Maestro Filter Sample and Hold

[Fp-www.Tonepad.com]

New project:

http://www.tonepad.com/project.asp?id=50

I will need some of the experienced builders (Mark comes to mind) to check this adaptation of the Maestro FS1 using LM13600 otas and including Tonda's mods.

Fp

[Bill Bergman]

Cool!! Wish I didn't already have one...... although it only the 2 knob model.

[Ry]

Dang, I built one from the GGG site too.  I may have to build this one just to see if it sounds any different.  I really love this pedal.

[Fp-www.Tonepad.com]

Please DO

And report back. As I said, not only is the layout not verified, but the circuit uses different chips and adds a bunch of mods.

I've checked it out many times and haven't found errors... but then again... who knows.

Fp

[StephenGiles]

Hi Francisco - buenos tardes, correct me if I'm wrong but the filter section is a normal triggered filter? If so what is the ripple like in the control voltage?
Stephen

[Fp-www.Tonepad.com]

Stephen,

Look at the DPDT switch that selects between filter and s/h modes:

the "A" spdt biases the pnp transistor. (when in s/h mod you set it with the trimpot)

the "B" spdt selects the source for the CV from the envelope detector (highly modified by Tonda to include direction, attack and decay) and the sample and hold which is the noise and LFO.

So, to my understanding, the filter section (made with the two otas and surrounding parts) is a band pass filter, controlled by the CV. So yeah, I would say it's a normal triggered filter.

This brings us to the next experimentation: since the circuit blocks are so well defined, one could use other CV filters (or anything CV controlled), or use other CV sources to control the filter. (manual control for example)

Fp

[moosapotamus]

Wow, that looks great, Fp! Beautiful work, as always.
I just might have to build myself another, too.

~ Charlie

[Fp-www.Tonepad.com]

HEY CHARLIE!

Remember how long it was when we started working on this!!?

What? two years??

Fp

[wui223]

why there are two transistor join together after the input stage?
why is Q4 missing ?
the mark '9' '10' at the transistors near the input stage meaning what?

[moosapotamus]

What? two years??

What? ohmygod. That long, really? :shock:
Wow, I'm almost as fast as you at finishing up projects. :lol:

Thanks
~ Charlie

[Ry]

I might have to give it a go.  I have yet to etch a board, though...this may not be my best bet as a first try!  I will report back if I get up the courage to try it out.

Ry

[Transmogrifox]

why there are two transistor join together after the input stage?
why is Q4 missing ?
the mark '9' '10' at the transistors near the input stage meaning what?

a)  This is called a darlington connection.  Using transistors like this significantly increases the input impedance.  The late Crybabies use the darlington input transistor.

b) I don't know why Q4 got skipped, but it may have been something present early in the design that had been determined to be unecessary and was removed, and it was easier not to rename everything else after the layout had been finished.

c) The LM13600 IC chip includes a darlington transistor built-in which is used as a buffer for OTA's, in fact, it is designed to be used how those JFETs are used.  It must have been determined by the designer that the filter had a warmer tone using FETs instead of the built in darlingtons.

Since the darlington pins would otherwise be useless, why add another seperate transistor to the design?  Simply use the extra unused transistor for an input buffer.

The short answer is this:  Pins 9 and 10 refer to the base and emitter of the darlington built in to the OTA chip.

[wui223]

thanks i check out the datasheet and give me the same explanation. but can i ask why the MAX1044 usage in Geofx differ from tonepad's schematic? the Geofx use two caps and tonepad use one only.
geofx connect pin2 to 4 by a 10uF cap but tonepad ground the cap and ground the pin 4 as well.
Can someone explain this, which one is right?
In fact i built this effect before but i didnt works, i wonder why

[Joe Viau]

Also, R. G. had a schematic for a CMOS pseudo-random generator on his site:

http://www.geofex.com/FX_images/pseudorandom2.pdf

Has anyone tried this with the old S/H and is anyone planning on trying it with the new revision? I've read threads about how persnickety the noise transistor is.  I have schematics for an ARP 2600, and I may look to see how that synth did its noise source.  It had pretty good S/H, too.

[Fp-www.Tonepad.com]

Transmogrifox got it right.

I'll re-check the max1044 part of the ckt.

Fp

[wui223]

i think the one in tonepad is right coz i try the connection shown in schematic, and i get -9V from pin5. But i use a 7660 instead of MAX1044
How to choose for the noisy transistor? i couldnt get one that is working right. For the FET can i simlpy use any FET like 2N5458?

[Fp-www.Tonepad.com]

Wui,

I just checked and it does match the schem at geofex. It's correct.

You can use the 2N5458, just make sure the pinout matches the pads of the layout.

About the noisy transistor, get a few 2n2222 or nte123ap and use a socket to try different ones.

Fp

[soggybag]

I made the FSH from the layout at uStomp. I used 2N3904 for the noise transistor. I hooked my scope to the noise opamp that amplfies the noise and tested a handfull of transistors until I found one that worked well.

I remember reading a few posts, as I was building mine, where people had tried several different varieties of transistor.

I'm a total novice, but if I understand what's going on the only two leads of the transistor are used and it is acting like a revese biased diode. So I would guess that you might even be able to use a diode here.

[Fp-www.Tonepad.com]

I remember RG once said something about transistors becoming more noisy after being biased in a certain way... maybe it's time to search the archives. Perhaps we can make a noisy transistor noisier.

Fp

[LP Hovercraft]

Good work FP!  On a previous thread I saw somebody mention that a leaky NPN germanium like a not-so-to-spec NTE101 might do the trick.  I'd be curious to see if anybody has tried it.  As for this layout, this makes a lot more sense to me than a bunch of icky 741's and the now "obsolete" CA3080.  Though I'm in the middle of perfboarding the GGG version,  I'm sure I'll make another using this version eventually.