Dyna comp questions

[Kipper4]

Since the lm13700 is essentially two ca3080s in a 16 pin package plus Darlington buffers.
And Ca3080s are getting harder to find and are silly prices.
Why couldn't I make a Dyna Comp using the much more available and better priced Lm13700?
Despite the larger chip size being an obvious downside vis a vis the pcb size.
I did consider using the buffers but I don't think it would work.
Would I need to ground out / Vb the unused pins ?
Opinions invited as I'd like to develop a diylc plan for this.
Rich

[R.G.]

Since the lm13700 is essentially two ca3080s in a 16 pin package plus Darlington buffers.
And Ca3080s are getting harder to find and are silly prices.
Why couldn't I make a Dyna Comp using the much more available and better priced Lm13700?

I'm pretty sure you could adapt the LM13700 to work.

[Kipper4]

I wasn't thinking of adapting the chip RG.
Is this what it would take?
Maybe I'm mistaken about the lm13700 being a dual ca3080?

[Mustachio]

You can leave the unused pins unconnected. It will work as a ca3080 replacement . Check out the Ibanez CP-9 at tonepad it uses the lm13700. They leave the second side of the OTA open for mods etc maybe you can do something with the other side and buffers.

I believe there where a few madbean projects that used the buffers of the lm13700 might wanna take a look for some ideas on how to use em. I was drawing up something similar a while back, will have to look for it again.

[Mark Hammer]

One of the smartest thigs to do would be to use one of the OTAs on the chip for a compressor, and the other one for a downward expander downstream to control noise.  Presumably, the same rectifier circuit would inform/control both.

[R.G.]

I wasn't thinking of adapting the chip RG.
Is this what it would take?
Maybe I'm mistaken about the lm13700 being a dual ca3080?

It's a dual improved and hot-rodded ca3080. There are differences. Fortunately, it will likely work in the same circuit, but tinkering it a bit results in better performance.

- It has predistortion diodes to increase the signal level it can take
- it has a different current mirror on the gain current pin, so that pin rides at two diode drops, not one
- it does not die suddenly until Iabc gets to 2ma, not 1ma like the CA3080
- it has built-in darlington buffers like the CA3094, but is dual unlike the 3094
- the transconductance is a tad different from the 3080s
- it's a bit more prone to RF interference

You can use it by ignoring one side - if Iabc is zero, that side has no operating current and is off. Really off. You can't use the darlington buffer for the dynacomp circuit, as you don't have access to the collector. So you still need all the transistors on the output of the OTA in the dynacomp.

Mostly, be prepared to tinker a bit if you go for this. As always, be very suspicious and decouple the power supply right at the chip, make all connections short and direct, and keep output well away from the inputs.

[PRR]

> Presumably, the same rectifier circuit would inform/control both.

One of Barry Blesser's papers shows why that is possible in math but difficult in real systems.

In particular: the limiter needs good level detection at high levels and the noise gate needs good detection at super-low levels. Trying to get both in one is extended precision, most of it pointless. OTOH a linear booster for the NG detector allows both detectors to be quite crude.

Audio Dynamic Range Compression For Minimum Perceived Distortion
Barry A. Blesser
IEEE Transactions On Audio And Electroacoustics
Vol. AU-17, No. 1 March 1969

The model in Fig. 1 shows the gain-computer section monitoring the input and output signals, although either one of them contains sufficient information to determine the same gain. Let us define a set of functions for the three cases when the gain is controlled by input, output, or both:

G = F,(Xo)
G = F2(Xi)
G = F3(Xo, Xi).

If the input and output are measured in dB, then

Xo = Xi + G.

For each of the three possible cases, the gain is the same for a given input (and hence the outputs are the same) if the 3 functions are related in the following way:

F1(X) = F2(X - F1(X)) = F3(X, X - F1( X )) .

If each of these functions exists and contains no singularities, then a compression system could be built using either the input, output, or both for controlling the gain. In practice, however, it is much easier to implement a feed-forward system for compression ratios less than one, and a feedback system for ratios greater than one. For this reason, the compression and limiting section are shown with the output signal controlling the gain, and the expansion section is shown with the input signal controlling the gain. Being able to implement an arbitrary gain function, however, does not help to determine which gain function should be used.

