gml opamps

[verbos]

actually, that reminds me of something...

Audio design seems to be one of the great black arts. It's like the whole thing is built on secrets. Do you think that the technology would progress further if it was all "open source" ? I'm just curious. I feel like the only way to know about this stuff is to see the way others did theirs. Then from that you can see what works and what doesn't. I'm surely not gonna buid a GML EQ for myself,but I do really want to know why it is so much better than the others.

So much in the audio industry is people repeating what they heard someone else say, it's interesting to see when a certain box actually does bring something new to the table. Obviously George invented the Parametric, so he has brought something to the table. However, is his EQ the best today because of that?

Seeing the actual facts always proves a point to me about what is really better and what is a bunch of advertising hype. I don't think, however, that just because something is simple it's not good. Well, you know....

The inside of that GML didn't look too simple. I am shocked to see SMT parts though. Many people have claimed that they "don't sound as fat" or whatever Hi-Fi nonsense.

Rock on guys, big respect George, loved your presentation at the Digidesign area at AES!!!

Mark

 

[jrmintz]

I think the point is that the rest of the world makes it so very difficult for small manufacturers to survive - absolutely everything is stacked against them - that we have an obligation not to make life harder for them. Particularly since these are friends, people we respect, whose work we admire and whose gear we use. We just need to make sure we don't hurt them in the process of satisfying our curiosity. Winston O'Boogie made a great point when he said he didn't think that anyone who posts here would make inappropriate use of the knowledge they gain here, but that this could be a fertile place for lurkers with no scruples - if we're not careful.

:thumb: :sam:

 

[22busy]

This is probably an heretical position to take on a diy forum, but I'd just feel a little uneasy about making a direct copy of a GML eq even if it were for purely personal use. What would happen to these talented designers if they can't sell enough product to keep them in business? Would they abandon designing and thus deprive the rest of us of nifty stuff in the future?

But, it appears to me to be quite different regarding companies like Sontec, that once produced what are still considered some of the best eq's ever made, but who have essentially abandoned their customers who need service on the older units and who appear to have ceased production of new eqs?

GML eqs are great ( I bought and still own one), and so are Sontecs. Unfortunately, as much as I'd love to have a Sontec, I wouldn't buy a new one (even if it were available) because of the lack of support. This seems a perfectly appropriate place for a clone or knockoff. Am I misguided?

BB

 

[jrmintz]

Unfortunately, as much as I'd love to have a Sontec, I wouldn't buy a new one (even if it were available) because of the lack of support. This seems a perfectly appropriate place for a clone or knockoff. Am I misguided?

That's really the tough question, BB. I can certainly understand your position. Of course, I'm not sitting here with an expensive EQ that doesn't work so it's bloody easy for me to say what I think in a vacuum. I think if you need to fabricate parts to keep an original unit working and you can't get them from the manufacturer, then it's OK. You have to understand how the part works to make it correctly. I think, unfortunately, each user has to do this for him or herself. Like copying songs, you can make as many copies as you need for your own use. It's when you reproduce something for other people that the line gets crossed. That's my $.02, anyway.

 

[Winston OBoogie]

...

 

[Learner]

[quote author="SSLtech"]

I might have something to offer the group on this: A while ago I built a working PWM "chopper" gain element with fairly accurate gain reduction indication on a regularly marked VU meter. -If anyone would like to take the cigarette-packet sketch that I made of how I saw it evolving into a compressor, I might be talked into 'donating' it into a group project, as long as there was reasonable sincerity in seeing it through... I'd hate to unveil it and have it "die on the vine" as it were.

Keith[/quote]

Sounds like an interesting idea Keith, inspired from PWM psu? class d amp? cranzane type?

Don't have enough technical knowledge to offer here but would love to hear your idea on the PWM gain reduction method, can also lay the PCB circuit for personal use........ :green:

:guinness:

 

[Peter Simonsen]

[quote author="jrmintz"]I think the point is that the rest of the world makes it so very difficult for small manufacturers to survive - absolutely everything is stacked against them - that we have an obligation not to make life harder for them. Particularly since these are friends, people we respect, whose work we admire and whose gear we use. We just need to make sure we don't hurt them in the process of satisfying our curiosity.
:thumb: :sam:[/quote]

jrmintz,

What a great post. I agree with every word.

