AjohnL, keep in mind that a capacitor multiplier does not replace the energy stored in high value capacitors, the instant energy has to come from somewhere upstream. You can only see it as a voltage drop with higher time constant. The capacitance is not really amplified by the gain of the transistor, although it might seem so. On a class AB amplifier, the capacitance multiplier has virtually zero effect. It is useful in class A when the current is virtually stable it will reduce PSU ripple. While used on an AB amp it may reduce hum at very low volume, but the amp sound will typically fall apart when high power peaks are a sudden requirement. I would suggest using a well selected reservoir cap that can produce the peak current, followed by a capacitor multiplier to smooth out the ripple.
Last edited:
I'm not convinced about the use of capacitance multipliers. Connecting a capacitor to the base of a pass transistor could, under fault conditions, dump its charge into the transistor, which would then most likely pass a huge current which might destroy it.
I agree with Nico too that it needs a current to establish a useful working state, but for reliability I'd recommend good old large capacitors.
I agree with Nico too that it needs a current to establish a useful working state, but for reliability I'd recommend good old large capacitors.
Capacitance multipler do not choke, sieve. They do smooth only.
And: A higher capacity, afterwards, will again be audible: The e.g. different internal resistances of multipler and high capacitance is obviously perceptible differently.
And: The choice of components is also elementary for good sound with active power supplies. The influence of different e.g. transistors is, my experience, even much greater than that of different coils of similar size.
And: A higher capacity, afterwards, will again be audible: The e.g. different internal resistances of multipler and high capacitance is obviously perceptible differently.
And: The choice of components is also elementary for good sound with active power supplies. The influence of different e.g. transistors is, my experience, even much greater than that of different coils of similar size.
I know that actually as have worked in electronic areas for rather a long time.
What I don't know though is how clean those laptop supplies are. I noticed that they may be useful when they started to appear. Actually people around me were working in improving them when 20khz ones were found to not be as silent as suggested as some can hear that pretty easily. I'm also aware people may disregard new ideas for rather a long time.My impression of hood's amp when a number of us knocked one up in the lab was the same as yours.
Can be used to power an amplifier (excluding the output transistor).
The top transistor of the output stage will act as a filter.
Attachments
Rather than listening to any emotional argument, take one SMPS load it and look at it, and like any good engineer/technician fiddle with the output filter until you have the results you want. This is a hobby, not a production development. You are doing it for yourself and your enjoyment. Take the plunge, just do it, I bet that any high speed computing device would be more susceptible to rubbish on the power supply than our ears that cuts off at 20 kHz. Years ago I had problems with writing hash into E2 just because of noise generated on the battery supply in a single board computer. If you don't satisfy your own curiosity, then asking is not going to get you anywhere.
Last edited:
John_ellis, I would agree with your statement and yes you see many amplifiers that decouples the front stages from the output and sometimes even have a higher voltage on the front than the output since these are very sensitive to any voltage ripple. The output normally has a gain of less than 1 would you hear the 70mV ripple modulating the output running from a simple capacitive filter, I think not.
cnbb, if you set up the power supply as suggested, then there are two obvious problems. The 1st is that you will never swing the output high enough to get the designed power from it, the 2nd is that any modulation on the output section will affect the driver section adversely. Unless you intend to use two separate power supplies, which is not a bad idea. Therefore, the frontend should have the higher voltage so even if the power stage drags the supply down, it would not cause the front to run into clipping. I hope it makes sense.
Post 8.966 I want to recognize as a reference to the feasibility of "filtering" - with limits, compared with coils - by feedback. By "negative feedback".
Experience shows that orientation on "power" - determined by distortion - is not a good idea to develop a clean and dynamic sounding amplifier.
I would never build a JLH: too complex. And three gain-stages exceeds the limit I set: two stages. An original JLH is sonically a disco amplifier. Two stages sound like party. One stage is hi-fi, high end.
Experience shows that orientation on "power" - determined by distortion - is not a good idea to develop a clean and dynamic sounding amplifier.
I would never build a JLH: too complex. And three gain-stages exceeds the limit I set: two stages. An original JLH is sonically a disco amplifier. Two stages sound like party. One stage is hi-fi, high end.
What I have done many times with amp power supplies is the exact same as mentioned above, but I always used a toroidal transformer since it is very easy to add a couple of windings (which is thin since it is low current) to generate the higher voltage for the front and driver end, Usually about 10V, in other words create a 10V tap higher than the power section supply. You will notice a marked difference in amplifier distortion as a bonus.
Maybe become a PASS fan with a single stage amplifier, there is no reason not too. I am led to believe they sound great!
I would use the e.g. 46 VDC and adjust the resistors accordingly.
May be the better sound by "create a 10V tap higher than the power section supply" is result of more noise due further parts such as diodes and capacitors and connections. It is usually necessary to counter the flat and woody character of speakers. E.g.
May be the better sound by "create a 10V tap higher than the power section supply" is result of more noise due further parts such as diodes and capacitors and connections. It is usually necessary to counter the flat and woody character of speakers. E.g.
That is why Hood suggested it but obviosly there will be a voltage drop.
It's not my fault that some one called them capacitance multipliers. How they work is pretty obvious really. The load on the smaller cap is just that needed by the darlington arrangement, For this use it's a pretty sensible option. The alternative would be a regulator,