Question about LM13700 offset

bugbrand · Posted: Sun Jan 31, 2010 5:02 am Post subject: Question about LM13700 offset

Hi there,

I've been using the LM13700 for plenty of designs, but have a bit of a head-scratcher regarding offset voltages..

Here's an example little circuit:

Posted Image, might have been reduced in size. Click Image to view fullscreen.

Posted Image, might have been reduced in size. Click Image to view fullscreen.

A pretty standard setup!
- the exp. converter is a standard 'simple' approach used widely for simple applications
- the LM13700 part is straight from the datasheet - a simple LoPassFilter (I've been trying out for glide)

My question::
The Offset trimmer is, of course, necessary to null out any offset and ensure that there is no DC offset at the output.

BUT!

The DC offset changes with the control current?!
-- when the current is at its maximum (ie Manual control up to V+ -- less than 1mA current) the offset changes -- its only a few millivolts and I've not accurately measured it, but for pitch control it is noticeable as a slight offset in pitch.

If anyone can give an ideas to why I'm experiencing this I'd be very grateful!
Is it, perhaps, that the Exp Converter is too simple?

Big thanks!

gwaidan · Joined: Mar 07, 2009 Posts: 53 Location: Sydney, Australia

In the OTA the difference in apparent input voltage resulting from offsets (as well as from the TL072 in the feedback loop) gets amplified by the control current just like any signal input. If you have it trimmed to zero then it should (ideally) stay zero regardless of control current-you've kept input impedance identical between channels so no problems there...

bugbrand · Posted: Mon Feb 01, 2010 7:48 am Post subject:

Well...

Looking a bit more closely.

I was initially using a TL072 -- this would give an offset range of around 300mV -- that's quite a lot!

The only other dual opamp I've got in SOIC8 at the moment is an LM358 -- changed over to this and the offset range falls to around 100mV. Hmm, maybe getting somewhere..

I think I'd better get some 'special' opamps with especially low offset voltages and try out like this.

Probably partly down to my overly standardized use of mainly TL07x for most applications!

frijitz · Posted: Mon Feb 01, 2010 7:52 am Post subject: Re: Question about LM13700 offset

bugbrand · Posted: Mon Feb 01, 2010 8:00 am Post subject:

Thanks for chiming in Ian - I was hoping for input from you!

Ahhhh-shhhh...

So the trouble is likely from the LM13700 rather than the opamp (TL072)?

That's a bit of a spanner for me then!

Sound · Joined: Jun 06, 2006 Posts: 842 Audio files: 1

Could it be related to the characteristic described in page 4, first graphic, on the LM13700 documentation?
Could it be solved restricting the span from 5 to 200 or 300mA?

Edited again the span according the graphic Very Happy

frijitz · Posted: Mon Feb 01, 2010 8:27 am Post subject:

bugbrand · Posted: Mon Feb 01, 2010 10:43 am Post subject:

Sound · Joined: Jun 06, 2006 Posts: 842 Audio files: 1

Just thinking on this, only an idea without check, is it wild think -if the two OTAs of one LM13700 are matched -and the ratio Iabc/offset is same in both- use one of them -same input Iabc, without input signal- just for cancel the offset of the other, just inverting it?

bugbrand · Posted: Mon Feb 01, 2010 11:39 am Post subject:

Hey, that's a really interesting idea to investigate!

gwaidan · Joined: Mar 07, 2009 Posts: 53 Location: Sydney, Australia

Re the TL072-I've found the LF412 is much much better in this regard altho a bit more expensive-worth a try.