Very well said

Kind regards

Peter

 

[22busy]

[quote author="Winston O'Boogie"]Anyway BB, our conversation the other day regarding GML and Sontec inspired me to snag these when they were offered to me so I could take a listen.
John.[/quote]

John, I'm interested in your impression of the GML stuff. But, I'm more interested in trying to convince you that you'd be the perfect person to design something to fill the Sontec void in the marketplace. You've got the ears and the chops to do it and no one else seems to be picking up the baton. :idea: :thumb:

Bruce

 

[Kev]

[quote author="Consul"] I have often wondered why, with all the incredible talent we have here, there aren't more original designs being proposed. [/quote]

I have kept out of this thread ( I think :roll: ) till now but I now feel it time to point out that the only reason Group DIY was focused on some .... and I repeat some of the classics ... was as a learning tool and because popular demand pushed us that way.

It was always the intention to bring NEW ideas and new variants to these old staples. Projects 1 and 2 are prime examples and so from these two projects it might be possible to ...
for example
enhance the SSL which is very much from the application notes of THAT with a ClassA or Gain Block driver stage.
A Passive EQ with ANY gain make up driver stage.
An opto comp like mine with modern I/O ... yes mine was only using 2143 and 2142 with LM833 buffers but an original side chain (LA4). A beginners project that in the single channel form is about as easy as it could get.

As for a genuine new projects ... I have one thing to say,
G9
that's why Byron and I leaped to test this thing out for Jakob.
I wanted to go around one more time on development BUT still this is a very cool project.

As pointed out above ... PRR and Kent brought that comp idea to all.
Fantastic stuff.

Instead of talking about it, head for the workshop and make something new or re-apply an old idea with modern components.

...... :shock:
sorry Kev will shut up now

 

[NewYorkDave]

A Passive EQ with ANY gain make up driver stage.

Done.

Well, at least one guy built one up based on my sketch, and seems to like it :green:

One of these days, I'll get around to building my own (the thing was designed with paper and slide rule, so to speak, and I never got a chance to make an actual prototype), but my implementation is going to be much simpler than Bluebird's awe-inspiring six-channel behemoth.

 

[Kev]

yes Dave
excellent work

you know I was generalising, sorry if some got caught by a very wide statement ...
many guys here have put their head down and done some great stuff.

Steff is one
seriously state of the DIY art
he doesn't say much and lets his pictures speak for him
a true DIY Jedi.

and there is more
look to the deep, original core of Group DIY.

:shock:
I said I would shut up ... again, sorry
This time I will ... Kev has left the forum :wink:

 

[Scenaria]

just one brief comment...

theres something to be said about reversing how someone or some company designed a product.... this affords guys like us to learn from the mistakes that others might have made... Im not talking about "cloning"..

It is always positive to create and design based on ones own ideas however the learning process also comes from observing the way others have done things both correctly and incorrectly...

If I had never worked as a second I would have been miles behind the competition in the studio.... I would have missed out on alot of important aspects to the studio world... the way I learned these things was by observing others doing it.. and then *copying* what they did... sometimes incorrectly and sometimes correctly... I developed my own opinions, thoughts and ideas buy going through this process.

I feel the same about audio design

 

[Winston OBoogie]

...

 

[Bear]

Damn, the 312 flavored preamp project that everybody's doing just about everything with is a prime example of all that can be learned from "cloning", since there's issues of the aftermarket or DIY opamp, issues of transformer selection, going dual and/or using servos (thanks to Joe Malone), and so on. I think there's a lot of general principle to be derived from this instance of "cloning".

Bear

 

[SSLtech]

Plexibreath,

-Is that pic of you... or Harry Shearer?

Anyhoo... the PWM was inspired by a Studer PWM compressor that I used which was quite reasonable, considering... it led me to thinking, which led to a wish to investigate the various GR topologies used.

Eddie Ciletti's site mentions that PWM can be very tricky , expensive and generates a lot of heat. (from memory... forgive me if I'm misquoting, Ed!) -Anyhow, I wanted to see if it could be done at a reasonable price, using readily available parts, and without roasting the rack.

