I think "FuriousD" tested it on the UCD and swwwwwore by it, likely one of the reasons we got dual (tripple) pads for it now.
He reported a much improved bass and silky mids/highs?
The only bad reports I've seen on it were those who didn't pamper it properly, including the decoupling of and more notably constant current biasing it.
I know it's far too expensive for me to try it, but I'm not that unhappy with the AD op amp either.
I think at most there's a few small ceramics bypassing the op amp of my version ~4 board.
I'm tempted to try some 220uF panasonic FM's .... possibly too big... and too low an impedance to damp resonance.
He reported a much improved bass and silky mids/highs?
The only bad reports I've seen on it were those who didn't pamper it properly, including the decoupling of and more notably constant current biasing it.
I know it's far too expensive for me to try it, but I'm not that unhappy with the AD op amp either.
I think at most there's a few small ceramics bypassing the op amp of my version ~4 board.
I'm tempted to try some 220uF panasonic FM's .... possibly too big... and too low an impedance to damp resonance.
I wouldn't bother with the OPA627 guys, I've used them (not smd) in various things, they are not bad but not that great either IMO, you do really have to give them a good supply or the sound becomes muddy and much too coloured, I had to use 100uf per rail very close to the actual chip with 100nf multilayer ceramics right on the op-amps pins with the caps legs cut short to get decent results.
The OPA627 sounded best when used with ALW super regs
I'm sure the UCD deserves something better, you may like the OPA627 but I find they have a signature similar to the AD8620 which I dislike.
It seems like the LM6172 may be the best option IF it won't cause any problems, these are probably the only op-amps I have not tried yet
The OPA627 sounded best when used with ALW super regs
I'm sure the UCD deserves something better, you may like the OPA627 but I find they have a signature similar to the AD8620 which I dislike.
It seems like the LM6172 may be the best option IF it won't cause any problems, these are probably the only op-amps I have not tried yet
t. said:
I searched the forum, and pretty much came to the same conclusion.
I've been staring at the UcD400 layout. It really sucks for the op amps by the way. about an inch of trace before the bypass caps.
Anyways, I think you can drill through the board to get a good ground close to the IC so it can likely be made to work. The 6172 is actually a stable op amp, but at that speed it can't be hack or it will be an issue.
I'm ordering some today.
Mike
Nice one Mike, we'll look forward to reading your results, there seems to be hardly any decoupling for the op-amps on my UCD180
Dear Mike:
And the rest of the spectrum??
IF you noticed less extended HF, more laid back and warm sound and the opposite with the toroids *maybe* you are experiencing psycoacoustical effect about wich Bruno spoke: a fraction of a dB in HF response is cappable of influencing all the perceived spectrum, even the bass. Plus, I believe, the toroids being so efficient and high bandwidth, amplifie HF noise. This further enhances the former aspect.
That's why I am so happy with my isolation TX 😀
A link to R-core TX (FYI)
http://eshop.diyclub.biz/product_info.php?cPath=152_74&products_id=60
Cheers
M
And the rest of the spectrum??
IF you noticed less extended HF, more laid back and warm sound and the opposite with the toroids *maybe* you are experiencing psycoacoustical effect about wich Bruno spoke: a fraction of a dB in HF response is cappable of influencing all the perceived spectrum, even the bass. Plus, I believe, the toroids being so efficient and high bandwidth, amplifie HF noise. This further enhances the former aspect.
That's why I am so happy with my isolation TX 😀
A link to R-core TX (FYI)
http://eshop.diyclub.biz/product_info.php?cPath=152_74&products_id=60
Yesterday I saw a post on "My_ref" thread who swares that double-mono PS configuration wired "out of fase" enhanced significantly the bass. I'll try to find it.
Cheers
M
Double mono wired out of phase...... huh? You wire bi-phase to make more efficient use of the supply. Can't do it with dual mono supplies.
Seriously I had appreciable sonic differences with wiring bi-phase, nor any downside, so I left it that way as it at least helps do away with pumping and "should" be better.
I'd think if you do notice a big difference you're likely working with a ..... low grade PSU?
Seriously I had appreciable sonic differences with wiring bi-phase, nor any downside, so I left it that way as it at least helps do away with pumping and "should" be better.
I'd think if you do notice a big difference you're likely working with a ..... low grade PSU?
Hi Chris:
This is not mine but forum member *rudi*'s observation on "my_ref" thread.
I found it interesting.
