PSU from article dated 2006

[bazabarry]

I've recently built a psu from an article on this site dated 2006. It used a 15v +15v transformer output to supply two circuits. One is a 90v ht supply and the other a 2v constant supply.
I would now like to use the same method to build a 120v and 60v psu but do not have the knowledge to modify the circuits.
I assume this would be possible using the build method in attached diagram i.e.one 15v supply for the 120v and the other for the 60v. Hope this makes sense and look forward to your comments and layman circuit changes.
This would then give me all supplies for my cossor 234 radio (except the grid bias)

Barry.

[TrevorG3VLF]

A 40V-0-40V transformer will give 60 and 120V or thereabouts. Small 40V transformers do not seem to be readily available.

The voltage doublers could be enhanced to voltage triplers to get more voltage.

You could add a boost convertor (cheap) to your existing circuit but may have interference problems.

I have used the slight HT drain to produce 9V negative for grid bias
My filament supply was set to 2V and a series transistor connected as a diode was used to drop to 1.4V when necessary.

[kellys_eye]

C1-4 along with D1-4 form a voltage multiplier circuit to derive the ht.

You can either remove a stage (D3/C4) or add a stage to lower/increase the output.

If you look carefully you can see how the sections are 'repeated'.

[TonyDuell]

That voltage multiplier looks to be a quadrupler to me, so presumably the HT output is 4*sqrt(2)*(RMS value of transformer secondary voltage). With the 15V transformer, that would give 84V (near enough to 90V). With a 21V (rms) transformer (or, easier to find, a 20V one), you'd get around 120V output. I would increase the capacitor voltage ratings, 63V is a bit marginal then. Say use 100V capacitors.

The 'bottom' end of the transfomer secondary in that circuit would appear to be the mid-point of the supply. So you could take your 60V from there. Add RC smoothing (similar to the output circuit on the 120V line) between that point and the HT-ve connection.

The only problem is that if you use a transformer with 2 20V secondary windings, the input to the LT regulator will also increase, meaning that IC will disipate more power. It might be better to use 2 transformers, a 15V one for the LT circuit (and keep that as it was) and a 20V one for the HT. What a pity those wind-your-own-secondaries kits are no longer available.

[Radio Wrangler]

How about two transformers back to back?

There are some nice small ones with dual primaries 0-110-120v where you connect the 0-120 taps of both in series to get a 240v primary. dual 0-6v secondaries are commonly available. use the secondaries in series to get 12v with a 6v tap.

Take a second identical transformer and power its series connected 0-6v and 0-6v windings from the 12v output of the first transformer. you can now use the 0-110 winding of the second transformer to run a bridge rectifier and a high voltage regulator to give your 120v HT. Add further regulators for 60v, 90v if you need them. The HT power taken by these sets is low, so regulator efficiency isn't terribly important.

For the heaters, pick power off of one of the 0-6v secondaries, bridge rectify and smooth, use the regulator design you already have to make 2v regulated.

You still have a 0-110v winding unused on the second transformer. rectify, smooth and regulate and you have a grid bias supply.

Done this way, the LT, HT and GB supplies are all isolated from each other, just like three batteries would be, so the power supply doesn't place limitations on how the set connects them together.

A diagram would be so much better, but I've not got software handy at the moment.

David

[bazabarry]

Hi and thanks for all your suggestions. Sorry but most of it has gone over my head. Tony's idea of using a 20v – 20v transformer and using the same method as my first psu, ie a quadrupler circuit.
(4x20=80 x 1.414 =113.12) Not quite 120v but possibly ok for my cossor radio. Its how to use the other 20v secondry to produce 60v that's beyond me. Any one able to explain in a not to technical way would be very much appreciated. To sum up a 20v-20v transformer to produce one 120v and one 60v supply.
Barry.

[TonyDuell]

If the 60V supply can have the same -ve connection as the 120V supply (which would be normal, if the 60V supply is say for the screen grid of one of the valves) then you can build the circuit as shown (with a 20V transformer) and take 60V (or actually 56V) from the +ve side of C3 in that circuit. I would make a second copy of the R1-R3,C5-C7 circuit, connect the -ve side of that to the -ve side of the existing C6 and the 'input' of the new R1 to the +ve side of C3, to improve the smoothing. That only uses one 20V secondary.

If you want a separate 60V supply then use the second 20V secondary winding on your mains transformer to feed a second copy of the voltage doubler C1,C4,D2,D3. There will be 60V across the copy C4. Again add a copy of R1-R3,C5-C7 to improve the smoothing.

[bazabarry]

Just what I was looking for. Building separate quadrupler and doubler circuits should meet my needs. Thanks to everyone for your input.

[Mr Moose]

Hello,
Don't worry if your voltage is a bit low. A typical 120V HT battery would start at about 125V and then rapidly decline to about 110V and then slowly decline to about 90V and then decline more rapidly again to about 60V. Batteries were expensive so most pepole would use a 120V battery down to at least about 80V so you can usually get away with a lower HT voltage although some sets are more tolerant than others. As the HT voltage drops it can be advantageous (or sometimes essential) to reduce the grid bias for best results (always turn the set off when adjusting the grid bias.)
Yours, Richard