UCD separate opamp supply?

Hi,

I don't recall anyone mentioning having done it. Nor has there been much discussion of it. Your'e a trailblazer.

I thought of it myself but found I just don't like the idea, it seems less than ideal to me and a good regulator would be more... feasible.

Does your source operate with battery power as well, and if not, is there a point to the UCD input stage being battery powered?

A few issues that come to mind:

Limited listening time, leading to multiple batteries so one is being charged/conditioned while the other is in use.... would 2 be enough? How long should one be allowed to run before the rise in internal impedance becomes problematic.. switching to the other would give a higher voltage, seems regulation is still required. Temperature dependancy is rather ugly with them as well.

Doing something like this right seems like a huge undertaking when you consider the level of house keeping circuitry required for seamless and safe operation, far from elegant.

I'm not sure it would be worth the effort considering the output stage is still running off an unregulated raw dc supply. You'll have to decide all this for yourself. It quickly lost it's appeal to me, being that it's nothing simpler than what would already offer excellent performance.

Personally I'd keep them for your source if anything. Maybe someone who's actually done the battery supply thing before can chime in.

Regards,
Chris

I have built a few SLA powered DACs so I have a little experience with those things. First thing is that you do need a regulator, since the SLAs start at about 13.2V fully charged and the slowly drop to about 11.5V before they go rapidly dead. So that's a 2 volt swing before the thing is dead. Do the UCDs care about a specific voltage range or is clean power in a certain range all they need there?

If you need a regulator and 12V, you can't use a 12V battery, because the regulator just won't kick in unless you are at about 15 volts input level. Taking two 12V batteries in series gets you to 24 volts, which depending on the power draw of what you want to power with it may be ok, or far too much (mega heat sinking to shave off that extra power you don't need).

I tend to use SLA power for powering CD transport clocks and DACs, which is where it seems to make a huge difference. Not sure if the effort is worth it for the power amps, but it could be a pretty cool tweak. If the power draw of the opamp isn't very high, all you need is two SLAs and a 24V smart charger, a few parts for a clean regulator and a switch that toggles the battery between smart charger and amps. a relay could do that, too, combined with the overall on/off of the UCD module or softstart

Peter

Hi,

I think if anything it could benefit from the flatter frequency response. The regulator will still be imposing at least some ripple, right?

Look at the "audiophile" regs on the UCD700's supply and ask yourself do I need this, really?

Another big drawback is the size and weight of this whole thing. This led me to the discovery of those 23A ultra small 12V cells. They're not even 100mAh, but hey, ~1$ a piece!! Being alkaline I'd imagine they'd die a slow death, memory effect, etc. You could conceivably run as many in // as you like to extend the play time, they're so small and so cheap, who cares?

Might be a cheap and easy /throw-away means to give cell power a try if you don't already have SLAs.

I'd think the level of voltage seen at the op amps mostly is optimised for their THD performance, the UCD itself won't really care if it's a few volts higher or lower, on the AD op amps though the "max" is 12V, so they'd have to be regulated to no higher than that.

Still it would affect your amp's gain as time plays on and internal impedance goes up/voltage sags. That's why I don't think you can really allow that to occur and should switch out one pack before it becomes noticeable. The op amp stage is something like half the gain of the amp.

Seems like a SMPS would be a good regulator for the front end and yet if you're going to do that you really don't need a battery.

Just to be silly, (really), had the thought for a perpetual amp machine, making use of a dual voice coil to charge the supply. Being over 90% efficient it seems doable. I know, I know, how would it start you ask? Clap On!

Regards,
Chris

Battery PS:

If I recall well, our forum mate Rha61, who has a respectable experience in PS options for the modules, used a battery to power the opamps. He reported improvement in sound.

Regards
M

I've been thinking about trying the ALW super regs or building a some LM317/337,tl431 regs to power the AD8620's in my UCD180's

Appart from cutting the track on the UCD board is there another way of fitting the external regs? I don't fancy hacking at the track to be honest

I am still not 100% happy with the sonic signature of the AD8620 even with the coupling caps removed, it feels like these are not letting the UCD amp show its full potential.
I'm trying to find a discrete op-amp circuit to build, this way we could build the regulation on the same board
Any idea's?

t. said:

I've been thinking about trying the ALW super regs or building a some LM317/337,tl431 regs to power the AD8620's in my UCD180's

Appart from cutting the track on the UCD board is there another way of fitting the external regs? I don't fancy hacking at the track to be honest

I am still not 100% happy with the sonic signature of the AD8620 even with the coupling caps removed, it feels like these are not letting the UCD amp show its full potential.
I'm trying to find a discrete op-amp circuit to build, this way we could build the regulation on the same board
Any idea's?

