davidsrsb
pfm Member
There will always be capacitors near the outputs or you will never get stability.
If they are the main reservoirs, you get the 100 Hz charging pulses causing a buzz
Interesting New Amplifier from Thomas and Stereo. Large capacitor banks good or bad?
- Thread starter OzBrit Audio
- Start date
If they are the main reservoirs, you get the 100 Hz charging pulses causing a buzz
martin clark
pinko bodger
As @davidsrsb said; and - Steve McCormack really knew what he was about; it was distributed bypass capacitances - suited to the local task, and, decoupled: NOT - a pile of smallish caps, in place of PSU reservoir.
' How would you rather plough a field? With two Bulls, or a thousand chickens in harness..?'
I mean - any decent design takes care of local bypass&decouple, as-required.
.. so I'll take the two-bulls approach, everytime.
I am reminded Scott McNeally's (of Sun Microsystems) rhetorical question -
' How would you rather plough a field? With two Bulls, or a thousand chickens in harness..?'
I mean - any decent design takes care of local bypass&decouple, as-required.
.. so I'll take the two-bulls approach, everytime.
Last edited:
OzBrit Audio
pfm Member
Well interesting discussion here
I started with observing the Thomas and Stereo amplifier recently released that had 200000uf Capacitor banks. This has been very well reviewed, particularly in the bass region
Feedback seems split between a smaller singular capacitor bank and much larger microfarad multiple capacitor banks. The chap who makes banks of multi farad caps to with a wire and plug to add to a range of existing amps made me laugh I have to admit!
But seriously I don't think there is a right or wrong answer to this, the answer depends on the implementation. The thedesign of the power amplifiers. how they are wired, their PSRR and the overall amplifier earthing arrangements all make a difference. I do think the caps as close as possible to output stages has to be a good thing and I have yet to hear an amplifier that doesn't benefit from a regulated front end for the sensitive input stages.
But at the end of the day it's all about the sound. If it sounds good well thats all that matters!
Throwing in my ten peneth for what it's worth. I have two amplifier implementations here which I regularly swap boards between. The first is my main box which is dual mono, Avondale minicap6 boards (6 x 10000uf cap with Schottky rectifiers) and a 600VA twin secondaries torroid from Canterbury Windings. The minicap6's are located close to the NAP boards so wiring is kept nice and short
The second is my test rig, that has a 800VA, normal 35A bridges with four secondaries, two for the grunt and two for regulated front ends. Caps are all singular with 22000uf for the power and 10000uf for the regulators.
Here switching NAP sized boards such as the NCC220 and SE200 the minicap6 (multiple caps) sounds consistently better over the large single cap solution of my test rig. The sound has better nuance and probably more get up and go when installed with the minicap6.
Thats another vote for the multiple cap solution I think
At the moment I am listening to the SE400 boards which I am currently bedding in (on my test rig). They do sound really good, lots of rhythmic drive, bass goes deep and drives the music along nicely. I do intend to update the single caps with boards of 2 x 10000uf with Schottky rectifiers and shorter wiring. I'll see what difference it makes and if I prefer the sound. So same cap size just a different implementation.
I started with observing the Thomas and Stereo amplifier recently released that had 200000uf Capacitor banks. This has been very well reviewed, particularly in the bass region
Feedback seems split between a smaller singular capacitor bank and much larger microfarad multiple capacitor banks. The chap who makes banks of multi farad caps to with a wire and plug to add to a range of existing amps made me laugh I have to admit!
But seriously I don't think there is a right or wrong answer to this, the answer depends on the implementation. The thedesign of the power amplifiers. how they are wired, their PSRR and the overall amplifier earthing arrangements all make a difference. I do think the caps as close as possible to output stages has to be a good thing and I have yet to hear an amplifier that doesn't benefit from a regulated front end for the sensitive input stages.
But at the end of the day it's all about the sound. If it sounds good well thats all that matters!
Throwing in my ten peneth for what it's worth. I have two amplifier implementations here which I regularly swap boards between. The first is my main box which is dual mono, Avondale minicap6 boards (6 x 10000uf cap with Schottky rectifiers) and a 600VA twin secondaries torroid from Canterbury Windings. The minicap6's are located close to the NAP boards so wiring is kept nice and short
The second is my test rig, that has a 800VA, normal 35A bridges with four secondaries, two for the grunt and two for regulated front ends. Caps are all singular with 22000uf for the power and 10000uf for the regulators.
Here switching NAP sized boards such as the NCC220 and SE200 the minicap6 (multiple caps) sounds consistently better over the large single cap solution of my test rig. The sound has better nuance and probably more get up and go when installed with the minicap6.
Thats another vote for the multiple cap solution I think
At the moment I am listening to the SE400 boards which I am currently bedding in (on my test rig). They do sound really good, lots of rhythmic drive, bass goes deep and drives the music along nicely. I do intend to update the single caps with boards of 2 x 10000uf with Schottky rectifiers and shorter wiring. I'll see what difference it makes and if I prefer the sound. So same cap size just a different implementation.
martin clark
pinko bodger
I don't disagree.
Note though, the Minicap6 has small -value inductors, c 10uH, in series so the input from the rectifier is filtered C-L-C-L-C.
This knocks a large amount of HF noise off the raw supply from any sources, rectification or the mains ac input. That helps a lot, because PSRR falls-off with increasing frequency. If [for a given amplifier design] PSRR is not-so-great in the audio band - of which, the 0-10Khz bit is the most important by far - then the PSRR becomes truly-terrible out at >80Khz and above [because the amplifier open-loop gain is falling away, so reducing available feedback to correct for this]: and that potentially leaves the door wide open to HF inter-modulation problems; which can be manifestly -audible.
That - IMO - is very likely the reason for the observable difference: 'big cap' vs 'minicap board' is not a like-for-like case of 'One large vs three modest caps' at all.
ETA 16.03.2024: small edits in brackets, to make the sense clearer.
Note though, the Minicap6 has small -value inductors, c 10uH, in series so the input from the rectifier is filtered C-L-C-L-C.
This knocks a large amount of HF noise off the raw supply from any sources, rectification or the mains ac input. That helps a lot, because PSRR falls-off with increasing frequency. If [for a given amplifier design] PSRR is not-so-great in the audio band - of which, the 0-10Khz bit is the most important by far - then the PSRR becomes truly-terrible out at >80Khz and above [because the amplifier open-loop gain is falling away, so reducing available feedback to correct for this]: and that potentially leaves the door wide open to HF inter-modulation problems; which can be manifestly -audible.
That - IMO - is very likely the reason for the observable difference: 'big cap' vs 'minicap board' is not a like-for-like case of 'One large vs three modest caps' at all.
ETA 16.03.2024: small edits in brackets, to make the sense clearer.
Last edited:
OzBrit Audio
pfm Member
Hi Martin
I don't disagree with any of the above regarding HF noise.
But to make things a little more complicated my minicap6's don't have the inductors. Originally I had them in but in the end took them out as I felt there was in impact on dynamics with them in.
I think there was a thread on here where people voted their preferences for with or without. Quite a few said they felt their amplifiers sounded better without. For the low current front end power supplies I think the consensus is they are definitely a good idea, for the back end grunt not so.
I think the boards I decided were better without inductors were the SE200's. The Minicaps do have Schottky rectifiers over the 35A bridges on my test rig. Also 30000uf, with 3 caps a side on the minicap versus 22000uf single cap on test rig. Much shorter wiring with the minicaps too.
Transformers are also different Canterbury windings for the minicap, Tiger toroid for the my test rig. Whether that makes much difference to SQ difficult to tell.