MUFF WIGGLER Forum Index
 FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 
WIGGLING 'LITE' IN GUEST MODE

Simulating LM13700 designs.
MUFF WIGGLER Forum Index -> Music Tech DIY  
Author Simulating LM13700 designs.
dogfaceaudio
I've been experimenting with various LM13700 OTA based circuits and was wondering if anyone else has had success in doing circuit simulations using the LM13700. I grabbed MultiSIM Blue from Mouser, plugged the basic triangle/square VCO from the datasheet into the schematic capture and wired up the virtual oscilloscope, but I'm not getting anything on the output. I also tried LTspice, but the interface is pretty awful so I didn't get too far. What are people using for circuit simulation? I'd prefer something free or open source.
Dr. Sketch-n-Etch
I use Multisim and have never had any problem simulating 13700 circuits. Of course, I haven't built a 13700 circuit for many years, since I only use 2164 for all voltage control applications now.
dogfaceaudio
Do you by any chance have an LM13700 project file I could download and look at? Maybe I'm just hooking things up incorrectly.
windspirit
I was trying to simulate an lm13700 oscillator using cadence pspice and had no luck, using the model from texas instruments and setting up the circuit as I had in the datasheet, bias voltages and currents didn't converge. Let me know if you have any success. The Pspice interface is god-awful (like my mouse stutters every time I move it awful) so I might have a look at multisim.
slow_riot
I use the NatSemi model which is probably the same thing everyone else is using. It works OK for capturing high level effects.

Weirdly I tried to build one in SPICE with discrete transistors to get low level effects but it wasn't working at all, although for true high accuracy one would need to use models for embedded BJTs which have some nuances in behaviour that are different to discrete.

Out of the freely available sims LTSPICE is the one I would stick with.
L-1
I simulated LM13700 VCA using a model from oficial site.
Model not shows an offset which real chip creates.
Also i used fake LM13700 from ebay at first, it made about twice more gain. But I solved this when bought a chip from normal shop.
neil.johnson
Don Sauer's site has lots of info on simulating the LM13700:
www.idea2ic.com

Neil
dogfaceaudio
I managed to get the single amplifier VCO from the LM13700 datasheet working as a simulation under MultiSIM Blue. Here's the schematic I input if anyone else is curious to try it out.

dogfaceaudio
I'm still unable to get the triangle/square VCO circuit from the LM13700 datasheet simulating correctly under MultiSIM. Here's the schematic I'm inputting into MultiSIM, I've checked it a pile of times but maybe there's something I'm missing:



But when I run the simulation the only output I get on the scope is this:

emmaker
I used LT Spice and it worked for me with the LM13700 model I used. The frequency is pretty high so you want to change the scope sweep to 5, 2 or 1 us per div.


    .SUBCKT LM13700b IAB LIN INP INN OUT VEE BIN BUF VCC
    QN1 IAB VN2B VEE npnv 3
    QN2 VN2B VN2B VEE npnv 3
    QN3 VN3C IAB VN2B npnv 3
    QN4 VP3B INN VN3C npnv 3
    QN5 VP6B INP VN3C npnv 3
    QN6 LIN LIN INN npnv 3
    QN7 LIN LIN INP npnv 3
    QN8 VN10B VN9B VEE npnv 3
    QN9 VN9B VN9B VEE npnv 3
    QN10 OUT VN10B VN9B npnv 3
    QN11 VCC BIN VN12B npnv 15
    QN12 VN12B VN12B BUF npnv 3
    QN13 VCC VN12B BUF npnv 150
    QP1 VP3B VP2B VCC pnpl 3
    QP2 VP2B VP2B VCC pnpl 3
    QP3 VN10B VP3B VP2B pnpl 3
    QP4 VP6B VP5B VCC pnpl 3
    QP5 VP5B VP5B VCC pnpl 3
    QP6 OUT VP6B VP5B pnpl 3

    .MODEL npnv NPN(
    *==========================================================
    +IS=1.1E-18 NF=1.005 BF=220 VAF=130 IKF=5e-03
    +ISE=9.15E-15 NE=2
    *==========================================================
    +CJE=2E-12 CJC=2E-13 CJS=3E-12 TF=.6E-9 )

    .MODEL pnpl PNP(
    *==========================================================
    +IS=1.1E-18 NF=1.8 BF=5 VAF=170 IKF=6e-03
    +ISE=3E-15 NE=3
    *==========================================================
    +CJE=4E-12 CJC=6E-13 CJS=5E-12 TF=70E-9 )

    .ends


emmaker
Looking at the MultiSim schematic again saw the transistor buffers aren't tied to V+. Could that be it?

Also don't know how MultiSim works but you might also have to hook the power supply up to the second OTA used as a comparator.
Dr. Sketch-n-Etch
emmaker wrote:
Looking at the MultiSim schematic again saw the transistor buffers aren't tied to V+. Could that be it?

Also don't know how MultiSim works but you might also have to hook the power supply up to the second OTA used as a comparator.


Yes, you are not actually powering the 13700s! You have to connect the transistors to +V.
dogfaceaudio
I just modified the schematic symbol for the LM13700 to match the datasheet symbol, the V+ for the transistors is actually tied to V+ for the IC. U1A has the +/- 15V supplies which is why U1B has the red X's on pins 6 and 11. I've attached the MultiSIM project file, maybe someone can take a look and see if I've neglected to do something obvious.
Paul Perry
L-1 wrote:
I simulated LM13700 VCA using a model from oficial site.
Model not shows an offset which real chip creates.
Also i used fake LM13700 from ebay at first, it made about twice more gain. But I solved this when bought a chip from normal shop.


I doubt it would be worth faking a LM13700.
I seem to remember from the data sheet that the gain can vary by 2 to 1 anyway.. which might be the reason why apparently identical filters constructed from LM13700s can sound a bit "different".

I'm not a fan of sockets normally, but with the LM13700, it might make sense.
Andreas
Paul Perry wrote:

I doubt it would be worth faking a LM13700.


Just wanted to chime in, although the thread is about four years old. I got some fake LM13700M from China. Legs have absolutely no function at all. Buffer-Transistors don't have any voltage drop. 6 and 11 do not conduct at all, which the original ones do at 6MOhms.

So i think i got some nice plastic packages with legs.

Markings are engraved/lasered: 41AHY1UG3 (G may be C) LM13700M
Two circular markings on the Bottom with engraved: C D / 70
MUFF WIGGLER Forum Index -> Music Tech DIY  
Page 1 of 1
Powered by phpBB © phpBB Group