Hello there!
I have assembled two capacitance multipliers and with BD139+TIP3055 it is doing its job. No hum No buzz no etc. etc. I made only the positive side coz my amps are output cap coupled single supply type.
LINK->Capacitance Multiplier Power Supply Filter
Life is good with these multipliers. But the time of change has come and I'm confused.
Two days ago I discovered One shop here in Kolkata surprising me with "Yes" when I asked "Do you have these transistors?". 😀 The list contained about 10 different FAST power BJTs specially made for audio applications. ft>=20MHz.
I bought two 2SC5200s, costing about 100 rupees (~$2) per device. Fake? Don't know.
I have all the main amplifier circuits finished, tested and ready to be put inside the cabinet. So the only place I could think of placing these two monsters are the multipliers.
Under the above circumstances I have some confusions and questions about the influences of the devices used in the multiplier on the sound quality and bandwidth of source follower type single ended class-a amps.
1. I heard that the output signal chain runs all through the power supply for these kind of amps. Is this true and can someone please explain how?
2. Does a large value( >1000uF) capacitor at the output of the multiplier electrically hides (bypasses) the multiplying devices from the amp? If yes, then does it help with a varying current class-a amp like these, and how much?
3. Does putting a fast gain device like 2SC5200 in the multiplier make it sound better than with TIP3055?
4. Is there any chance of oscillation in the multiplier when made with fast transistors?
5. Can the multiplier power devices be paralleled without an emitter resistor?
Thanks in advance.
I have assembled two capacitance multipliers and with BD139+TIP3055 it is doing its job. No hum No buzz no etc. etc. I made only the positive side coz my amps are output cap coupled single supply type.
LINK->Capacitance Multiplier Power Supply Filter
Life is good with these multipliers. But the time of change has come and I'm confused.
Two days ago I discovered One shop here in Kolkata surprising me with "Yes" when I asked "Do you have these transistors?". 😀 The list contained about 10 different FAST power BJTs specially made for audio applications. ft>=20MHz.
I bought two 2SC5200s, costing about 100 rupees (~$2) per device. Fake? Don't know.
I have all the main amplifier circuits finished, tested and ready to be put inside the cabinet. So the only place I could think of placing these two monsters are the multipliers.
Under the above circumstances I have some confusions and questions about the influences of the devices used in the multiplier on the sound quality and bandwidth of source follower type single ended class-a amps.
1. I heard that the output signal chain runs all through the power supply for these kind of amps. Is this true and can someone please explain how?
2. Does a large value( >1000uF) capacitor at the output of the multiplier electrically hides (bypasses) the multiplying devices from the amp? If yes, then does it help with a varying current class-a amp like these, and how much?
3. Does putting a fast gain device like 2SC5200 in the multiplier make it sound better than with TIP3055?
4. Is there any chance of oscillation in the multiplier when made with fast transistors?
5. Can the multiplier power devices be paralleled without an emitter resistor?
Thanks in advance.
each of the paralleled devices needs it's own emitter resistor.
The only exception could be a Lateral MOSfet.
Hi shaan, are we talking about a charge pump here?
IE are you charging one capacitor from one rail & then switching so that the capacitor is now on top of the original rail voltage & thus nearly doubling it?
No we aren't. Sorry, just took a look at the link 😀
3. I very much doubt it as the base will be tied to a rather large capacitor which is deliberately designed to slow things down, hence the title "capacitance multiplier"
4. Again i doubt it'll oscillate simply because of 3.
5. As AndrewT says, you'll only get away with paralleling devices that have a negative temperature coefficient without an emitter resistor.
IE are you charging one capacitor from one rail & then switching so that the capacitor is now on top of the original rail voltage & thus nearly doubling it?
No we aren't. Sorry, just took a look at the link 😀
3. I very much doubt it as the base will be tied to a rather large capacitor which is deliberately designed to slow things down, hence the title "capacitance multiplier"
4. Again i doubt it'll oscillate simply because of 3.
5. As AndrewT says, you'll only get away with paralleling devices that have a negative temperature coefficient without an emitter resistor.
Last edited:
Member
Joined 2009
Paid Member
In most amplifiers the signal goes through the power supply chain (the capacitors). But it depends on the exact topology of the amplifier. People report better sound when they use large capacitors and better quality capacitors in the PSU because both reduce capacitor distortion.
There are topologies that offer the potential to keep the signal out of the power supply. One of them is a single ended amplifier with a constant current source load. i.e. it is NOT a push-pull Class A amplifier (or an Aleph current source). The output capacitor and speaker are in series and connected across the active output device of the amplifier. The signal current then theoretically doesn't flow through the PSU, which sees only the constant current source load current. You can do something similar with push-pull using a virtual earth. In this case you can more or less keep the signal out of the Capacitance Multiplier, but you need large value low ESR capacitor on the output of the amplifier to pull this trick off (as you suggest in point 2.) and wiring/layout is also important.
I've no idea, it would be good to experiment. TIP3055 is a modern device, not like the old 2N3055. I'm not sure a Cap multiplier takes advantage of the qualities of the 2SC5200 either.
- Status
- Not open for further replies.