Unbelieve power caps for Aleph 3

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Adding some extra capacitance will always improve the amplifiers performance. 8 x 22000uF seems enough to me since I've placed 'just' 2 x 22000uF and this works perfectly.
But be aware of the high inrush current, not every rectifier is pleased to feed a dead short (decharged capacitor)
8x22000uF can be to high


I am not quite sure if this is really better. By parallelling 4 caps, you get a leakcurrent that is for times greater. This will mean that more noise is generated by the caps. Such a large capacitance has to be used IMO with a softstartcircuit, because otherwise not only your rectifiers will suffer, but also the transfo. This can lead to a transdo that resonates.

Of course there are also great benefits. The impedance lowers, much more stable regulation (the voltage almost doesn't drop) and a lower resistance....

What I want to say is that experimentation is really needed to see what is best. If the amp doesn't draw alot of current (like a class AB amp with high efficiencyspeaker) then it is really a disadvantage to use such a large reservoir. On the other hand is you use it as a subwoofer amplifier, the caps will really tighten the bass.

An interisting link to read about advantages/disadvantages of caps can be found here :http://www.capacitors.com/pickcap/pickcap.htm

Hope this info helps


I wonder HOW true it is in an amplifier that draws only slightly more at full output than it does at idle. The power follower 99's make use of a relatively small main filter cap as it has a steady state 3 amp draw running class A up to its rated output. Doesn't the Aleph series basically work the same way? Why would more capacitance improve the sound quality? We're really not relying on peak current available as in a class AB or sliding bias scheme are we?
Mark Gulbrandsen
Salt Lake City, UT
The Aleph amplifier circuit current draw is not steady; it varies with the signal. That's how it's able to be more efficient than other class A designs. Any time the current is varying, you've got a good possibility that augmenting the power supply capacitance will help.
Now, the thing I still haven't figured out is why addition of film cap bypasses to the main power supply doesn't make that big a difference to the sound quality of the Alephs. Even a relatively small film bypass on the Thresholds was immediately audible in the higher frequencies. My tube amps use a large bank of film caps for the second leg of a PI filter stage to good effect. Every circuit I've ever added film caps to sounded better...except the Aleph power supply. No change.

"Every circuit I've ever added film caps to sounded better...except the Aleph power supply. No change. " That's very strange.Power supply bypass caps are my 'secret weapon'.Most people are stunned and amazed at what $4 worth of caps added to the power supply will do.I have only seen three amps EVER that came from the factory with bypass caps.Am also curious about the Aleph.
The one and only
Joined 2001
Paid Member
Of course if you really want to go to town on
the power supply, build a "pi" filter with
two banks of caps and an inductor in between. This
takes out the ripple and improves the noise floor.

You can get pretty crazy with the amount of capacitance
since the Thermistor in the power supply holds the
inrush to a dull roar.

As far as constant draw for the Alephs go, you pretty
much have to be into clipping or driving a low impedance
load for the line draw to go up significantly, and the
amps were not designed with either in mind....
GRollins is wrong about the Aleph's current draw: IT MUST BE STEADY! I completely agree with Nelson Pass, the current is not supposed to change a bit until the amplifier clips.
And when the current changes with the input signal it results in an awful sound, I can't imagine GRollins hasn't experienced that.
The one and only
Joined 2001
Paid Member
So that I'm clear here, let me say that the
Aleph's power draw looks constant if you hook
up an amp meter to the AC power line. In this
case, the draw is pretty constant until you get
up to the linear limits of the amp.

This does not mean, however that the draw is constant
on an instantaneous basis. This is, after all an AC
circuit. The power supply caps will see the charging
ripple current, and also the the output current to the
speaker. It averages out as a constant.

Part of this is the Aleph's current source, which acts
as a negative impedance equal to twice the load impedance,
meaning that the positive supply sees 1/2 the AC current
to the load as a Class A circuit.

The negative supply sees signal current reflecting the
full amount of the current through the load, twice that
of the current source for AC, but again, this looks
constant averaged over 100 milliseconds or so.

