Aleph 2 LC powersupply ...question..


2001-03-06 3:07 pm
i got hold of some pretty nice 1,1mH inductors (they way about 0,7kg each) and i want to use them in my ALeph 2 ps...witch "hook up" would be best?

Rectifier --- Inductor---Capacitor ---Aleph

Rectifier --- Capacitor --- inductor ---Aleph

would i gain much with?

rectifier --- capacitpr --- inductor --- capacitor ---- aleph

even though the first cap will be a smaller and of less quality than the later one?... (due to size)

Grey shuffles in the door wearing his best aw-shucks look...
(No one [okay, <i>almost</i> no one] thinks evil of a Southern boy when he's got on his 'aw-shucks' look.)
Actually, it's a teeny bit more complicated than that. Depends on what voltage/current you want out of your power supply.
#1 will give you less voltage--in theory about .9 x VAC--but more current capability
#2 will give you something on the order of 1.4 x VAC, but with a bit more smoothing (think of it as a filter for a speaker--12 dB/oct)
#3 will give you the same voltage as #2, but with even better smoothing (18 dB/oct)
Given those three options, and assuming that you can make the voltages come out the way you want them to, I vote for #3, same as Francois.

Aleph inductors

I think the first would be the best since the inductance between the rectifier and the capacitor will tend to smooth out the pulse of current that goes into the capacitor on AC peaks. I would NOT use the second one. The amplifier wants to have a low impedance power supply and having the inductor between the capacitor and the Aleph raises that impedance. If the inductors have an iron core you might consider


Iron can take a "set" if there is current going only one way through the coil. With the arrangement above you force the magnetic field in the core to reverse each cycle which might make the inductor a little more effective. This will not happen with an air core coil. I know this is a poor explaination, but I am not an engineer. Maybe an EE can bail me out. This idea originally appeared in Audio Amateur Four/87.
more on inductors

Grey got that post in while I was typing mine. In order to get the .9 X AC voltage you have to have a certain amount of inductance called a critical inductance. THAT is about the DC output voltage divided by the output current in milliamps. I don't think a millihenry is enough.

Aw-shucks Grey, you do a great job with this forum and I hate to be the Yankee boy who disagrees with you.
hi folks !!!

very interesting discussion !

i have a tube amplifier with hv unregulated.
what design do i have to consider ??

power t + inductor + diode +cap

power t + diode + inductor + cap

power t + diode + cap + inductor + cap

i want the best ripple rejection but i want a stabler current
since hv is not regulated.

thank you folks

It depends

The "optimal" solution depends on the application. Aleph pretty much draws a predictable current and so impedance of supply is not really important. I would expect number 2 below might be "more optimal" than some of the others.

TR - Ind - Cap - Aleph: "Best" if you have a small transformer as you will be using it more effectively

TR - small cap - ind - cap - Aleph: Kind of best of both worlds

TR - large cap - ind - large cap - Aleph: You don't really get to use transformer more effectively and you are still spiking current draw and attempt to filter it out later.

My favourite is:

-----small_cap --- ind --large cap --- ind large cap
-----small_cap --- ind --large cap --- ind large cap

You can have as many sections as you like, but 1 section should be enough. If you prefer, you can do the following:

Tr--small cap--ind--large cap--ind--large cap--aleph

All these scenarios are most easily simulated using PSUD2

Remember, to have maximum effect out of an inductor, you need to have ripple across it!

Could be anything you desire.

A very small one, less than 1 uF, preferably bypassed by a ceramic (!) would basically take spikes which is what we want. Watch voltage rating of this cap!

The larger you go, the less efficiently you use the transformer and the less strain you put on the inductor. The output voltage also increases towards the peak instantaneous voltage (1.41*Vtr RMS)

The smaller you go, the more voltage spike you get on it.

The cap needs to have very high current capability. Perhaps 10 foil caps in parallell would not be so bad (+ Ceramic bypass of course).

Try it out with PSUD2, it is most educational.

OOOPS! Henries

In my previous post I should have stated that the critical inductance in henries is approximately equal to the output voltage in volts divided by the minimum output current in milliamps. Sorry about that. The Aleph is (I think) a class A amp so the current draw on the power supply is constant. With class AB amps the current varies so you must pay attention to the minimum current to make sure that is enough to make the inductor "work". If you can meet the minumum current requirement for critical inductance from your inductor a choke input power supply will give you the best voltage regulation, but the output voltage will be 0.9 X VAC. It is possible to add a shunt regulator on the output of your power supply to be sure that you draw enough current to make the input choke "work". This will give you the very best voltage regulation, but in a way the shunt regulator wastes power since it does not contribute directly to music output. Hope this helps.
The .9 multiplier for the voltage isn't a sudden, either/or kind of thing. You don't increase current draw (or inductance) to the critical value, then suddenly get a 1.4 to .9 voltage drop. There'd be a gradual lessening of rail voltage until the .9 figure is approached. All in all, it'll need to worked out in conjuction with the transformer secondary voltage, current draw, etc. so that the voltage coming out of the power supply is somewhere in the vicinity of what you want. Unless, you want to regulate, of course, in which case you're free to have most any input voltage you want as long as your regulator is robust enough to take the voltage drop/current draw/heat dissipation. Maybe one of those capacitor multipliers people talk about from time to time?
The Alephs are class A, but the draw per rail varies. (The draw from the wall will be pretty steady, though.) The only way to get an absolutely steady draw would be with a current source (the Zen kind where it's constant--the Aleph current source varies). So there is some benefit to be derived from a steady, quiet rail. No, I haven't gone that far; my Aleph 2s just have big cap banks like the production units had. I have wire on hand, but haven't taken the time to wind chokes. Not enough hours in a day.
Now class AB amps...definitely a cool idea. Class B? They're prime territory for a tidy little setup like Petter had above with multiple filter stages.



2001-03-06 3:07 pm

this is the setup wich coil should i use?

900VA Toroid

1,1mH coil DCR 0,1ohm Rated at 10Amps or?
25mH coil DCR 0,4ohm Rated at 3 Amps
47000µF caps

Which coil shoudl i use? the first should give -26db suppression and the 25mh -57db.... but i havent taken the DCR into account...

Find out how much current you need


You need to find out how much current you will draw (depending on Aleph model).

Then you check if your inductor can handle the current ripple (depends on inductor -- if it is air-core, don't worry, if significantly air-gapped probably OK). It is very likely that the low valued device will handle it.

If your are unsure about the large value units ability to handle ripple, go with the smaller one, put some capacitance (for example 10.000 uF) behind it and use the second one as a second stage.

If you need more than 3 A steady state, you cannot use the large coil safely.

The coil's resistance is usually not all that important. It will assist with damping and suppression, but also make the supply less stiff. For a quasi Class A amp like the Aleph, it is unlikely to be a problem at these levels. For a Class AB amp, a high coil resistance will likely cause modulation of the power supply by the audio signal.

Watch the voltage rating of your rectifiers!

Place the inductors so that they don't interact (right angles and large distance is usually good, depending on topology). Also try to place them a fair distance from the audio circuitry if possible.