Portlandmike said:
I think you have problem in your setup, micro/macrodinamics and control is what , after living 10 months with ucd, still impress me. Every component is very chain dependant and react very differently in the same system.
Caps and stuff
T,
I am using 2-2.2uf @ 200 volt Auricaps per module to enable the use of the 7162. Things have never sounded so lifelike. I wouldn’t hesitate to recommend them even though I don’t make a penny on sales and being associated with Audience has nothing to do with how well they work and sound. Some times it will startle you if a sound comes along you aren't expecting. The impact and shimmer of cymbals is particularly note worthy. This is usually the first thing I notice about a live performance. Always thinking I wish I could get cymbals to sound like that at home, now if the recording is any good I am close. Things like telling a Strad from a Yamaha and a Steinway from a Bösendorfer have always been there but now they have more of their real character. Getting closer all the time! I will have to try a discreet front end design to be sure. It will be a surface mount update of a successful opamp circuit I did for CD and DVD player mods but to tell the truth I don’t think the 7162 will give up much to it. It too is an all bipolar design so will also require the double input caps. I am not sure what you mean by “folded cascade”?
Chris,
I think you are right as I feel the same way. I feel the lack of dynamic compression is one of the UcD’s outstanding strong points. This could be some sort of bad amp/speaker/cable interaction.
Roger
T,
I am using 2-2.2uf @ 200 volt Auricaps per module to enable the use of the 7162. Things have never sounded so lifelike. I wouldn’t hesitate to recommend them even though I don’t make a penny on sales and being associated with Audience has nothing to do with how well they work and sound. Some times it will startle you if a sound comes along you aren't expecting. The impact and shimmer of cymbals is particularly note worthy. This is usually the first thing I notice about a live performance. Always thinking I wish I could get cymbals to sound like that at home, now if the recording is any good I am close. Things like telling a Strad from a Yamaha and a Steinway from a Bösendorfer have always been there but now they have more of their real character. Getting closer all the time! I will have to try a discreet front end design to be sure. It will be a surface mount update of a successful opamp circuit I did for CD and DVD player mods but to tell the truth I don’t think the 7162 will give up much to it. It too is an all bipolar design so will also require the double input caps. I am not sure what you mean by “folded cascade”?
Chris,
I think you are right as I feel the same way. I feel the lack of dynamic compression is one of the UcD’s outstanding strong points. This could be some sort of bad amp/speaker/cable interaction.
Roger
Roger
My experience with the 6172 from other applications are the same as yours soundwise - very good !
What you are saying is that with the 6172 you will have to place a dc blocking on the input as well as on the output, right?
How about biasing it into class A?
Another good candidate for good sound is the AD797. The real drawback here is of course that it is only a single OPamp so you will have to make an extension board. It is also a little hard to tame the HF instability. But it sounds very good.
koldby
My experience with the 6172 from other applications are the same as yours soundwise - very good !
What you are saying is that with the 6172 you will have to place a dc blocking on the input as well as on the output, right?
How about biasing it into class A?
Another good candidate for good sound is the AD797. The real drawback here is of course that it is only a single OPamp so you will have to make an extension board. It is also a little hard to tame the HF instability. But it sounds very good.
koldby
koldby said:
AD797 is more for pre-amplifier as RIAA or microphone where we need high gain at low level. Use it as a input stage buffer in the UcD is not ideal.
Re: Caps and stuff
Thanks Roger🙂 Realism is the thing I want and if fitting the LM6172's will get me closer that is great because with the AD8620 my UCDs is probably my third favourite amp, I want it to be my best one😉
I think the discrete circuit Bruno uses is a folded cascode, theres a few op-amps thats based on this like the AD826
sx881663 said:
Thanks Roger🙂 Realism is the thing I want and if fitting the LM6172's will get me closer that is great because with the AD8620 my UCDs is probably my third favourite amp, I want it to be my best one😉
I think the discrete circuit Bruno uses is a folded cascode, theres a few op-amps thats based on this like the AD826
Re: Re: Re: opamps?
Thanks Stef,
I do not like the Silmic caps at all but the BG NX Hi-Qs are not bad, I've never tried the Rubycon NA but I have the ZA and ZL which is my current favorite decoupling cap, never though of trying those as coupling🙂
stef1777 said:
Thanks Stef,
I do not like the Silmic caps at all but the BG NX Hi-Qs are not bad, I've never tried the Rubycon NA but I have the ZA and ZL which is my current favorite decoupling cap, never though of trying those as coupling🙂
Rubycon NA are bipolar non polarised caps as BG N or NX.
Rubycon ZA or ZL are for PSU (very good). It's better to do not use them for coupling.
Rubycon ZA or ZL are for PSU (very good). It's better to do not use them for coupling.
patriz said:
I'm happy for you. My experience is that with another custom built "quad diff" A/B bipolor, hand built, it was much superior, not as laid back. I agree there are many things the UcD does right, but what it lacks (like all cept one amp I've had) is the live dynamic element. The drums and precussion, they are way back in the stage, and don't convince me they are real, like when I'm at a small venue live preformance. I've had that, and prefer it over stellar staging, which incedentally, I usually don't hear at live concerts, but I sure like it in my system.
