2KW solid state linear amp using two MRFE6VP61K25H devices

[ka1tdq]

My plan is to use one 1400 watt transformer with a big series resistor during startup. The resistor will be shorted during normal operation. I did weld the transformer to a big steel plate. This plate will have a layer of heat sink compound on the bottom and be mechanically mounted to the top of my mega-transformer (must weigh 200 lbs) for additional heat sinking.

I just purchased a lot of 10 electrolytic capacitors from eBay. 2400uf at 450 volts each. This should provide enough storage during voice peaks.

Jon
KA1TDQ
 

[WBear2GCR]

Hang on a second. The 200 pound transformer is only rated at 1400watts?? Something amiss there. Perhaps you lost a zero? 14,000 watts?

I have some VERY large Signal Transformer Co isolation xfmrs that weigh between 75 and 100lbs and they are rated at 10kva or more... They are I'd guess better than 2/3cuft in total dimension, something between that and a full cuft... BIG.

Got a picture?
How many volts are these devices supposed to run at??

Welded to a plate?? The core welded or the mounting feet? Heat compound, not sure that is indicated. How is the heat going to get off the plate? Air flow is usually the cure for heat on a large xfrmr, unless it's in oil and the heat is carried to an outer case, etc...

           

[ka1tdq]

This picture should clarify what I'm planning.  The HV MOT transformer is thermally and mechanically mounted to the large step-down transformer.  The 102 vac taps on the secondary of the large transformer is connected to the 6.3 vct @ 24 amp filament transformer (I got it for $20).  The large transformer also serves as the heat sink for my microwave transformer.

A house wall switch will short the series dropping resistor going to the capacitor bank. 

[W3RSW]

A 120 vac house rated wall switch in the 3000 volt HV line?

Are you crazy or what am I missing?  Yes the delta v across the resistor is not a full 3kv, but...

Well, yes it is for the first half cycle or so.  Remember that Q = C V.  The capacitor bank represents a short in initial uncharged state.

Switches at 3kv and 1 amp or more have arc chutes for G's sake.  At least put a .01uf at 10 kv cap across it.

Mount the switch on very insulated plate and run an insulated toggle arm or have a remote rod to the actuator if your really going to do this.

Naw....  

[ka1tdq]

Thank you for checking my schematic.  I've made a change according to your recommendations.  I've included a remote shorting relay for the HV line, and I've corrected the diode and capacitor configurations so that now the MOT chassis is ground.  I'll gladly take any comments or suggestions on this current schematic.

Jon

[W3RSW]

Didn't mean to come across so strong but HV can be nasty to say the least.
Usually step start is placed in the primary.   -Relay or contactor controlled.  I've used SS relays At 240 VAC, a voltage for which they're designed.  Many industrial contactor's are available surplus.

If you place a step start resistor in the primary across the relay contacts of say 10 to 20 ohms at 50 watts and across each live leg, you'll be able to auto step start HV and your filaments briefly. Center conductor to ground of course in a three wire feed.

Auto timing of the contactor's or relay's coil done with a pick off diode to make DC and a resistor and capacitor with values such that the time constant is about a second or two.  You don't have to invent the wheel here.  There are numerous step start circuits available from ARRL handbooks over many years and on the web.

Looks like your having fun designing. Good luck with your project.

[ka1tdq]

I'm not sure why I was thinking about the resistor in the secondary.  I had inline resistors for he values you mentioned for my last amplifier in the HV primary.  It does make things a lot less dangerous. 

So, awesome!  Thanks for the input on this.  I'm glad I was prompted to put a schematic online and rethink the idea.  I could've seriously arc welded the cat to the chassis if I had gone with the original idea.  I'm curious to see how well this MOT will hold up under continuous duty (an evening of being on the air) being attached to a hunk of metal. 

Jon

[Opcom]

Actually, before I venture into Class E stuff I should probably start with something small like a single fet rig.  To get on the air now though, I will stick with what I know. 

I'll build a single 3-500 amp with the original plan of 2 microwave transformers sitting in a mineral oil bath for heat dissipation (in the cooer in th picture). I will use the big transformer's 100 volt taps so that I can use a center tapped 6.3 volt filament transformer. 

All I need to get is a tube and electrolytics for about $200 (same as a class E rf deck, but I don't need to experiment).

