We did our event this weekend and it was a lot of fun. This is the first of two posts where I talk about how it went.
Gear involved
I am running one of the Miro AD1862 DAC's (LINK). My build features Antek transformers and 78xx + 79xx regulators as per Miro's PSU1. The AD1862 chips each have their own independent bipolar 5 and 12v supplies, all op amps run on a shared independent bipolar 12v supply, and the JL Sounds USB to I2S III board runs on its own +5v supply. In my op amp tests, all opa's were given plenty of current supply. I can't speak to how something like a Burson would interact in a circuit with less current on tap.
I included a low pass Sallen Key filter in the DAC, so each channel in the DAC sees an I/V stage and a single op amp SK filter. I have adjusted the stages to put out a voltage peak approaching 5v so that I can power my F5 power amp directly without another preamp. Most parts are Wima, Dale, or Nichicon.
Power stage is a Firstwatt F5 built dead stock from a DIYaudio kit.
The chassis are all matchy matchy
Speakers are Altec 604's with Mantaray horns, in custom built 620 dimension cabinets and Jeff Markwart crossovers using jantzen and solon parts. If it matters to you, I'm using about 10 ft of oxygen free copper 12awg cabling per side.
The room is almost 20x20ft and it sounds OK, although it could use some material to address a little bit of sibilance and slapback.
We tested four op amps. Each was scoped out before we began listening and confirmed stable.
In no particular order:
TL071 - IPE4 (MF # TL071IPE4). JFET inputs, pretty ubiquitous, $0.69 per chip.
NE5535 AP (MF # NE5534AP) Fully bipolar, we ran these with compensation caps as detailed previously. $1.32 per chip.
Burson V6 Vivid, Fully JFET, $70 for a pair of singles on the DIYaudio store before shipping. So about $40 per chip. Quite a bit more than say a TL071, but in the same ballpark as purchasing NOS OPA627's
ADA 4627 - 1 BRZ (MF # ADA4627-1BRZ) JFET input, if I understand correctly these are a contemporary OPA627, which is the chip Miro's DAC is originally designed around. $11.71per chip, but if you want to swap them into DIP8 packages they need adapter boards and pins so we'll say closer to $14 per.
Also let it be known that Burson did send me one pair of V6's, but I purchased the second set myself in order to swap out all 4 sockets. I mention this to highlight that I did receive free stuff from the manufacturer, but also did experience the sting of setting out $80/pair after shipping.
Gear involved
I am running one of the Miro AD1862 DAC's (LINK). My build features Antek transformers and 78xx + 79xx regulators as per Miro's PSU1. The AD1862 chips each have their own independent bipolar 5 and 12v supplies, all op amps run on a shared independent bipolar 12v supply, and the JL Sounds USB to I2S III board runs on its own +5v supply. In my op amp tests, all opa's were given plenty of current supply. I can't speak to how something like a Burson would interact in a circuit with less current on tap.
I included a low pass Sallen Key filter in the DAC, so each channel in the DAC sees an I/V stage and a single op amp SK filter. I have adjusted the stages to put out a voltage peak approaching 5v so that I can power my F5 power amp directly without another preamp. Most parts are Wima, Dale, or Nichicon.
Power stage is a Firstwatt F5 built dead stock from a DIYaudio kit.
The chassis are all matchy matchy
Speakers are Altec 604's with Mantaray horns, in custom built 620 dimension cabinets and Jeff Markwart crossovers using jantzen and solon parts. If it matters to you, I'm using about 10 ft of oxygen free copper 12awg cabling per side.
The room is almost 20x20ft and it sounds OK, although it could use some material to address a little bit of sibilance and slapback.
We tested four op amps. Each was scoped out before we began listening and confirmed stable.
In no particular order:
TL071 - IPE4 (MF # TL071IPE4). JFET inputs, pretty ubiquitous, $0.69 per chip.
NE5535 AP (MF # NE5534AP) Fully bipolar, we ran these with compensation caps as detailed previously. $1.32 per chip.
Burson V6 Vivid, Fully JFET, $70 for a pair of singles on the DIYaudio store before shipping. So about $40 per chip. Quite a bit more than say a TL071, but in the same ballpark as purchasing NOS OPA627's
ADA 4627 - 1 BRZ (MF # ADA4627-1BRZ) JFET input, if I understand correctly these are a contemporary OPA627, which is the chip Miro's DAC is originally designed around. $11.71per chip, but if you want to swap them into DIP8 packages they need adapter boards and pins so we'll say closer to $14 per.
