Small Signal Circuits vs Discrete Opamp

[ruffrecords]

+26 dB headroom above -8dBu, peaks around +18dBu which is not exactly whisper level, but I concede driving 600 ohms will require more current and result in higher distortion. Active differential outputs adds 6 dB more level making that +18dBu a respectable +24 dBu while if you are driving 600 ohms that makes it dBm.

JR

1970s Neve consoles used a single 24V rail and their max output was +26dBm.The output stage and transformer provides the additional 8dB of gain but this means the output stage has to drive the same power in to about 150 ohms on the primary which requires a significant current drive capability.

Cheers

Ian

 

[user 125886]

Mr. Ruffrecords I am just amazed at the level of intimate knowledge you have about the neve circuit and it's history, I really mean that. I'm just an armature hobbyist. Can you recommend any books about this specific subject matter, so I can truly understand how audio circuits work, and why they work? At this point I only understand about half of what your saying. There are so many books out there, not sure which ones are the real deal.

 

[ruffrecords]

Mr. Ruffrecords I am just amazed at the level of intimate knowledge you have about the neve circuit and it's history, I really mean that. I'm just an armature hobbyist. Can you recommend any books about this specific subject matter, so I can truly understand how audio circuits work, and why they work? At this point I only understand about half of what your saying. There are so many books out there, not sure which ones are the real deal.

Thank you for the kind words. I worked at Neve in the 70s so I was immersed in these circuits every day. The classic textbook these days is Doug Self's "Small SIgnal Audio Design". Ir covers some transistor circuits but mostly is concerned with op amp circuits. As far as I know there is no textbook that discusses circuits like the early Neve ones in detail.

Cheers

Ian

 

[alphasnk]

Thank you for the kind words. I worked at Neve in the 70s so I was immersed in these circuits every day. The classic textbook these days is Doug Self's "Small SIgnal Audio Design". Ir covers some transistor circuits but mostly is concerned with op amp circuits. As far as I know there is no textbook that discusses circuits like the early Neve ones in detail.

Cheers

Ian

You guys should write that book, no one is gonna do it apart from you guys who lived during that time! I encouraged Paul Wolff to start a youtube show as well haha

 

[JohnRoberts]

You guys should write that book, no one is gonna do it apart from you guys who lived during that time! I encouraged Paul Wolff to start a youtube show as well haha

Have you explored reference books here ?

In my judgement it is a little late in the game for discrete audio circuit design. I sure wish we had all these books back decades ago when I was trying to find my butt.

Over the years I bought a bunch of text books to teach myself various aspects of design. My complaint was that serious texts were like jumping into the deep end of the pool without a life guard. The book I wanted to find was a comprehensive everything about everything covering passive components to SOTA ICs. Of course this book would be obsolete before the ink dried, but still worthwhile IMO.

I would like to see the book organized with chapters beginning with the most basic entry level information, and progressively getting deeper into each topic. The self learning student could skip over information they know, but can always back up as needed when confused. You read until your head hurts then jump to the next topic. Later if you need more information to understand some problem you crack the book open again and use it as a reference as needed.

The book would have to engage different specialized experts for the most technical sections of the sundry chapters. It is impossible for one person to be expert about everything while some may think they are.

JR

 

[alphasnk]

The book would have to engage different specialized experts for the most technical sections of the sundry chapters. It is impossible for one person to be expert about everything while some may think they are.

That's exactly it, just the way you guys share your knowledge here! We all learnt from you, it would be awesome to compile all this knowledge, it would help preserving it..

 

[matriachamplification]

That's exactly it, just the way you guys share your knowledge here! We all learnt from you, it would be awesome to compile all this knowledge, it would help preserving it..

I am a huge fan of the director Werner Herzog who has a masterclass on filmmaking. Years ago I took the online course and immediatley connected the premise to GDIY learning.

I would very much like to see a masterclass lecture from GDIY contributors. We have been jamming ideas on this for a while just never followed up with any direct conversations regarding who would be willing to talk for 45 minutes. Maybe it's time.

Wall

 

[PhilipMarlowe]

1970s Neve consoles used a single 24V rail and their max output was +26dBm.The output stage and transformer provides the additional 8dB of gain but this means the output stage has to drive the same power in to about 150 ohms on the primary which requires a significant current drive capability.

Cheers

Ian

Thanks to people like you and posts like these I'm slowly getting the hang of the technical aspect. I'm starting to understand why certain pieces of gear sound as good as they do. Thank you for laying it out like that. From what I've gathered a lot of what sounds pleasing to my ears specifically comes down to headroom and consequentially harmonics/saturation generated by intentionally overdriving circuits with high voltage with minimum crossover distortion.
Does this maximum +26dbm of headroom from vintage desks you refer come from the the output level before audible distortion or is it a hard limiting (well not really) ceiling? I've read a few times how big consoles have so much headroom on the mixbus specifically, that it would "crap out" long after your AD would clip due to a too hot signal. I've never mixed on the big boy consoles like an 80s series Neve or a vintage API but even my small format austrian mixer seems to have such a high level on the mixbus that it can sometimes clip my mastering AD converter which boasts +24db as dbfs with a dynamic range of +121db.