I don't know if the technical name I gave is correct. Please see his drawing:
http://www.diyaudio.com/forums/attachment.php?postid=737780&stamp=1128518691
A quote from a friend of his:
Sorry to interrupt the flow of the discussion
Regards
M
This is not mine but forum member *rudi*'s observation on "my_ref" thread.
I found it interesting.
I don't know if the technical name I gave is correct. Please see his drawing:
http://www.diyaudio.com/forums/attachment.php?postid=737780&stamp=1128518691
A quote from a friend of his:
Sorry to interrupt the flow of the discussion
Regards
M
Hmmmm, I couldnt' recommend doing that and only see disadvantage to it, namely that of giving up the advantages of having dual mono in the first place. Bad idea.
This is not wiring the amps in opposing phase either... I'd go so far as to say improvements of doing this were dreamt.
This is not wiring the amps in opposing phase either... I'd go so far as to say improvements of doing this were dreamt.
maxlorenz said:Dear Mike:
And the rest of the spectrum??
IF you noticed less extended HF, more laid back and warm sound and the opposite with the toroids *maybe* you are experiencing psycoacoustical effect about wich Bruno spoke: a fraction of a dB in HF response is cappable of influencing all the perceived spectrum, even the bass. Plus, I believe, the toroids being so efficient and high bandwidth, amplifie HF noise. This further enhances the former aspect.
That's why I am so happy with my isolation TX 😀
A link to R-core TX (FYI)
http://eshop.diyclub.biz/product_info.php?cPath=152_74&products_id=60
Yesterday I saw a post on "My_ref" thread who swares that double-mono PS configuration wired "out of fase" enhanced significantly the bass. I'll try to find it.
Cheers
M
M,
Everything was better. Its more dynamic, stages better, mids and highs more present and has a more convincing "you are there" aspect to it. Not just more bass. Larger bypass caps have that, and trash the rest of the spectrum. (That might just be lower ESR, I'm not sure yet.)
It could be 1) differnt typass caps, dual bridge wiring, or IE cores or all of the above. Also better RCA and binding post could be a factor.
I agree with Bruno though. It doesn't take much to really tilt the balance. I once designed a podium mic that got beat up by the sales guys as really sucking compared to brand A. I measured it and it had a 2dB bump at 100Hz. (I thought flat was the goal, fool me!) I put a 3dB bump and they just raved about it. 🙂
maxlorenz said:Hi Chris:
This is not mine but forum member *rudi*'s observation on "my_ref" thread.
I found it interesting.
I don't know if the technical name I gave is correct. Please see his drawing:
http://www.diyaudio.com/forums/attachment.php?postid=737780&stamp=1128518691
A quote from a friend of his:
Sorry to interrupt the flow of the discussion
Regards
M
I agree with Chris, in that I can see absolutely no benifit, and only disadvantages. Unless I'm missing something it seems like a good way to loose the advantage of dual mono. There could be some advantage in situations were one channel was being driven much harder than the other, other wise its going to couple one amp to the other more directly, which isn't a positive.
Also note his improvement was vs a single transformer!
That then explains that it would likly be better, but likely better yet with seperate dual mono.
regards,
Mike
JoshK said:
Simple trick. On a PSU that's driving stereo channels, you simply invert the inputs of one channel, and the outputs of that same channel.
Since recorded bass signals are somewhat "mono" for both channels, it causes them to draw power from opposing rails for low frequencies instead of having them both load the same rail.
Therefore you can expect better supply sharing for low frequency "stereo" signals and also a cancelled pumping effect.
Since you're sharing the rails more equally as each rail sags equally (ideally) it enforces the opposing rail, you're using the pumping effect to your advantage.
The outputs remain in proper phase, and it really has no effect on the highs at all.
It's a very simple tweak for anyone using a stereo single supply setup.
Mike,
In a word, No. I’ve been using the 1000ufd as sort of a compromise between the 470ufd M cap and the FC cap, in a sense trying to get the best of both worlds. But after extended listening with several different caps, I’ve come to the conclusion the UCD was designed for the Yageo/M cap, and I plan to put the 470ufd M cap back this weekend. I have chased this dog enough. I don’t think the answer to the dynamics issue lies here.
