Hi Julien,
I noticed yesterday while giving this thread a complete read that you're using solide silver wire in your PSU.
Would stranded silver plated copper have an edge over this... I'm obviously thinking of skin effect here.
Thanks,
Chris
I noticed yesterday while giving this thread a complete read that you're using solide silver wire in your PSU.
Would stranded silver plated copper have an edge over this... I'm obviously thinking of skin effect here.
Thanks,
Chris
Hello,
"Hotrodding" hmmmm..... I'll go with the assumption that means tweaking tips and not custom design, like a discrete input stage etc.
Has anyone tried the constant current bias tweak on their op amps yet??? Just seems a heck of alot easier/cheaper and far more "tweak" like for the hotrodder than to consider seperate supplies /regulators.
Cheers,
Chris
"Hotrodding" hmmmm..... I'll go with the assumption that means tweaking tips and not custom design, like a discrete input stage etc.
Has anyone tried the constant current bias tweak on their op amps yet??? Just seems a heck of alot easier/cheaper and far more "tweak" like for the hotrodder than to consider seperate supplies /regulators.
Cheers,
Chris
classd4sure said:
Hi Chris.
I only used solid silver for my main PS capacitors connections. 12ga for the ground center rail, and 14ga for the + and - voltage rails. I used Teflon tubing for insulation between connections.
All other power supply wiring was done using 14ga Teflon insulated silver plated copper stranded wire.
I also use this wire for speaker cables and power cords. Once broken in, this wire really sounds transparent and smooth.
Hope that helps.
Sorry Yves but there's more aspects to be concerned with than simply capacitors.
I stand by the question and thanks for the answer Juliene, regardless of if I agree with it or not.
Yves, what does a post of that nature have to contribute? Not much?
Regards
I stand by the question and thanks for the answer Juliene, regardless of if I agree with it or not.
Yves, what does a post of that nature have to contribute? Not much?
Regards
I'm sorry but this forum has always been very informative - all tweaks & tips so far seem to have a good scientific foundation and I would be nowhere without this forum.
I just hope this doesn't turn into another forum where stones are put on CD-players and plastic dots are placed on speaker cabinets - I left many nice forums due to things of this nature...
As for cables, I too prefer a nice gauge quality cable between my speakers and amplifier, and I do understand that in the PS all quality parts add to the final product.
However, no offense to anyone here, I will apologize for my last post and abstain from further comments on cabling matters.
In fact I'll just delete the post if I still can.
I just hope this doesn't turn into another forum where stones are put on CD-players and plastic dots are placed on speaker cabinets - I left many nice forums due to things of this nature...
As for cables, I too prefer a nice gauge quality cable between my speakers and amplifier, and I do understand that in the PS all quality parts add to the final product.
However, no offense to anyone here, I will apologize for my last post and abstain from further comments on cabling matters.
In fact I'll just delete the post if I still can.
Yves that's kind of why I asked about skin effect when I saw solide core wire used...
I don't think I'll be using solide core myself, but I didn't think it hurt to ask.
No need to delete my friend.
I don't think I'll be using solide core myself, but I didn't think it hurt to ask.
No need to delete my friend.
Questions of content
These are legitimate issues and I see no problem with them being posted. I will also agree that the last thing we want to do is detract from the basically scientific nature of what we are doing. As we all are strapped with non ideal reference systems it is foolish to put in a lot of other unprovable or immeasurable parameters. The important thing is to make only one change at a time so you really know what it is doing. It is also important to try to determine if so called improvements are actually compensating for other deficiencies elsewhere. This is the difficult one and can waste a lot of time when you find a coloration and realize other changes were compensating for it. I have been down this path so many times that I now will only use parts that I have personally proven to be the most neutral available within a reasonable price range. Of course new stuff will have to be checked out and this is where this thread is to be of value. Lets all try to keep it that way.
