Discrete sounds better than integrated? a possible reason

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
5v333 said:
well in that case i guess it must have been a factor if Rupert didnt isolate his VUs. did he?

sorry  if i was dragging the thread away from the main point that Ian was making about feedback/gain setups.
You are right, VU meters were not isolated/buffered in those days. However, when Neve mixers were tested, there was a standing order to 'switch any VU away" from the output being tested.

Cheers

Ian
 
Heres an interesting (short)article , its around a long time now ,

https://www.audionote.in/the-negative-effects-of-feedback
 
Hello

ruffrecords said:
You are right, VU meters were not isolated/buffered in those days. However, when Neve mixers were tested, there was a standing order to 'switch any VU away" from the output being tested.

Cheers

Ian

5v333 said:
thats called cheating, right? :)

The nevegate  ;D
Faked THD emission measurement for aproval  8)
I guess the one programming the software that recognise the test bench situation and minimise compromising data will have trouble wih regulation :)

Best
Zam
 
Tubetec said:
Heres an interesting (short)article , its around a long time now ,

https://www.audionote.in/the-negative-effects-of-feedback
I put it in the same category as Matti Otala's Transient InterModulation, which as long since rejoined Zenon of Elea's paradox, where the arrow can never attain the turtle.
 
This is a very old debate often from marketers pimping some new design.

Test equipment has progressed measurement resolution way beyond what I can hear, and I suspect everyone else here, but I try not to argue with people on the internet about that they say they can hear.

I am a fan of null testing. If you suspect an audible difference between IC and solid state designs, the first step is objective test bench measurements. If this doesn't reveal an obvious difference, a null test will isolate and reveal differences, while it won't tell you which one is causing the difference. 

If this is an old argument does that make it better. :)

JR
 
abbey road d enfer said:
I put it in the same category as Matti Otala's Transient InterModulation, which as long since rejoined Zenon of Elea's paradox, where the arrow can never attain the turtle.
I remember TIM and the alphabet soup of "new" distortions, pretty much all variants on slew limiting phenomena. I don't know what it is about about audio electronics that makes people think it is not a mature technology, rich for new discoveries.

JR
 
Could it be the case that many monolytic op amps are just more demanding when it comes to implementation?


I have noticed very audible differences with correct vs. incorrect (chassis) grounding (cable shield only connected to chassis at the in-/outlets and audio ground only connected to chassis via the PSU ground input vs. other schemes). Let RF into the circuit and the op amps won't perform as good as they can.

Same with decoupling, for instance there are many ways to create resonances in the impedance response with the wrong combination of PSU decoupling caps.
 
Would this not be easy (at least in concept) to test by using an uncompensated op-amp in a given circuit?  Test A would be compensated with a dominate pole down around 10Hz (as per normal), and test B would be uncompensated, with feed-forward capacitance with a roll-off at 75 kHz?
 
Matador said:
Would this not be easy (at least in concept) to test by using an uncompensated op-amp in a given circuit?  Test A would be compensated with a dominate pole down around 10Hz (as per normal), and test B would be uncompensated, with feed-forward capacitance with a roll-off at 75 kHz?
One wonders why it's not been done before...
 
abbey road d enfer said:
One wonders why it's not been done before...
I'm trying to learn as much as anyone else here...if you think it's a dumb idea, you can just say it's dumb, and I promise I won't be offended.  :) Ian seemed to be postulating a reason why two supposedly equivalent topologies might sound different, and I offered what I thought might be a (hopefully simple) way to answer his question, but it's also possible I just swung and missed.
 
I encourage honest inquiry but am repeating myself, this is mature technology.

There are no dumb questions, but unfortunately lots of ASSumptions that live up to the shorter spelling.

JR

PS: Test what exactly..? Dominant pole capacitor compensation is just one of multiple factors affecting a given circuit/s transfer function.
 
