LM13700 - CA3080 adapter board

[armdnrdy]

I though that this might be of some use for adapting a LM13700 to an existing 3080 project.



Here's the PDF with info and PCB print:

https://dl.dropboxusercontent.com/u/53299166/DIYstompboxes/LM13700%20-%20CA3080%20adapter%20board.pdf

[haveyouseenhim]

Nice. Now we just need to make an adapter board for all of the fake LM13700 chips from china. 

Good work!

[armdnrdy]

You know what's funny....

Before I started working on this I wanted to make sure that there wasn't one available on the net....so I searched around....

Do you know what I found? An adapter board to combine two 3080s to make a 13700! 

[Mustachio]

Nice work Armdnrdy! really nice pdf !

Haha what 2 3080 adapter board! crazyness!

I was lookn at it  and wasnt sure about the resistors. Pull down resistors from power and input ? That works good for most applications ?

I been trying to figure out how to use the buffers on the 13700 , found a few projects that used them but they where for synth modules I think.  Got any ideas for implementing the buffers ?

[armdnrdy]

Thanks Jim.

The two resistors make a voltage divider on the non inverting input. The inverting input is tied to the output.

I found a post on Texas Instrument's site where the question was asked, "How to deal with the un-used cell for LM13700?"

The answer from a T.I. engineer was: set up the unused portion as a unity gain follower with it's input at half supply.

I figured for the price of two 10K resistors, (next to nothing) and a couple millimeters of board space...it can't hurt.

As far as using the buffers...I don't really see the buffers used in the project files that I have. The designer of the LM13600/700 stated in an article that they were added because 3080 designs generally use buffers.

I am not aware of such designs.

The buffers collectors are tied to V+......so that limits the use as a local general purpose NPN.

Hopefully R.G. will pop in! 

[armdnrdy]

Here's a good read by Ray Marston. Article on 3080s and 13700s.

Understanding and using OTA Op Amp ICs:

http://www.nutsvolts.com/uploads/magazine_downloads/11/April%202003%20Ray%20Marston%20-%20Understanding%20and%20Using%20OTA%20Op-Amps.pdf

[midwayfair]

Haha what 2 3080 adapter board! crazyness!

I actually need one of those ...

[Mustachio]

Haha what 2 3080 adapter board! crazyness!

I actually need one of those ...

Blasphemy! Your in luck I believe thats called an LM13700    

Nice link armdnrdy! I looked up the next edition with the lm13700 info.
http://www.nutsvolts.com/uploads/magazine_downloads/11/May%202003%20Ray%20Marston%20-%20Understanding%20And%20Using%20OTA%20OP-Amps.pdf

Check out the bottom vcf's looks like their using the buffers. I gotta try bread boarding something up with that when I get time I love vcf's!

:::Edit:::

Oh yeah when I first read the url I read it as Nuts And Volts and lol'd! that would be a great site but nutsvolts is great too! 

[armdnrdy]

I have a combined copy saved in my OTA folder.

I thought the link that I posted had both articles. 

Good you found it! Great articles!

[midwayfair]

Haha what 2 3080 adapter board! crazyness!

I actually need one of those ...

Blasphemy! Your in luck I believe thats called an LM13700    

Madbean's Sharkfin (FSH-1) uses two 3080s instead of an LM13700. Brian said it was because it fit better on the layout, but the two 3080s are right in line with each other, so I'm a little confused by that reasoning. I bought the 3080s, but every time I look at them, I think "eewwwww."

[R.G.]

Madbean's Sharkfin (FSH-1) uses two 3080s instead of an LM13700. Brian said it was because it fit better on the layout, but the two 3080s are right in line with each other, so I'm a little confused by that reasoning.

It has to do with where the pins are on the chip for a given function. I prefer two dual opamps to one quad opamp for the same reason. The signal flow on the PCB layout can be improved. I suspect that Brian has read my "PCB Layout for Musical Effects", where the topic is addressed. I've posted the concept here, too.

[jdub]

That, my friend, is very useful.  Thanks!

[Fender3D]

Buffers are actually separated from OTAs, you can use them for anything you need:
Vref buffering, extra filtering stage, OTA's Vc etc..

You can make 4 LP/HP filters out of a 13700 using buffers, or double buffer a dynacomp...

[midwayfair]

Madbean's Sharkfin (FSH-1) uses two 3080s instead of an LM13700. Brian said it was because it fit better on the layout, but the two 3080s are right in line with each other, so I'm a little confused by that reasoning.

It has to do with where the pins are on the chip for a given function. I prefer two dual opamps to one quad opamp for the same reason.

So ... you're saying we can't win? Noisy 3080s or awkward 13700 signal flows?



BTW, RG, the smallbear link for your book on your site seems to be broken. Maybe the link changed. http://www.smallbearelec.com/servlet/Detail?no=679

[StephenGiles]

What about the SSM 2164 quad VCA? Not OTA I know, but only Ashdown pedals seem to use them.

[R.G.]

Maybe Gilbert cell.

It is possible to make an OTA with monolithic pairs of transistors. I think I posted that here some years ago. The tracking and such won't be as good, but you *can* make them.

The input diffamp is what's critical for matching, and everything else needs matched only locally to the pair, as the others are all current mirrors. The mismatch from mirror to mirror is much less important than mismatch within the pair, and matched pairs are what you can get. There are no resistors or caps in an OTA - only active devices. It's all silicon.

[armdnrdy]

The input diffamp is what's critical for matching, and everything else needs matched only locally to the pair, as the others are all current mirrors. The mismatch from mirror to mirror is much less important than mismatch within the pair, and matched pairs are what you can get. There are no resistors or caps in an OTA - only active devices. It's all silicon.

Is it essential that the Darlington pair in the emitter follower buffer of both the 13700 and 3094 be matched?......or....not so much?

[R.G.]

Matching doesn't mean all that much for Darlington pairs. In fact, it's usually counterproductive. The first transistor in the cascade (as opposed to pair) is guaranteed to be working at a collector current that's 1/HFE of the second. So you want a first device that's optimized for gain, bandwidth, linearity, etc. at a much lower current. This is one reason that monolithic darlingtons are generally better performing than discrete - a lot of thought went into optimizing them for the situation.

[armdnrdy]

So not matched - but selected for a particular performance. Got it. Thank you

[DIMstompboxes]

If your tired of using FET for MXR noisegate I recommend this http://hammer.ampage.org/files/Expander.gif
I got this way back years on Mark's website and I built this using separate TL061 and 3080 and its verified. I just copied the PCB design and scale to fit my case.
Although the PCB is not for LM3700 but I think it can be done. If only someone can redesign the PCB into LM3700