Thanks all,
Roger
These are legitimate issues and I see no problem with them being posted. I will also agree that the last thing we want to do is detract from the basically scientific nature of what we are doing. As we all are strapped with non ideal reference systems it is foolish to put in a lot of other unprovable or immeasurable parameters. The important thing is to make only one change at a time so you really know what it is doing. It is also important to try to determine if so called improvements are actually compensating for other deficiencies elsewhere. This is the difficult one and can waste a lot of time when you find a coloration and realize other changes were compensating for it. I have been down this path so many times that I now will only use parts that I have personally proven to be the most neutral available within a reasonable price range. Of course new stuff will have to be checked out and this is where this thread is to be of value. Lets all try to keep it that way.
Thanks all,
Roger
Funny thing is the circuit is basically one big interraction, everything matters to a certain degree.
As a DIYer I refuse to ever fall prey to throwing a single penny away on baseless myths, because it's a mistake.
I feel I should point out the obvious here that while skin effect is science based we all know it's one of the first things the snake oil guys lean on when selling their magic wires so the concerns in all the posts above are totally understandable.
There's too many good people here to allow the myths to overule the science of things and I for one couldn't bring myself to imagine the loss of a source such as this. I'm also glad to see it seems to get better all the time.
Can't even tell you what I've learnt about the influence of different caps in here and to have it related to the science is just..... well... sniff
As a DIYer I refuse to ever fall prey to throwing a single penny away on baseless myths, because it's a mistake.
I feel I should point out the obvious here that while skin effect is science based we all know it's one of the first things the snake oil guys lean on when selling their magic wires so the concerns in all the posts above are totally understandable.
There's too many good people here to allow the myths to overule the science of things and I for one couldn't bring myself to imagine the loss of a source such as this. I'm also glad to see it seems to get better all the time.
Can't even tell you what I've learnt about the influence of different caps in here and to have it related to the science is just..... well... sniff
Science?
Chris,
Not to despair! Caps are science or rather can be. They also can be the biggest snake oil around, what is your current flavor of the month? I appreciate and use caps that are the best testing in all parameters for their type including some very easily done microphonic tests which may have the greatest influence of all. Take a ½ dozen of your current favorite caps but be sure to include an Auricap. Charge all of them up to 100VDC. Leave all of them for an hour and come back. Quickly measure them all, throw away any below 85v as too leaky. If the charge is ok pinch it firmly, nearly as hard as you can and watch what the voltage does. Write it down and go on to the next one. When you are all done select the best testing one to use in your system as it will also sound the best meaning most neutral or least colored. It will be the Auricap. I have done this against all the others except metal case ones and they have other serious problems. Auricaps were checked sonically against a solid piece of quality wire for qualification. Some of the really expensive caps have other issues like very high dissipation factor or very high ESR. The Auricap excels in these parameters as well. This whole cap issue is why I associated with Audience in the first place. I could go into several other areas thought to be snake oil as well and show with test results that some have actual merit. Of course the film caps are only one link in a long signal chain.
There are an enormous number of things that have a sonic influence that haven’t been brought to the test bench yet. I think for the good of our business/hobby/passion we owe it to ourselves to try to figure out some of these things and make our systems the better for it.
Roger
Chris,
Not to despair! Caps are science or rather can be. They also can be the biggest snake oil around, what is your current flavor of the month? I appreciate and use caps that are the best testing in all parameters for their type including some very easily done microphonic tests which may have the greatest influence of all. Take a ½ dozen of your current favorite caps but be sure to include an Auricap. Charge all of them up to 100VDC. Leave all of them for an hour and come back. Quickly measure them all, throw away any below 85v as too leaky. If the charge is ok pinch it firmly, nearly as hard as you can and watch what the voltage does. Write it down and go on to the next one. When you are all done select the best testing one to use in your system as it will also sound the best meaning most neutral or least colored. It will be the Auricap. I have done this against all the others except metal case ones and they have other serious problems. Auricaps were checked sonically against a solid piece of quality wire for qualification. Some of the really expensive caps have other issues like very high dissipation factor or very high ESR. The Auricap excels in these parameters as well. This whole cap issue is why I associated with Audience in the first place. I could go into several other areas thought to be snake oil as well and show with test results that some have actual merit. Of course the film caps are only one link in a long signal chain.