The Gain reduction stage measured fairly well, worked reliably, tracks extremely accurately for stereo (better than FET and optos, by a long way, and almost as well as VCAs, with rather less matching and so forth required) and uses things like a 555 timer as the main oscillator...

keith

 

[rafafredd]

So, will you ever post it as a DIY project or do you have other plans for this one?

 

[Learner]

[quote author="SSLtech"]

Anyhoo... the PWM was inspired by a Studer PWM compressor that I used which was quite reasonable, considering... it led me to thinking, which led to a wish to investigate the various GR topologies used.[/quote]

Wooo...nice, please do tell us more!? studer schematic? :green:

Eddie Ciletti's site mentions that PWM can be very tricky , expensive and generates a lot of heat. (from memory... forgive me if I'm misquoting, Ed!) -Anyhow, I wanted to see if it could be done at a reasonable price, using readily available parts, and without roasting the rack.

Is that the interview with paul wolff?

The Gain reduction stage measured fairly well, worked reliably, tracks extremely accurately for stereo (better than FET and optos, by a long way, and almost as well as VCAs, with rather less matching and so forth required) and uses things like a 555 timer as the main oscillator...

keith

please tell us more, may be a schematic perhaps? can't expect an idea to work without any actual experimentation and implementation to test how it works.... :thumb:

:guinness:

 

[verbos]

You'll need to run that clock at a vary high frquency to be able to filter out the clock and not destroy th audio in the process.

That being said, the was to do it is by making a sawtooth oscillator that feeds a comprator and sending the control voltage into the other input of the comparator. The output of the comparator to the fet, you can get a very linear and predictable gain control element.

I've always avoided putting audio through one, but they're great for control side applications. You'll have to run that oscillator at more then 50 kHz for this to even be workable. Probably significantly higher than than even. MHz?

However, in many cases you need more than one, and you'll only need the one oscillator. So it gets simpler with multiple channels. So for quad or 8 channel units, this could be ideal.

Mark

 

[SSLtech]

[quote author="verbos"]You'll need to run that clock at a vary high frquency to be able to filter out the clock and not destroy th audio in the process.[/quote]
...how about 300-40kHz. That's what mins is running at, and it can go as high as half a MHz, but the FET spends more time in the transition stage if it goes too fast, and the distortion starts to creep up (still a small fraction of a percent, but every little bit should be fought!)

That being said, the was to do it is by making a sawtooth oscillator that feeds a comprator and sending the control voltage into the other input of the comparator. The output of the comparator to the fet, you can get a very linear and predictable gain control element.

Which is indeed exactly how it's done. The 555 produces a sawtooth across the charging capacitor. You have to put a high-impedange, low capacitance buffer after it so that whatever you feed it into doesn't alter the waveform & frequency. (capacitance at the far and of a small resistance will start to affect the "sharp" edge of the sawtooth)

You can use a triangle instead of a sawtooth... you can use anything you want apart from a square wave. The steepness of the slope at any given point afects how 'eagerly' the GR attacks the signal... a sine wave for example, would have a gentle beginning, a gentle tail and an aggressive middle section.

Actually, for a correctly-tracking stereo/quad/5.1 compressor, you don't just use a single oscillator, you'd use a single comparator too! Only the number of chopper FETs would have to change. There's one for each channel and one for the metering.

Ah... I've already given so much of the principle away, -Howzabaout I start a new thread and post the schematic if anyone is willing to take it further?

I don't have any web hosting, so if anyone wants I can do a PDF of the rough that I have. The GR element was working unshielded on my bench with THD+N of a fraction of 1%... The peak detection, filtering and the attack/release time constants are just sketched-in for now, but that's just stuff to tinker and fiddle with... takes more time than I have in the forseeable future.

Keith

 

[Gus]

I think two of the MXR petals were PWM based a filter one and a phaser. IIRC they use cmos ther are schematc on the web need to find them.

There are some problems with 555s when it switches it turns on hard and can make an almost inpossable to remove power glitch. I believe some of the newer versions might take care of that problem.

You could have the pwm section optical coupled there are photofet optos. Or you could use a LA x type photocell and have the pwm change the intensity of a led. I have some Forest Mims the 3rd books with PWM optical transmitters in them using the 555 with a small amout of parts.

Or yopu could go right off of pin 3 and modulate the signal.