Rolling Op-amps in Gustard H20 Headphone Amplifier

[watchnerd]

I would love to see ASR do an assessment of tube rolling.

 

[JohnYang1997]

I agree with the message that don't do this blindly. Without measurements, without unbiased critical ab comparison, it's pretty much a joke. At least you have to know the surrounding components values and the purpose of the opamp.
However opamp is just like any other components. It's a choice, a selection according to the need. Of course it will make a difference in some applications.
Also just because sparkos is expensive, doesn't mean that "best opamps" has to be expensive. On the other hand, 49720 is already a pretty expensive opamp. opa1612 more so, opa2211 even more so.
You hear pop when inserting your headphone into the plug because of offset. How many opamps have offset under 0.1mv? Is it shown on most popular measurements?
I think amirm shouldn't indicate that opamps are all the same. But that's what most people would think after reading these.
How about testing imd of opamps without load and with load? How about assesing the open loop gain and stability of opamp circuits? How about swapping opamp in objective 2 which is output stage. How about using 12db gain? What if you have an 100k potentiometer in the front?
There are many things that come to opamp selection, and no they are not the same just like any other components in the circuit.

 

[trl]

I guess these are the opamps from the voltage-amplification gain and the max. gain would probably be somewhere around 5X or so, right?
AFAIK, the opamps from I/V stage are more prone to sound change "effects" and artefacts, but also the ones from low-pass filter too, so too bad you can't test these two stages. Hope f=in the near future you'll be able to test this in another DAC.

I guess that if the measurements are about the same with both opamps, then the PCB is well-done and the PSU is indeed low-noise; perhaps bypass caps are well-placed and of enough capacitance. I'd love to see if pre & post ringing occurs with 20KHz squares when switching from Sparkos to TI.

BTW, is it me or the graph from this balanced measurement looks much better than the one from the initial review of H20 (THD+N: 116dB vs. 90dB)?

 

[JohnYang1997]

I guess these are the opamps from the voltage-amplification gain and the max. gain would probably be somewhere around 5X or so, right?
AFAIK, the opamps from I/V stage are more prone to sound change "effects" and artefacts, but also the ones from low-pass filter too, so too bad you can't test these two stages. Hope f=in the near future you'll be able to test this in another DAC.

I guess that if the measurements are about the same with both opamps, then the PCB is well-done and the PSU is indeed low-noise; perhaps bypass caps are well-placed and of enough capacitance. I'd love to see if pre & post ringing occurs with 20KHz squares when switching from Sparkos to TI.

BTW, is it me or the graph from this balanced measurement looks much better than the one from the initial review of H20 (THD+N: 116dB vs. 90dB)?

There is powersupply noise difference here. I'm not sure it's the difference between channels or the opamps.

 

[trl]

[...]
You hear pop when inserting your headphone into the plug because of offset. How many opamps have offset under 0.1mv? Is it shown on most popular measurements?
I think amirm shouldn't indicate that opamps are all the same. But that's what most people would think after reading these.
[...]

Good point with the pop thing when inserting the cans, but also worth adding the same pop noise when switching the gain adjust switch. I usually get that pop when I switch the gain button if the opamps installed in voltage amplification gain are having too much DC on the outputs (>10mV). Also, if DC-servo exists on the circuit and the output DC is too high to handle (or the DC servo opamps have been "upgraded" with an incompatible one).

There's a general opinion among audiophiles that swapping opamps will increase the performance of a DAC or amplifier. The measurements done by Amir are only proving that this is wrong and that we need to measure the performance after rolling opamps and then get the final conclusion. In the case of H20 makes no much sense to swap opamps from the voltage-gain stage, which is a very good thing actually, from a money-saving perspective.

 

[JohnYang1997]

Good point with the pop thing when inserting the cans, but also worth adding the same pop noise when switching the gain adjust switch. I usually get that pop when I switch the gain button if the opamps installed in voltage amplification gain are having too much DC on the outputs (>10mV). Also, if DC-servo exists on the circuit and the output DC is too high to handle (or the DC servo opamps have been "upgraded" with an incompatible one).