Mike
koldby said:
I don't think you need to do it twice. Once should be good enough. Likely at the input since (if your lucky enough to have a UcD400 6.1, which has an output offset trim pot. You also don't need 100k input resistors to ground, it could be 10k. The input topology of the UcD op amp is fully balanced, so offsets bias should cancel to a large degree, and the input resistance is very low on the one side, like under 500 ohms.
I'm not sure you'd actually need AC coupling for sure. I'll look at it more closely. (since there is trim!)
Also, UcD400 6.1revs have sockets provisions for single op amps and current source loading from either rail.
Regards,
Mike
ghemink said:
Thanks, I suspect your right. Then again, you could add it in series with the cap😉
Mike
classd4sure said:
Chris,
Not sure about any of those.
At first, I had issues with my discrete single ended preamp, but after some good tips from a expert friend on electrolytic coupling, and a few weeks of break in, its awsome. More of the live dynamics than before, but still showing up short.
Differential........ It just needs to work, and need not be true differential.
A UcD less the op amp is a diff amp input. All that matters is that both inputs are low impedance. Having them be equal and opposite may be best, but only if you get there without causing more problems.
You lucky folk with transformers might want to try this if you can get the gain. In fact, a step up transformer driven by a simple follower might be ideal. Care about offsets needs to be had, but it may be a very good solution, although I've never heard one.
Fairly high bandwidth. That for sure I don't agree with in that the UcD bandwidth is so limited. Also, and more importantly, there is likely advantage in NOT having it be high bandwidth since a) its not in the loop, and b) out of band stuff will just modulate the inputs.
Just my opinions though.
Mike
LM6172 could work w/0 AC coupling.
I've looked at this op amp for the UcD circuit closely, and I don't think the offset bias current will be much an issue as long as all is DC coupled, or symetrically AC coupled. (Yes, if you need to AC couple a single ended preamp, then the same thevinin AC coupling needs to apply to the other input.)
That said, to a very large degree the offsets cancel.
The input offset current is high, 20nA, but the impedances are low too, like sub 500ohms on the inverting inputs and on the non-inverting inputs it will depend on AC coupling of not.
The bias offset current will be 20nA x 100k 2mV at a gain of 2k/560=3.6 or 7.2mV. That's typ numbers.
The rev 6.1 trim range is at least 30mV, maybe more.
For non-rev 6.1 applications.... I think the coupling caps should be left in, or put back.
My bigger concern would be oscillation as this is a well behaved op amp but its got gobs of bandwidth 100Mhz. There is no way
to roll this off much either in the circuit configuration given in that the best a feedback cap could do is reduce the differential gain to 1.
I'm ordering some from digikey. There what's in my SACD 777 by the ton, and I know they have good sound.
Mike
I've looked at this op amp for the UcD circuit closely, and I don't think the offset bias current will be much an issue as long as all is DC coupled, or symetrically AC coupled. (Yes, if you need to AC couple a single ended preamp, then the same thevinin AC coupling needs to apply to the other input.)
That said, to a very large degree the offsets cancel.
The input offset current is high, 20nA, but the impedances are low too, like sub 500ohms on the inverting inputs and on the non-inverting inputs it will depend on AC coupling of not.
The bias offset current will be 20nA x 100k 2mV at a gain of 2k/560=3.6 or 7.2mV. That's typ numbers.
The rev 6.1 trim range is at least 30mV, maybe more.
For non-rev 6.1 applications.... I think the coupling caps should be left in, or put back.
My bigger concern would be oscillation as this is a well behaved op amp but its got gobs of bandwidth 100Mhz. There is no way
to roll this off much either in the circuit configuration given in that the best a feedback cap could do is reduce the differential gain to 1.
I'm ordering some from digikey. There what's in my SACD 777 by the ton, and I know they have good sound.
Mike
Toriod vs standard xfrmer findings
I've read that some people don't think toriods are as good as Rcores. I'm not sure what an Rcore is, but I just put some well broken in UcD400's that were in toriodial monoblocks into a chassis with good quality "old fashon" type transformers. I'd call them IE, but..... who knows.
Anyways, The amps are stock, with removed coupling caps. The bass improvement was more signficant than my experience with switching out bypass caps and it had none of the negatives. Maybe its true toriods aren't best!
Its not quite apples to apples though.
The other difference to be noted was the IE core chassis was wired with dual bridges, one per rail, and one power supply for two amps wired out of phase. The bridges were 50A standard, not fast recovery.
The toriod solution was a monoblock with centertap, one bridge per amp in the toriod solution. The bridge is standard, but not has high of current, don't know, but its much smaller.
Mike
I've read that some people don't think toriods are as good as Rcores. I'm not sure what an Rcore is, but I just put some well broken in UcD400's that were in toriodial monoblocks into a chassis with good quality "old fashon" type transformers. I'd call them IE, but..... who knows.
