HV transformer voltage drop

[scottdarling1221]

Hello all,I Am building a multiband 3cx3000a7 amplifier for use mainly on 75 meter am.I have all of the parts i need i just had to buy a new 20 amp roller inductor.

My question is i have a 2400 volt 4.2 ccs transformer that im going to use with a fullwave voltage doubler.This will give me 6768 volts dc with just a filter cap and no regulation.I am wondering how much the voltage will drob down to when im running the legal limit on am.

I know the transformer is over kill but i already had it so why not use it and save some cash.

The amp is capable of 1500 watts or so keydown on am with the roller inductor that i will be using of course i would never transmitt with that amount but i like to have headroom.

I would like to have an idea of the voltage drop at the 1500 watt level modulated.

I have over 240 volts coming into the shack with a dedicated 100 amp service.Came with the house.

Any help would be greatly apreciated.

[KE6DF]

You can get some idea of how large of caps you will need by playing with the following free power supply simulation program:

http://www.duncanamps.com/psud2/

The trouble with it is it only simulates the case where the supply is loaded with a fixed resistor.

And a linear has a varying load based on audio amlitude.

So, the dynamic regulation under an audio load (ie, when you are talking) would be more of an issue than just the voltage output under a fixed resistance load.

The only good way to deal with that is to use plenty of output capacitance.

[KM1H]

Since high C 450V caps are rather common these days and even available surplus Id go with a minimum of 25uF total. Higher will give better regulation once you decide to fess up and use that tube the way it was meant to be   but at 1500W PEP the voltage will barely budge with that iron. The popular values among linear amp builders these days are 470 and 560uF with a few going into strings of 1200-2400uF.

At that voltage and with any C you need a serious step start and real diodes. Id go with 6A10's and remember to leave the leads long for heat dissipation (same with 1N5408's in a smaller PS).  Run a .001@ 10kV or better across the entire string of diodes and another to ground at the B+ out point; these protect the diodes from ultra fast spikes. Install a 1N5408/6A10 reverse connected across each cap to protect them in case of a PS diode string failure; they will short and your safety devices will open. You dont want to be around a 6-7kV PS when the caps blow from raw AC....in an instant.

You also need a real HV fuse in the PS and a 30-50 Ohm 50W minimum HV glitch resistor at the base of the plate choke. This is no time to get sloppy as Ive seen in way too many HB PS.

Carl
KM1H

[Detroit47]

 I've built a lot of big boxes. I never use strings of 450 volt caps that is half assed. Either use a large oil filled job like a power factor correction job or http://cgi.ebay.com/HIGH-VOLTAGE-CAPACITOR-60uF-4000V-Lot-2pcs-New-/160430225428?cmd=ViewItem&pt=LH_DefaultDomain_0&hash=item255a62f814. I have had great success with these. I have used these for Dielectric work as well as RF amplifiers.

73 N8QPC

[scottdarling1221]

I agree.I usually use the 5kv 40 uf version i havent tried the 60uf caps yet.

[Detroit47]

I've ran these at 5500 vdc no problems and have been in service for several years.

73 N8QPC

[KE6DF]

Of course you need two capacitors to make a voltage doubler work. So two 60's would be nice giving the equivalent of 30 mfd at the output of the supply.

I like big oils much better than a string of electrolytics.

It's just inherently more reliable to have a single point of failure rather than several.

But two of those big guys would be expensive.

[Detroit47]

I would put a step start or soft start circuit in your supply to keep from tearing up diodes. The first couple of cycles are going to put a huge load on diodes. I like to use a step start with a resistor in series with the primary of the transformer.

73 N8QPC

[scottdarling1221]

i always use a step start.Or as i usually call it soft start.Thanks for the advice though.

[Opcom]

You can get some idea of how large of caps you will need by playing with the following free power supply simulation program:

http://www.duncanamps.com/psud2/

The trouble with it is it only simulates the case where the supply is loaded with a fixed resistor.

...

A solution is to run several simulations and put the data into excel to get the expected regulation data. It is tediuos but works fine. I have asked Duncan for a multi step current, or a current sweep, but I don't think he is developing the program further.