The upshot is that the Aleph's are not very sensitive to
AC line impedance and tend to have a constant load on the
transformer and the rest of the power circuit, but in fact
they do benefit from more capacitance and filtering, at
least in reducing ripple and signal modulation on the
film capacitors across electrolytics in aleph power supplies

No difference with film capacitor in aleph power supplies?
I have different experiences with film capacitors in aleph power supplies.
They improve the sound as much as in any other amplifier. It is even easy to here the differences between the types of filmcaps like Paper in oil, MkP, MKS, or whatever i tried.
Does someone make the same experience?
Aleph supply caps

Absolutely! The Aleph circuit responds to every parts upgrade very audibly in my setup. I just put 100uF ELNA Cerafine bypass in my Aleph with improments fron the bass to the very top end. I used four and as short a leads from the output and current souce boards the the central power supply board ground as possible.

About bypassing electrolitics with plastics i d'ont share the opinion
that seems generalisated in this forum...But if all the people think the some there are no need of discussion...:)
The idea of putting one plastic capacitor across a electrolitc is to by pass his indutive behavior in the highs frequenties...ok...so far so good...
But when we put a capacitance across a indutance we got a ressonante circuit...i have made some experiencies...and it is what hapen...some frequencies became exagerated.
In the sound it seems like a time smear and a sensation of the timing becoming slower...
Thats my opinion...sorry y knew that it is not "politicsly correct" but i have to tell what i think even when is not the some as others...


Political Bypass

Hi Tube_Dude,
I agree that adding a film cap can cause a power supply to become resonant, and indeed I have heard amps where this is detrimental.
I usually add an additional RC damping network across the rails to help avoid/reduce this condition.
Also the supplies need to be bypassed with several values of film caps to have proper good effect.

Regards, Eric.
Right on!

Jorge and Eric,

You guys nailed it . Too many people think that bypassing power supply electrolytics with films help.

In most cases they hurt the sound. A change is too often assumed that it is for the better. All that glitters is not gold!

If I need a bypass I use a smaller electrolytic.

I'm glad Mr feedback that you also have find that the supply become ressonant...now we are two!! ;)
And there is also the case that many people put plastics bypass
for every place...forgoten that the ground is also a signal path...
this is amazing hobbye because thera are very much hiden efects..
and decodificate all the interactions...is as dificult as to decodificate the "human genoma"....
Sorry for my English! ;(


Mrfeedback and Jam

Now we are 3...

This is the beggening of a revolution???? ;)

Yes the sindrome "is diferent is better"... is complicated!!

I have begun making amplifiers with 13 years old...now i'am 48
a long way...in the beguining i use scematics from the magazines...after i began designing myself some amplifiers.

I only began having good sound when i forgot all that i hear and read...and began making my ones experiences...

for instance...for me the interface betwen ampli and speakers is very important for the sound of the amplifier...we c'ant forget that
the Emf from the speaker via de feedback path to the input can make hard times for the input stage!

well we try arder!!!

Is a plesure to talk with people that share the some hobbye...

Some more food for thought.

My experience is that some changes based on sound techncial fact will help (but to create a change is not necessarily better).

But my recent experience suggests on advise from Mr Pass that parrelleling some smaller power identical capactiors will improve performance to near that of say some exotic types (cerefines etc).

I have 4 x 15,000 uf 63volt Samhwa brand in parrellel for +-60,000 for my stereo Aleph 5. The effect of film capacitors is less obvious with parrelled electroyletics for me.

Be careful though, I tried doubling the above value on one bridge and it squashed the sound literally so I figure the rectifier was hurting with 240,000 uf.

I then used independant bridge rectifiers for seperate left and right filter banks of +-60,000 from a common transformer and whammmmo, baby she bangs, baby she bangs!

The improvement in dynamics was sensational.

All this works much better than 44,000uf x 2 Sprague, there is vastly less audible hum, and it sounds cleaner and tighter.

Previously adding polypropylene films to the Spragues helped the top end, so I assume the Spragues were crap to start with.

The impact of decoupling capacitors on the boards is less obvious with the parelleled 60,000 bank also, as I removed the 100uf Cerefines last night.

But the results of various earthing schemes have an obvious effect on audible ripple and sound quality, particularly bass impact, imaging and dynamics. (using one common transformer)

My best result is as per text book of Doug Self and follow the Aleph schematic.

I use a star earth but run a short thick cable spur (say 50 - 75mm)from the centre tap/capacitor connection and use this as the star point for the board earth and speaker common.

The capacitors need to be connected via plenty of copper to avoid ripple. I used copper foil soldered to a pcb which works well.


This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.