I face the same problem, and that's quite a dilemma by the way...
Rectifiers must be together with the caps. High AC fluctuations (I mean in the cable between the amp and PSU) produce more EM-interferences, than high DC fluctuations. DC heats up wires more than AC with the same rate... Then this star-ground part...
For now I've decided to keep my tranny (2500 VA, weighs 20KG) with the rest of the stuff. Just have to find a proper power cable
Rectifiers must be together with the caps. High AC fluctuations (I mean in the cable between the amp and PSU) produce more EM-interferences, than high DC fluctuations. DC heats up wires more than AC with the same rate... Then this star-ground part...
For now I've decided to keep my tranny (2500 VA, weighs 20KG) with the rest of the stuff. Just have to find a proper power cable
i dont know an answer to the question sorry, but whatever you do watch out for parsaitic inductance and capacitance in the long PSU lines it really can cause nasty oscillation (use the thickest wire you can, preferably stranded to avoid skinning).
Also check out snubbers on DIY audio if you do get oscillation
That is one -mutha- of a transformer where'd you get it and whats it going to power?
What current can it put out continuously?
Craig
Also check out snubbers on DIY audio if you do get oscillation
That is one -mutha- of a transformer where'd you get it and whats it going to power?
What current can it put out continuously?
Craig
The caps must be near the amp of the length of the leads will introduce too much inductance leading to instability and poor bass response.
The bridge must be near the caps because long wires will emit significant interference with the short charging pulses. Ideally the transformer should be close as well.
The bridge must be near the caps because long wires will emit significant interference with the short charging pulses. Ideally the transformer should be close as well.
I already know about the cable trick, and i think I'll use something like 3x ø 3mm cross-wound for +- and ground. I hope nothing will oscillate, cause the wire will only be about a meter long... If yes, I was going to use snubbers on DC-inlets anyhow
The transformer is custom-wound 2500 VA toroid with 2*50-60-70-80-90-100 V outlets. I'm intending to use it in a test stand for different 8-Ohm power amplifier projects. The thing is, that it can only supply about 13 Amperes without overheating. In future I'm planning to wind a 5000VA version for 4-Ohmic loads (who knows, maybe I'll use it for arc-welding as well).
The transformer is custom-wound 2500 VA toroid with 2*50-60-70-80-90-100 V outlets. I'm intending to use it in a test stand for different 8-Ohm power amplifier projects. The thing is, that it can only supply about 13 Amperes without overheating. In future I'm planning to wind a 5000VA version for 4-Ohmic loads (who knows, maybe I'll use it for arc-welding as well
).Hi,
If I were you I would recitify and use a few capacitors at the transformer end, then use more capacitors inside the amp to null any effects of the cable resistance/inductance. This way there is minimal AC on the umbilical cord, which means less chance to induce any noise into nearby circuitry (especially from the sharp current peaks from charging the capacitors at the top of each cycle).
Chris.
If I were you I would recitify and use a few capacitors at the transformer end, then use more capacitors inside the amp to null any effects of the cable resistance/inductance. This way there is minimal AC on the umbilical cord, which means less chance to induce any noise into nearby circuitry (especially from the sharp current peaks from charging the capacitors at the top of each cycle).
Chris.
Konnichiwa,
The answer was given elsewhere, but not quiet outlined well in terms of reasoning.
1) You MUST have a reasonable amount of capacitance near the actual amplifier circuit (how much depends on many issues).
2) You must keep rectifier and reservoir capacitor in close proximity, to keep the current loop tight.
This implies to place the rectification and the bulk capacitance either in the amplifier or in the supply box, but under all conditions a considerable amount of capacitance should also be in the amplifier chassis.
An obvious option is to use multiple reservoir capacitors with some extra choke filtering inbetween, where rectifier, input cap and filter chokes are in the PSU chassis and the amplifiers PSU capacitors in in the Amplfier chassis.
Sayonara
poolorpond said:
The answer was given elsewhere, but not quiet outlined well in terms of reasoning.
1) You MUST have a reasonable amount of capacitance near the actual amplifier circuit (how much depends on many issues).
2) You must keep rectifier and reservoir capacitor in close proximity, to keep the current loop tight.
This implies to place the rectification and the bulk capacitance either in the amplifier or in the supply box, but under all conditions a considerable amount of capacitance should also be in the amplifier chassis.
An obvious option is to use multiple reservoir capacitors with some extra choke filtering inbetween, where rectifier, input cap and filter chokes are in the PSU chassis and the amplifiers PSU capacitors in in the Amplfier chassis.
