13600 vs 13700 once again

[bioroids]

Hi!

So, datasheets and posts point to a better performance using the 13700, regarding control current bleedthrough.

What I wonder is how big this difference may be. I'm working on some phaser stuff, and still have some clicks, wich go away with apropiate filtering.  I'm using JRC13600. But this is gonna be built using LM13700 (wich are already ordered, but hadn't got here yet).

The question is, should I expect noticeable less noise when switching the ics (wich will allow me to get away with less filtering)? Or it will perform practically the same. I'm using the internal buffers, where the difference resides...

Thanks in advance!

Miguel

[toneman]

My bet is U won't notice any audible difference.
Data sheet says 13700 is "slightly better for audio" due to
differences in buffer gain control.  With a phaser, it will  B
very hard to tell any difference.
just my .02
tone

[gez]

What I wonder is how big this difference may be. I'm working on some phaser stuff, and still have some clicks, wich go away with apropiate filtering.

Is the clicking due to current spikes on the rails from a LFO, or are you modulating the OTA, via the Iabc pin, with a square wave?

[bioroids]

I'm modulating the OTA via the Iabc pin... is kinda squarish waves, like from a random or sequenced LFO. It behaves pretty good, with the adecuate filtering, but I want to use the less filtering posible, because I dont want to loose the egde of the modulation.

Still I have a few weeks for the 13700 to get here! I'm getting the "waiting for the order" anxiety

Right now I'm drawing the schematic to post as soon as I get the whole thing finished.

Luck!

Miguel

[gez]

I'm modulating the OTA via the Iabc pin... is kinda squarish waves, like from a random or sequenced LFO. It behaves pretty good, with the adecuate filtering, but I want to use the less filtering posible, because I dont want to loose the egde of the modulation.

I don't think the LM13700 will be any better.  It's difficult getting tick-free operation when modulating an OTA with a straight (as in vertical) edge.  You can use an active integrator designed to clip so that the edges are slightly sloped (isosceles trapezoid).  It's possible to get very steep sides so that the thing has all the feel of a square wave, but minus tick.

[bioroids]

Ok, that is what I was afraid of.  

Thanks for the integrator tip. This thing is getting a little huge for my standard board size anyway, I'm not sure if I could squeeze more components in it!

Is your web page up? I just tried to access but I get nothing...

Regards and thanks

Miguel

[gez]

Is your web page up? I just tried to access but I get nothing...

It should be up but it seems that the server is down. They're usually pretty good and this rarely occurs, but it happens now and again. :roll:

There's no DIY stuff there (just spam) so you're not missing anything!  

[gez]

Thanks for the integrator tip. This thing is getting a little huge for my standard board size anyway, I'm not sure if I could squeeze more components in it!

You can do it quite easily with half a LM13700 a few resistors and a cap. The other half can even be used for audio if you're careful with layout.

[bioroids]

Or maybe I could use the other half for an exponential converter and get an hypertriangular wave... this goes on top of 3 13700 for a six stage phaser, plus the 4013 and the 4029 for the waveform generator... maybe I should use a double sided board for this one!

In fact the generated waveform has a good feel, I dont think it need more hypertriangularity (!)

Anyway you've been very helpfull Gez, maybe I should name this Gez's Phaser :oops:

Luck!

Miguel

[Paul Perry (Frostwave)]

There are three places the click could be from:
1. coupling via the power supply.. decouple the LFO, or whatever the clock is.
2. pulse from clock/LFO going along ground wire & beign coupled into rest of circuit.. use star grounding.
3. edge of LFO is too square, which means, if there is any DC offset going into the LM13600, then this is getting chopped & giving a, well, 'click'.
Adjusting input offset of the LM13600 might help, making the LFO more rounded might help.

[gez]

Adjusting input offset of the LM13600 might help, making the LFO more rounded might help.

Yeah, the trimpot thing on the inputs does help minimise it (see OTA article at Geofex), but sloping the sides of a square (if you can get away with it - large frequency range can make it difficult) can eliminate the problem totally.

[Paul Perry (Frostwave)]

One way to slope stuff would be by a slew rate limiter. I havn't tried it, but perhaps a 4 diode bridge steering current thru a transistor wired as a constant current diode could do it.
A 'real' slew rate limiter has too many parts, I think.

[gez]

One way to slope stuff would be by a slew rate limiter. I havn't tried it, but perhaps a 4 diode bridge steering current thru a transistor wired as a constant current diode could do it.
A 'real' slew rate limiter has too many parts, I think.

Well, the circuit I came up with could be regarded as a slew limiter (sort of   ).  If you connect the darlington buffer to the OTA's output then wire the whole thing up as though it were an op-amp follower (connect emitter to -ve input) but with a cap at the OTA's output, you get sloping.  Bias the +ve input with a divider and feed it with your square wave via a resistor.

The square is divided down and presented to the +ve input.  With positive swings the output ramps up til the -ve input is at the same voltage, then the output levels out.  When the input goes negative the output ramps down until both inputs are at the same voltage.  By varying the resistance from the LFO to the +ve input you can control depth.  By varying Iabc current you can control slope.  Output waveform is a trapezoid (gives tick-free operation, and with steep sides has all the characteristic feel of a square wave).

[bioroids]

There are three places the click could be from:
1. coupling via the power supply.. decouple the LFO, or whatever the clock is.
2. pulse from clock/LFO going along ground wire & beign coupled into rest of circuit.. use star grounding.
3. edge of LFO is too square, which means, if there is any DC offset going into the LM13600, then this is getting chopped & giving a, well, 'click'.
Adjusting input offset of the LM13600 might help, making the LFO more rounded might help.

I think I can discard options 1 and 2. I have the LFO generator on a separate breadboard with a good decoupling.

I'm not using the input trimpots, I was trying to get away without them (because good quality ones are not so cheap around here, and bigger than a dual opamp!). I will install them now.

Gez, your idea seems very interesting for a tremolo. Should use a separate OTA for the audio part or you think it can handle it?

Thanks again for your suggestions

Miguel

[gez]

Gez, your idea seems very interesting for a tremolo. Should use a separate OTA for the audio part or you think it can handle it?

No need for separate OTAs, use one half of a LM13700 for the trapezoid shaping and the other half for audio (modulate this half from the output of the other).  You need to be a little careful with layout, but it's not much of a problem.  In a trem it does everything from choppy to full-on pseudo-echo/stutter.

[bioroids]

That sounds like what I'm looking for the trem! This OTA thinguies are pretty cool. I'm glad I finally got into trying them.

I just got to access your page. Are you planning to add more effects?
I like the all metal look, specially the knobs!

Nice sounds by the way, you should add more samples (if bandwidth allows it of course).

Luck

Miguel

[gez]

That sounds like what I'm looking for the trem! This OTA thinguies are pretty cool. I'm glad I finally got into trying them.

The trapezoid thing can be used for all sorts of effects.  You can use it to shape a triggered pulse which in turn sweeps a filter.  With a bit of thought you can control attack hold decay (or whatever the terms are, it's been a while   ), but it does work very well in a trem (tick-free stutter).

I just got to access your page. Are you planning to add more effects?  I like the all metal look, specially the knobs!

Nice sounds by the way, you should add more samples (if bandwidth allows it of course).

Thank you.  I have a few effects I'd like to add eventually, just need to iron out a few kinks to get them to my liking.  Other stuff I've designed is too run of the mill to bother with really.