A friend used to put uWave oven transformers in oil for insulation. At one point a dozen of them to power a Tesla coil. The results were that the transformers did not short out despite this abuse and the arcing as well. I like your schemes a lot, both the UHF FETs on HF, and the transformers in the oil bath. This was done outside.

Beware if mineral oil may be flamable and displease both the XYL and the cat if there was a problem. There is maybe some transformer oil that is safer. Someone suggested Shell Diala X oil for use in a 2KVA audio output transformer with a 5KV supply. Great work though, very interesting to experiment all round.

[ka1tdq]

The power supply works, but not as well as I had hoped. The first time I powered it up without a load, I had to turn it off a few seconds after I turned it on so I wouldn't exceed the 450 volt rating of each electrolytic. It neared 5000 volts and I was thinking of ideas to limit it.

Then I gave it a load test with a large 1k resistor across the full electrolytic bank and ended up getting only 390 volts, as shown in the picture.  I suppose the idea would work, but you'd need like 5 transformers to handle the current.  Also, they make a loud hum which would be picked up by the microphone.  Having good audio does nothing if you hear a loud, "Bwaaaaaah!" in the background.

So, I have most of the parts mounted for the single 3-500 amp.  I'm just going to use a regular old HV transformer for the B+.  I can't think of any other novel way to get the high voltage DC other than blanketing the backyard with solar panels (I do live in Arizona), but then night time operation would be limited.

Jon

[AB2EZ]

Jon

Just a note of concern:

The ad-hoc physical layout of the transmitter that you just posted suggests (to me) that you do not fully appreciate the dangers associated with this transmitter's high voltages, and the large amounts of energy that will be stored in the power supply's capacitors.

One mistake is more likely to kill you than it is to just get your attention (as it would in the case of a boatancher with a 600V plate supply).

A sloppy/ad-hoc physical layout invites careless operation.

Stu

[AB2EZ]

Duplicate of my previous post

[AB2EZ]

Sorry, another duplicate

[ka1tdq]

I couldn't agree with you more about the dangers of high voltage. This was the testing phase of the project and i ensured all electrolytics were discharged completely before I finished.

As for the future of the amp, my wife gave me a great idea of using what you see in the picture as the front of the amp behind a layer of plexiglass. My plan was to use the plain, brown back of the filing cabinet as the front. It will look much cooler, especially if I add blue led's, but above all being enclosed and safe.

Since I can't get the HV I need from an MOT, my new plan is to grid-drive a pair of 4CX400's using the big transformer as a voltage quadrupler from a 240 VAC supply.

Jon

[KD6VXI]

Something is amiss.....  I've seen much better regulation from MOTs before.

Measure pri and sec v out of your buck transformer.  Of it's sagging you've found your problem, maybe.

You have 5kv resting and 400 volts loaded  (at 5Kv, your 1kohm resistor Is loading your supply to 25kw.  At 400 volts, it's only 200.).  I would think you have something miswired, otherwise your 1kohm resistor should have exploded.  Most MOTs I've dealt with will hold up to a 400 to 750 watt ICAS load.  Your pair should hold up to at least a kw, or close.

I hope that 4 diode bridge looking xdvice center right front isn't the only rectifiers...  And is that a kickstart type cap wired in series?

I haven't looked at your schematic, and viewing this on a 4 inch phone screen, so excuse me of I'm off in left field

Shane
KD6VXI

Screwed up by sufficiency (autocorrect) on my Android phone.....

[ka1tdq]

I'm sure something is amiss too. An MOT transformer probably could be tweeked to perform as a HV power supply, but it is very noisy. Even with working the bugs out, my plan would have been to use a smaller HV transformer to keep the electrolytic bank topped off during receive periods and then relay in the MOT during transmit (like an after burner).

The more I think about using my new plan of a pair of 4CX400's grid-driven with the big transformer providing the B plus in quadruple mode, it seems much cleaner and quieter. I can throw in extra B plus on top with another transformer if needed. I can get a pair of Russian tubes fairly cheap and the filaments are 6.3 volts.

I stopped at Fry's today and picked up a cold cathode light strip for cool lighting effects inside the amplifier chassis ( a feature Ameritron or Alpha have yet to catch on to). I will use additional 6.3 volts taps from the filament transformer to power the light strip and supply the grid bias transformer. I picked that up today at Goodwill for 12 dollars including about 100 feet of low voltage exterior wire. I will back feed the transformer 12 volt side and negatively rectify the 120 volt side.