There's a general opinion among audiophiles that swapping opamps will increase the performance of a DAC or amplifier. The measurements done by Amir are only proving that this is wrong and that we need to measure the performance after rolling opamps and then get the final conclusion. In the case of H20 makes no much sense to swap opamps from the voltage-gain stage, which is a very good thing actually.

Yes.

 

[trl]

Wrap the fuses with aluminum foil ... that ensures it won't blow and keeps your $ 150.- fuses safe.

You know that some might actually do that?
Also, you need to specify if the foil needs to be 22K gold or the 18K should do as well?

 

[maxxevv]

Also just because sparkos is expensive, doesn't mean that "best opamps" has to be expensive. On the other hand, 49720 is already a pretty expensive opamp. opa1612 more so, opa2211 even more so.

Depends on which flavour of the OPA2211. The price swings from top to bottom for the 8-pin variants is a big gap.

 

[solderdude]

I think there is a substantial difference between swapping op-amps used inside a circuit where it doesn't drive the load directly and when it is used as the actual output device, when it is driving loads they really weren't designed for .

Nulling 2 op-amps is relatively easy to do (but getting perfect nulls takes some effort) if you REALLY want to find out if there are audible differences and what these are.
The 'discussions' of how op-amps 'sound' would really be very different IF all golden eared folks could do this, is all I have to say about 'audible differences' between op-amps.

 

[amirm]

I would love to see ASR do an assessment of tube rolling.

I did one test on that: https://www.audiosciencereview.com/...be-rolling-of-schiit-saga-pre-amplifier.6522/

 

[JJB70]

Even by the special standard of audiophile idiocy the idea of spending $$$s on audiophile fuses is particularly dumb.

 

[Addicted to music]

Even by the special standard of audiophile idiocy the idea of spending $$$s on audiophile fuses is particularly dumb.

You see it on every audio/DIY forum, not just fuses, but IEC power cords too where they swear there’s a difference...
Then there’s ones that claim that they can hear the effects of the solar inverter next door come on line that effects SQ...... really? ..... and yet they say they’re electrical engineers Gee, I’m severely deformed..... I must have genetically shcitt pair of ears....

 

[Pluto]

Even by the special standard of audiophile idiocy...

Brilliant turn of phrase. ROFL.

 

[Nango]

Certainly a fuse I the weakest (literally) part of an amp so one might think it matters if I strength specially these parts.

 

[3vilgenius]

........ Shall I.............. No I'm going to decline this one....

go on @March Audio let it all out

 

[RayDunzl]

Real amplifiers use Circuit Breakers.

 

[Topher5]

Hey there, I'm the guy who loaned the H20 and Sparkos. Thanks again for the review and op-amp comparison Amir!
When I unpacked the Sparkos, they certainly had the look and feel of a quality product. However not long after I popped them in, I knew I had wasted my money. After giving some listens, and comparisons swapping them back and forth, I knew I could not tell a difference if given a double-blind test. I'm not saying these op-amps might not improve other amplifier applications, but certainly not this one. Even before sending them to Amir, I sort of knew what the results would be. The only surprise was the overheating! I never play even close to max volume (more like the first few notches of volume) so I dont think I had anything to worry about.

 

[Ralf Stocker]

Hello! Please explain. In the previous test we had SINAD 90dB. In this test 116dB. What is the difference?

 

[amirm]

Hello! Please explain. In the previous test we had SINAD 90dB. In this test 116dB. What is the difference?

Former was "unbalanced" output. The latter, balanced. As I noted, the unbalanced output is poor in this system. So for rolling op-amps, I resorted to dashboard view in balanced mode so we had maximum amount of room to see any improvement/degradation.

 

[Ralf Stocker]

Thanks!