 

[abbey road d enfer]

Does this maximum +26dbm of headroom from vintage desks you refer come from the the output level before audible distortion or is it a hard limiting (well not really) ceiling?

Maximum output level should always be quoted for a specific distortion level.
E.G. a power amp my be spec'd at 100W for 0.1% THD, and be capable of delivering about 120W for 10% THD.

even my small format austrian mixer seems to have such a high level on the mixbus that it can sometimes clip my mastering AD converter which boasts +24db as dbfs with a dynamic range of +121db.

Do you mean your mixer is capable of delivering more than +24dBu? This is not unheard of, but operationally, there is no reason to do this.
I suspect the meters on your mixer must be pegged...

 

[PhilipMarlowe]

Maximum output level should always be quoted for a specific distortion level.
E.G. a power amp my be spec'd at 100W for 0.1% THD, and be capable of delivering about 120W for 10% THD.

Do you mean your mixer is capable of delivering more than +24dBu? This is not unheard of, but operationally, there is no reason to do this.
I suspect the meters on your mixer must be pegged...

Thanks. Yes, i forgot the THD specs but I’m not sure how to read those. Do they imply a point of crunch or what kind of distortion would that even be? Not all distortion is created equally. Certain types make a sound bigger while others make it smaller.

And wouldn’t it operationally have the reason to increase S/N ratio? The more dynamic range the final stereo print has, the less noise it would in theory pick up. It is ultimately bottlenecked by the AD converter.

 

[abbey road d enfer]

Thanks. Yes, i forgot the THD specs but I’m not sure how to read those. Do they imply a point of crunch or what kind of distortion would that even be?

It depends...
Tube circuit are known to have a "gentle" distortion rise vs. level (soft clipping), when solid-state exhibit a quite sudden rise with a very small transition range (hard clipping). In reality, it is not so clear cut.
It is well known that guitar amps with global NFB (Marshall, most Fender) have a more restricted transition range than those without global NFB (Vox), and that some SS amps include a soft-clipping scheme.

And wouldn’t it operationally have the reason to increase S/N ratio? The more dynamic range the final stereo print has, the less noise it would in theory pick up. It is ultimately bottlenecked by the AD converter.

When an ADC has a decently implemented 24-bit converter, I don't call that a bottleneck. I usually track at about -15/-20dBfs, which leaves about 90dB DR, the media is not the bottleneck.

 

[iomegaman]

I find it refreshing in a mildly shocking way when an engineer/producer who has the chops to be listened to makes comments that reveal how much we've romanticized some of these designs...for instance Greg Wells who is no slouch once commented that he keeps a Mackie 1202 ProVLZ mixer around for recording drums because it is an incredible preamp with very low noise...
This is a guy who could probably use anything he wants but the lowly Mackie which got a bad rap for indeterminate reasons is actually a good piece of kit to have around...I remember when I first got one (I still have it) I thought it was remarkably quiet and clean but because it was "opamp" and not discrete I relegated it to "lower class"...

I have since decided the song is more important than whether or not the signal path is discrete or some chip infused design.

 

[Ricardus]

I find it refreshing in a mildly shocking way when an engineer/producer who has the chops to be listened to makes comments that reveal how much we've romanticized some of these designs...for instance Greg Wells who is no slouch once commented that he keeps a Mackie 1202 ProVLZ mixer around for recording drums because it is an incredible preamp with very low noise...
This is a guy who could probably use anything he wants but the lowly Mackie which got a bad rap for indeterminate reasons is actually a good piece of kit to have around...I remember when I first got one (I still have it) I thought it was remarkably quiet and clean but because it was "opamp" and not discrete I relegated it to "lower class"...

I have since decided the song is more important than whether or not the signal path is discrete or some chip infused design.

I still have my American made Mackie 1202 and I find it super useful to have in a pinch. I podcasted with it for 5 years, and it I use it as a utility mixer all the time. Although it has some issues and I will be doing a total rebuild one of these days.

 

[abbey road d enfer]

There are many circumstances where utility takes over performance. I read an article where George M. said he used several GT "The Brick" in a master class. I think it's probably because there was not enough good preamps available since the Brick is an absolute disaster. Note I don't say that often of equipment.