Click to expand...

Woah, no trace hacking required. Once you remove the pass transsitor that's there, and whatever else they said to remove in the hotrod thread someplace, continue on with the directions and solder your new power wire to the emitter pad of the pass transistor you removed. No hacking whatsoever.

The AD op amp took some getting used to for me. They're very precise and that tight precision does seem to feel like it's being held back. The again it's not the only ingredient in the soup either. Try driving the comparator portion directly and see what you think of that sound.

There's only so much you can do with it though. What op amps do you like the sound of? Could always try another too.

Regards,
Chris

Hi Chris,

Well thats good regarding not having to cut any track, I don't suppose there is a schematic for the UCD input stage and its psu layout.
If we did use an offboard discrete op-amp I suppose I can remove the wire links where the coupling caps was and connect it there so the AD8620 is left disconnected

To be honest theres not many op-amps I like, my external dac uses a fully discrete I/V and buffer stage, when I was modding a cdp's I/V and output stage I tried a number of op-amps including the AD8620, to be honest it had that same sonic signature there as what it does in the UCD input stage, I also don't like OPA627
They sound too artificial for me, I know the NE5532 is far from the best and can be a bit gritty etc in the highs and needs coupling caps because of its DC offset but to me it sounded more open and less coloured.
I've got a few more op-amps including AD826,OPA2132 and others, the problem is its going to be hard removing the AD8620 without damaging it, these things don't like a lot of heat on the pins, I don't want to try some different types which would sound even worse and be stuck with them

I've been thinking about trying the ALW super regs or building a some LM317/337,tl431 regs to power the AD8620's in my UCD180's

Click to expand...

I am new to ALWSR. Just a few days on my superDAC's 5V supply (D & A). I'm not sure if it is the way to go. Maybe overkill and sticked on a certain sound signature.
I supose trying LM3x7 regs + separate R-core Tx + the PS caps you like the best is a wiser approach, if you have the space.

A funny thing is that I thought my UCD400 with BB opamp would sound "darker" and it is not the case, at least in the first few days of playing. I will let more time to burn-in to post my findings
(UCD400ST monoblocks; DC coupled; Sikorel caps; OCC cooper hook-up wire; )


Good luck and courage 😀

Mauricio

t. said:

Hi Chris,

Well thats good regarding not having to cut any track, I don't suppose there is a schematic for the UCD input stage and its psu layout.
If we did use an offboard discrete op-amp I suppose I can remove the wire links where the coupling caps was and connect it there so the AD8620 is left disconnected

To be honest theres not many op-amps I like, my external dac uses a fully discrete I/V and buffer stage, when I was modding a cdp's I/V and output stage I tried a number of op-amps including the AD8620, to be honest it had that same sonic signature there as what it does in the UCD input stage, I also don't like OPA627
They sound too artificial for me, I know the NE5532 is far from the best and can be a bit gritty etc in the highs and needs coupling caps because of its DC offset but to me it sounded more open and less coloured.
I've got a few more op-amps including AD826,OPA2132 and others, the problem is its going to be hard removing the AD8620 without damaging it, these things don't like a lot of heat on the pins, I don't want to try some different types which would sound even worse and be stuck with them

Click to expand...

Hi,

No matter what I can't help but compare my 180's to the sound of the homebrew I made, as it was the discrete comparator driven directly, and an air core coil in the filter.

Aside from the level of hiss you'd expect from such a beast, the sound from it was/is so pure, it instantly became my benchmark to compare even the Hypex modules to.

I know exactly what you mean about the 5532. I prefer the 8620 though because DC coupled it's far superior in alot of ways, but yeah, what you said, it's just a little bit thin I think.

Good thinking on the coupling caps 🙂 I believe that will work nicely.

There's no such schematic but the parts for the regulators are clearly stated someplace in the hotrod thread by Jan-Peter, should not be hard to locate them once you find that post.

First I'd try the biasing tweak and see what you think, it's so cheap and all you have to do is solder it on, the pads are already there. I've still neglected to order those parts, but intend to sometime soon. That should liven it up more I think.

Mauricio, interesting observations.

