REDD 47 Power Supply

[Gearsix]

I am becoming to realize this preamplifier (p2p) and I'd like to know if someone can help me to find the schematic for a definitive 290V dc power supply without the tubes
I know that tubes contribute to the sound, but I prefer a traditional power supply (I have seen in a photo that the Revolution version of Redd 47 is without the tubes)
My project is to realize 4 channels so I need an adequate power supply
Thank you in advance for the help

 

[alexc]

For the HV part of the PSU, if using tubes or making your own pcb is not an option, I would do a turret board point-point implementation of the simple regulator circuit from the Orange86 schematic.

Charley's documentation for the Orange 86 psu is https://dl.dropbox.com/u/9818263/O86_Build%20Docs-v1.5.zip

It consists of a 4 stage R-C pi filter feeding a zener-diode-string-referenced series-pass-transistor.
Quite straight foward and easily scaled up to provide the required 120mA or so.

You could use a t03 metal can transistor on  a heatsink easily enough with wires to the turret board where the caps, resistors and zeners live. You get the regulated, loaded dc voltage you want by adjusting the pi-filter resistor values.

The phantom +48V power could be easiest done with a commercial pcb, for example, a JLM 3-rail PSU pcb.
It can also provide an additional pair of +/- rails for lights/relays etc.
There are probably others out there that can provide regulated +48V with or without doubling/tripling or additional rails.

In the case of the JLM, it accepts ac voltages from 15 to18V (voltage tripling) or 22 to 25V (voltage doubling), that would mean you need a  psu transformer something like this :

250vac to 275Vac at 150mA for HV
15 or 24Vac at 50mA for phantom  (doubling or tripling) or 50Vac for non-doubling-tripling


Heater power could be ac, so you would need 6.3Vac winding (either centre tapped for 'hum pot' or dc raising or non CT for the simplest case) at 4x 500ma-600mA (depending on tube choice) - so 2.5A minimum. Some extra current rating is a good idea. Say 3A or 4A if possible.

You could also do dc heater. Everyone has an  opinion on this. I prefer ac heaters. You might not.

If you don't want ac heaters :

For unregulated, no problem. You need a good sized rectifier bridge and around 25000uF, 25V of capacitance for smoothing and probably a  25W 0.5R or so dropping resistor. (actual value determined in the circuit using a variable rheostat to find the correct value)

Regulated heater is another issue again - probably a commercial pcb would be best. And a big heat sink. Has the advantage of not needing to use a rheostat to determine the correct dropping resistor.

You could also do a turret board thing similar to the hv regulator. But I wouldn't bother.

So there it is for a roll-your-own psu.

Find a transformer to suit.

I can't think of any single toroids that can supply the required 120mA. 100ma yes - they are available. That could be used if you adjust the plate resistor of the driver tube some to lower the current to a total of 25max mA per preamp.

You could use 2x toroids - people do it all the time for the G9 inside the chassis. No problem especially for stand-alone psu.

There are some on ebay that would suit. For example http://www.ebay.com/itm/Toroid-Transformer-T30-55W-for-Tube-Amplifier-Amp-/280416409885?pt=US_Home_Audio_Amplifiers_Preamps&hash=item414a1ed91d

For e-core style, Edcor would probably have something to suit.

You would also need some connectors and cable suitable for carrying quite a decent amount of power.
2.5A of LV dc and 120mA of HV dc.

If you like to do this sort of thing, you could end up with a really great unit at a fraction of the cost of something commercially sourced.

Good luck with it.

 

[Gearsix]

Thank you, the schematic of the Orange 86 PSU seems really easy to realize
The only thing I don't understand is why it uses a double secondary instead of only one
What's the difference if I use a single output 275 Volt secondary?
There is a reason because it uses two different types of diodes in the bridge?

 

[alexc]

The 'double hv' is an additional 'hv expansion' connector - you can ignore it.

The different diodes is a typo. Should be same type. 1N4007 is fine.

Yes - it is quite simple, and it works very well. And it doesn't require a centre-tapped HV.

 

[Gearsix]

So, for 4 channels I need a power transformer with a secondary of 275/280 Volt (200mA) and a 6,3 Volt C.T. (4A), is it right?
Usually in low voltage PSU I use to put "around" the diodes bridge 10nF polyester capacitors (as in the circuit attached).
Do you think it may be useful in this circuit, too?

 

[ruffrecords]

So, for 4 channels I need a power transformer with a secondary of 275/280 Volt (200mA) and a 6,3 Volt C.T. (4A), is it right?
Usually in low voltage PSU I use to put "around" the diodes bridge 10nF polyester capacitors (as in the circuit attached).
Do you think it may be useful in this circuit, too?

Yes. I have had direct experience of unsuppressed diodes creating hash in the HT line of a mic pre sufficiently large to be measurable at the output. I have not used caps across each diode but a single 100nF cap across the ac input to the bridge and sited right next to the bridge. That seems to do the trick for me. Attached is a not very good scope picture of the noise itself

Cheers

Ian

 

[alexc]

I would say 200mA is a bit too high - 120mA to 175mA abosulte max.
The size of the transformer will jump quite a bit going to 200mA. Not really a good thing if it can be avoided.

So my recommend is for 270Vac (non CT) @150mA, 6.3Vac 4A  (CT or non CT)
and a winding suitable for your 48V pcb choice.
Could be 15Vac-18Vac or 22Vac-25Vac (double/tripling) or 50Vac (non double tripling)
Doesn't hurt to have an extra rail for utility functions like leds, relays, meter buffers etc.

----------------

Interesting - I've not used the caps on diodes, but I will for my current pair.

Thanks!


Also, to clarify a comment I made earlier - the regulated dc voltage is basically set by the zener diode string.

The adjustment of the pi-filter resistors is to ensure sufficient voltage under load is provided to the series pass transistor to allow regulation.

ie. the voltages will drop under load, so you need to adjust the resistor  values you select so as to make sure you allow enough voltage into the transistor so it regulates, but not so much so it must dissipate too much power.

 

[alexc]

While on the subject of clean psu, may I ask Ian, if you have seen benefits to using an IEC mains filter?

I've used them on my tube amps, good ones I believe, and certainly ended up with good, quiet psu rails.
But I didn't compare to the case without them so I can't say they were in particular, beneficial.

Just wondering if it's worth using them on my current Orange86 build.
They mostly filter out high-freq hash I think?

Also, if you have any comment regarding the usefulness of cheaper ones over more expensive ones.
(they can go quite high in price!)

Cheers

 

[Gearsix]

Thank you Alexc and Ian
Just another question. Have you compared the 5:1 output Edcor transformers with the 7:1 Sowter?
I have used Edcor transformers in different projects and I have found them really good

 

[alexc]

I've not used the 7:1 sowter. I have compared the 5:1 Edcor and 5:1 sowter in an la2a. I found both to be fine.
Each had extended frequency response.

I think I can hear something I can best describe with the word 'graininess' in the edcor.
A bit like a solid state amp from the 70s straining with high levels

The Sowter didn;t exhibit this to my ears.

I would say the Sowter is 'smoother' or less noticable distortion at higher levels.

But really, not very much in it.

In fact, I prefer the Edcor in applications where I am trying for more subtle coloration.

I do think 7:1 is too much for the Redd47 circuit. I prefer 4:1 or 3:1 even!

However, your milage may well vary    They are all good options.

 

[Fablab]

Another output transformer may be the Carnhill VTB 2291?
While for the input may be interesting to know how is the EA2622 from Classic Audio