B+ from GZ34 Heater CT ?

[sgerus]

I have a question about the best way fo get B+ from the GZ34 rect.

I started with the B+ coming off pin 8 of the GZ34 and T1 was getting very “Warn”.
Now I have the B+ off the ct of the 5V heater windings for the GZ34 and T1 is a lot cooler.

So its working, same B+ and T1 is cooler.

Almost all of the power supplies I see have the B+ right off the rectifier.... Yeat this B+ off the heater CT seemed to work better in my case.

Any thoughts?

[SY]

The GZ34 is not a filamentary rectifier, it has a cathode. The cathode is connected to the heater at pin 8. If it got too warm, you might look down stream and see what was causing that ( a guess: are you using a C-input filter?). Running off the heater CT increases the series resistance which degrades regulation.

[sgerus]

Tetsujin,

Thanks for the reply. First of all I was wrong about the amp running cooler (I was testing with the amp on it's side).

After 4 hours of use T1 was still warm.... 102 deg F.

Your right about the C input, this is a CLCLC filter.
50uF 10H 50uF 7H 30uF 33,000R 30uF

To get the voltage down to what I need, I have two 300R on the 750vac output from T1 before the GZ34 ( I need to drop about 20v)

What I'm thinking of doing is change C1 to say 4uF and see how much that lowers the B+

OR swiitch to a choke input pus and just run the 300B at less power.

[Sheldon]

Originally posted by sgerus

What I'm thinking of doing is change C1 to say 4uF and see how much that lowers the B+

OR swiitch to a choke input pus and just run the 300B at less power.

Before you do it physically, model it in PSUD. My experience is limited, but I have found it quite accurate at predicting PS parameters, if you can give it accurate input data.

Sheldon

[kevinkr]

Hi Sgerus,
If you measured only 102 F on the core of the transformer after several hours of operation this is just fine. As a rule if you can actually touch the core (say 60C or 128F) you are operating at an acceptable temperature level.

[Eli Duttman]

Sgerus,

Your power trafo is 375-0-375. If you use a pseudo choke I/P filter (the cap. that follows the rectifier is < 1 muF.), you will retain good regulation of B+ up to a certain point. DC O/P voltage = AC RMS is no problem. As the cap. increases in value, the DC O/P voltage rises and regulation worsens. You have to work things out empirically at the bench. Get rid of those heat generating dropping resistors and experiment with the small 1st filter cap. value until you get the rail voltage you want.

[SY]

Sgerus, if you use an input cap, 50u is a bit too big for that rectifier. 40u or less should be safe.

And I also have been amazed at how well PSUD2 models supplies. It's free, it's easy to use.

[sgerus]

Two more questions:
With the PDUS2 program, do I use the RMS voltage?

I understand if I remove the cathode resistors I reduce some heat;

but is it a good idea to use say 2 - 100R to lower the voltage surge into the GZ34 at start up time?

( the data sheet I have for the GZ34 shows 2 x 125R for a 400V Capacitor input filter)

[Sheldon]

Originally posted by sgerus
Two more questions:
With the PDUS2 program, do I use the RMS voltage?

I understand if I remove the cathode resistors I reduce some heat;

but is it a good idea to use say 2 - 100R to lower the voltage surge into the GZ34 at start up time?

( the data sheet I have for the GZ34 shows 2 x 125R for a 400V Capacitor input filter)

Yes to RMS.

I believe the resistance quoted is the total input resistance per plate. My data sheet for a 5AR4 shows Rs = Rsec+N(squared)Rpri+Ra, where Rs = plate supply resistance per plate, Rsec = transformer secondary resistance, Rpri = primary resistance, Ra = added series resistance per plate, and N = transformer step up ratio per section. In this case it was with a 40uf cap. With a smaller cap total series resistance could be less. also if series inductance is present. Mainly you don't want to exceed the maximum transient peak current. You may not need the extra series resistance, depending on your transformer DC resistance. PSUD calculates peak current for you (just look at I(D) for the first few cycles), so you can check that you are within safe limits.

Sheldon