Battery eliminators

[john_jmec]

Having recently got some old Pye radios with 2 volt heaters and a range of additional supplies required I'm considering making up a small PCB switching regulator to generate 1.4/2 volts up to an Amp, 60/90/120v and the grid bias outputs. Any interest in this if I make the PCB available, and what sort of voltage ranges (and simulatanous values) would people want? The whole thing could run off a 6 or 12 volt lantern battery or rechargeables or car battery or a wall socket PSU. It would need properly screening to suppress the interference.

Or is there something already does this - I didn't turn anything up on a brief search, except mains ones.

Apologies if I'm repeating stuff already in existence

John

[Ed_Dinning]

Hi John, I've produced a few of these based on old video mains transformers, these have lots of windings that can be seriesed up to give the HT (and voltage doubled if requred), as well as LT windings that will supply the 2 or 4V for the heaters through a linear regulator.
There have also been designs published on this forum and you may find some info in the "LINKS".

As a professional Engineer who has done a lot of SMPS work, I would say not to go that way as the interference and hash problems are horrendous. Some sets tolerate this better than others.

Ask yourself if you really need these sets to be portable; if so use PP3's and a cyclon or similar battery. If not, make up a dummy battery case (see what Battery Maker has done) and fit a mains derived PSU in there.
The LT supply needs to be well smoothed, filtered and stabilised. The HT supply only needs good smoothing and filtering. The GB, if required can be made fron individual cells as there is no drain.

PM me with you e-mail if you want a copy of my circuit.

Ed

[jay_oldstuff]

I can reccomend Ed's sugestion of useing an old VCR transformer i built my first eliminator based on his circuit and it works very well useing a LM317 to regulate the LT. it gives very little if any interferance unlike some other eliminators i've tryed.
I will admit though i have just built a replica 120v HT battery filled with pp3's for that total authentic look.

Jay

[georgesgiralt]

Hi John !
I've played a little with this subject.
Getting 90V HT with little or no interference is not so easy as it seems. (if you also want high efficiency...)
The most efficient design I've got is using a Kosbo.com inverter (PWM using IRC a Maxim chip) but it is unsuitable as it radiate in a large spread of frequencies all in the receiving band. Unshieldable...
At the other end, I've got a German design (published in Elektor) using a 50 Hz square wave oscillator driving a plain mains transformer reversed. No interference but a very low running time...
The one I use is Dr Jirmann's design (sine osc driving a home made tranny (RM8 core), self starting by current sensing and built in interference suppression. Compact design, transistor based good efficiency easy to shield. I've joined the schematics on this forum, search for Jirmann should give a link to it.
If you plan to design yours, use standard cores and wind it yourself. I've done it by hand, easy and not so time consuming (secondary use 120 turns IRC, in one half of an hour you've a working transformer... Re-using old parts is subject to finding one....
As far as the LT, if you use a NiMh or Nicad cell, (today you've D size NiMh up to 8.5 Ah) you do not need regulation at all. The LM317 will eat some power best kept for the valves. (of course if you plan to use 2V heaters, .....)
Hope this helps.

[kalee20]

I'm doing exactly what john jmec has in mind. I have an eliminator design which provides HT, LT (regulated) and GB, from a 6V input. Output voltages are stable until the input drops below 3.8V. Overall efficiency is 80%, for a prototype optimised for 90V 9mA, 1.4V 125mA, and -7V. I've just re-vamped my PCB layout, the first units will be built up Feb / March. It's an enhancement of this.

Biggest challenge is interference - I have copious amounts of filtering on input and output, and the prototype is mounted in a diecast box. With this, it operates a Vidor attache case portable with no squeals.

My own needs are as above; also one at 90V 15mA, 1.4V 250mA, and -7V; and another at 120V 15mA, 2V 600mA, and about -8V. HT regulation could be an issue as one of the radios has a QPP output stage. But, as the LT supply is derived from a low-voltage winding on the same transformer as the HT and GB outputs, and the LT output is sensed and used to control the switching stage, I've got my fingers crossed that the voltages will track well enough not to be a problem.

[john_jmec]

Thanks all - I had a look at the screened inverter - copper is great for suppressing electric fileds, but not so good for magnetic fields, which of course is what a frame aerial (or ferrite rod) picks up. This could mean that if the inverter was correctly orientated there might be a null in the intereference - but this would probably be different for the harmonics.
I had in mend something to supply the heaters and grid bias using an ADP2105 (see http://www.analog.com/static/importe..._2106_2107.pdf ) and the whole thing including some additional supression chokes (all surface mount) could fit on a DS PCB about the size of two postage stamps.
It switches at about 1 MHz - basically your burying a 1MHz 2 watt transmitter in your radio, so how screenable it is should be fun. I don't
really want to succumb to LF mains transformers.
I'll make one up over the next few weeks and give it a try, and in the meantime use lots of lab power supplies and the attached instant grot Grid Bias battery.

J