I modified my intermediate mosfet amp for a little over 10 watts out carrier, but will need to tame that down again so I don't overdrive the tube amp.

I typed this on my phone too so forgive the mistakes.

Jon

[K1JJ]

Jon,

There MUST be a ham locally in AZ who could give you a decent, quiet HV transformer rated at 2-3KW.  Forget the voltage quadrupling stuff.   If you were able to pick it up, I have one here that weighs about 70 pounds and could handle your requirements easily.  You could have it for free.  Pick up only.

Put out a "wanted" post looking for a HV transformer and I'll bet you find one locally ASAP... for free.  Many OT's would love to get some of that big iron out of the cellar and save their backs.

Good HV regulation is absolutely critical to a clean linear signal, so might as well get something worth all your hard effort.

T

[K1JJ]

Jon,

From your latest "wanted" posting, looks like you found a great transformer from Dennis / W7TFO.


Good going!   It often pays to axe around.   And TNX to Dennis.

T

[Opcom]

I'm glad a proper xfmr has been arranged. You can use MOTs, but not like you showed, except in certain circumstances.
for one thing that usually make some 2200VAC, so a properly working doubler is going to make too much voltage. It looks like a doubler in the circuit you drew, same as in the oven circuit. The oven, though, uses only 250mA to the magnetron at about 4KV. One trans is not really enough for a 3-500Z.

The cap in the doubler has to be the same size as the output capacitance for the thing to work well. So, just in case I missed it, what is the value of it? Something looks funny about that circuit. In a microwave, the full current also passes through the cap as it is charged an discharged each time, I think (been a while, maybe I'm wrong on this).

Now as to the layout, it does not bother me at all because I don't have to work on it, but I have worked with crazy HV schemes for many years and possess at least the minimum necessary fear of death, which saves me from my own doom. Also note that the plastic sleeves of the electrolytics might not withstand the voltage from the case to GND. That is, any imperfection is an opportunity for problems.

One piece of advice, get a shorting stick. Run its ground strap through a couple of 200-500 Ohm 100W resistors in parallel to get a polite discharge. The two res. are so that if one goes bad the thing still works. Some folks would say to be rid if the resistors entirely. Steve Haney of Haney Electronics tc0654@mesh.net usually has some NOS military shorting sticks, very nice with long braided heavy ground straps inside a sleeve, and a quite thick hook and long 1 FT handle. I have one installed in each transmitter. I can always make sure the HV is gone, each time, and every time. Please do this, get a shorting stick. I'd hate to no longer enjoy your posts about building stuff. "Jesus is only a prayer away, and a few hundred volts is all it takes to meet Him". or (insert your beliefs here, to be fair)  Hey just be careful with that stuff.

I saw a guy work on a uWave oven for 'no heat' and didn't discharge that little 2uF cap. It was full (maggie's fil. connection was open) and it blew a chunk about 1/4"x1/2" out of the side of his thumb, most spectacularly and made a little mess in the oven too, which I had to wipe up as he had been taken to the emergency room with a blood soaked towel on his hand and the customer was due in later. I think he was lucky it only went through his hand.

8-)

[WBear2GCR]

Ok, I got lost here.

Thought this was about a solid state linear?

Then it is a 3-500 linear?

Next it is a 4CX400 linear, grid driven?

Is it for AM or SSB?

This HV stuff really requires some experience and/or someone with experience to oversee.
It has the bad and nasty habit of reaching out and grabbing you.
Often there are no second chances. FYI.

IF you have yet to build up any tube stuff with <1000vdc B+, then it would be my suggestion to first experiment and build up something, maybe not even a linear, as in interim learning project. Alternately, repairing a "basket case" tube rig is not a bad way to get ur feet wet. But even something like an Apache or DX-100, Valiant is not without danger. Even 400-600vdc can stop ur heart.

Quite frankly HIGH VOLTAGE is scary stuff.
A lot of hams don't want to put their hands inside big HV rigs for that reason.

Another thing that makes sense is to follow a known recipe for a construction project.
Less chance of a random and fatal building flaw or error.
For example an unintentionally floating chassis or subchassis can kill you dead, fast.

A least define ur project goals and then see what you can do to meet them. My 2 cents.

                       

[wa3dsp]

This is what happens when you mess with high voltage!

http://www.youtube.com/watch?v=H4hSMQQa0Kk

But all kidding aside do a search for  'arc flash'  on YouTube. It will give you great respect for electrical power.