Cheers

maxlorenz said:

I am new to ALWSR. Just a few days on my superDAC's 5V supply (D & A). I'm not sure if it is the way to go. Maybe overkill and sticked on a certain sound signature.
I supose trying LM3x7 regs + separate R-core Tx + the PS caps you like the best is a wiser approach, if you have the space.

A funny thing is that I thought my UCD400 with BB opamp would sound "darker" and it is not the case, at least in the first few days of playing. I will let more time to burn-in to post my findings
(UCD400ST monoblocks; DC coupled; Sikorel caps; OCC cooper hook-up wire; )


Good luck and courage 😀

Mauricio

Click to expand...

Thanks Mauricio

I have both the ALWSR's and some diy lm317/337,tl431 I built up with schottky's etc onboard so I may try those instead of the SR's first
Regarding the R-cores I'm tempted to buy some 600vA rated ones later for the actual UCD modules too if I can find some with a suitable voltage, the only ones I could find had 2 x 0-27 v secondarys, they are not exactly cheap but should hopefully be a worthwhile upgrade at a later date

classd4sure said:

Hi,

No matter what I can't help but compare my 180's to the sound of the homebrew I made, as it was the discrete comparator driven directly, and an air core coil in the filter.

Aside from the level of hiss you'd expect from such a beast, the sound from it was/is so pure, it instantly became my benchmark to compare even the Hypex modules to.

I know exactly what you mean about the 5532. I prefer the 8620 though because DC coupled it's far superior in alot of ways, but yeah, what you said, it's just a little bit thin I think.

Good thinking on the coupling caps 🙂 I believe that will work nicely.

There's no such schematic but the parts for the regulators are clearly stated someplace in the hotrod thread by Jan-Peter, should not be hard to locate them once you find that post.

First I'd try the biasing tweak and see what you think, it's so cheap and all you have to do is solder it on, the pads are already there. I've still neglected to order those parts, but intend to sometime soon. That should liven it up more I think.

Mauricio, interesting observations.

Cheers

Click to expand...

Thanks Chris,

Its strange because the UCD180 with the AD8620 still sounds ok but I don't seem to be wanting to listen to it as much now so I've got to sort out something, it actually sounds more like my other amps espcially the modded Naim Nac pre with NCC200 now since replacing the op-amps which is ok but I was hoping to get the UCD to walk all over my other amps if you know what I mean😉 it does have the potential
I presume the bias tweak involves adding SST511 on the outputs of the AD8620? has anybody else actually tried this? I've read a few posts mentioning it but can't remember reading anything about anybody trying it out

t. said:

Thanks Chris,

Its strange because the UCD180 with the AD8620 still sounds ok but I don't seem to be wanting to listen to it as much now so I've got to sort out something, it actually sounds more like my other amps espcially the modded Naim Nac pre with NCC200 now since replacing the op-amps which is ok but I was hoping to get the UCD to walk all over my other amps if you know what I mean😉 it does have the potential
I presume the bias tweak involves adding SST511 on the outputs of the AD8620? has anybody else actually tried this? I've read a few posts mentioning it but can't remember reading anything about anybody trying it out

Click to expand...

Could be a number of things, wondering what your supply is like? I've found the odd little change in the wiring can make mean the difference from it just sounding OK to becomming musical with the very same components. How you ground it is a big one.

As far as the bias tweak yep that's it. I've no idea why people are so hesitant to try it, it's defininatly way over there on the science side of things, it _will_ improve the sound quality and I expect rather noticeably at that. I plan on doing that long before an aux supply, considering cost /complexity, it's a no brainer. Soon as I'm motivated I'll order a few.

Regards,
Chris

classd4sure said:

Could be a number of things, wondering what your supply is like? I've found the odd little change in the wiring can make mean the difference from it just sounding OK to becomming musical with the very same components. How you ground it is a big one.

As far as the bias tweak yep that's it. I've no idea why people are so hesitant to try it, it's defininatly way over there on the science side of things, it _will_ improve the sound quality and I expect rather noticeably at that. I plan on doing that long before an aux supply, considering cost /complexity, it's a no brainer. Soon as I'm motivated I'll order a few.

Regards,
Chris

Click to expand...

