Building a home brew linear using a GS35b

[AB2EZ]

Hi

I'm going to build a grounded grid linear amplifier for 75-80 meters (single band to start, multi-band upgrade later) using a GS-35b. I just have an urge to build a home brew, high power, RF, tube-based project... and this is probably the simplest one.

Putting aside the RF deck, for the moment (I think I've pretty much done my homework on that)...

I have a simple question regarding the power supply. I want the plate supply to provide approximately 3000 volts at 800 ma when the linear amplifier is at peak power.

I'm going to use a capacitor-input design, with about 90 uF of capacitance.

I'm trying to decide whether to use a 4800 VAC CT transformer to drive a full wave rectifier, or a 2400 VAC transformer to drive a bridge rectifier.

It seems to me that it shouldn't make much difference, either way. I'm inclined to go the 4800 VCT route.

Either way, I'm going to purchase a new transformer from Peter Dahl... so availability shouldn't be an issue.

Comments?

Stu
AB2EZ

[W8ZAP]

Hi Stu,
I have built a couple of GS-35B HF amps and always opted for the full wave bridge method. That tube will safely run at around the 4000v area, so 3000 to 3500 should be a good head room voltage. I used the "Block" type rectifiers from K2AW, but obviously that is just one style you could use. It sounds as though you have already got a HV.PS layout some what arranged, but I used the PIV rating of 1.41 above the desired plate voltage for my rectifier selection. Also, with the GS-35B, being that it is an "indirectly heated cathode" tube design, you can use a regulated DC power supply for the filament. The one I use is the Power One 12.6vdc@3.4amp with an adjustable voltage and surge/current limiting. Sounds as though you have a good project going.
Jerry
W8zap

[WA1GFZ]

Stu,
I used a pair of BC610 plate transformers. One either side of a 240 line.
in my old 4-1000A linear.
I had 16 uF cap filter. The 610 transformer has 2500 either side of CT.
I got about 3000 VDC on the high tap . There is also a 2000 volt tap but never used it. I think the BC610 transformer is rated for 550 ma ccs. I did a pair of full wave rectifiers in diode or mode.
I suggest you buy a transformer with a couple different voltage options since they are big bucks.
Also the T368 transformer is a good one. I think the output is higher.

[W2VW]

Hi Stu,

      The answer depends on your goals for this design. Stage gain? Cost of parts? Form factor? Ease of duplication by other experimenters? All these things make a difference.
      Right off the bat 3000 volts should be the absolute minimum plate voltage for this tube in cathode driven service. 3500 would be better. It does not directly compare with an 8877 which does fine below 3000 volts. Your proposed supply will not make 1.41 X AC volts under load. Even a reasonable bleeder will reduce that. Look at power supplies for military linear amplifiers of the same capability for ideas. Most use resonant choke input filters. That subject has been covered here so the archives will help. A voltage doubler circuit with a transformer designed to have a low resistance secondary winding is more cost effective than full wave rectifier capacitor input supply.
      I've recently had good luck using turbocharger hoses for tube chimneys. There are lots of sites which sell high temperature silicone hoses to hot rod types. Much cheaper than Teflon and the hose can extend all the way to the top of the amplifier cabinet to duct all the heat away from the other components. Alpha used that design. Check out the Alpha 76A for pictures. I assume that you have seen this: http://gs35b.com/

[W2VW]

Also the T368 transformer is a good one. I think the output is higher.

Most T-368 transformers are 6336 volts secondary 110 primary and WILL withstand full wave bridge configuration without the unused centertap failing when used with choke input. Same is true of the secondaries of the later BC-610 oil filled iron. IIRC the late BC-610 iron was 4800/5720V CT. A pair of primarys either side of the 240 line works well. Surplus 3 phase variacs using 2 phases plays well for voltage control.

[WA1GFZ]

T368 transformers would be a better choice with the extra voltage.
I could get 1500 watts out of a 4-1000A but not much more. I think the voltage was 3200 or 3300 VDC under load. gfz

[AB2EZ]

Hi!

Thanks to all who provided their much-welcome advice.

I have decided to use a bridge rectifier design, with a 2400 volt 1.5A CCS-rated transformer. I have a large 240 volt input variac, which I can use to adjust the DC output voltage up or down a bit. I'm assuming (as a rough rule of thumb) 1.25 DC volts output per rms AC volt input, under load. Thus I am planning to operate the tube at 3000V, even though it can handle more. I'll also use a small variac on the filament supply to keep the voltage close to the specified value of 12.6 VAC.

Best regards
Stu