subbing an lm13X00 for a pair of ca3080's

[pinkjimiphoton]

hey guys,
built me another honeydripper... found an ancient board brian burned for me years ago for the original 3080 variant.

built the sucker. no quack. voltages all good. first guess is shitty counterfeit 3080's. i bought some from jameco, who used to have decent parts. turned up as ca3080n's. no bueno.
pretty sure that's why the honey ain' drippin.

but i DO have lm13600/700's in stock. so i figured i should be able to just use one of them, and cover the two 3080's i need til hopefully some legit ones come in in a week or so. i had a whole sleeve of the sons of bitches, but i never thought i'd use 'em so i gave 'em all away <way to go, pink, you effin moron...>

so anyways, i had a google for pinouts, and tried my hand at a conversion table. i believe this is good to go, but have a couple questions still




the pins marked nc on the 3080's i'm not worried about BUT do i need to ground the pins for the buffers.... pins 7 and 8, and 9 and 10? or can i leave them floating? i'm not real savvy with this stuff.

also, i'm assuming stuff that's common to both chips, i only need to hook up to one, correct?  i mean pins 4 and 7 on the 3080's.

the plan is to make one 16 pin socket for the 13x00, and have the pins wired to two other 8 pin sockets to plug in where the ca3080's go. either it will work... or it won't.

but if it DOES, then i don't need to wait for the new parts to arrive and i can get this beeotch quacking today, which would be good cuz i'm outta cheeb and need something to occupy my one frazzled brain cell til my brethren gets back from jamaica.

i know, dumb questions from an idiot mind in chaotic times. but regardless of the tomfoolery out there, i wanna make this suckah quack!

thanks in advance yo

hopefully helmut will forgive me for reposting this from his defunct but awesome website, courtesy of the wayback machine...

[Slowpoke101]

I've wired LM13700s up like that as 3080s before for fun and they worked.
Try it. Should be fine.

[pinkjimiphoton]

thank you, ian, sir!

i DID wire it up, got good voltages even, but it didn't work in this application unfortunately.
i've kinda-sorta got it working with the crummy 3080's a little if i gotta fuzzface cranked up in front of it
and get my guitar just right.. but it's nothing like the other ones. tried a bunch of chips in it, it should be quacking better by now. all that's left to swap out is the real deal jelly beans when they arrive this week sometime.

[R.G.]

It should be fine. The differences in operation of the 3080 OTA and the OTA section of the 13700 are trivial. They work the same way. Match the pin functions and it goes.
Here's a run down of the differences:
Amplifier bias current pin sits at two silicon diode drops (~1.1V) instead of one silicon diode drop (~0.5V) because of the improved current mirror inside. This generally doesn't cause any problem.
The diode bias pins can be ignored. Or if you trickle some current into them with a 15K to 22K resistor, they raise the input signal level you can use before it distorts.
The darlington buffers can be ignored.
That's about it. Generally if you got a circuit to work with a 3080, the OTA part of the 13700 will just work.

[pinkjimiphoton]

thanks for the confirmation rg!
it worked out about as well as the bogus 3080's. you can hear a bit of quack, but its pretty faint.
we'll see how the intersils work out when they come in. rock on!

[PRR]

It should be fine. The differences in operation of the 3080 OTA and the OTA section of the 13700 are .... ..... Amplifier bias current pin sits at two silicon diode drops (~1.1V) instead of one silicon diode drop (~0.5V) ....

And the plan which PJP posted *has* a D112 noted as 'voltage shift for LM13x00'. Adds one diode to the Iabc voltage (misusing a current for a voltage). AFAICT, he only needs a pinout hack and to follow the fine-print comments.

[Slowpoke101]

And the plan which PJP posted *has* a D112 noted as 'voltage shift for LM13x00'. Adds one diode to the Iabc voltage (misusing a current for a voltage). AFAICT, he only needs a pinout hack and to follow the fine-print comments.

Paul has very sharp eyes for finding that "fine print" info. I totally overlooked it. I then though I should have had a closer look and I found another mod' that also needs to be done when using LM13X00 chips - D117. It is used to raise the voltages of the 10V, 5V and the 2.5V rails by about 0.6V. This may be why Jimi's unit (along with D112) isn't playing nice. Perhaps this may help but Jimi probably already knows all this.

[pinkjimiphoton]

And the plan which PJP posted *has* a D112 noted as 'voltage shift for LM13x00'. Adds one diode to the Iabc voltage (misusing a current for a voltage). AFAICT, he only needs a pinout hack and to follow the fine-print comments.