There are an enormous number of things that have a sonic influence that haven’t been brought to the test bench yet. I think for the good of our business/hobby/passion we owe it to ourselves to try to figure out some of these things and make our systems the better for it.
Roger
Hi,
I'm not yet sold..but great post and I largely agree.
I question the validity of the squishy test. So long as their temperature settles I cant' see anything much influencing case pressures in a similar manner as physically squishing it. No level of SPL I could imagine generating would possibly aproach the amount of force I "think" (never having measured myself) required to deflect the "plates" enough to have a substantial affect.
I do agree with careful selection based on parameters and often most aren't always given, which I find limits the available choice as I don't consider them.
I certainly can appreciate the ESR test and will try it in the future.
Thanks,
Chris
I'm not yet sold..but great post and I largely agree.
I question the validity of the squishy test. So long as their temperature settles I cant' see anything much influencing case pressures in a similar manner as physically squishing it. No level of SPL I could imagine generating would possibly aproach the amount of force I "think" (never having measured myself) required to deflect the "plates" enough to have a substantial affect.
I do agree with careful selection based on parameters and often most aren't always given, which I find limits the available choice as I don't consider them.
I certainly can appreciate the ESR test and will try it in the future.
Thanks,
Chris
Squishy test
Chris,
Yes the test does greatly exaggerate the microphonics of the part, that is the whole idea. This can be a subtle and insidious coloration as it may be very high Q. Very tiny signals generated due to microphonic caused charge fluxatuion can form positive or negative feedback. Both are wrong and very undesirable. I am convinced this is why some of the high end pieces using these guilty parts are also full of all kinds of damping material. Something to think about anyway.
Roger
Chris,
Yes the test does greatly exaggerate the microphonics of the part, that is the whole idea. This can be a subtle and insidious coloration as it may be very high Q. Very tiny signals generated due to microphonic caused charge fluxatuion can form positive or negative feedback. Both are wrong and very undesirable. I am convinced this is why some of the high end pieces using these guilty parts are also full of all kinds of damping material. Something to think about anyway.
Roger
Hi,
So it's more the casing resonance you're concerned with?
Could you put a charged cap on a scope and play a test sweep "at" it while stepping the SPL? If I saw a resonant Q there I'd be convinced.
I'd try it myself but am ill equipped.
Actually if that's the case you'd think such a thing would appear in their data sheets.
Thanks,
Chris
So it's more the casing resonance you're concerned with?
Could you put a charged cap on a scope and play a test sweep "at" it while stepping the SPL? If I saw a resonant Q there I'd be convinced.
I'd try it myself but am ill equipped.
Actually if that's the case you'd think such a thing would appear in their data sheets.
Thanks,
Chris
Summarising the UCD tweaks
Hi,
I've tried to summarise this thread with the popular tweaks (feedback welcome):
1) Replace the bootstap capacitor with a 220u/16V oscon
2) Replace the supply bypass caps with Panasonic FC 22uf 50v
3) Look to use a better power supply for the input circuitry (+-12v)
4) Replace the filter cap with a small form factor audiophile cap (experimenting still continuing, no concete conclusion yet).
5) Short the coupling caps between the opamp and the power stage (in the newer UCDs) but if you have a DC input problem you will need coupling caps on the inputs (2.2uF auricap is a good choice).
6) Panasonic FC caps to replace the 2 big 470uF/100V caps with 680uF/100v (tight fit).
A great power supply design is the one used by JM (see http://www.audiocircle.com/circles/...ame=gallery&file=index&include=view_photo.php). Maybe consider the use of 10,000uF 63V Aerovox T-Power 4 pole capacitors if your budget can stretch.
This might be a stupid question, but regarding DC protection for the amp, if you are using true balanced inputs (ie. balanced outputs on the preamp), then any DC differential between the source and the amp will be common to both inputs, cancelled by the differential input on the op amp, so you will never need a coupling cap to protect against DC if you use differential inputs. Is this the case in reality?
sx881663, how are you going with your experiments with the filter cap? Do we have any news from Audience regarding a better filter cap that has the right form factor for the UCD?
Also, when is the UCD700 due?