Well I have a pair of 2 x 0-30v 600vA Holden and Fisher transformers, dual MBR10100 rectifiers for each module with zobels across, a pair of 6800uf BHC Aerovox ALP20 on each +/- rail, all the grounds are separated and join at the UCD modules, I had to fit the two earth wires together soldered onto a single spade plug which connects to the UCD module, at the psu end they are separated.
The inputs are XLR, interconnects are balanced one end using XLR plugs and the earth and - signal are joined at a phono plug at the other end which plugs into my 22k stepped Attenuator
There is no hiss or hum, infact the amp is totally silent🙂
Oh and the secondary wiring is twisted into pairs
I did try MUR860 diodes but much preferred the MBR10100, I may try the MSR1560 next

I'll try the bias tweak as soon as I can find some of those SST511's, they are easy to remove just incase
If I can get rid of that signature it now seems to have I'll be happy🙂

Just been looking through my stash of components, I've got 4 x 22uf 35v Rubycon ZA's spare, would it be worth changing the standard AD8620 supply decoupling caps to the ZA's?

It's hard to envision your grounding at this hour for me.

What I got from that is that you have a dual mono construction with seperate supplies in the same case, and have connected their common point at the modules themselves? But seperated at the supplies? Makes a huge loop huh? Area wise at least, still bad. That's going to mud things over.

If it is what I'm thinking, glad you brought it up, because it's the exact kind of "just a little wire change" I was talking about. Doesn't matter if it sounds good if isn't musical, just a little something that brings out that feeling.

If you really feel the need to give the supplies a common point it should be done nearest to the transformers/caps as possible (I think), forming the smallest loop area as you can. Gives em a more solide foundation so to speak and aside from that you won't have a huge inductor coil running around the amp with heavy ground current flowing through it. Run a short thick wire between the two supplies common, pick a spot on the middle of that wire and solder both amp grounds to it.

Far better still, don't connect them at all, and let that float from the chassis and from one another. Then make the signal grounds common, and earthed, then hack the rest of your setup to make that work.

That way your supplies and rails will all be fully isolated from one another and can't interfere with your signal ground. That will get you dancing. No dung!

BTW I've been through all kinds of wiring and a few PSU renditions.. siiiigh. Throughout all that, I've found many combinations which produced dead silence and next to no audible hiss at all. Right now I have it as close to ideal as I can get it. But the hiss is both loud and disgusting, I don't hear it from across the room, I hear it from three feet away though. It's a very ugly, dirty, switching, electronic hiss.

The reason for this is because I drive it from my soundcard, which has a big problem like most cards do with less than ideal earthing. So I floated the entire PC. Now I'm hearing that noisy ground from the SMPS as the soundcard is still connected to it. I'm this >< close to cutting some traces to cure that.

That aside, it's not loud enough to be audible from where I sit, and with the amp grounding properly referenced to the signal input and all I said above it's never sounded so good. I can make it better but I'm happy with it as is.

Moral is that hum and hiss just don't tell ya you got it right, but when you do you'll know it, at least from the people laughing at your dance moves.

Ummmm yeah, those caps. I don't see any electrolytic bypass caps for the op amps on the board, must have made em ceramic or something. There's 22uF's on mine, vishay's, 63volts, on either side of T sink next to the mosfets, they're high ESR and need to remain so, so don't change those.

I was just looking at my board and noticed my bootstrap cap on one module is at least a mm off the board. Finally a true complaint with these modules

BTW my homebrew benchmark didn't have a toroid but an E core, I can test it with a toroid sometime.

Interested in your progress. Sounds like a really good setup.

Those current reg. diodes can be found at digikey.

classd4sure said:

It's hard to envision your grounding at this hour for me.

What I got from that is that you have a dual mono construction with seperate supplies in the same case, and have connected their common point at the modules themselves? But seperated at the supplies? Makes a huge loop huh? Area wise at least, still bad. That's going to mud things over.

If it is what I'm thinking, glad you brought it up, because it's the exact kind of "just a little wire change" I was talking about. Doesn't matter if it sounds good if isn't musical, just a little something that brings out that feeling.

If you really feel the need to give the supplies a common point it should be done nearest to the transformers/caps as possible (I think), forming the smallest loop area as you can. Gives em a more solide foundation so to speak and aside from that you won't have a huge inductor coil running around the amp with heavy ground current flowing through it. Run a short thick wire between the two supplies common, pick a spot on the middle of that wire and solder both amp grounds to it.

Far better still, don't connect them at all, and let that float from the chassis and from one another. Then make the signal grounds common, and earthed, then hack the rest of your setup to make that work.

That way your supplies and rails will all be fully isolated from one another and can't interfere with your signal ground. That will get you dancing. No dung!

