would recommend using current sensing via one or more current : transformers (CT. If you direct drive your secondary, use base current : sensing (via a simple current transformer). If you are driving a tank : circuit, use a CT to detect primary current. In either case, you want to : drive switch the output of your amp(s) at (or very near) the current zero : for ZCS mode of operation. : : Finally, you'll want to control burst width and burst-burst timing for : various effects, and to prevent current into your primary tank circuit : from climbing to excessive values. This is where digital control becomes : essential, especially when using half or full drive to your tank circuit. : You may need to add a current monitoring circuit that will automatically : shut down the input drive signal once a predetermned amount of peak : primary or secondary current has been reached. Some good ideas for current : monitoring, burst control, etc. are on Steve Ward's site. BTW, IGBT's are : actually better suited for this mode of operation than MOSFETS. : : http://www.stevehv.4hv.org/ : : Finally, have a supply of replacement output transistors available - : chances are quite good you'll need them... :^) : : Good luck, : : Bert : : Hmmmmm, It looks like I've bitten off more than I can chew, I'm just a hobbyist who knows enough about electronics to be Dangerous.
I actually built the amp about15 years ago when I swindeled a job as a electronics tech in a small Mom + POP opreation servicing the Pub and Club industry here in Melbourne The boss traded an old 250 watt unit with 1 blown channel and gave it too me, and said if I can fix it I can have it! He also said I could use the workshop and anything in it to fix up the old clunker. Taking that as cart blanche I gutted it and put everything from the current series of amps inside the old case! I don't think that was what he had in mind but I cant look a gift horse in the mouth! ( and the pay was lousy so I considered it a bonus!) The upshot of all that is, I never took the schematics when I left so I would have to rely om 15 year old memories, as well as all that I have had a Stroke about 10 years ago so I've got fat chance of rembering anyway!
Maybe I'll put it on the back burner for a bit while I learn more about our passion!
Cheers Mick. ;-))
P.S. I have started a blog about my T/C adventures! here is a link if you are interested.