Also let it be known that Burson did send me one pair of V6's, but I purchased the second set myself in order to swap out all 4 sockets. I mention this to highlight that I did receive free stuff from the manufacturer, but also did experience the sting of setting out $80/pair after shipping.
Last edited:
Shootout Procedure
All parts were dropped into a closed box. DAC had something covering the top so I could not peek into the vents. Friend 2 and I would go to the kitchen to refill wine glasses and mix a roux, cut bread etc while Friend 1 would swap opa's and put everything back together before calling us back in. None of us discussed what we heard until we had listened to all 4 opa's. We all kept written notes on notepads and listened to the same sequence of 6 tracks.
Friend 1 is an audio engineer and admitted to having probably the most bias about op amps to begin with, but also obviously the most bias because they were the participant who was not blinded in the comparison. More on this later. Here's what we thought:
ADA 4627
Huge and dense low end presence, darker EQ profile. Image felt tall and wide but not particularly deep. Felt like being in the center of the mix. Little bit of compression on snare hits and vocal peaks. Smushy during loud, dense passages. To me it felt a little bit like the T27/Bx00 KEF speakers do: little bit soft and compressed but the textures felt good.
I rated the ADA 4627 dead last and couldn't wait for it to be pulled out. My friends rated it #1 and loved it. In my opinion this chip sounded like what people are looking for in tube gear - little bit "vintage", little bit dark, "organic", etc but I personally didn't like it.
NE5534
All of us agreed that this chip was boring. Shouty, uninvolving, bland but passable. Engineer friend says "this is what I expect off the shelf consumer gear to sound like. I would be pissed if I built something nice and it sounded like this".
In my opinion, this isn't a part I could live with. I thought it sounded like garbage. That's my personal preference. Soundstage was all pushed way forward with no sense of dimension, EQ was aggressive.
Burson V6 Vivid
We were split on the Bursons. I rated the Bursons #1 by a mile, and my friends rated it dead last. The Burson sounded extremely different from the ADA 4627. As soon as we powered on and started with the Bursons, I thought "Ah there it is. That's what I'm looking for", and my friends immediately thought "The ADA's sounded 10x better than this".
We agreed that the Bursons sounded: more controlled, more like you were sitting in front of the staging instead of inside it, brighter EQ, less "exciting", and clearly had the most dynamic peaks.
Here's why I liked the Bursons: I am a guitar player and have owned enough small shop dreadnaught's to know how they should sound with monel 13's. the Bursons nailed Tony Rice's inflections. I loved that the vocals and spotlight instruments were pushed forward. Saxaphones sounded brassy and popped out. Everything stayed nice and clear during dense passages. Echo and decay sounded right to me with the Bursons.
My friends did not like this type of sound-staging, but I did. "It felt like the singer was in front of the band with a spotlight on her". Engineer friend says "These feel similar to the Sennheiser 6-- headphone line". I agree. I really liked my Sennheiser 6xx and felt like their presentation was just right. They gave theirs away to a buddy because they disliked them.
TL071
The TL071 is the real winner here. We universally enjoyed this chip. Exciting, textures were good. EQ profile was engaging without screaming (looking at you NE5534). Good low end presence (more than the Bursons, less than the ADA 4627). Good sizzle, cohesive soundstaging, felt wild like live music. And less than $1 per chip.
After our discussion:
We compared results and had a meal. I told my friends I would be keeping the Bursons in the unit, and talked about why I liked them. Engineer friend remarked that they probably had the clearest bias against the Bursons and asked to listen once more before the end of the night. They requested Aja, because it's an eye roller but we all know it well and com'on it's Steve Gadd -- and through the track the drums popped forward just like I thought they would while the rest of the band punctuated behind them. Engineer friend says something along the lines of "Ok, I get it. Not bad."
Conclusions:
My highly subjective conclusion: I will be running the Bursons. They sound good. I would also heartily recommend TL071 / TL072 op amps in hifi gear and I could live happily with these parts. The price difference for 4 single Bursons VS 4 TL071 is about $150. Are the Bursons $150 better sounding than the TL071?
Also my friends loved the ADA 4627 - 1.
Skip the NE5534
Thanks to @John Burson for the pair of amps, in my opinion they sound great and this was a lot of fun.
I think next we may shoot out pairs of output transformers in a KT88 amp.
All parts were dropped into a closed box. DAC had something covering the top so I could not peek into the vents. Friend 2 and I would go to the kitchen to refill wine glasses and mix a roux, cut bread etc while Friend 1 would swap opa's and put everything back together before calling us back in. None of us discussed what we heard until we had listened to all 4 opa's. We all kept written notes on notepads and listened to the same sequence of 6 tracks.