 

[JohnRoberts]

I wrote an article about this (regarding console technology) back in 1980. An inconvenient reality for high end purveyors is that low cost off the shelf op amps by then were much better than needed for reproducing line level audio (faster and quieter). As I noted back then the heavy lifting for console designers was mic preamps (because of high gain with low noise), and sum buses (because of high noise gain and loop gain margin requirements).

Numerous manufacturers, including value SKUs like Mackie, used Cohen topology for mic preamps (I didn't know it was called Cohen back then despite using it myself for years). These typically delivered NF (noise figures) of <2dB or within 2dB of perfectly noiseless. Few customer's believed that inexpensive preamps could possibly be that good. What's worse if we bundle multiple such decent preamps inside a single SKU the perception is that they must be cheap crap.

Over 15 years selling inexpensive SKUs I only encountered a handful of famous professionals with enough self confidence to listen for themselves without preconceived notions. There may be more but few were willing to publicly declare observations that differ with the general consensus. The vast majority listened with their eyes, or worse.

JR

PS: I've shared this anecdote before and I won't mention specifics (because I don't recall exactly which competitor's mixer model it was), but a subtle trick to make a preamp appear quieter than it is involves undersizing the capacitor in series with the gain resistor. At low-mid gain settings the preamp delivers full LF response (good for the measured specifications). However at maximum gain the undersized capacitor forms a HPF cutting off the very LF response. In mic preamps at very high gain, the noise floor can be contaminated with 1/F noise. This disingenuous trick can fool users into thinking the preamp is quieter than it really is. End users rarely bench test their mixers as rigorously as a competitor would.

 

[Ricardus]

There are many circumstances where utility takes over performance. I read an article where George M. said he used several GT "The Brick" in a master class. I think it's probably because there was not enough good preamps available since the Brick is an absolute disaster. Note I don't say that often of equipment.

Now I must have one!

You'd think George M would have access to lots of preamps!

 

[k brown]

Thoughts on this? Seems rather different from the common 'ring of three'.

Advent MPR-1 accessory mic pre.

 

[k brown]

I wrote an article about this (regarding console technology) back in 1980. An inconvenient reality for high end purveyors is that low cost off the shelf op amps by then were much better than needed for reproducing line level audio (faster and quieter). As I noted back then the heavy lifting for console designers was mic preamps (because of high gain with low noise), and sum buses (because of high noise gain and loop gain margin requirements).

Numerous manufacturers, including value SKUs like Mackie, used Cohen topology for mic preamps (I didn't know it was called Cohen back then despite using it myself for years). These typically delivered NF (noise figures) of <2dB or within 2dB of perfectly noiseless. Few customer's believed that inexpensive preamps could possibly be that good. What's worse if we bundle multiple such decent preamps inside a single SKU the perception is that they must be cheap crap.

Over 15 years selling inexpensive SKUs I only encountered a handful of famous professionals with enough self confidence to listen for themselves without preconceived notions. There may be more but few were willing to publicly declare observations that differ with the general consensus. The vast majority listened with their eyes, or worse.

JR

PS: I've shared this anecdote before and I won't mention specifics (because I don't recall exactly which competitor's mixer model it was), but a subtle trick to make a preamp appear quieter than it is involves undersizing the capacitor in series with the gain resistor. At low-mid gain settings the preamp delivers full LF response (good for the measured specifications). However at maximum gain the undersized capacitor forms a HPF cutting off the very LF response. In mic preamps at very high gain, the noise floor can be contaminated with 1/F noise. This disingenuous trick can fool users into thinking the preamp is quieter than it really is. End users rarely bench test their mixers as rigorously as a competitor would.

I was always amazed at the quality of the pres in even the cheapest Soundcraft mixers, even though they used the hated-by-audiophiles TLO72. To my ears considerably better than early Mackie.

 

[JohnRoberts]

I was always amazed at the quality of the pres in even the cheapest Soundcraft mixers, even though they used the hated-by-audiophiles TLO72. To my ears considerably better than early Mackie.

over the years I have used truckloads of tl07x and ne553x op amps... nothing wrong with 10V/uSec slew rates and microvolt level input noise.

The soundcraft mic pre was probably a little higher distortion but a variant on the Cohen topology*** so suitably low noise. I suspect there may be some perception bias.... Don't ask me about the blind monitor speaker listening test held in the bay area, where a couple studio guys, dropped out, after they learned that they had chosen Peavey (AMR) monitor speakers.

JR

*** there may be more than one "soundcraft" mic preamp.

 

[k brown]

I was directly comparing Folio Notepad with 1st gen Mackie 1202; Mackie rather veiled-sounding, but smooth. Rather doubt there was expectation bias, because I was expecting the exact opposite; I assumed the robust Mackie would sound better than the little, plastic Notepad - after all, famed Classical engineer Marc Aubort used the 1202 pres (direct out) on a string quartet session I sat in on.

It's been claimed all Soundcraft mixers used same pres.