BTW I've been through all kinds of wiring and a few PSU renditions.. siiiigh. Throughout all that, I've found many combinations which produced dead silence and next to no audible hiss at all. Right now I have it as close to ideal as I can get it. But the hiss is both loud and disgusting, I don't hear it from across the room, I hear it from three feet away though. It's a very ugly, dirty, switching, electronic hiss.

The reason for this is because I drive it from my soundcard, which has a big problem like most cards do with less than ideal earthing. So I floated the entire PC. Now I'm hearing that noisy ground from the SMPS as the soundcard is still connected to it. I'm this >< close to cutting some traces to cure that.

That aside, it's not loud enough to be audible from where I sit, and with the amp grounding properly referenced to the signal input and all I said above it's never sounded so good. I can make it better but I'm happy with it as is.

Moral is that hum and hiss just don't tell ya you got it right, but when you do you'll know it, at least from the people laughing at your dance moves.

Ummmm yeah, those caps. I don't see any electrolytic bypass caps for the op amps on the board, must have made em ceramic or something. There's 22uF's on mine, vishay's, 63volts, on either side of T sink next to the mosfets, they're high ESR and need to remain so, so don't change those.

I was just looking at my board and noticed my bootstrap cap on one module is at least a mm off the board. Finally a true complaint with these modules

BTW my homebrew benchmark didn't have a toroid but an E core, I can test it with a toroid sometime.

Interested in your progress. Sounds like a really good setup.

Those current reg. diodes can be found at digikey.

Click to expand...

Thanks again for the advice, its much appreciated, my UCD180 is currently bolted to a piece of wood until I can get the case sorted out🙂
I'll have a play about with the earthing, just as long as it does not buz or hiss too loud, my room is only smallish and my ribbon tweeters seem to make any hiss very noticeable🙂
I can easily etch a psu board with the common joined together if it would be better, I then can use just a single earth wire going from the psu to the UCD module, what ever sounds best is fine

Your right regarding the decoupling caps for the Ad8620's, I read a post where JP recommended its worth experimenting with C28/29 regulator decoupling caps, I just looked at mine and couldn't find any so I presume that they are on the UCD400 not the 180

Now go and get some sleep😀 Its not so late here in the UK

t. said:

Thanks again for the advice, its much appreciated, my UCD180 is currently bolted to a piece of wood until I can get the case sorted out🙂
I'll have a play about with the earthing, just as long as it does not buz or hiss too loud, my room is only smallish and my ribbon tweeters seem to make any hiss very noticeable🙂
I can easily etch a psu board with the common joined together if it would be better, I then can use just a single earth wire going from the psu to the UCD module, what ever sounds best is fine

Your right regarding the decoupling caps for the Ad8620's, I read a post where JP recommended its worth experimenting with C28/29 regulator decoupling caps, I just looked at mine and couldn't find any so I presume that they are on the UCD400 not the 180

Now go and get some sleep😀 Its not so late here in the UK

Click to expand...

You're very welcome, glad to help if I can.

You can expect to run into some mean buzzing at times while experimenting but it can and should be cured, you'll be enjoying the same silence you are now. My setup had some unique requirements, and I still have options to further it, but they mostly require hacking my PC and soundcard hardware, that can wait for now.

I don't think they're on the 400 either anymore, I remember him saying they were removed "for the tweakers", so I guess we're left to experiment with that.

In actual fact I really think it would be best to isolate them from one another by all means available. Take full advantage of your dual supplies in this way.

Maybe you can do some fancy jumper work to test it both ways before you go ahead with another PCB for it. I've never actually tried that particular setup as I have a single supply only.

Sleep? I've heard of it. It's early enough here too.
Cheers,
Chris

classd4sure said:

... As far as the bias tweak yep that's it. I've no idea why people are so hesitant to try it...

Click to expand...

Chris-
You and I chatted about the bias tweak over on the hotrod thread. I was all worked up to do it, but found the CRD is already installed in the 400s that I have. I wish I could tell you how I'm liking the sound, but through an incredible bit of late night stupidity I torched one of my modules. (Too embarassed to elaborate... suffice it to say that I'm now waiting for a package from the Netherlands.)

On the grounding discussion, I'm doing monoblocks. The XLR is of course connected to the chassis at the same point as the PSU ground. It doesn't feel right to leave the mains earth out, but I recollect that this was the recommendation from the main 180 or 400 thread.

Any thoughts?

-Bill