Friend 1 is an audio engineer and admitted to having probably the most bias about op amps to begin with, but also obviously the most bias because they were the participant who was not blinded in the comparison. More on this later. Here's what we thought:
- Right away it was apparent to all of us that there are enormous sound differences between all 4 parts. None of them sounded alike and the differences were not subtle.
- We were able to agree on all of our subjective descriptions regarding how each part sounded, but even though we experienced the same things, our conclusions and preferences were much different.
ADA 4627
Huge and dense low end presence, darker EQ profile. Image felt tall and wide but not particularly deep. Felt like being in the center of the mix. Little bit of compression on snare hits and vocal peaks. Smushy during loud, dense passages. To me it felt a little bit like the T27/Bx00 KEF speakers do: little bit soft and compressed but the textures felt good.
I rated the ADA 4627 dead last and couldn't wait for it to be pulled out. My friends rated it #1 and loved it. In my opinion this chip sounded like what people are looking for in tube gear - little bit "vintage", little bit dark, "organic", etc but I personally didn't like it.
NE5534
All of us agreed that this chip was boring. Shouty, uninvolving, bland but passable. Engineer friend says "this is what I expect off the shelf consumer gear to sound like. I would be pissed if I built something nice and it sounded like this".
In my opinion, this isn't a part I could live with. I thought it sounded like garbage. That's my personal preference. Soundstage was all pushed way forward with no sense of dimension, EQ was aggressive.
Burson V6 Vivid
We were split on the Bursons. I rated the Bursons #1 by a mile, and my friends rated it dead last. The Burson sounded extremely different from the ADA 4627. As soon as we powered on and started with the Bursons, I thought "Ah there it is. That's what I'm looking for", and my friends immediately thought "The ADA's sounded 10x better than this".
We agreed that the Bursons sounded: more controlled, more like you were sitting in front of the staging instead of inside it, brighter EQ, less "exciting", and clearly had the most dynamic peaks.
Here's why I liked the Bursons: I am a guitar player and have owned enough small shop dreadnaught's to know how they should sound with monel 13's. the Bursons nailed Tony Rice's inflections. I loved that the vocals and spotlight instruments were pushed forward. Saxaphones sounded brassy and popped out. Everything stayed nice and clear during dense passages. Echo and decay sounded right to me with the Bursons.
My friends did not like this type of sound-staging, but I did. "It felt like the singer was in front of the band with a spotlight on her". Engineer friend says "These feel similar to the Sennheiser 6-- headphone line". I agree. I really liked my Sennheiser 6xx and felt like their presentation was just right. They gave theirs away to a buddy because they disliked them.
TL071
The TL071 is the real winner here. We universally enjoyed this chip. Exciting, textures were good. EQ profile was engaging without screaming (looking at you NE5534). Good low end presence (more than the Bursons, less than the ADA 4627). Good sizzle, cohesive soundstaging, felt wild like live music. And less than $1 per chip.
After our discussion:
We compared results and had a meal. I told my friends I would be keeping the Bursons in the unit, and talked about why I liked them. Engineer friend remarked that they probably had the clearest bias against the Bursons and asked to listen once more before the end of the night. They requested Aja, because it's an eye roller but we all know it well and com'on it's Steve Gadd -- and through the track the drums popped forward just like I thought they would while the rest of the band punctuated behind them. Engineer friend says something along the lines of "Ok, I get it. Not bad."
Conclusions:
- My hypothesis: I would be able to consistently tell that the op amps were different. Yes.
- Hypothesis: I would rank the NE5534 as a peer. NOPE!
- Hypothesis: Bursons will have the largest soundstage: Kind of: the ADA 44627 featured a much larger scale, but the Bursons had a more defined sense of forward to back.
- Hypothesis: I will not have a strong preference. Wrong, I had a strong preference for the Bursons and the TL071, didn't vibe with the ADA 44627, and hated the NE5534
- Friend's hypothesis: All op amps will sound good. Nope.
- Burson's website specifically markets the V6 Vivid as a replacement for the 5532 / 5534. In my opinion, they are significantly better sounding provided your power supply can handle a less efficient chip and you have the physical space.
- Engineer friend "They sounded a LOT different. I learned something new."
My highly subjective conclusion: I will be running the Bursons. They sound good. I would also heartily recommend TL071 / TL072 op amps in hifi gear and I could live happily with these parts. The price difference for 4 single Bursons VS 4 TL071 is about $150. Are the Bursons $150 better sounding than the TL071?
Also my friends loved the ADA 4627 - 1.
