My polymoog battle. Tl072?

[sameal]

Would substituting a TL072 for the mc1458's draw too much current?

(I did replace the l.e.d.s with modern ones, so i got back a little wiggle room on that)

I'm thinking putting in TL072's where the MC1458's are will drop noise significantly. I did it already with NE5532 and it was great!(see link below) but im afraid of over taxing the supply going with 5532.


Backstory: my polymoog since i bought it many years ago has never been right. It was pretty close, but not enough. I replaced the caps in the supply immediately and that helped a little. I then replaced the faratron supply with a moog supply from the 280, and recapped. Still pretty noisey

Just recently, i replaced all diodes and semi conductors on the power supply, because two diodes burned up after i replaced the mc140007 chips with cd4007ube. All of that seems fine now. The synth is sounding better, but the noise is still pretty crazy. Im also replacing all sockets with machined pin types, as the old sockets are brittle, dirty and causing issues.

Here is the other thread i have going:
https://www.gearslutz.com/board/geekslutz-forum/1083158-my-polymoog-battle.html


Take a look at the last post for sound files

 

[benb]

I've read a little about the Polymoog design, as mentioned in the thread it uses electronic-organ-style tone generators, and a typical design of the time can let the tones come through without keys being played. I had a Farfisa MiniCompact as a teen that did this as designed, the tones were, I'm guessing, 40dB down from when a key was pressed. As I learned more about electronics I dug into it and fixed it (in many ways), but it switched the tones through mechanical switches, a different design and story from the Polymoog.

I've read the Polymoog allegedly uses a "chip for every key" design, and it was supposedly a custom chip. If whatever I read meant the CMOS 4007, that's definitely not custom, that's a jelly bean chip that's been made for as long as the 4000 series. Maybe there's another "special" chip that generates the attack-release voltage for each key. Maybe I should look up a schematic.

Your mention of the 4007 part immediately brought to mind using FETs as variable resistors (I may have seen this in Popular Electronics or the like in the 1970s), as the chip has several FETs with all the terminals available on pins. It's a weird, bizarre design for a 4000 series chip, or for any "logic" chip. Here's some circuits that do just this - the Polymoog almost certainly uses something like (but probably not the exact same schematic) the VCA schematic in the lower right:
http://www.schmitzbits.de/vcf4007.html

That chip tester you had might work for "fully digital" chips, but this particular one is, without a doubt, used in analog mode in the Polymoog, and I wouldn't trust an average 4007 to work right in this circuit even if that thing DID say it passed.

Different brands (or even different production runs) of this chip may have slightly different gate voltages where the FET starts to turn on, and that may let too much signal pass through when the key is off. Also, whatever generates the control voltage to the FET gates may be sitting at a "wrong" voltage with no key pressed, letting too much of the signal through.

A 5532 or even TL072 is surely better than a 1458 dual op-amp as far as noise, but op-amps are surely not the cause of the noise you described. How many of these does it use? I do recall the 5532 actually gets warm on +/- 15V. Get the tone feedthrough thing in control, then maybe you can detect some opamp noise go away when replacing the 1458.

 

[benb]

One more thing that may be of interest, this guy recommends replacing ALL 4000 series chips in older synths - we've heard about replacing electrolytics a million times (he also goes into that), but this is the first I've read of replacing chips:

http://www.oldcrows.net/~oldcrow/synth/tips.txt

 

[sameal]

Yeah, that i.c. tester was junk. Failed every chip, no matter what it was.


I read that old crow articule, and i think it has merit. I've been replacing chips and sockets all over this thing and it keeps getting better.

The presets always bled like that, even before the chip swaps.  I have a file someplace im trying to find of the moog before i started working on it. I had to cut out the quiet sections on the moog track so it didn't sound like a spaceship landing.

So the 5532's would probably be a bad choice im guessing? There's about 20 of them throughout.

In the last post on my other thread, the sound file of the 5532, the bled was about cut in half. Im thinking if i replace the 1458's in the circuit before the preset cards it might just do the trick.

Otherwise, is there a better chip then CD4007UBE?

 

[JohnRoberts]

IIRC the CMOS CD4007 is just some uncommitted  mosfet pairs inside a dip package that can be reconfigured and used like discrete devices.

Anything different will not work the same.

JR

 

[sameal]

IIRC the CMOS CD4007 is just some uncommitted  mosfet pairs inside a dip package that can be reconfigured and used like discrete devices.

Anything different will not work the same.

JR

This is true.

I used them to replace the M140007 chips that originally came in the poly.

The cd4007's seem to be working as they should be.

 

[Monte McGuire]

The 1458 is spec'ed at 45nV/root Hz, so even a TL071 at 18nV/root Hz will be quieter, and a whole pile of other IC op amps will  quieter still. The TL071 will have a ton less current noise too, since it's a JFET input amplifier. But, the noise of the entire circuit is more than just from the amplifiers - the VCA-like things may dominate. And, the slow, sludgy dynamics of the 1458 are probably part of why that keyboard sounds the way it does. While you might reduce the noise, it won't be a polymoog anymore, so tread lightly...

Best of luck!

 

[Monte McGuire]

Forgot your first question: the power supply spec of the 1458 and the TL072 are nearly identical, both around 3mA for both amplifiers, so they should not cause a problem with the power supply.

 

[sameal]

Well i have never known a polymoog in 100% condition, so i don't know what that sounds like anyway.

The other thoughts i had were metal film resistors, but i can't imagine that having a great impact.

There is a handful of chips i have not replaced. Most of what i have replaced has been control type things. 4011/140007/7401/3080

The 3080's are on their way, but i can't see them making things a whole lot quieter. I think im going to hear a difference, but not a whole lot of difference.

So those changes anyway shouldn't affect the original sound.

 

[sameal]

I also wonder, and im not a master at these things so forgive me if this is silly, or already implemented in the system, if i use decoupling caps on the 1458's individually if that would make them quite enough.

The other idea, keep the 1458's in the preset cards and only change the 1458's on the input of the top boards.

 

[fanwander]

Hello,

Having repaired and modified several Polymoogs (and destroyed one ;-) ), I can go with benb: the major noise of the Polymoog is the "ring through" of the 72 tones. And even if you neglect this, then the cause of the noise is the so called Polycom on the voice cards. The 1458 in the  bus boards are neglectable in regards of the noise for sure. My final solution was a Drawmer DS201 noise gate...

Florian