Amplifier power supply upgrade ideas

[czag]

A few ideas were suggested by a fellow a Sherbourn when I asked about raising the capacitance storage to improve the dynamics of my 5/1500a. The amp has 5 400va toroidal transformers (1 for each channel), 8 output transistors per channel and 20,000 microfarads per channel. I was told that each channel was capable of drawing 7.2 amps when the amp is driven near clipping (which I will probably never do).He suggested these:

1. Upgrading to a 10awg to 12awg power cord, no brand suggested just the awg.

2. Put in a dedicated line from the breaker box for the amp. It is now on a standard 15a circuit that is shared by the whole living room.

3. Raising the capacitance storage from 20,000 microfarads up to a max of 33,000. he said beyond 33,000 I probably would never notice the difference.

I am looking for someone to share their experience if you have tried any one of these or maybe one of the engineers on this forum can put in your 2 cents worth as to the validity of any of the ideas and which one might be more worthwhile than the others.

I'm not looking to start another "AC power cord" type of thread, I am just looking for some of you to share you knowledge and experience. I had posted this at the very end of the "AC power cord" thread, but the thread was locked. Deward and Pete both chimed in with some insight, maybe a few of you could add more.

Also, Penngray, that "AC power cord" thread has made you a little paranoid. Nothing in my post was directed in any way at you. That was a very rude and uncalled for response, again, nothing in my post was even remotely aimed at you. I am not interested in p!ssing matches. The "real" engineers just meant that I know there are some real electrical engineers here on the forum and want to hear from them, but anyone with knowledge or experience on/with the ideas presented is welcome and encouraged to give their thoughts.

Thanks,
Chris

[bangleiii]

On the capacitors what you need to consider is the esr of the cap, A larger cap may not charge/ discharge as fast as a smaller one.
So what would be a good upgrade would be to add some low esr caps to the caps you already have.

The next weak link is to add wire to the traces on the pc board where the traces are thin, from the caps solder positions to the diodes and output transistors, On many amps the bass can be tightened up by doing the same from the output transistors to the speaker jacks, some even will wire the speakers cable directly to the output transistors.

[Face]

Bypass large electrolytics with small film caps.

[Æ]

All of them are good suggestions. Always good to have plenty of power supply "smoothing" capacitance. The more the merrier up to a point. The only way you'll ever know for sure how much, would be to do a before and after measurement of the ripple. I'm not sure what percentage of ripple factor is considered acceptable, but the lower the better.

Along the same lines as the power cord thread, it makes no sense to have 6 feet of 10 gauge wire when the other 60 feet from the breaker are only 14 gauge. So the idea to put in a dedicated line/breaker is excellent, then you can go with a larger gauge wire all the way and not just the last 6 feet.

8 output resistors? Didn't you mean transistors?


A few idea were suggested by a fellow a Sherbourn to when I asked about raising the capacitance storage to improve dynamics of my 5/1500a. The amp has 5 400va toroidal transformers (1 for each channel), 8 output resistors per channel and 20,000 microfarads per channel. I was told that each channel was capable of drawing 7.2 amps when the amp is driven near clipping (which I will probably never do).He suggested these:

1. Upgrading to a 10awg to 12awg power cord, no brand suggested just the awg.

2. Put in a dedicated line from the breaker box for the amp. Is is now on a standard 15a circuit that is shared by the whole living room.

3. Upping the capacitance storage from 20,000 microfarads up to a max of 33,000. he said beyond 33,000 I probably would never notice the difference.

I am looking for someone to share their experience if you have tried any one of these or maybe one of the engineers on this forum can put in your 2 cents worth as to the validity of any of the ideas and which one might be more worthwhile than the others.

I'm not looking to start another "AC power cord" type of thread, I am just looking for some of you to share you knowledge and experience. I had posted this at the very end of the "AC power cord" thread, but the thread was locked. Deward and Pete both chimed in with some insight, maybe a few of you could add more.

Also, Penngray, that "AC power cord" thread has made you a little paranoid. Nothing in my post was directed in any way at you. That was a very rude and uncalled for response, again, nothing in my post was even remotely aimed at you. I am not interested in P!ssing matches. The "real" engineers just meant that I know there are some real electrical engineers here on the forum and want to hear from them, but anyone with knowledge or experience on/with the ideas presented is welcome and encouraged to give their thoughts.

Thanks,
Chris

[czag]

On the capacitors what you need to consider is the esr of the cap, A larger cap may not charge/ discharge as fast as a smaller one.
So what would be a good upgrade would be to add some low esr caps to the caps you already have.

The next weak link is to add wire to the traces on the PC board where the traces are thin, from the caps solder positions to the diodes and output transistors, On many amps the bass can be tightened up by doing the same from the output transistors to the speaker jacks, some even will wire the speakers cable directly to the output transistors.


I never considered the esr, thanks. Adding wire seems like a good and cheap upgrade, thanks again.

Face, what values would I use on the film caps?

Thanks,
Chris

[czag]

8 output resistors? Didn't you mean transistors?

Yes, just edited the post. Also, thanks for the input.

Chris

[bangleiii]

I never considered the esr, thanks. Adding wire seems like a good and cheap upgrade, thanks again.

Face, what values would I use on the film caps?

Thanks,
Chris

When adding wires to the traces be sure to folow the trace fairly close els you can get some humm or worse.

[MagicO309d]

Even if you are running full-range mains, set your processor setting to "small" to off-load more base duty to your sub(s). If you are already doing this, try higher crossover points and moving your sub(s) if needed so that you will gain more headroom for your amp/mains combination.

