"Boutique" opamps from chinese vendors. Anyone tested?

[abbey road d enfer]

Probably why I sadly skip coming yo tjis place no more...

If only...

 

[leadbreath]

I dont realky care about my credibilty....so fuck you!

 

[leadbreath]

Fuck you too JR fucking asshole...

 

[JohnRoberts]

Fuck you too JR fucking asshole...

Not sure what I said to incur this wrath (I did ask people to be nice).

It looks like leadbreath got what he was asking for (forum death by moderator).

Too bad I liked some of the pictures but the Tourettes is against the rules.

JR

 

[Ricardus]

Wow. Usually it's newbs with 12 posts resorting to that language.

I bet the poster talking about the holy grail op-amp he designed is bogus, too.

 

[Ricardus]

It looks like leadbreath got what he was asking for (forum death by moderator).

 

[JohnRoberts]

lets not pile on.. He cant respond.

"its nice to be nice"

JR

 

[abbey road d enfer]

I bet the poster talking about the holy grail op-amp he designed is bogus, too.

I'm not so sure... Let's him have a chance of dissipating ambiguity by answering the pertinent questions it generated.

 

[Newmarket]

I used truckloads of 5532 and TL07x often inside the same SKUs. The 5532s were better when heavy lifting was involved because of their stronger drive capability. The TL07x were fine for line level stages.

Typical of the UK designed/built desks I was familiar with in the 90s would be a mic pre using NE5534 for the op amp then most other stages a TL072. NE5532/4 or TL072 for outputs (IIRC) dependent on spec' / cost.

 

[inkster]

Wrt reduced stray reactive impedances - do you mean in comparison to thru hole discrete or to ICs ?

If you're reacting to my post, it is my belief and experience as a professional studio tech that through hole circuits when compared directly to identical discrete smd circuits often lack detail and imaging among other desirable traits...that said I'm still a big fan of Neve, API and other classic circuits, but even those can benefit from nice smd equivalents.
In those consoles I typically only use smd versions in the center section(control monitoring) depending on the client's needs.
As one studio engineer said to me "it's like you removed the walls and ceiling", I took that as a compliment.

 

[abbey road d enfer]

If you're reacting to my post, it is my belief and experience as a professional studio tech that through hole circuits when compared directly to identical discrete smd circuits often lack detail and imaging among other desirable traits...that said I'm still a big fan of Neve, API and other classic circuits, but even those can benefit from nice smd equivalents.

Do you have examples of mixers or processors available in both TH and SMD versions and otherwise identical? That's what would be needed for a fair comparison.

 

[JohnRoberts]

If you're reacting to my post, it is my belief and experience as a professional studio tech that through hole circuits when compared directly to identical discrete smd circuits often lack detail and imaging among other desirable traits...that said I'm still a big fan of Neve, API and other classic circuits, but even those can benefit from nice smd equivalents.
In those consoles I typically only use smd versions in the center section(control monitoring) depending on the client's needs.
As one studio engineer said to me "it's like you removed the walls and ceiling", I took that as a compliment.

I prefer bench test measurements over anecdotal listening impressions but there is a known benefit for SMD in smaller, tighter PCB layouts.

Back in my old day job at Peavey (last century) we made many budget mixers on large single sided PCBs. When we brought SMD technology in house we were able to layout mic preamps almost completely underneath the XLR mic input jacks (we were able to put smd parts on the solder side of single sided PCB). I can't say that I noticed any difference in measured specifications (for better or worse) but making the layout tighter reduces the area for potential noise pickup.

As I recall we had some issues at first getting the very low noise bipolar transistors in SMT packages because the low Rbb die were larger than small signal transistors, but when buying truckloads at a time, the vendors tend to bend over backwards to accommodate you. They found a package that worked.

JR

 

[thor.zmt]

I bet the poster talking about the holy grail op-amp he designed is bogus, too.

One needs to be VERY careful what one reads into things.

If you use +/-36V Rails and a competent circuit, you can easily exceed older IC Op-Amp's in any number of areas. Monolithic matched duals are available for input pairs.

Make for example a CFB SE Input using the dual smt versions of 2N4403/4401 and a few mA current that has minimal input offset current and under 1nV|?Hz Ein.