Skip the NE5534
Thanks to @John Burson for the pair of amps, in my opinion they sound great and this was a lot of fun.
I think next we may shoot out pairs of output transformers in a KT88 amp.
Last edited:
Do you normally trip and fall down after 4 oz of Bordeaux at 2pm? Or is it mostly blacking out and falling asleep?
Yes although sometimes after the first one I get pissed and accidentally pour closer to 53ml -- but hey it's a partyYou mentioned refilling wine glasses, do you drink 50ml at a time?
More seriously, before opening the Bordeaux we split the remainder of a Rioja between me, my wife, and friend #2, so the first glasses were very light pours aimed at tasting. So maybe a total of 10 oz of wine over the course of 4 hours for me. Friend #1 drank only water all night.
If I am ripping drunk during an audio review I will make sure to specify so, like this: "During op amp #3 I spilled my 5th cocktail and wet myself"
Hey Anthoney,
FWIW,
I think you’ve had a great attitude throughout this shoot out. I’ve currently got some 4627, 797 and 6171 made up on break out boards but haven’t completed the dac yet. Initially I’ll be using a stock psu1, with jlsounds and no filter. Curious , we’re there any changes beyond installing different opamps? I’m guessing not, maybe the way you’ve implemented individual power supplies makes an improvement , but I’m not sure. All in all, nice effort
Thanks!
On my board I installed the optional SMD cap in the IV feedback loop as well as the optional current limiting resistors in case one of my plugs got grounded. But for this exercise nothing changed, just plug in / plug out.
I've run this DAC without the filter too and really prefer it with the filter in although I know Miro and others have different preferences
On my board I installed the optional SMD cap in the IV feedback loop as well as the optional current limiting resistors in case one of my plugs got grounded. But for this exercise nothing changed, just plug in / plug out.
I've run this DAC without the filter too and really prefer it with the filter in
although I know Miro and others have different preferences
The fuss with the OPA627 may stem from the fact (I believe) that it's basically Nelson Pass X amp (or XA I'm not sure) on a piece of silicon. The design is I think licenced from Passlabs. It's one of the most musical things made, then, or even today.
The nearest thing I can find to Nelson licensing an op amp is where he sold a patent to TI here; https://www.diyaudio.com/community/threads/texas-instruments-licenses-pass-patent.77568/
I can't find any information about what this product may have become if anything
Maybe people like the opa627 because they sound dark and compressed
I can't find any information about what this product may have become if anything
Maybe people like the opa627 because they sound dark and compressed
Maybe that's because of your kit, in mine (Cambridge Audio CD3 (4 x tda1541A. S1), or Micro Seiki CD-M100, Kenwood KP9010 with Sony XL44L, Krell KRC2, Passlabs Aleph 5 or Sony TA-8650, and Mezzo Utopias) they are open, musical and very detailed. I'm using solid silver with PTFE interconnects.
Regarding the licence, find the 627 tech design and compare it with the X or XA you will see the similarity. Anyway, you like what you like, I'll do the same. As your tests show, it's all subjective anyway.
Decades ago I used to build amplifiers that were based on a TL071 driving 2SJ50/2SK135 in high bias. Then the TL071 was easily bettered by OPA627/637. NE5534 did not cut the cake then as well.
We generally did not believe in burn-in in active devices but judging the OPA627/637 within minutes was thin ice. So, we concluded, religion is best practiced in church.
We generally did not believe in burn-in in active devices but judging the OPA627/637 within minutes was thin ice. So, we concluded, religion is best practiced in church.
Last edited:
No, I think you are thinking of opa1632, which is an FDA (Fully differential amplifier) and i'm not sure that was ever confirmed; though Papa did consult on the project afaik. the opa627 existed long before the X-amps came to be and there is little in common between them..
Last edited:
Ho ho! FDA = Full Digital Amplifier The new standard in audio that we all will meet one day. I have a few and like the concept better than the multi conversion chains that existed before them. Nice is that one does not need any opamp/DAC chip/analog source selection/analog volume control anymore which is a logic consequence in a mostly digital world. A new challenge is feedback and how it is done.
The new standard in audio that we all will meet one day. I have a few and like the concept better than the multi conversion chains that existed before them. Nice is that one does not need any opamp/DAC chip/analog source selection/analog volume control anymore which is a logic consequence in a mostly digital world. A new challenge is feedback and how it is done.
Last edited:
No. FDA also stands for Fully Differential amplifier. Also Fully differential amplifiers are an actual thing, unlike 'Fully Digital Amplifiers' Which is nothing but a marketing term.