The AM Forum

I'm designing a power supply, and want to use a pair of 3B28 rectifiers.

The problem is, I don't have a source of 2.5v.

Instead, I have a winding that gives me 5v at 6amp -- but no center tap.

Would it work to connect the two rectifiers in series, and take the DC off the mid point where the filaments are connected?

Or would I likely run into problems with uneven filament voltage?

Have others done this and seen it work?

Dave

Not sure, but I would think it is a questionable way to run the filaments.  2.5v filament xfmrs should easy to come by.

Fred

FWIW,

I had an old Motorola low-band base station years ago that did just that.
It had a pair of 4-125s for finals and a pair of 866s for the rectumfryers.

They used a 5v fil transfoma for the 866s and pulled the HV off of the center link between the filaments.

I have a second related question:

With MV rectifier tubes, the designer is cautioned to shield the tubes from strong RF fields.

Does this issue come up with Xenon rectifiers?

Dave

On the series string, I've done it and never noticed any difference in tube life.

On the RF question, any gas-filled tube will act up in the presence of hot RF.

73DG

Your filament xfmr will need to have a winding insulated for at least the working voltage of the rectifier. Otherwise it should work if the rectifier cathodes were to be connected together anyway.

I'm looking to build a small transmitter for 10M. Planning to build it on a single 17x14x3 chassis and house it in an old black wrinkle Bud 8 3/4 cabinet.

I plan to use a large TV replacement transformer with a 400-0-400 winding at 210ma.

I was thinking of using a 4D32 as the final, and a couple 6883 (12V 6146s) in AB1 as the modulators.

The tube choice is based on what I have and the goal is about 50W out -- conservative for the tube choice, I realize.

The TV tranny has tons of filament windings including three 5V windings, two 6.3V windings, and another 212-0-212 winding. It weighs about 15 lbs, so I think it has enough capacity for such a TX.

I need to use cap input to get high enough voltage (around 500V) and don't want the voltage sag of a 5R4.

Hence the idea of using 3B28s.

But, there is likely to be lots of RF inside the cabinet, so perhaps that would be a problem.

An alternative is a high-current TV damper tube like  6CJ3s.

Those have such low plate resistance that the voltage sag would be no more than an MV or Xenon tube.

I realize I could use silicon diodes, but I like tubes.

Dave

I would skip the tube rectifiers and use diodes.  Your TV xfmr will need all the help it can get.  No load on the 5V filaments will help.  Put all three 5V filaments in series and use a FWB rectifier to make a 12-15V relay supply.  I've used many TV xfmrs from the 1950s and know just about what they can do.  I use two of them in my 50W 6146 rig.  I still have many NOS monster RCA TV xfmrs that date from the early 1950s.

Maybe you can post any part numbers on that TV xfmr and I'll look it up to find the ratings.

Fred

It's a Halldorson P9727

I have the catalog and looked it up:

400-0-400 @ 220ma
212-0-212 @ 90ma

6.3V @ 10amps
6.3V @ 2.6amps

5V @ 3amps
5V @ 3amps
5V @ 2amps

weight 14 3/8

Actually, I made a mistake in my last post.  The lighter 6V winding probably was for the damper.  The damper winding has a higher voltage insulation, reason it was a separate winding.   With only 220ma not sure why they used two 5U4s.  Two 5U4s were used in TVs but most always it was only one.  I see I have a P-9908 which I think is a Halldorson.

The 2A 5V winding probably was for the 200V winding using a 5Y3.

The xfmr should be OK for a 50W rig,  not sure you'll get 50W with only 500VDC.

Use the 200V windings for all the low level stages and the 400V windings for the HV B+

You'll do much better using SS diodes instead of the tube rectifiers.  IMO

Fred

According to my catalog, the Halldorson P9908 has the following ratings:

320-0-320 @ 150ma

5V @ 3amp

and three 6.3V windings at 3, 8, and 1 amp.

I didn't see it mentioned, but.....

if you are going with gaz rectifryers, you will need to use a choke input before
the filter cap...

Pete

I have heard of people using MV and Xe rectifiers with cap input successfully.

Here is my understanding of this:

Rectifier tubes (and semiconductor too) have four key specs that come into play.

Average current - The PS output must be less that twice this number for FW CT
Repetitive peak current - This is the peak current on each ac cycle
Non-repetitive peak current - This is one time, occasional peak during turn on.
Peak inverse voltage - we all know about this one.

Anyway, it you keep everything within these parameters, you should be OK.

You may need a series resistor if the dc resistance of your power transformer is low to keep the repetitive current within spec.

You may need a soft start to avoid exceeding the non-repetitive spec.

But, if you keep to the specs, MV and Xe rectifiers should work with cap input.

Note that the 83 MV rectifier has specs in the manual for use with cap input filters.

Dave

Why even bother with the P.I.T.A. of MV rectumfryers for that low voltage? ?
A couple of 1N5408s in series for each leg and you're good to go. And, it keeps it simple. The 5408s prolly cost around 10 cents a piece (I buy them by the 100)
And I use them for just about everything low voltage. My general rule for rectumfryer application is as follows:

1N400x (usually use 4007s for all applications)  Low voltage, low current applications   
            (like 12 or 24v control circuits, biass rectifiers, low voltage B+, etc)

1N5408  Medium to high voltage B+ (stack as needed) medium to high current
            applications (100mA to the 3 amp rating of the diode) (stack as needed)

HV diode blocks   Call them what you like, (K2AW diodes, microwave oven diodes, etc)
                        I usually keep a few of these around for those serious HV, high
                        current applications, like big TX plate supplies.

With just keeping the the 4007s and 5408s around, you can cover 90-95% of any application that you will need a rectumfryer for. I usually buy them by the 100          (at least) when I reorder them..

So with 2 or 3 basic types of diodes, you can pretty well cover just about all of tyou rectumfryer needs. (at least for tube type gear) And it wont cost much $$ to keep the rectifyer drawer of your parts cabinet "stocked up"

Because they are lonely and all they want to do is glow for us :'( :'( :'( :'(

I have done it with Xeon rectifiers but with MV, your playing with disaster.

I agree with the Slab.  Use the 1N5408 diodes, for 500VDC use at least two in series in each leg.  Using tube rectifiers only adds to the project in a circuit that can be done much easier.  The tubes also eat up valuable space on your chassis.

Fred