Make the output Class A for loads beyond where the canned Op-Amp's can compete.

Use the extra high rails for cascodes on inputs and make another CFB Stage followed by a classic Push-Pull VAS, buffered to minimise loading effects (the 5532/34 is also 3-Stage).

You can get an extremely linear (inherently) circuit that can drive low impedance loads at levels no canned Op-Amp can even manage without clipping and with high gain and wide bandwidth, again, way past what canned Op-Amp's manage.

You could make a single "Op-Amp" Microphone Pre using input & output transformers that has ultra low Noise, ultra high gain (being CFB adjusting gain adjusts open loop gain keeping loop gain constant with varying gain) and vanishingly low distortion, way beyond what a canned op-Amp can do and at very, very low cost.

Drawback? PCB size, Power consumption and the need for a real analogue engineer to design the circuit. Can we use a bunch of Op-Amp's instead and achieve similar performance? With enough circuitry, sure.

Which Microphone Pre will be better? Who knows.

I know which would interest me more as a design project for a paying customer.

Thor

 

[thor.zmt]

through hole circuits when compared directly to identical discrete smd circuits...

Ohhh, beware. There are many differences between SMT parts.

A lot of older gear used carbon film resistors. Often there are many coupling capacitors that can be nixed using more modern IC op-amp's.

It may seem there are minimal variables, but that is generally not that strictly speaking true. I do find that many modern SMT passive parts are excellent and superior in objective and subjective performance, yet equally there are many very poor quality ones.

Good thing, the most expensive SMT resistor or capacitor by the reel, for the kind of things we use them for is still dirt cheap.

And it's easy to use for example 10pcs 2% SMD C0G capacitors to get a required value, voila a 0.63% tolerance capacitor, actually I find in practice even using nominally 5% capacitors that way gets me way better than 1% tolerance in practice.

Assuming equivalence between SMT and TH is not reasonable. And often the differences have less to do with TH vs SMT, but all with specific characteristic of the parts used, that would have been able to be found in TH, but perhaps at very diffrent costs.

Thor

 

[user 37518]

I dont realky care about my credibilty....so fuck you!

Fuck you too JR fucking asshole...

Classy...

 

[user 37518]

If you're reacting to my post, it is my belief and experience as a professional studio tech that through hole circuits when compared directly to identical discrete smd circuits often lack detail and imaging among other desirable traits...that said I'm still a big fan of Neve, API and other classic circuits, but even those can benefit from nice smd equivalents.
In those consoles I typically only use smd versions in the center section(control monitoring) depending on the client's needs.
As one studio engineer said to me "it's like you removed the walls and ceiling", I took that as a compliment.

While SMD does make a difference due to lower inductance/capacitance at RF frequencies, I have absolutely no clue as to how you think that would make a difference at audio frequencies. Again, can you provide any non-anecdotal evidence of what you are saying? Or are we again back to quoting Goethe like the other posters who claim that ICs have "bad sound"?

 

[Ricardus]

One needs to be VERY careful what one reads into things.

I agree 100%, but the forums (all forums I'm on) are being joined by accounts that eventually reveal themselves to be fake to cause trouble, to make unsubstantiated claims, or to just post ChatGPT stuff.

As Carl Sagan once said (I think?):

"Extraordinary claims require extraordinary evidence."

 

[JohnRoberts]

While SMD does make a difference due to lower inductance/capacitance at RF frequencies, I have absolutely no clue as to how you think that would make a difference at audio frequencies. Again, can you provide any non-anecdotal evidence of what you are saying? Or are we again back to quoting Goethe like the other posters who claim that ICs have "bad sound"?

There are known audio frequency issues from too tiny SMD resistors, of course experienced audio designers avoid them for premium paths. It is a huge leap to suggest that "all" smd designs suffer from such component issues, so I am not suggesting that.

As usual with any design the details matter, audio is a pretty mature technology. SMD is just a variant component packaging method.

JR

 

[user 37518]

There are known audio frequency issues from too tiny SMD resistors, of course experienced audio designers avoid them for premium paths. It is a huge leap to suggest that "all" smd designs suffer from such component issues, so I am not suggesting that.

