Updated SSI2130 VCO Datasheet available...

[neil.johnson]

The EXPO FREQ pin is a ground-referenced current input, limited to the range 0 to 200 uA. No negative current. As to what that means depends on the voltage-to-current resistors and the current going into the SCALE pin. The example given in the datasheet is for 20uA/octave (49.9k resistors), which is about the maximum you would want to go. Better would be 10uA/octave (100k resistors) and adjust SCALE accordingly.

Neil

Could you please explain what the danger is of a negative current going into pin 7?

I think in eurorack it's very possible to see a -10v input going into the 1v/oct input, so that's -10v over 100k = -100uA

Would that be a problem? Is there any maximum allowable negative current?

thanks!

If it goes negative then the VCO will stop, or run very slowly.

You, as the designer, are responsible for sanitizing (range limiting, scaling, offsetting, etc) the input signals to suit the parts you are using.

Neil

[kybernetik]

If it goes negative then the VCO will stop, or run very slowly.
You, as the designer, are responsible for sanitizing (range limiting, scaling, offsetting, etc) the input signals to suit the parts you are using.

Neil

Ok thanks, yes all inputs are sanitized, I just wanted to make sure there's no chance of with a small negative current

Like what happened with the grounded mix inputs

[neil.johnson]

Well, try to avoid any negative current since to do otherwise would mean you are operating the part outside of it's specified operating parameters.

Neil

[JanneI]

Your DIP adaptor looks very similar to the one from SSI, but with the timing capacitor included:
http://www.soundsemiconductor.com/downloads/dab2130.pdf

Neil

Yes, I actually added the BW COMP 267R&10nF too. And hopefully, the bigger PCB works as intended, so that I don't need to use those adapters at all :-) The mission with this is trying to find out the "most musical implementation" by combining 2-3 of these VCOs, before advancing to the next objective. There's sure plenty of options to choose from. I'll report back!

[neil.johnson]

You might want to try other values of TCAP. 3n9 is ok for audio, you could try a much larger value for LFO duties, or tweak the smaller value for different control range.

As you say, plenty of options!!

Neil

[lickspittle]

As you say, plenty of options!!

Neil

Just for laffs I lashed together a 2130 Maximus* mock-up...

*with all due respect to Thomas Henry

[neil.johnson]

Nice

[burdickjp]

I'd love to see a few of these put together in a way like FSS OSC2 Recombination Engine or
Rossum Electro-Music Trident:
https://www.modulargrid.net/e/future-so ... ion-engine
https://www.modulargrid.net/e/rossum-el ... c-trident-

I've done PCBs from scratch before in KiCAD. I haven't done anything synth related. Where should I look to start?
I'm guessing power supply is something I can lift straight off of other open source modules. Any one in particular?
What else can I pull from open source?

[KSS]

[BugBrand]

Got my 2FW design working yesterday - still a couple of bits to tweak out for sure as I haven't gone deep yet...

- Sine section - following the DC coupled datasheet example, I need to figure how to trim it a bit though want to try to avoid having to add trimmers. I had a little play around the suggested 4k02 & 12k7 resistors but didn't quite catch a solution yet.

- 3-way waveform control (again from datasheet) - nearly there on this exciting part - proto has so far needed a little tweak or offset to go fully to triangle (which leads to some overtones or noticeable glitch in LFO mode) - again, trying to ensure that any design can be done without using trimmers or excessive numbers of different resistor values.

[forestcaver]

trying to ensure that any design can be done without using trimmers or excessive numbers of different resistor values.

I ended up with an excessive number of resistor values!

[BugBrand]

trying to ensure that any design can be done without using trimmers or excessive numbers of different resistor values.

I ended up with an excessive number of resistor values!

Ah! To get a decent sine or for the waveform mixing?
On the sine-shape side - did you try with several builds as I imagine there might be some slight chip-to-chip variance. On old sine-shapers I always implemented shape & offset controls - could do here, of course, but would love a set & forget approach.

The waveform mixing is so close already... (though I scaled it up to 5V rather than the suggested 1v - played about a bit in Multisim & noticed that the control outputs go funny outside the particular input range - ie. for my 5V approach the input needs to be limited 0 to +5V)

[forestcaver]

trying to ensure that any design can be done without using trimmers or excessive numbers of different resistor values.