Which leads me to a probably long-winded explanation of where I stand now, which is in line with the current discussion of the Op-amps. Over a year ago I thought I would try the UCD boards based on things I read here and other places. I happened to have an old amp with a transformer of the right voltage so I picked up a pair of the original ST boards and built up the amp. My impressions where very positive, but one of my first thoughts was where’s the beef? I naturally concluded it needed a better power supply. So I built up a nice dual mono supply with 625VA trannies, dual rectifiers and 60,000ufd per side. Fired it up and thought, much better but, where’s the beef? I have amps with less power that sound more dynamic than this. So the next logical step was the Op-amp. I installed the AD8620 having used it with good results in other designs. Fired it up, and again thought better but, where’s the beef? Next I was thinking about biasing in to class A as I have done before. That’s about when Chris brought this up. So when the CRD thing came up I thought of course, and jumped on it thinking this is it. Once again better, but WTB? In the mean time I started playing with the various caps and other things on the board. Of which the two that stand out are the input caps, which I removed and the Wima’s in the output filter. The next step was installing separate power supplies and regulators to the Op-amps. This cleaned things up nicely and brought the sound up to a new level, but still no beef.
When I received the new V6 boards with the extra Op-amp pads I thought right away of the BB OPA627’s. But on further reflection I really don’t think chasing the Op-amp monkey is the solution either. Yes, the AD and BB chips sound different than each other and changing them will change the overall sound. But based on the differences I heard between the ST and AD boards I don’t feel this will solve the dynamics issue. I have made numerous comparisons between the BB and AD chips over the last few years and found in a nutshell the AD chips to be a little on the cooler/analytical side and the BB to be on the warmer/smoother side, if that makes sense. Both are very good and depending upon the application they are in one can sound better than the other. Looking at the specs though between the AD8620 and the OPA627 the AD8620 smokes the OPA627.
At this point I’m not sure I have any solutions, although I suspect Mike may be correct that some of the solutions may be under the epoxy. In other words the “tube like” sound may be built into this amp with no way to really change it. This is not really a bad thing and before anyone gets the wrong idea I really do like this amp. In many ways it is far and away the best overall sounding amp I have owned. I just have a sense that it is capable of more. In other words it could be a real giant killer or poor mans reference amp if you will. Which in a way is what I have been trying to get out of this amp. Perhaps if we buy Bruno a Bier he would shed some light on this. Discreet inputs (I have been waiting for some to go with tubes here J) may be the solution, I don’t know.
Anyway, just my random thoughts. Hope this helps.
“Sorry to take up all your sweet time. I’ll give it back you one of these days.” J.H.
Mike2 said:
Mike2,
Perhaps it is just "in there" and we can't touch it. If it were easy, the first place I'd go would be the input. My experience is front ends can win or loose the beef. This is a comparator, so maybe all bets are off on this, don't know until its tried, which isn't going to be trivial.
There is lots to be gained in linear amps by having the input pair perfectly balanced with zero delta in operating current. The current source for the pair also can make a significant difference in linear amps.
Resistors can also make a differnce and to be honest, I've never had to use surface mount ones. I've got some good surface mount resistors on order for the front end, but they are back ordered. Mouser PFC-W0805LF-03-1001-B.
I'm pretty sure the surface mount resistors could by themselves be responsible for eating the beef!
Problem is again that some of them are under that epoxy.
Why why why do that? Just to make it hard?
I will see if the LM6172 can be make stable this week. Longer to put them in the amp for other reasons.
I like that they are bipolar inputs too, in that when done right, I find them better.
Regards,
Mike
I'm sure of that will make some difference, though I'd argue that an AC coupled bipolar input has to be worse than a DC coupled Jfet....
I will simply say that the two changes that brought the beef out in my system were the low impedance supply .... having enough capacitance but not excessive, and the FC decoupling caps.
If putting in FC caps did nothing at all for low end grunt I'd have to think there's a bottleneck further upstage... like in your supply?
Two obvious reasons for the epoxy.... 1. Keep is somewhat secret.. (III know....but the odds of it costing you to see it are increased)
2. Better temp tracking than not having it. (bonus)
Unbalanced/matched pairs and such would do more harm on microdetail where this amp seriously excels than in the low end grunt, and it is fully differential anyway, so it's not much of a problem.
Experiment with your supply more, shouldn't have to dig into it to that extent to get some bass from it.
I will simply say that the two changes that brought the beef out in my system were the low impedance supply .... having enough capacitance but not excessive, and the FC decoupling caps.
If putting in FC caps did nothing at all for low end grunt I'd have to think there's a bottleneck further upstage... like in your supply?