As usual with any design the details matter, audio is a pretty mature technology. SMD is just a variant component packaging method.

JR

John, what you mention is a different thing and has to do with the component itself, you could have the same component with through hole leads and it would be the same thing. What I was referring to is about the difference between having a leaded component that goes through the PCB and one with shorter leads that solders on top of the PCB, all other things being equal; since the argument being proposed here is that the parasitics caused by the way the component is mounted and the lead lengths create big enough differences as to make someone claim, and I quote:

..."it's like you removed the walls and ceiling"...

 

[Newmarket]

There are known audio frequency issues from too tiny SMD resistors, of course experienced audio designers avoid them for premium paths. It is a huge leap to suggest that "all" smd designs suffer from such component issues, so I am not suggesting that.

As usual with any design the details matter, audio is a pretty mature technology. SMD is just a variant component packaging method.

JR

tbf the post that Dual was replying to was one saying SMT was advantageous.

 

[JohnRoberts]

John, what you mention is a different thing and has to do with the component itself, you could have the same component with through hole leads and it would be the same thing. What I was referring to is about the difference between having a leaded component that goes through the PCB and one with shorter leads that solders on top of the PCB, all other things being equal; since the argument being proposed here is that the parasitics caused by the way the component is mounted and the lead lengths create big enough differences as to make someone claim, and I quote:

I'm going to need to tap out soon...

I saw differences in very small smd resistors that seemed worse that through hole, while I have seen distortion in through hole resistors too when exposed to large voltage swings (like high power audio amps, across the feedback network). That's why we read the data sheets.

Last century in connection with my day job, I toured a major capacitor factory in Mexico. They were literally attaching leads to tiny chip capacitors and then encapsulating the finished cap with a dip in epoxy(?). As I recall the capacitor chips were manufactured in South Carolina or somewhere like that, then shipped to Mexico for the labor intensive lead attachment. This was before SMD packaging was popular. Now SMD parts just forgo attaching the wasteful leads, that only get cut off later.

tbf the post that Dual was replying to was one saying SMT was advantageous.

IMO SMD has benefits and faults that competent design engineers have to manage, and they generally do.

JR

 

[thor.zmt]

As usual with any design the details matter, audio is a pretty mature technology.

Is it now? Nothing new since 1950 then?

SMD is just a variant component packaging method.

Nope, it's a bit more complex than that.

Failure to understand the whole picture of how components differ between both common and less common SMD and TH variants can lead to all sorts of fun and games and measurable performance degradation, in the oh so mature technology called audio.

More often than not, common grade SMD parts are markedly worse than TH variants, HOWEVER there are also excellent option that often exceed the objective performance of the parts common in 1980/1990's gear made strictly TH style.

As with the discrete vs IC Op-Amp this is a lot more nuanced and actually interesting than the simplistic view "it all sounds the same" suggests.

I recommend reading "linear audio" among other publications to actually understand passive components (and many other interesting things) better, going even as far as D. Self writing an article on capacitor "burn in" that I guess he was not prepared to publish in his usual venues...

Thor

 

[thor.zmt]

I saw differences in very small smd resistors that seemed worse that through hole, while I have seen distortion in through hole resistors too when exposed to large voltage swings (like high power audio amps, across the feedback network). That's why we read the data sheets.

And the voltage dependence of resistance is not in the datasheet. It also applies to SMD parts of course, in fact often especially with small format parts it is very large.

A simple trick is to make up a larger value resistor that sees a lot of voltage swing from multiple smaller values. It reduces the voltage modulation resistance modulation by lowering the voltage.

Many capacitors too have voltage dependent capacitance, how dependent? Well, as with other factors, it depends (duh).

And if not using ppm level tempco resistors it is essential to watch thermals. While it is implicit in the datasheet, it is easy to miss. Again, here SMD resistors are commonly at disadvantage, but it is easier to multiple units in series and we can get to a point where we easily outperform classic TH.

I come back to my point that each technology has distinct limitations and drawbacks that require nuanced design approaches to get the desired end result, not tarring with broad brush and then generously applying feathers.

Thor