Paul has very sharp eyes for finding that "fine print" info. I totally overlooked it. I then though I should have had a closer look and I found another mod' that also needs to be done when using LM13X00 chips - D117. It is used to raise the voltages of the 10V, 5V and the 2.5V rails by about 0.6V. This may be why Jimi's unit (along with D112) isn't playing nice. Perhaps this may help but Jimi probably already knows all this.

nope, in fact, jimi is an idiot

i just gotta add a diode to pins 1 and 16 of the 13x00? does it matter what kind of diode? i am but an egg with most of this stuff.
monkey with a breadboard, opposable thumbs and too much time on his hands and a strange hankering for the magick smoke of burning silicon... its an alchemical thing i think.
i've got 914's and schotkes of all persuasions and voltages guess i need to try and grok this a bit better.

again, thank ye, gennlemenz


if i get a chance to mess with it later, i'll see if i can get it to quack. right now i'm messing with a continuation of a joke pedal.
i cobbled together a simple treble booster using literally only parts on my benchtop.
stupid little one transistor dealio, no knobs, just off and on. gives ya about a 4db bump in your upper mids and treble.

called it the talent boost.

so i made the talent boost deluxe model, with input and output levels. after much dicking around, realized it sucked baalz and had to correct it. during the course of experimentation, it went germanium only, and has no connection from e of the weird npn ge that runs it to ground other than thru the 47uf cap to ground there from the original rc network i was playing with. the leakage thru the cap makes the damn transistor bias up nice and get a real nice growl to it. kinda almost like the first stage of a harmonic perc.

its weird, and may just be a fluke, but if it persists, i'll make some video happen. its kinda weird, but cool. sadly its a one-off. i've tried similar gain and leakage transistors, and only this weird 404 style deal out of an old thomas organ makes it shine.

anyways... different sortie for a different daze...

thanks again guys

[Slowpoke101]

You only need two 1N914 diodes or any standard silicon diode will do. You are only looking for the 0.6 to 0.7 volt drop across the diode junction. This is partly to do with the extra diode junction that R.G. mentioned that is in the LM13X00's current mirror - usually this doesn't bother us too much but it seems to annoy this effect unit.

I hope that you have a circuit diagram relevant to the circuit board (PCB) that you are using. There are three diodes in series that connect to 0V (GND). You need to install another 1N914 in series with this lot so you have a total of four. I think Helmut refers to this as the "Bounding Network" section..Not sure of that.

Next. Find the two 5V1 zener diodes - then identify the one whose anode (non banded end) connects to 0V. Desolder that end from the PCB and lift it up. Feed the cathode of the remaining 1N914 into the PCB hole and resolder. Solder the anode of the zener and the anode of the 1N914 together and that's it, finished. Check voltages are 18V, 10.6V, 5.6V and 3.1V (or so).
Finding the 2.5V (3.1V) rail may be fun so follow the 5V line to were the 22K resistor is. One side of it is 5V and the other is the 2.5V rail. Good luck.

I had a look at Helmut's site via the Wayback Machine. Very interesting info indeed. Apparently you can't just drop an LM13X00 into one of these things and expect it to work straight away. It just sounds terrible. Modifications must be done.

Edit: But if we look at Madbean's Honeydripper doco (it's in his archive section), he's just plopped a LM13700 in there and it worked just fine. Hmm.... I might just build this for fun to see what differences there are - as if I don't have plenty enough to do anyway. But none of that is fun but this would be. Plus I get to play with another effect. Bonus! 

[PRR]

> does it matter what kind of diode?

You are semi-cancelling a plain Silicon small transistor drop. A plain Silicon small diode is the obvious choice, 1N914, 1N4148, etc.

The OTA is the least part of this plan. Are you sure the signal processing makes a good 'ploop' when you strum?

[pinkjimiphoton]

> does it matter what kind of diode?

You are semi-cancelling a plain Silicon small transistor drop. A plain Silicon small diode is the obvious choice, 1N914, 1N4148, etc.

The OTA is the least part of this plan. Are you sure the signal processing makes a good 'ploop' when you strum?

nope, have to run a strong fuzz into the front of it to get minimal response. others i didn't have this issue. so far, most voltages seem what i'd expect.
i undid the germanium i had in the first detector stage (d3 and 4) and went back to 914's there.  r 5 is a trimmer i set to where it seems closest to unity.
r21 is a 10k trimmer. again, trimmed by ear to the sweetest spot.
but none of it seems to be quacking correctly. assumed it was shitty ca3080's. had that issue in the past i think with the nte "equivalent" in the first build years ago. bought a sleeve of nos intersil's from a guy with good feedback, we'll see what happens.

i don't want to mod the circuit til i get it quacking... so first, i guess i'll wait for the jellybeans to come in, and try the correct part first.

thanks guys. its a gas circuit... when it works.

[pinkjimiphoton]

so added 914's to the 13x00 last nite... slightly better quack, still kinda weak tho

if i get a chance will add another to the suggested spot in the power supply and see if it works.
forgot to bring the paperwork down to the dungeon last nite.

thanks for the help and advice!

[anotherjim]

Ouch, that's one complex build. You wouldn't want doubt & uncertainty with the OTA quality.
Possibly dumb question -  is the expression jack fitted? Do Mode1 & Mode 2 short together without a pedal? I don't even know if your board has this option, but you know, it's the silly things like that...

[pinkjimiphoton]

hi jim!
i just did the madbean board. i MAY do the exp mod once i get it working! i'm still kicking myself in the ass for getting rid of the first one, and the last one i had built... them suckers had some SQVACKKKKK!!! to them