I suppose really what we want to test is how long is the recovery time of your circuit after a transient event .
Maybe an initial pulse just to get your circuit into 'tail wags dog' territory followed shortly after by a more nominal level steady state signal , not really sure if REW can be set up for that kind of measurement ,maybe more of a job for a Digital storage scope .
 
ruffrecords said:
I think this should be required reading for anyone designing with op amps like the TL072:

https://www.ti.com/lit/an/sloa020a/sloa020a.pdf

Cheers

Ian

That was not published until decades after I started designing with TL07x op amps (in the 1970s).

I did discuss amplifier stability concerns in my 1980 article about console design.

Again not exactly breaking news to those skilled in the art.

JR 

PS: The TL07x was hot stuff back in the 70s, now we have better devices available. But properly applied the TL07x still does not suck today.

[edit= sorry this sounds a little dismissive.  Tomorrow I'll share a story about NF /edit]
 
Tubetec said:
I suppose really what we want to test is how long is the recovery time of your circuit after a transient event .
Maybe an initial pulse just to get your circuit into 'tail wags dog' territory followed shortly after by a more nominal level steady state signal , not really sure if REW can be set up for that kind of measurement ,maybe more of a job for a Digital storage scope .
Just run two circuits in parallel, feed them signals of opposite polarity, then mix the output and trim the levels to minimize the output. What's left is the difference. AKA null testing as JR has already said.

Note that the LF transient / modulation I mentioned before was not in reference to individual amplifiers like op amps or the discrete equivalent. I personally do not believe in that sort of thing at all (how op amps "sound" and such). But when you consider the entire circuit that the amplifiers are in, with filter networks, asymmetric and / or rectifying elements, there could be transient behavior that gets overlooked by folks doing simple THD measurements with a 1kHz tone.
 
ruffrecords said:
i do not know if it was deliberate or not, but this technique of varying open loop gain at the same time as closed loop gain was used in all Rupert Neve's three transistor class A designs of the late 1960s and in the Helios and Cadac three transistor and Calrec four transistor designs of the same era.

Thank you so much for this thread Ian, and for your explanations, what you wrote in the first post makes sense and I actually think you might be right there.

Just wanted to add up that as far as the Helios goes, I really think besides the electronic circuit, Helios achieved a cult status also because of all the amazing records from the 70s that were recorded or mixed in Helios.
Led Zeppelin, Rolling Stones, Deep Purple, Bob Marley, Jimi Hendrix, Joy Division, Black Sabbath, Eagles, Eric Clapton, The Who....

I love Helios stuff, but I ask myself sometimes it's the fact that it was used in so iconic albums or the circuit that makes the cult...
Could an inferior circuit with the same credits achieve a cult status also?

 
Tubetec said:
I suppose really what we want to test is how long is the recovery time of your circuit after a transient event .
Maybe an initial pulse just to get your circuit into 'tail wags dog' territory followed shortly after by a more nominal level steady state signal , not really sure if REW can be set up for that kind of measurement ,maybe more of a job for a Digital storage scope .
Recovery is about how a circuit reacts after it has been submitted to a signal that puts it out of its linear operation, typically after an overload or a slew-rate limitation.
Regarding the former, some discrete circuits have an edge here, if they take advantage of bootstrapping or staggered rails, which maintains normal operation of some elements even when the rails are hit.
About slew-rate limitation, it is long known that it should be avoided by placing a passive LP before the input.
It is what should define the final power BW of the compound.
The consequence is that the actual power BW of the active stage must be higher than that of the passive filter.
Failure to achieve this is what gave some circuits bad rep.
The "tail wags dog" situation you describe became an issue when audiophools tried to make their power amps capable of passing AM radio frequencies. Problems that could be put in evidence in the lab were non-existant in the field, whatever they said at the time.
Indeed, some discrete circuits, using local NFB or degeneration and very little or no global NFB, having a much higher slew-rate, seldom show slew-rate limitation problems.
 
Back
Top