I bought a Vivid V6 single opamp with my own money - definitely not a paid shill like many other Burson reviews I see!
I've measured its distortion performance using the QA403 audio analyzer and here are the graphs, first distortion v. level for both inverting (gain=-1) and non-inverting (gain=+1) topologies:
Clearly this opamp has decent (but not world class) distortion performance. Powered at +/-15V. 600 ohms load seems to be handled in its stride.
Alas the power consumption is a massive 13mA with no signal (the only pitiable specimen of a datasheet I could find claimed 8mA). That same datasheet only claimed 0.02% for distortion, which perversely under-sells it - perhaps the datasheet is not for the Vivid V6?
Little bit let down at higher levels, but most practical uses will be somewhere in the 0 to 10dBV range.
Frequency response:
No signs of performance dropping at HF.
19kHz/20kHz IMD:
Again decent performance in line with other graphs.
So the only elephant in the room is the cost - there are many opamps with similar or better performance that are much cheaper - for instance OPA1642 will do rather better with only 1.8mA per opamp...
The high power consumption makes it run at 50C or so which isn't great.
The packaging it came in was not anti-static which is worrying.
Why on earth don't Burson make a proper datasheet for each of their opamps - its not like they have anything to lose by stating its true performance which is much better than the existing terse effort: https://drive.google.com/open?id=1JyV88oA-33QHSHbGYF16A3FRW1mDRV1N
I guess my conclusions are:
1) If you have one, don't worry its pretty decent performance.
2) If you don't then think twice and look at cheaper and cooler-running alternatives like OPA1688, LM4562, OPA1656, ADA4625, OPA1612, OPA1642, OPA1652, OPA1662 and more... (NE553x if you can still get them!).
3) Its fussy about decoupling, you can't necessarily get away with just 100nF between V+ and V-, make sure both rails are decoupled to ground as well.
[ Oh one other thing, my unit came with the polarizing groove on the base and the case 180 degrees apart - turns out the case was correct and the base was wrong, but that's rather disconcerting... ]
I've measured its distortion performance using the QA403 audio analyzer and here are the graphs, first distortion v. level for both inverting (gain=-1) and non-inverting (gain=+1) topologies:
Clearly this opamp has decent (but not world class) distortion performance. Powered at +/-15V. 600 ohms load seems to be handled in its stride.
Alas the power consumption is a massive 13mA with no signal (the only pitiable specimen of a datasheet I could find claimed 8mA). That same datasheet only claimed 0.02% for distortion, which perversely under-sells it - perhaps the datasheet is not for the Vivid V6?
Little bit let down at higher levels, but most practical uses will be somewhere in the 0 to 10dBV range.
Frequency response:
No signs of performance dropping at HF.
19kHz/20kHz IMD:
Again decent performance in line with other graphs.
So the only elephant in the room is the cost - there are many opamps with similar or better performance that are much cheaper - for instance OPA1642 will do rather better with only 1.8mA per opamp...
The high power consumption makes it run at 50C or so which isn't great.
The packaging it came in was not anti-static which is worrying.
Why on earth don't Burson make a proper datasheet for each of their opamps - its not like they have anything to lose by stating its true performance which is much better than the existing terse effort: https://drive.google.com/open?id=1JyV88oA-33QHSHbGYF16A3FRW1mDRV1N
I guess my conclusions are:
1) If you have one, don't worry its pretty decent performance.
2) If you don't then think twice and look at cheaper and cooler-running alternatives like OPA1688, LM4562, OPA1656, ADA4625, OPA1612, OPA1642, OPA1652, OPA1662 and more... (NE553x if you can still get them!).
3) Its fussy about decoupling, you can't necessarily get away with just 100nF between V+ and V-, make sure both rails are decoupled to ground as well.
[ Oh one other thing, my unit came with the polarizing groove on the base and the case 180 degrees apart - turns out the case was correct and the base was wrong, but that's rather disconcerting... ]
Last edited:
Nice to see some measurements on them!
I still have my units I was sent, and I have no compelling reason to use them over regular jellybean chips. Might end up putting them up for trade.
I still have my units I was sent, and I have no compelling reason to use them over regular jellybean chips. Might end up putting them up for trade.
I suspect not, that would be a gift to the knock-off merchants... Its pretty impressive performance for a discrete opamp, but you'll never beat the best IC opamps with discretes these days so unless it was cheaper I don't understand the attraction. I can understand amateur discrete opamps being designed for the challenge, but selling such commercially without a datasheet full of actual data I cannot figure...
Well, t.b.h.:I would not be surprised in case the opposite would appear be true: the last 50 years a bunch of discrete opamp type circuits have been published: it is also more or less how Douglas Self got started in WW&E. Add or leave out a CCS, a Current Mirror or a stage here and there, and ready you are.. Anyway, thanks for the measurements: I never understood the fascination either.
I've just compared the Burson to a discrete opamp (basically a Blameless style circuit, 10 transistors) that I knocked up on a breadboard using KSC1845's and KSA992's:
Mine above clearly isn't so good with 470R load at high levels, but is generally lower distortion and much less noise - from this graph's noise-limited segment I estimate the noise of mine as -123dBV, the Burson -102dBV (although I'm not sure of whether the QuantAsylum window noise bandwidth is factored into these values, so the absolute value is probably different).
