Help a noob with respeccing a transformer for silicon rectifiers.

xephon

Member
2005-08-08 5:52 am
So I am ready to build my first tube amplifier after years of MOSFET's, IGBT's and other silicon new-fangelry.

I always liked building things with CRT's, VFD's and nixies and the like (even a Crookes tube or two..) so now I am ready to build the amplifier linked below.

So my question is this.

If I were to replace the selenium rectifiers with silicon (maybe Cree Schottkys), how can I calculate the new voltage required from the transformer to maintain the correct voltage after the rectifiers?
I'll be getting the transformer custom wound, so I may as well get it wound to suit silicon rectifiers.

Here are the specs of what is specified now:

Transformer: 400 - 0 - 400V @ 340mA (I'll probably spec about 500mA to give some headroom).
Rectifiers: 25mA Selenium

Can someone help me to work this out? :D
 

Attachments

  • Electronics-World-1961-03 pg29.png
    Electronics-World-1961-03 pg29.png
    486.7 KB · Views: 79
  • Electronics-World-1961-03 pg30.jpg
    Electronics-World-1961-03 pg30.jpg
    579.7 KB · Views: 77
  • Electronics-World-1961-03 pg31.png
    Electronics-World-1961-03 pg31.png
    669.4 KB · Views: 78
  • Electronics-World-1961-03 pg32.png
    Electronics-World-1961-03 pg32.png
    551.2 KB · Views: 70
  • Electronics-World-1961-03 pg85.png
    Electronics-World-1961-03 pg85.png
    240.5 KB · Views: 69

ericj

Member
2008-12-08 10:24 pm
Short answer: You don't have to change the transformer. Change filter components instead. It's my understanding that when guitar amp techs replace dead selenium rectifiers with modern silicon, they just add a series resistor.

Also, simulate it in PSU Designer 2:

PSUD2
 

xephon

Member
2005-08-08 5:52 am
Short answer: You don't have to change the transformer. Change filter components instead. It's my understanding that when guitar amp techs replace dead selenium rectifiers with modern silicon, they just add a series resistor.

Also, simulate it in PSU Designer 2:

PSUD2

I was hoping to do away with a resistor so things were kept cooler and more energy efficient, seeing as I will have the transformer designed and built to my exact needs anyway.


I had a look at that PSU designer, but it won't allow me to design the PSU section of my schematic, it's really clunky and difficult to use...
 
Last edited:

xephon

Member
2005-08-08 5:52 am
Ah, yeah, sorry. I'm still learning all the tube terminology...

So directly replacing the selenium bias rectifiers with silicon will work without any other changes needed? Do I need to reduce C25/26 and/or R53/54 as Mooly suggested?

Should I install bypass capacitors in parallel with the replacement silicon diodes?
 
Last edited:

Mooly

Administrator
Paid Member
2007-09-15 8:14 am
Reducing the cap value slightly (0.05uf to todays 0.047uf) will reduce the bias voltage a little... its a non problem though because you have a pot at the end of it all anyway to set the bias voltage. Its all such high impedance that I doubt the change from selenium to silicon will make much difference anyway.

A cap in parallel. I don't know what the audiophile thinking is on that for this application although given that the rectifier is fed from a cap anyway (which has significant reactance), well I doubt it would make any difference tbh.
 

Koonw

Member
2013-04-09 9:37 pm
Ah, yeah, sorry. I'm still learning all the tube terminology...

So directly replacing the selenium bias rectifiers with silicon will work without any other changes needed? Do I need to reduce C25/26 and/or R53/54 as Mooly suggested?

Should I install bypass capacitors in parallel with the replacement silicon diodes?

Here is an article you can read about replacing selenium rectifier. You need to know how many stacks it has originally, as each stack will drop about 1V DC.

If you're to build this amp you should know all the value of components, it'll help you or others try to sort it out. The component values are not labeled directly along with parts, some maybe missing, making it very difficult to work on.
 

xephon

Member
2005-08-08 5:52 am
The values are all there in the list, but I am currently working on redrawing the schematic so it will be much easier to follow.

The only specs for the selenium rectifiers is 25mA, and nothing else at all... I'm not sure how to work out how many plates that will have? I that link you provided (I already have it printed out! :D ) and also on Wikipedia that most plates will handle 20V reverse voltage on average, I guess this means that for 400V, it will have 20 plates, so it will drop about 20 volts?

@Mooly, I do have 0.047uF caps etc, so I guess I'll suck-it-and-see and use that pot to trim it back to where it should be.
 

Koonw

Member
2013-04-09 9:37 pm
The values are all there in the list, but I am currently working on redrawing the schematic so it will be much easier to follow.

The only specs for the selenium rectifiers is 25mA, and nothing else at all... I'm not sure how to work out how many plates that will have? I that link you provided (I already have it printed out! :D ) and also on Wikipedia that most plates will handle 20V reverse voltage on average, I guess this means that for 400V, it will have 20 plates, so it will drop about 20 volts?

@Mooly, I do have 0.047uF caps etc, so I guess I'll suck-it-and-see and use that pot to trim it back to where it should be.

Yes, but why connect to such a high voltage? You can do it another way: get -18-20V with a low voltage transformer output 15V AC. If you have to connect to 400V, this is my initial sch. If you label all components I can plug in his value to check.
 

Attachments

  • bias supply-2.png
    bias supply-2.png
    127.1 KB · Views: 22
  • bias supply-1.png
    bias supply-1.png
    140.2 KB · Views: 18
Last edited:

Mooly

Administrator
Paid Member
2007-09-15 8:14 am
@Mooly, I do have 0.047uF caps etc, so I guess I'll suck-it-and-see and use that pot to trim it back to where it should be.

I'd be confident that doing that would be fine. In the unlikely event the voltage was that bit to high (and so outside the trim pot range) then you only have to alter either the cap or more easily, one of the resistors.
 

xephon

Member
2005-08-08 5:52 am
Yes, but why connect to such a high voltage? You can do it another way: get -18-20V with a low voltage transformer output 15V AC. If you have to connect to 400V, this is my initial sch. If you label all components I can plug in his value to check.

I guess because the schematic says to do it with 400v. I don't know enough yet to go changing stuff up that much....

I'm fine with transistors and inverters, but valves are new to me....
 

Koonw

Member
2013-04-09 9:37 pm
I guess because the schematic says to do it with 400v. I don't know enough yet to go changing stuff up that much....

I'm fine with transistors and inverters, but valves are new to me....

Oh ya sure certainly it has to do HT of 400V, it's called semi-auto bias, when HT drops, the bias will also drop, imaging if the bias is rigid, the higher bias will reduce the current and power when the output increased. But this type arrangement maybe too rough, the bias bounced up and down very quickly as HT varies with output. Anyway, you can find out how it's later.