I'm not an HT expert, just parroting that advise that I've read about setup from recognized experts. It seems that mixing main speaker bass duty and LFE together is supposed to always be the ideal HT configuration.

Otherwise, start looking at high-efficiency pro-driver or line array based speakers. How many dB will you pick up from caps or a new power cord compared to a 3 dB or more speaker upgrade?

[Deward Hastings]

So far all good suggestions. To recap what I posted in the "other thread":

Dedicated circuit. Mine is 240V with a step-down/isolation transformer and commercial noise filters.

Power cord doesn't matter, as long as it's 14 gauge or heavier.

Bigger power supply caps have to respect the limits of the rest of the design, because they will increase the surge currents in the rectifiers and transformer.

And decent amplifier circuit should have 60-85 dB power supply rejection . . . "ripple" is rarely a problem. The amplifier should be stable if the supply "wilts" up to 10% too . . .

As noted in comments above sometimes there's benefit from beefing up the amp's internal wiring (but it shouldn't be necessary in a well designed amp). If you deviate from the existing layout you run the risk of noise and hum pickup. Large electrolytic caps tend to have high internal inductance . . . they may have no better "surge" capacity than lower values. The speaker should return to the power supply star, not to the amplifier circuit board. There may be some filter components on the amp output . . . mess with or modify them at your own risk, but with some loads (that do not compromise stability) removing them may improve performance. If you mod an amp (including power supply "improvements") treat it as a new design . . . all the components interact.

[czag]

Excellent ideas.

Some opinions on which one would you implement first if money was no object, and which one would be the best bang for the buck?

Thanks,
Chris

[bangleiii]

Excellent ideas.

Some opinions on which one would you implement first if money was no object, and which one would be the best bang for the buck?

Thanks,
Chris

How loud do you listen? what are the speakers?

[A9X]

Excellent ideas.

Some opinions on which one would you implement first if money was no object, and which one would be the best bang for the buck?

Thanks,
ChrisBasically, I agree with Deward, except for the point about amp output inductors; unless you really know what you're doing, you could end up with an oscillator instead of an amplifier. Leave the amp stages themselves alone unless you have the expertise to design an amp from scratch.

I've done a lot of amp design and building over the years, including some with huge amounts of capacitance (one had nearly 1F) and the require soft start circuitry etc. Once you get past a certain point, which varies from amp to amp, there's no point in going higher as it doesn't make any measured or audible difference. As this is a manufactured amp, vs DIY where you can spec it out however you want unless you are willing to make changes to transformers and other PS parts, stick to the manufacturer's recommendation for capacitance increase. The difference isn't large, so don't expect much. Ditto for increasing the gauge of the power cord, it will help though.

Where I would make the first change would be to have a new sub main run from your switchboard to your AV gear location, in the biggest gauge that's practical and via the shortest route. Remember to tell the sparky where you want it run or he'll likely do it the easiest, but it may not be practical. US member would need to advise on the details as I'm not familiar with the NEC, but am with AS/NZS 3000 and 3008, my local equivalent.

[czag]

How loud do you listen? what are the speakers?

Right now it is a small 3 way with an 8" woofer. I use a Marchand active xover for the lowpass @ 400hz. So I use 4 of the 5 channels, 2 for the 8" woofers and 2 for the mid-tweeter in the pair of speakers. But, I foresee a larger floorstander with possibly 2 8" or 10" woofers with an MTM configuration for the mid-tweeter section. Hopefully I will have that completed in the spring.

I am not really going for extra sensitivity, just want higher dynamic range (headroom) to hit the peaks. I usually listen at around 85-90dbs, but I like the peaks to actually peak.

Also, I will take a serious look at adding wire where suggested, but I do not have enough (any) electrical engineering knowledge to have the confidence to take components out.

Thanks,
Chris

[billfitzmaurice]

Dedicated circuit. Mine is 240V with a step-down/isolation transformer and commercial noise filters.Balanced? The single best thing you can do is go from 110-0-110 balanced street supply to 55-0-55 transformer balanced for the dedicated line. Transformer isolation in and of itself doesn't do that much, it's separating current delivery from ground that zaps noise. Larger caps for cleaner DC filtering aren't required if line noise is never present to begin with.

[Æ]

Balanced? The single best thing you can do is go from 110-0-110 balanced street supply to 55-0-55 transformer balanced for the dedicated line. Transformer isolation in and of itself doesn't do that much, it's separating current delivery from ground that zaps noise. Larger caps for cleaner DC filtering aren't required if line noise is never present to begin with.

Which caps are you talking about? Are you talking about the main power supply caps? Their main function is to smooth the rectified wave, otherwise it would be quite choppy. Hopefully nothing AC passes, regardless.

http://upload.wikimedia.org/wikipedia/en/7/7e/Smoothed_ripple.svg

[Deward Hastings]

Balanced? Unfortunately not . . . it was a "price is right" acquisition . . .

I do pay particular attention to floating the "neutral" in all my equipment power inputs, though, so the equipment (and signal) grounds can be handled properly. I have considered putting a center tapped choke across the 120 to "force" balance to ground . . . so far it hasn't seemed necessary.

[czag]

Thanks for all the ideas and opinions. If there are more out there, feel free to share.

Chris

[billfitzmaurice]

Unfortunately not . . . it was a "price is right" acquisition . . .
I hear you on that. But if one does have the $2k or more it takes and wants to spend it on something that actually works a balanced supply is the way to go. For those not familiar it's explained here:
http://www.equitech.com/articles/enigma.html

Which caps are you talking about? Are you talking about the main power supply caps?I didn't suggest removing them, just that you might have no need to add more capacity to them if you drop the noise floor by 12dB or better with a balanced supply.