So the V6 Vivid is clearly a noisy beast (voltage noise to be specific) if its 21dB worse than a bog-standard Blameless input section...
I might rerun the Burson tests to check I didn't make a mistake as its so different.
Mine above clearly isn't so good with 470R load at high levels, but is generally lower distortion and much less noise - from this graph's noise-limited segment I estimate the noise of mine as -123dBV, the Burson -102dBV (although I'm not sure of whether the QuantAsylum window noise bandwidth is factored into these values, so the absolute value is probably different).
So the V6 Vivid is clearly a noisy beast (voltage noise to be specific) if its 21dB worse than a bog-standard Blameless input section...
I might rerun the Burson tests to check I didn't make a mistake as its so different.
Ah - there was something definitely wrong with my last Burson measurements - here's non-inverting unloaded graph I've just remeasured - much more like it, noise is NOT an issue with this device (not sure where it came from last time though...perhaps inadequate decoupling or pickup from surroundings?)
And for completeness redone the plots for unity gain inverting as well as non-inverting with various loadings, plus a THD+N plot.
Which looks pretty respectable for noise... Needless to say I used 4 decoupling caps dotted around the breadboard this time!
Which looks pretty respectable for noise... Needless to say I used 4 decoupling caps dotted around the breadboard this time!
For comparison, same measurements on same rig with a T.I. NE5534A:
So the Burson can't really match the 5534 for distortion at higher signal levels, but that's quite a high bar to exceed, but looks to be a bit quieter perhaps?
So the Burson can't really match the 5534 for distortion at higher signal levels, but that's quite a high bar to exceed, but looks to be a bit quieter perhaps?
Tube amplifier have much worst THD than solid state but they are still expensive and many people buying them. Another question is do you realy hear THD difference distortion between Burson and 5534. I think NO. I've used in my DAC cheap OPA1655 which is realy good, probably better than 5534 and Burson V6 classic. Definitely Burson is better. I think ear is better instrument than your QA403. Maybe Burson doesn't justify that price but it is not bad.
And I know its not. 0.01% is about the limit for human hearing pretty much (and even that takes something like the 7th harmonic of a 200Hz signal to get past the perceptual masking effect - for 2nd and 3rd harmonics the threshold is much higher). Ask an audiologist. The organ of corti simply doesn't have ppm dynamic range.
The only way humans can hear 0.0001% distortion is if there are two very loud ultrasonic tones producing an IMD product in the audible spectrum.
http://www.r-type.org/articles/art-143.htm
I could definitely hear the difference. I suppose it depends on the rest of the chain.
http://edsaudiopages.blogspot.com/2020/06/burson-audio-v6-classic-review.html
If you are playing a music source, the rest of the chain already contains dozens, if not hundreds of 5534s. You can hear one added one?
Well 5532's probably, and less true today with digital mixing, but it was the case for a long while. Dozens perhaps, not hundreds, for a signal chain. And probably a few TL072 for the EQ sections.
Wow, maybe I should swap my Burson op amps for those ne5532s that are now in the bottom of my junk bin?!
Thanks for the test regardless, have always been curious as to what the numbers were.
You apparently have more time than I do!
Thanks for the test regardless, have always been curious as to what the numbers were.
You apparently have more time than I do!
Mark T, thank you for the measurements! That is a lot of work, I know. So we have some numbers to refer to now.
Burson sent me some Vivid V6's to test out and I did install them as the output stage (replacing OPA1656) in my tube buffer. I enjoyed the sound and all, and thought they did a decent job of driving most line level loads. My main issue is how tall they are and to mount them to SOIC8 adapters makes them very fragile to being knocked off or peeling off the traces if they were bumped. I appreciated Burson sending me samples to test though.
The side topic here that is interesting is that the NE5532's are getting some love here. I had a board house assemble my latest headphone amp spec'd for OPA1642's, but they didn't have those in stock so I opted for NE5532's (plenty in stock). They actually sound great and measure well too. I was thinking of removing the factory installed 5532 and installing some 1642's, but maybe no point in it. Here is 3.0Vrms into 33ohms (NE5532 driving a diamond buffer output stage):
Burson sent me some Vivid V6's to test out and I did install them as the output stage (replacing OPA1656) in my tube buffer. I enjoyed the sound and all, and thought they did a decent job of driving most line level loads. My main issue is how tall they are and to mount them to SOIC8 adapters makes them very fragile to being knocked off or peeling off the traces if they were bumped. I appreciated Burson sending me samples to test though.
The side topic here that is interesting is that the NE5532's are getting some love here. I had a board house assemble my latest headphone amp spec'd for OPA1642's, but they didn't have those in stock so I opted for NE5532's (plenty in stock). They actually sound great and measure well too. I was thinking of removing the factory installed 5532 and installing some 1642's, but maybe no point in it. Here is 3.0Vrms into 33ohms (NE5532 driving a diamond buffer output stage):
Sorry, was that question to me? Or just in general.
No, i can not hear that. But that is beyond the point. Question was if you can spot the difference between two opamps. If you can not, fine.