linear amp for HV (TC) / Electromagnetic Radiation / Forums

Forums

4hv.org :: Forums :: Electromagnetic Radiation

« Previous topic | Next topic »

linear amp for HV (TC)

previous 1 2 3 next

Move Thread

LAN_403

aquilarubra

Sat Mar 20 2010, 07:06AM

Registered Member #2471 Joined: Wed Nov 18 2009, 07:23AM
Location:
Posts: 17

Thanks everybody for your ideas. Things start to become focused in my mind. A special thanks to Mattski, because I could move my search in another direction also, and I found some inexpensive pallet amplifiers that are almost fit for my purpose (example:

That is class AB though (I would rather use something in class A), and I still have to increase voltage. So, I expect a 50% loss if I use a transformer.

I know now that a simple, maybe slightly extended AM long to medium wave amplifier is simply what I need. So, I'm giving here the updated specifics:

Input: 110/220V
Input signal: from a generic function generator
Output signal: 100kHz-4MHz (or 2Mhz would be fine also)
Output power: the plasma tube really needs max 100W, but it should be high voltage; considered transformer losses, etc., I would say 200W; 300W optimum if connecting more tubes.

The increment of the output in stages is because I was thinking how it worked with valves. Since I can increase/decrease the output of the signal generator, I can have a higher or lower gain in the amp. So it is not a requirement any more.

Z should be a little bit variable, because it changes with the tube. I'm not sure what E and I are. Is E watts and I amperage? Probably a fixed amperage is the choice.

Steve Conner

Sat Mar 20 2010, 12:41PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

Someone mentioned a distributed amplifier. As far as I know that's what's inside those Amp Research things.

Someone on this forum posted a picture of the guts of one that he salvaged. It wasn't Amp Research, but it was a similar kind of thing. The inside was a matrix of about a dozen 4CX250B tubes. Very impressive. I can't remember who it is, but maybe he would sell it to you.

radiotech

Sat Mar 20 2010, 06:42PM

Registered Member #2463 Joined: Wed Nov 11 2009, 03:49AM
Location:
Posts: 1546

Here is a scan of the schematic for a HP 460 AR distributed amp
that is flat from 20 kHz to 100 mHz with -3db point at 200 mHz.

It is ultra-simple (however small) made of 10 pentodes and a transmission line. The Z is 200 ohms, howewver you probably could make a more beefy one out of TV line output tubes if you had the patience to wind the coils and come up with the layout. (edit)

Notice the anodes are operated below the screens ! ! !

The thing I remembered about his thing, is is worked with a dead tube or two.

Scott Fusare

Sat Mar 20 2010, 10:05PM

Registered Member #531 Joined: Sat Feb 17 2007, 10:51AM
Location: Burlington, Vermont
Posts: 125

Steve McConner wrote ...

Someone mentioned a distributed amplifier. As far as I know that's what's inside those Amp Research things.

Someone on this forum posted a picture of the guts of one that he salvaged. It wasn't Amp Research, but it was a similar kind of thing. The inside was a matrix of about a dozen 4CX250B tubes. Very impressive. I can't remember who it is, but maybe he would sell it to you.

'twas me. The rack was from a Maxwell Labs pulse amp from my employer's EMI test lab. It contained 24 4CX250Bs and was one of THREE racks contained in the amplifier. I have stripped the valves, sockets and chimneys from the rack so it no longer exists in it's original form. The intent is to offer them up for sale to the group if I ever get off my butt and build a test circuit to validate them.

Steve's stunning new Q.C.W. coil ought to spur some sales (tongue planted firmly in cheek).

Scott

aquilarubra

Thu Mar 25 2010, 09:25PM

Registered Member #2471 Joined: Wed Nov 18 2009, 07:23AM
Location:
Posts: 17

Thanks a lot to Radiotech and everybody. You have opened up many ways for me to investigate.

At the end I have found this amp:

It is practically all I need, except that I need to modify it to work from 100kHz or less. So, just extending the band downwards. I could even sacrifice part of the upper band, since I really need to work up to 4MHz.
Do you think that this design could be easily modified? Could somebody help me?

Thanks!

aquilarubra

Thu Apr 01 2010, 06:30PM

Registered Member #2471 Joined: Wed Nov 18 2009, 07:23AM
Location:
Posts: 17

I posted a link to the project here:
http://www.providenceuniversity.net/amp

Please, help! I'm still looking for a designer...

radiotech

Thu Apr 01 2010, 06:47PM

Registered Member #2463 Joined: Wed Nov 11 2009, 03:49AM
Location:
Posts: 1546

Here is a book to add to the resource list. Sevick, J. TRANSMISSION LINE TRANSFORMERS, 4th Ed. Scitech Publishing, Inc. Raleigh (2006)

Mattski

Thu Apr 01 2010, 07:00PM

Registered Member #1792 Joined: Fri Oct 31 2008, 08:12PM
Location: University of California
Posts: 527

If you want some references on modern distributed amps here are some lecture notes:
This one

(near the end)

These aren't pretty, being handwritten, but more complete:

And this one from MIT OCW looks pretty good

I would still recommend limiting the specs in some way, the combination of the wide bandwidth amplifier and getting such a high voltage out of it means it will be very hard for you. Tubes might actually be easier since you might be able to make the amplifier to directly output the 3kV.

You also never really said if you actually need all that bandwidth all at once. Are you just driving it with the signal generator? Because that's usually either a single tone, or you're using it to modulate a bandlimited signal. If that's the case then there's no reason to use a single amp to do everything.

Or if you're driving gas tubes and you do actually need all the bandwidth, what if you use multiple gas tubes each with its own smaller bandwidth amplifier, then drive the amplifiers with a crossover network?

aquilarubra

Fri Apr 02 2010, 06:54AM

Registered Member #2471 Joined: Wed Nov 18 2009, 07:23AM
Location:
Posts: 17

Thank you Mattski :)

I can limit the bandwidth in two or more steps. My ideal single and most used step would be 20kHz-4MHz. If that is difficult, I can also accept 20kHz-2MHz... but I have to build 2 of them them, if I need two sounds at the same time (one would be 4MHz).

I need that step of bandwidth because I need to sweep, and I have multiple tones in all that range. The signal generator will be controlled by a computer... so it is a bit complicate to go to the amp and manually turn a switch.

I need only to amplify specific waveforms, no modulation of sort. I have sines, square waves with different duty cycles, and also custom waveforms. One of the signal generators allows to draw your custom waveform. Basically, I have a sort of damped sine. Moreover, in a different experiment, I need to burst/pulse trains of damped sines. That's all.

Since I don't know much about electronics... I need somebody who can do it. I like the idea of the distributed amp... like those test equipments. However, I have been posting around internet... but I could not find anybody yet who has the expertise to take on the project... and possibly at an acceptable cost, since I am not backed by any company. So, please help, or give me some addresses please. Thank you.

aquilarubra

Fri Apr 02 2010, 07:23AM

Registered Member #2471 Joined: Wed Nov 18 2009, 07:23AM
Location:
Posts: 17

Ok, I just figured out that if I need to use the sound at 4MHz, I need anyways to have 2 amps. I do not want to do modulation, but only mix different tones directly in the plasma tube, in order to get a heterodyning effect. Thus, I could accept a step of 20kHz to 2MHz (but if it reaches 4Mhz, better).

previous 1 2 3 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob