op amp gurus

[peter purpose]

Oh masterful ones..... I have been foolish...... I have cobbled together bits from here and bits from there, yet have not found the path to the tone.

Please help me sort out my mess.

I think my problem lies with amps 3-7. They are dumping hash at extremes of cut and boost and are probably just plain wrong.
I suppose that ideally they should be fet input jobbies, so would you kind souls please impart the pearls.

As always... I bow before you.

peter

 

[tk@halmi]

Edit: Damn my eyes, C1 is 10pF! I need glasses.

I would have added a BJT emitter follower to the output of those. The performance improvement is worth another transistor. Better yet a constant current sourced emitter follower. Did you pluck this circuit out of the Zetex app notes?
Why not build the same citcuit on a little breadboard, add a feedback network (20k resista and 20k pot), add a 20k load to the output and hook it up to the scope. Increase C1 until it is stable at all gain settings (x2 to infinity).

 

[peter purpose]

Hi Tamas,

<<Did you pluck this circuit out of the Zetex app notes? >>
No, I just saw a simple op amp schematic and chucked what I had in there.

If I were to keep the exact same circuit, are you saying that increasing the value of c1 will eradicate the hash?

Question 2...
If I were to swap the input pair to 2N3819 jfets, would I need some serious resistor changes?

Thanks for your time.

peter

 

[bcarso]

Have you accounted for BPE?

 

[peter purpose]

BPE?..... Beer, pizza and erotica?

 

[bcarso]

Banana Proximity Effect

 

[peter purpose]

:green: :green: :green:

Ok... I'll try some kiwi and report back.

 

[peter purpose]

The kiwi didn't work, but I just made the bass amp into this.....

Dropped the voltage to 18V and all seems good on the scope.

peter

 

[Bo Hansen]

Peter,

It not a absolute need to use a discrete op-amp for this IC 3-7 gyrator circuits, use a good JFET op-amp, because this is not a real signal patch.
(this gyrator is similar as a tunable "passive" LC-filter connected to ground)
Ok, if the amp have very low slew rate and poor load marginal, it can have an effect on the sound, but the must important op-amps are IC 1 and 2.

I recognize this schematic, it is the eq-section from Alice ACM/ABCM consoles, Ted Fletcher`s baby.

--Bo

 

[peter purpose]

You are of course right on the money Bo. ACM2.

I just fancied having a go at making a discrete version.

The HA4741s that were in there have a slew rate of around a fortnight.

peter

 

[bcarso]

[quote author="peter purpose"]:green: :green: :green:

Ok... I'll try some kiwi and report back.[/quote]

Kiwi is particularly good with WiFi networks, I understand.

 

[clintrubber]

[quote author="Bo Hansén"]Peter,

It not a absolute need to use a discrete op-amp for this IC 3-7 gyrator circuits, use a good JFET op-amp, because this is not a real signal patch.
(this gyrator is similar as a tunable "passive" LC-filter connected to ground)
Ok, if the amp have very low slew rate and poor load marginal, it can have an effect on the sound, but the must important op-amps are IC 1 and 2.

I recognize this schematic, it is the eq-section from Alice ACM/ABCM consoles, Ted Fletcher`s baby.

--Bo[/quote]

Or perhaps just 'one-transistor gyrators' would improve things ? Haven't compared topologies, so am not sure these would work here. Have seen various circuits using a one-discrete.

 

[tv]

one-transistor ones are all over boss-stompboxes. I've posted something here too http://www.groupdiy.com/index.php?topic=166674#166674

could one-transistor thing sound *more alive* ? f.e.:
http://www.freeinfosociety.com/electronics/schematics/audio/pictures/bossmt22.gif (now obsolete ...)

I've built all sort of guitar stuff some decade (or more) ago and I liked sound of tda2320a and transistor circuits more than that of *better* stuff. It was more alive (then again tda2320a ain't exactly a bad chip - but 5532 is cheaper).

 

[clintrubber]

[quote author="tv"]one-transistor ones are all over boss-stompboxes. I've posted something here too http://www.groupdiy.com/index.php?topic=166674#166674[/quote]

Right ! These the ones I was thinking of, but was too lazy to provide a link to a schematic :wink: Also old graphic EQs doing this. IIRIC the discrete is used as an emitter follower, same principle as using an opamp wired as a buffer.

Have a look at Q7:
http://www.freeinfosociety.com/electronics/schematics/audio/pictures/bossbd2.gif

three more here:
http://www.freeinfosociety.com/electronics/schematics/audio/pictures/bossmt22.gif

 

[tv]

you just spoilt my post. eh one really can't go find a scheme and paste it into forum these days ...

but yeah, they've had also some different kindof lowpass one-transistor circuits - iirc some called them *simulated inductors* (if I find the schemes) with 3 caps and 4 resistors. I wonder if these can be *installed* somewhere in ss-pultec hi-boost/hi-q path. If I think more, that could provide *both* Q-control and boost level and switchable bands. The only drawback is how to get rid dc and - one would have to calculate the thing.

O.K. that was OT. end of interruption.

 

[clintrubber]

[quote author="tv"]you just spoilt my post. eh one really can't go find a scheme and paste it into forum these days ...[/quote]
Nah, we're just together messing up this thread with one transistor gyrators that Peter might not feel much for :wink:

But did we both thought of the same circuit as an example ? That's fun :thumb:
Or did you just copy a link and put it in your post-edit above ?! :grin:

 

[cuelist]

Peter,

This eq topology was described in some detail in Steve Dove's legendary series "Designing a Professional Mixing Console" (Studio Sound Magazine, early 1980's). Steve worked for Alice at the time.

I have these articles somewhere but from memory I recall that while the CAPS (constant amplitude phase shift) type bandpass filter is less taxing for the opamps than the integrators of a state-variable filter, in notheless requires some serious GBW and I doubt that 3-transistor circuit is suitable for the job. It really is a job for a "proper" opamp, discrete or not.

Nice to see some Alice circuits appear, they are lovely.

 

[tv]

But did we both thought of the same circuit as an example ? That's fun

I copied the adress, sure, but I went to the freeinfo site to get it - not from your post.

For the Watsons, here's another way to get to the same place:
http://filters.muziq.be/model/boss/compact/mt2?PHPSESSID=f28512edcb3016c19b361ec2823c9405

All in all one-tran gyrators seem like a sort of a blasphemy in *hi-end* enclosures. I, otoh wouldn't mind having a [ctepeo !!!] pultec clone size of a stompbox. With mos-phat gainstage

I mentioned a circuit (basically a lowpass) which some referred to as a *simulated inductor* - around Q3 in this schematic:
http://www.freeinfosociety.com/electronics/schemview.php?id=117
could this one be modded to go into a ss-pultec?

It really is a job for a "proper" opamp

I wonder if a simple 3-tran circuit could be extended with something like circuits from here:
http://www.tubecad.com/2004/blog0023.htm
so it could become a sort of a mojo-amp (***not your ordinary ...) maybe with irf510 or likes.
So it would be two bf256 (2n3819), bc560 and two irf510. Irfs could add to the mojo-loox as well.

 

[bcarso]

[quote author="cuelist"]Peter,

This eq topology was described in some detail in Steve Dove's legendary series "Designing a Professional Mixing Console" (Studio Sound Magazine, early 1980's). Steve worked for Alice at the time.

I have these articles somewhere but from memory I recall that while the CAPS (constant amplitude phase shift) type bandpass filter is less taxing for the opamps than the integrators of a state-variable filter, in notheless requires some serious GBW and I doubt that 3-transistor circuit is suitable for the job. It really is a job for a "proper" opamp, discrete or not.

Nice to see some Alice circuits appear, they are lovely.[/quote]

The phase-shift filter (another multiply-invented topology) is easier on the opamps in terms of not requiring as stringent adherence to single-pole-dominated open-loop behavior, compared to the feedback integrators of the conventional (Kerwin-Huelsman-Newcomb) state-variable filter. It has the disadvantage of requiring common-mode swing and hence has higher common-mode distortion.

If the gain-phase requirement is met the integrator-based filters actually require less open-loop gain for the same accuracy. Having said that I agree the three-Q opamp is a mite minimalist, especially when there is common-mode swing.

 

[bcarso]

[quote author="tv"]
I mentioned a circuit (basically a lowpass) which some referred to as a *simulated inductor* - around Q3 in this schematic:
http://www.freeinfosociety.com/electronics/schemview.php?id=117
could this one be modded to go into a ss-pultec?

[/quote]

Just looks like a standard three-pole unity-gain lowpass to me.

What is often used as a synthetic grounded inductor is the input to a Sallen-Key unity-gain two-pole highpass. The equivalent L has an R in series and another (generally different) R in parallel, so it's a bit awkward to use. With a C in series it forms a grounded R-L-C resonator, but because of the resistance across the L the impedance at high frequencies is never as high as at low.