I ended up with an excessive number of resistor values!

Ah! To get a decent sine or for the waveform mixing?
On the sine-shape side - did you try with several builds as I imagine there might be some slight chip-to-chip variance. On old sine-shapers I always implemented shape & offset controls - could do here, of course, but would love a set & forget approach.

The waveform mixing is so close already... (though I scaled it up to 5V rather than the suggested 1v - played about a bit in Multisim & noticed that the control outputs go funny outside the particular input range - ie. for my 5V approach the input needs to be limited 0 to +5V)

Both! The sine is definitely acceptable for me... it’s not got a perfect fundamental with zero harmonics (but the harmonics are pretty low power and acceptable to me - cant remember the attenuation off the top of my head). I’m happy with waveform mixing - I used +/- 1.2v and scaled with opamps... I was very careful with the input currents....

I’ve built five and carefully tested three with spectrum analysis etc

If you want, I’ll email you my files - I’ll be sticking them on github some time soon anyway...

I’ll be over your way as soon as shops/lockdown eases in a couple of weeks to go to one of the bike shops round the corner from you so could drop in if you like? (Maybe a long extension lead to the garden as well!) I need to try your cider!

[neil.johnson]

One of the nice features of the sine converter in the SSI2130 is that if you overdrive it the result is quite pleasant. A lot of sine converters sound really bad if overdriven so you need to trim them quite carefully.

Neil

[BugBrand]

Cool - cheers - I'm sure that it'll mostly come down to having some clear time to tweak values & investigate more.

Do pop by for a cider! - I'll make sure to save one of the last nice ones from a few years back (Frederick + various other bits) - workshop that is, near Stokes Croft.

[forestcaver]

Cool - cheers - I'm sure that it'll mostly come down to having some clear time to tweak values & investigate more.

Do pop by for a cider! - I'll make sure to save one of the last nice ones from a few years back (Frederick + various other bits) - workshop that is, near Stokes Croft.

Excellent - thanks - just waiting for the Specialized shop to reopen :-)
I’ve sent an email with a load of stuff in it....

[BugBrand]

Hoping to get some tweaks in today...
Here's a quickie of the 3-way waveform morphing (voltage controllable) + you can see the offsided nature of the sine at present..

[neil.johnson]

Wavemorphing looks good
Yeah that sine has a DC offset on it. Can you share your circuit?

Neil

[BugBrand]

Cheers!
Tweaked out the minor issue around that which was leaving a little smidge of ramp when it should have been tri only (which was overly noticeable in LFO mode)

The sine is DC coupled exactly as in the datasheet suggestion (4k02 input resistor from tri & 12k7 offset from vref) - will be tweaking these shortly even though it is tight in there on my board!

[BugBrand]

Increasingly thinking there's something screwy on this first build around that area...I tried adjusting values to no avail & just verified DC coupled sine on my older proto board so will proceed to build a 2nd of the new version to cross-compare.

[neil.johnson]

Worth checking the components are the correct ones? The required value of bias resistor according to the math is 12k79, but that's a non-standard value and 12k7 is close enough.

Neil

[BugBrand]

Yes, I did check I hadn't done a booboo on values or such but.. seems my issue must have been something on soldering the 2130 - part-way through building another & have just confirmed good sine with the suggested DC-coupled values.

[tIB]

Hoping to get some tweaks in today...
Here's a quickie of the 3-way waveform morphing (voltage controllable) + you can see the offsided nature of the sine at present..

Like the look of that three way morph!

[J3RK]

trying to ensure that any design can be done without using trimmers or excessive numbers of different resistor values.

I ended up with an excessive number of resistor values!

That’s why resistors come in 1% I think ALL of my circuits have way too many values. I just use what makes things look and sound good. Hardly scientific.

[Tonescape]

Hoping to get some tweaks in today...
Here's a quickie of the 3-way waveform morphing (voltage controllable) + you can see the offsided nature of the sine at present..

Oh yeah, looks great! Is this done with the example from the data sheet? Thanks