Sayonara
Looks like the best option up to now is to use 2 capacitor banks (or to build a 10U rack to fit everything 
).
The fact is that I've successfully tested a power amplifier design by my friend (a 300W sub amp) using a ~1m cable between the PSU and the DC inlets without any parasitic oscillations, although there were only two 4700uF caps on the PCB (and my PSU has 22000uF ones + 2 ultrafast bridges).
Anyway, I have the time to try all the available options and choose the best performing one. Life is hard without a proper test stand...
).The fact is that I've successfully tested a power amplifier design by my friend (a 300W sub amp) using a ~1m cable between the PSU and the DC inlets without any parasitic oscillations, although there were only two 4700uF caps on the PCB (and my PSU has 22000uF ones + 2 ultrafast bridges).
Anyway, I have the time to try all the available options and choose the best performing one. Life is hard without a proper test stand...
Pre amplifier external supply
Hi,
and what is your opinion in an external PSU for a preamplifier, that uses a regulated PSU? Still keeping trasformer-rectifier-cap in one enclosure, and more cap-regulated circuit in the pre chassis? Assume that the cable power between the 2 enclosures is very short, 20cm.
Regards.
Claudio
Hi,
and what is your opinion in an external PSU for a preamplifier, that uses a regulated PSU? Still keeping trasformer-rectifier-cap in one enclosure, and more cap-regulated circuit in the pre chassis? Assume that the cable power between the 2 enclosures is very short, 20cm.
Regards.
Claudio
How to completely negate cost of an external power supply?
A clarifying question:
Say you have an external power supply, a CRLC filter with about
100mF -> 200mOhm -> 100mF -> 2mH -> 100mF
There is a fat cable 50cm or less between the supply and the amp.
If I've understood this correctly one wants to have extra capacitors at the amplifier power input to negate having longer cables.
Now the question, how much capacitance would be ideal to completely negate the cost of having an external power supply?
Would it be enough to put a say a 33mF cap at the amp input or does one want more / less? A bigger cap followed by smaller caps or the other way around? Is it enough to only have additional caps at the amp or does one want something more / different?
Spread thy knowledge!
Thanks in advance,
Olle
A clarifying question:
Say you have an external power supply, a CRLC filter with about
100mF -> 200mOhm -> 100mF -> 2mH -> 100mF
There is a fat cable 50cm or less between the supply and the amp.
If I've understood this correctly one wants to have extra capacitors at the amplifier power input to negate having longer cables.
Now the question, how much capacitance would be ideal to completely negate the cost of having an external power supply?
Would it be enough to put a say a 33mF cap at the amp input or does one want more / less? A bigger cap followed by smaller caps or the other way around? Is it enough to only have additional caps at the amp or does one want something more / different?
Spread thy knowledge!
Thanks in advance,
Olle
Simulate it, You need to know more, the impedance of the transformer, the impedance of the cable and the frequency range of the amplifier. It is not possible to entirely negate the effects of a long cable with passive components, the amplifiers feedback loop will mask the low frequency effects, the bypass capacitors will ensure only low frequency components of the load appear on the supply cable.
Ah,
Is there a better tool to simulate this than PSUD2?
Should I calculate / measure the impendance / inductance of the cable and then enter it as an inductor?
And another question about the same subject, by what length from end cap in the psu to the input of the amp does the problems appear?
Say you instead a long cable use a ~ 5cm cable while optimizing layout so the total cable length from the last power supply cap to the amp pcb is less than 20cm, would it still then be a problem needing simulations to solve? When is the distance short enough that the effects are negligible?
It's not a very large amp I plan to make, I plan on using a 225-300 VA 2x12v transformer, don't know if it makes a difference but it can't hurt to mention it =)
// Olle
Is there a better tool to simulate this than PSUD2?
Should I calculate / measure the impendance / inductance of the cable and then enter it as an inductor?
And another question about the same subject, by what length from end cap in the psu to the input of the amp does the problems appear?
Say you instead a long cable use a ~ 5cm cable while optimizing layout so the total cable length from the last power supply cap to the amp pcb is less than 20cm, would it still then be a problem needing simulations to solve? When is the distance short enough that the effects are negligible?
It's not a very large amp I plan to make, I plan on using a 225-300 VA 2x12v transformer, don't know if it makes a difference but it can't hurt to mention it =)
// Olle
No Idea I have never used it, any simulator will work with an ideal source feeding the appropriate equivalent circuits. A short cable can be treated as an inductor, longer cables have distributed capacitance.
That depends on the amplifier and the cable, twisted pair has low inductance and should work fine even in lengths of a meter or so with monolithic bypass capacitors on the amplifier circuit board
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.