Two obvious reasons for the epoxy.... 1. Keep is somewhat secret.. (III know....but the odds of it costing you to see it are increased)
2. Better temp tracking than not having it. (bonus)
Unbalanced/matched pairs and such would do more harm on microdetail where this amp seriously excels than in the low end grunt, and it is fully differential anyway, so it's not much of a problem.
Experiment with your supply more, shouldn't have to dig into it to that extent to get some bass from it.
classd4sure said:
Chris,
Thanks. I'm not sure what I'm looking for can be discribed simply as bass. Its dynamics. I will say that since trying them with a different supply they have improved. I hear you that this might need more attention. Maybe I'll buy a case of T caps and see what it does.
On the bipolar op amps, not sure it necessarily means you need the coupling caps now that there is an offset trim option.
All this, and I keep forgetting that there may be something to be gained by the dead time adjustment too. Something I care not to play with out of a lab and good instrumentation.
I think the FC's are not for me though, unless changing the main supply can help them, and I don't see it helping. I think bruno designed the ESR of that bypass cap for the ferrites and the 5 mohm and ceramic bypass. To change that and not have things get worse will likely require all those to be tweaked, or at the very least, the ceramic cap.
Regards,
Mike
ghemink said:
Gertjan,
Good news on the 5 mohm and tweaking the bypass ceramics.
The 5mohm clearly is used to sense something, current limit likely as you suggest. The good news is that its divided down by a series R-R||C. Thus, one could increase this value, allowing the ceramic to be increased too, and the over current limit could then be corrected by changing the divider.
btw, the final ceramic bypass is 100nF.
Regards,
Mike
t. said:
yes in my experience , the 627 and 8620 are similar ( artificial treble , sound a bit compressed )
after trying a lot of op amps ( but not the 6172 ) with the ucd180 and a DAC output , i now use systematically the ad826 ( biased at 5mA ) with a little unregulated psu and i totally agree with Dejan conclusion :
http://www.zero-distortion.com/tests/x1.htm
(see "tweaking > choosing an op-amp"or "tweaking a yamaha cxx993"
Alain
rha61 said:
Hi Alain,
Appreciate the tip and the link to the article, an interesting read.
The AD826 does seem like it could be a decent choice, I think I'll give it a try if I can.
As per the article and his conclusions with respect to your own...
I dont' think I can agree 100% with his method of decoupling. First he seems to say results can't be predicted, then turns around and say's "just use this, don't think, do it", and it seems like it would rrrrrriiiiiinggggggg.
How do you then decouple them?
More issues arrise from the following:
"To give you an idea, I'll mention one such quirk on my current favorite, AD826. Look at its output and you will see it's not zero as it should be, but is in fact offset by 0.6V. The cure is to insert a standard diode (like for example 1N4148) from the output to the minus supply line, and presto, you have zero at output.
I honestly couldn't say if I heard any difference with or without that diode, but I know I wouldn't be able to sleep knowing an offset was there and not having done something about it.
Another popular trick is to force the op amp output into "pure class A" by connecting a resistor from the output to the positive supply line. While this may work, personally I'm dead set against such measures, which I feel are practically a rape of the op amp. Any forcing into any mode is not good, force never yields long term results.
Although, frankly I can afford to see it that way, since AD826's output stage is working in full class A anyway, so I have no reason to resort to tricks like that."
-An offset of 0.6V seems excessive does it not. Did you include the diode, or trim it? I guess in the INA like front it the symmetry ought to cancel this though. These seem borderline for DC coupling capability, have you tried it?
-The data sheet (according Bruno's gospel) would seem to suggest tying it to the negative rail with this op amp.. which is good as it suits the modules for the CRD (newest ones excluded as they have provisions for both rails I believe).
I trust you biased it to the negative rail as per the PSRR would seem to indicate is required?
One last thing, you're using an unregulated auxiliary supply? Can you go into any detail on that?
Cheers,
Chris
PS: At this point I'm rather curious to know if the sound of the FC's would perhaps be more natural with a more neutral op amp 🙂
classd4sure said:
Hi Chris
for the unregulated psu , i use a 30VA Rcore transformer with primary/secondary screen , a RURP860-based bridge and a CRCRC filter , the last C at the opamp pins is a 47uF ELNA Silmic 2 alone and the 826 is stable
the 826 is biased ( cascode 2SK170 FETs) by the negative rail of course
with a 10K input R at the input stage , i obtain a 50mV DC offset at the module output with no coupling cap
Alain
PS : Is the improvement important with the mkp Panasonic cap in the output filter ? Why have you choose a lower value than 680nF ?
- Status
- Not open for further replies.