Anyways, The amps are stock, with removed coupling caps. The bass improvement was more signficant than my experience with switching out bypass caps and it had none of the negatives. Maybe its true toriods aren't best!
Its not quite apples to apples though.
The other difference to be noted was the IE core chassis was wired with dual bridges, one per rail, and one power supply for two amps wired out of phase. The bridges were 50A standard, not fast recovery.
The toriod solution was a monoblock with centertap, one bridge per amp in the toriod solution. The bridge is standard, but not has high of current, don't know, but its much smaller.
Mike
Typically toroid are wound without IS shield meaning they have high capacitive noise coupling. With some power conditioning experiments, a friend and I determined that this has quite a negative subjective affect on the sound. EI tx's usually have much lower capactive coupling due to the way they are wound.
I recently purchased a pair of Plitron tx's with the IS shield option. I haven't tried them out yet, but plan to compare them to the UcD modules with standard tx's.
By the way, is a SOIC format small enough for the opamp on the ucd400? I am not familiar with all the sizes and such. I heard that the OPA2107au is the cat's pajamas of opamps and doesn't have stability problems.
I recently purchased a pair of Plitron tx's with the IS shield option. I haven't tried them out yet, but plan to compare them to the UcD modules with standard tx's.
By the way, is a SOIC format small enough for the opamp on the ucd400? I am not familiar with all the sizes and such. I heard that the OPA2107au is the cat's pajamas of opamps and doesn't have stability problems.
I haven't tried without but my plitrons also have static shield and there's zero hum, mechanical or otherwise, and no detectable line noise comming through. Well worth it.
JoshK said:
The op amp sockets on the UcD's are SOIC's SO-8 specifically.
The UcD400 Rev 6.1 has provisions for single or dual SO-8 IC's.
Interesting comment on the IE's. But doesn't explain the improved bass does it?
I was testing a toriod at high load. 35V AC, 15000uF bypass, 5 ohm load. I used a current probe to measure secondary currents. It limited at ~40A. I wonder if IE's do better and refresh the caps faster.
From looking at a toriod supply and my UcD's, I can't imagine doing better since the rails didn't budge a volt even under clipping with music.
The OPA2107 should be fine. You might try it. The LM6172 distortion is >100dB down at 20kHz though. Much better if it doesn't oscillate.
I know they sound good as my CD player has about 5 per channel in the path and it sounds great.
Mike
Hi,
If you have the provisions for the single op amps it may really be worth trying a opa627 ~! I know, they're not cheap.
If you have the provisions for the single op amps it may really be worth trying a opa627 ~! I know, they're not cheap.
Don't drop in the LM6172 just yet!
Just looked close at the layout of the UcD400 (rev 6.1)
the layout for the op amps bypassing is .... lets just say lacking.
I'd bet there isn't a chance in hell to just drop in a LM6172 and have it not oscillate.
There might be hope with some modding and additional bypass caps.
I'll let you all know as I have a dead UcD to try it on.
DO NOT drop in the LM6172 into the circuit yet. It might be capable of blowing up the amp if it oscillates.
Regards,
Mike
Just looked close at the layout of the UcD400 (rev 6.1)
the layout for the op amps bypassing is .... lets just say lacking.
I'd bet there isn't a chance in hell to just drop in a LM6172 and have it not oscillate.
There might be hope with some modding and additional bypass caps.
I'll let you all know as I have a dead UcD to try it on.
DO NOT drop in the LM6172 into the circuit yet. It might be capable of blowing up the amp if it oscillates. Regards,
Mike
OP627
I have had no experience with this new UcD version or with the 627 and wonder if it is as critical as the 6172 for decoupling. I doubt the decoupling caps will all be installed on the new module. It wouldn't hurt to do some research and get the best that will fit the supplied pads and also replace any that are there now with these better parts.
By the way on the input offset for the 6172, input resistors of 100k are fine as long as both +/- inputs are equal and the dc paths on both are broken with coupling caps. With single ended direct coupling the source impedances must be very low for there to not be a problem with the older modules offset. I don’t know if the adjustment range is enough on the new ones. Also this adjustment must be redone for any preamp change. This is fine for your own stuff but doesn’t work if the unit is for someone else with no tech savvy.
Roger
I have had no experience with this new UcD version or with the 627 and wonder if it is as critical as the 6172 for decoupling. I doubt the decoupling caps will all be installed on the new module. It wouldn't hurt to do some research and get the best that will fit the supplied pads and also replace any that are there now with these better parts.
By the way on the input offset for the 6172, input resistors of 100k are fine as long as both +/- inputs are equal and the dc paths on both are broken with coupling caps. With single ended direct coupling the source impedances must be very low for there to not be a problem with the older modules offset. I don’t know if the adjustment range is enough on the new ones. Also this adjustment must be redone for any preamp change. This is fine for your own stuff but doesn’t work if the unit is for someone else with no tech savvy.
Roger
classd4sure said:
Chris,
You know, I dismissed that op amp because in a seach of the forum, most said the 8620 bested it.
Maybe its worth a try. Its an 80 dollar test though.
Mike
- Status
- Not open for further replies.