Marconiphone T11DA power supply.

[Refugee]

Split by moderator from this thread.


My old aunt's radio is the same model as this. I used to do repairs on it as a teenager. I later inherited it and have forgotten most of the botched repairs I did at the time.

The dropper mains cable is the obvious problem and I do remember doing a bodge on it back in the 1970s.

So far I have visual checked the getters and measured the filament resistances with good results. I have looked on the valve museum website and added up all the voltages and found that the total matches the value given in a topic that is old and closed about replacing it with a capacitor dropper.

I have three solutions:

1) Buy a 110 volt transformer and add it as an eliminator style box like a modern gadget.

2) Do the same thing with a cut down and modified CRT printed circuit with a chip power unit with the sensor modified to regulate the heater chain voltage and the other coils connected with a doubler circuit if required. This risks a bit of noise on the reception. The set was made to run on DC from new anyway so this would not be an issue but has the advantage that there would no longer be a live or part live chassis.

3) I may already have an old printer eliminator that produced about 45V at quite a high current. Again this could be used with a doubler to get close to the heater chain voltage or a little below so that it might take a little longer to warm up.

What solution is best?

[Tractorfan]

Hi and welcome,
I'm thinking that option three, the printer PSU, will produce 45volts DC and you can't use a voltage doubler on DC. Any switch mode PSU will radiate lots of "hash" that could be very difficult to suppress, particularly with a radio.
I don't quite understand what's going on with option two, but you may, again, have problems with hash.
Personally, I would go for option one. Step down transformers are easily available from shops such as Maplin's, etc. But they don't isolate from the mains.
Cheers, Pete

[Refugee]

The old printer PSU I was thinking of if I have still got it was AC and had a non standard connector. If I can locate it again it would double up and with the lead changed for a 3-core power would go through the existing non repairable plug on the modified eliminator to supply the HT.
Even if it is glued I know how to make a cement that can join ABS plastic. Even if it was DC a rectifier is only 4 dabs with an iron and it is out.

[ukcol]

This link may be useful to you.

http://vintage-radio.com/repair-rest...per-calcs.html

It covers the normal resistive, diode/resistance and capacitive droppers.

Colin M

[Refugee]

The web link matches what I did on a bit of paper and the value of what remains of the dropper cable and gives me the voltage I need to apply. The doubler is best as if anything goes wrong it fails safe unlike the capacitor dropper.

[Adjuster]

Beware that if the printer supply output were rated at 45V rms AC, a voltage-doubling rectifier may give a lot more than 90V DC. The peak voltage could be of the order of 2.83 * 45V - (rectifier drops).

Although the average voltage would be brought down by around half the ripple voltage, the result could still be over 120V, particularly if you are generous with the smoothing capacitors, even more so if the transformer is less heavily loaded than in its original application.

It would be wise to measure the DC output into a suitable load (filament lamps?) rather than just connecting to the radio and hoping for the best.

Good luck.

[Refugee]

I was intending to run with smaller capacitors in order to "regulate" the voltage on the grounds that it is an AC/DC circuit and will tolerate the ripple. The transformer I have got looks more like 40v with two bobbins that can be connected to give 80v anyway so this is less of an issue now. I might just give it a try on plane AC first.

[Refugee]

What voltage should the DC be on the HT from the rectifier?
The transformer I have is powerful enough to make the set fully isolated so that the chassis can be earthed for safety perhaps with a booster diode from the half voltage tap I will end up with.
The component is 2 x 40V at 1.75A so it may creep up a little with a load of only 0.5A or so.
I will have another look at the spec of the audio output valve again I think.

[Herald1360]

HT current is about 72mA, speaker field L10 is 450R and screen volts on O/P valve is 115V so HT DC at rectifier cathode will be 115+(0.072*450)=150V approx.

Heater chain voltage required is about 120V AC or DC

A suggestion.....

Since using your transformer doesn't make for easy compromises, why not make up a line cord replacement using suitable power resistors in a grounded metal box. The circuit snippet below shows the values required.

The 220R should be rated about 20W, 35R about 5W and 190R about 10W. I would suggest using 25W and 10W ally clad chassis mount types- 220R, 33R are standard values as is 180R. If the odd 10R worries you, add a series 10R small ceramic WW type. If there's room inside the case, this lot could simply go on a heatsink in a spare corner.

[Refugee]

The reason for using the transformer is that it is laying around idle and also the energy consumption is lower by about 20w. If it needs more voltage the double windings allow me to double 40V and rectify the other 40V to get about 110V or so.

[Herald1360]

OK..... try this:

Voltage double rectify one of the 40V windings using 1N5407 diodes and 1000uF capacitors. That will give just about 110Vdc for 40V or a bit more if the TX regulation allows. That can feed the heater chain, dc negative to chassis.

Next, bridge rectify the other winding using 1N4007s and say 100uF. This will give about 52V. Stack this dc on top of the 110Vdc to give about 160Vdc. This isn't too OTT to just feed into the rectifier anode circuit to provide HT for the set.

[G8UWM-MildMartin]

If you don't power the rectifier heater, you're close enough to to the transformer's output.

[Herald1360]

True, but you still need to voltage double one 40V and bridge the other to get the right level of HT, and the voltage double is best done with a non-common terminal version for lower ripple and better transformer balance, which means connecting one end of a winding to chassis isn't possible.

[Refugee]

The transformer is 140VA so there is no chance of overloading it. The windings are on two bobbins on opposite "legs" of the core with primaries that can be series or parallel as are the secondaries on a "C" core. The isolation between the windings and the rating are both a little over the top but i have no other use for it anyway. I can test the set with two variacs first and see exactly what voltage works best and if required fit a silicon regulator if anything is a little high especially on the HT. The transformer was removed from a big dot matrix printer and the regulation of its output is very good.