Regards,
Dean
Hi,
I've tried to summarise this thread with the popular tweaks (feedback welcome):
1) Replace the bootstap capacitor with a 220u/16V oscon
2) Replace the supply bypass caps with Panasonic FC 22uf 50v
3) Look to use a better power supply for the input circuitry (+-12v)
4) Replace the filter cap with a small form factor audiophile cap (experimenting still continuing, no concete conclusion yet).
5) Short the coupling caps between the opamp and the power stage (in the newer UCDs) but if you have a DC input problem you will need coupling caps on the inputs (2.2uF auricap is a good choice).
6) Panasonic FC caps to replace the 2 big 470uF/100V caps with 680uF/100v (tight fit).
A great power supply design is the one used by JM (see http://www.audiocircle.com/circles/...ame=gallery&file=index&include=view_photo.php). Maybe consider the use of 10,000uF 63V Aerovox T-Power 4 pole capacitors if your budget can stretch.
This might be a stupid question, but regarding DC protection for the amp, if you are using true balanced inputs (ie. balanced outputs on the preamp), then any DC differential between the source and the amp will be common to both inputs, cancelled by the differential input on the op amp, so you will never need a coupling cap to protect against DC if you use differential inputs. Is this the case in reality?
sx881663, how are you going with your experiments with the filter cap? Do we have any news from Audience regarding a better filter cap that has the right form factor for the UCD?
Also, when is the UCD700 due?
Regards,
Dean
Re: Summarising the UCD tweaks
Dean,
This is too much to answer in one chunk. I will address the issues, separately, as time permits.
First off the power supply above is excellent but a couple of things need to be clarified. The .47 uf AC input cap should be across the transformers primary where it will do the most good and will be discharged by the DC resistance of the primary when power is off. This is an important safety issue as if you get shocked while carrying it you might drop it on your foot, OUCH! Also protection is provided by the input fuse at that point. This cap should be rated at 600 or 1000 VDC to be used across the AC line. The .1 uf cap across the switch may be too large and supply enough residual power to keep the amp turned on with the switch off. .01uf is normal and .047 uf is used for very high power switches. The .1 uf caps across the transformer secondarys might work better if it was larger. Try something like 1-2 uf and see how the noise looks. Although I have done this and noticed a sonic improvement I have never actually ran tests to prove it. It makes good sense that this will work as it helps cancel out the secondarys leakage inductance preventing large reactive spikes. The value would need to be selected by using a scope on the secondary and adjusting the value till the spikes are gone. This test would need to be done with near full power being delivered to be useful.
My next clarification is the connection of the bridge rectifier outputs. These need to be connected, each separately (+/-) right at the first cap terminals. This keeps the ripple current out of the grounds and lowers overall noise.
Like Bruno says “Its all in the details”.
More later,
Roger
deandob said:
Dean,
This is too much to answer in one chunk. I will address the issues, separately, as time permits.
First off the power supply above is excellent but a couple of things need to be clarified. The .47 uf AC input cap should be across the transformers primary where it will do the most good and will be discharged by the DC resistance of the primary when power is off. This is an important safety issue as if you get shocked while carrying it you might drop it on your foot, OUCH! Also protection is provided by the input fuse at that point. This cap should be rated at 600 or 1000 VDC to be used across the AC line. The .1 uf cap across the switch may be too large and supply enough residual power to keep the amp turned on with the switch off. .01uf is normal and .047 uf is used for very high power switches. The .1 uf caps across the transformer secondarys might work better if it was larger. Try something like 1-2 uf and see how the noise looks. Although I have done this and noticed a sonic improvement I have never actually ran tests to prove it. It makes good sense that this will work as it helps cancel out the secondarys leakage inductance preventing large reactive spikes. The value would need to be selected by using a scope on the secondary and adjusting the value till the spikes are gone. This test would need to be done with near full power being delivered to be useful.
My next clarification is the connection of the bridge rectifier outputs. These need to be connected, each separately (+/-) right at the first cap terminals. This keeps the ripple current out of the grounds and lowers overall noise.
Like Bruno says “Its all in the details”.