[Mark Hammer]

Now that I think about it more - and thanks for that prompt, BTW - one of the properties of many compressors is that the rectifier's output tracks the compressor's output, such that it reduces the gain of the OTA, which in turn results in a reduced amplitude from the rectifier, which modifies the OTA gain, etc.  That feedback, while appropriate for compression, is inappropriate for downward expansion.

[slacker]

There's some discussion here about being able to parallel the two halves of the LM13700 to get a better signal to noise ratio. http://electro-music.com/forum/post-347005.html The example they talk about is to allow the use of an LM13700 to replicate a particular "trick" you can do with CA3080s but I guess it applies to other configurations, no idea if there's any practical advantage for what we do over just using half the chip.

[Tubebass]

Korg CMP-1:

http://analogguru.an.ohost.de/193/schematics/Korg_CMP-1.gif

[haveyouseenhim]

I'll just leave this here......

http://www.diystompboxes.com/smfforum/index.php?topic=105341.0

[Kipper4]

Thanks guys schematic on the way.
If you care to look it over for me I'd appreciate it.
Also should I implement Larrys idea of the voltage divider on the open pins?
Thanks for the help
Rich

Edit; Heres the schematic with Larrys adaptation Version 2
https://www.dropbox.com/sh/7ma9n53c5jqdgsr/d-0TvwYXDL#lh:null-lm13700%20Dyna%20Comp%20V2.png

[Lurco]

Thanks guys schematic on the way.
If you care to look it over for me I'd appreciate it.
Also should I implement Larrys idea of the voltage divider on the open pins?
Thanks for the help
Rich

Edit; Heres the schematic with Larrys adaptation Version 2
https://www.dropbox.com/sh/7ma9n53c5jqdgsr/d-0TvwYXDL#lh:null-lm13700%20Dyna%20Comp%20V2.png

You mixed up the connections to pins 1 & 5 of IC2?

[Kipper4]

Thanks very much for the Eagle eye.
Schematic ammended and reposted in the above post
Appreciate that Lurco.
Now might be a good time to do a perf/pcb layout.!!!!!!

[PRR]

here

C16 is grounded at both ends (useless).

There's no timing cap (probably C16?).

C9 has no value.

Is this a copy or an interpretation? I'd move R2 to the input jack so that C2 has something to work against.

[Kipper4]

Thanks Paul Well spotted.
This represents hours of work and I am most grateful
reposted here

https://www.dropbox.com/sh/7ma9n53c5jqdgsr/d-0TvwYXDL#lh:null-lm13700%20Dyna%20Comp%20V2.png


I got the Ross Schematic i'm working from here at GGG

http://www.generalguitargadgets.com/pdf/ggg_dnr_ross_sc.pdf?phpMyAdmin=78482479fd7e7fc3768044a841b3e85a

Do I still need to move R2 to input?
Thanks

[Kipper4]

i found another mistake be right back

edit above ammended anew

I'm hoping i'm somewhere near now.

[Lurco]

Thanks Paul Well spotted.
This represents hours of work and I am most grateful
reposted here

https://www.dropbox.com/sh/7ma9n53c5jqdgsr/d-0TvwYXDL#lh:null-lm13700%20Dyna%20Comp%20V2.png


I got the Ross Schematic i'm working from here at GGG

http://www.generalguitargadgets.com/pdf/ggg_dnr_ross_sc.pdf?phpMyAdmin=78482479fd7e7fc3768044a841b3e85a

Do I still need to move R2 to input?
Thanks

Base of T5 still wrong (should connect to pin5 - not pin1). PIN-numbers of CA3080 and LM13700 are different!

[MrStab]

re. grounding unused inputs:

i'm pretty sure Jon/Midwayfair on these forums recommended grounding the unused inputs on the LM13700 as he found it had some effect on lowering noise. don't quote me on that, though. in any case, it seems to work whether they're grounded or not - would just be more ideal if they are.

[Kipper4]

Thanks Fellas
Changes made here.

https://www.dropbox.com/sh/7ma9n53c5jqdgsr/d-0TvwYXDL#lh:null-lm13700%20Dyna%20Comp%20V2.png


The only thing i havent done was move R2
I'm still waiting on PPRs advice now that I've posted the schematic im working from, to see if its still advisable.
Cheers
Rich