More later,
Roger
Hi deandob
That is a terrific summary. The JM PS is an exceptional PS. Perhaps instead of 4 x 4700uF caps, you could use 2 x 10,000uF caps, which would allow a little more free space within the cabinet.
I first saw this PS design done by Dejan V. Veselinovic, and having used it, know that it is excellent.
Anyway, well done and a terrific summary
That is a terrific summary. The JM PS is an exceptional PS. Perhaps instead of 4 x 4700uF caps, you could use 2 x 10,000uF caps, which would allow a little more free space within the cabinet.
I first saw this PS design done by Dejan V. Veselinovic, and having used it, know that it is excellent.
Anyway, well done and a terrific summary
Hi deandob
Sorry, forgot to say that PS details by Dejan V. Veselinovic can be found at 'Designing your own power supply' ... http://www.zero-distortion.com/start.htm
Sorry, forgot to say that PS details by Dejan V. Veselinovic can be found at 'Designing your own power supply' ... http://www.zero-distortion.com/start.htm
Re: Summarising the UCD tweaks
Dean,
Continuing on with this; Items 1- 6 are right as far as we know to this point. The Oscon is serious overkill but won’t hurt anything, just remember it is reported they take a long time to break-in.
Using the amp input as fully differential does not eliminate the DC problem only the component common to both +/- inputs referenced to ground. Any residual DC between the inputs is treated the same as the signal and is subject to the amps full gain. This is why only very small amounts are acceptable. I still consider coupling caps cheap insurance as long as they are of very high quality. In this case 2 would be needed.
I am just now starting to build the amps. To begin with, this will be a set of 4 identical monoblocks. This way I will be able to make single changes in one pair and do an A/B to the other with all else being identical. After a change is verified it will be incorporated in the other pair so they are identical again. I suspect this bootstrap process will continue for some time as I plan on trying a lot of different things and won’t leave many stones unturned.
From all that has been said and simple experiments done, I think the output filter cap will be one of the key improvements. The Audience part design has been verified for production suitability and is being built as a very small lot of samples for evaluation. Still much to do here before production goes into full swing. Future availability should not be a problem as there will be a rather large stock kept on hand. My guess is they are still 5-6 weeks away. I will check into this further and post what I find including the specs.
By the way, this new Auricap will also be a truly outstanding power supply bypass cap. This same cap should work very well except where very low ESR is problematic. In cases like this where snubbers are more appropriate, using the Auricap will allow precise tuning with an external resistor. This allows precise control and repeatability of the characteristics. The smaller values will not be available at first but are in the works.
Roger
deandob said:
Dean,
Continuing on with this; Items 1- 6 are right as far as we know to this point. The Oscon is serious overkill but won’t hurt anything, just remember it is reported they take a long time to break-in.
Using the amp input as fully differential does not eliminate the DC problem only the component common to both +/- inputs referenced to ground. Any residual DC between the inputs is treated the same as the signal and is subject to the amps full gain. This is why only very small amounts are acceptable. I still consider coupling caps cheap insurance as long as they are of very high quality. In this case 2 would be needed.
I am just now starting to build the amps. To begin with, this will be a set of 4 identical monoblocks. This way I will be able to make single changes in one pair and do an A/B to the other with all else being identical. After a change is verified it will be incorporated in the other pair so they are identical again. I suspect this bootstrap process will continue for some time as I plan on trying a lot of different things and won’t leave many stones unturned.
From all that has been said and simple experiments done, I think the output filter cap will be one of the key improvements. The Audience part design has been verified for production suitability and is being built as a very small lot of samples for evaluation. Still much to do here before production goes into full swing. Future availability should not be a problem as there will be a rather large stock kept on hand. My guess is they are still 5-6 weeks away. I will check into this further and post what I find including the specs.
By the way, this new Auricap will also be a truly outstanding power supply bypass cap. This same cap should work very well except where very low ESR is problematic. In cases like this where snubbers are more appropriate, using the Auricap will allow precise tuning with an external resistor. This allows precise control and repeatability of the characteristics. The smaller values will not be available at first but are in the works.
Roger
- Status
- Not open for further replies.