another CMOS logic switcher question

[celebutante]

Some of you may remember a post I put up re building a voltage patching device for my modular using CMOS logic. Got kinda of sidetracked, so I'm fooling with it again. The idea has evolved into something like the old Sequential model 700 programmer, minus the digital memory (instead having just four presets selected with a switch).

Mainly it will use 4052 chips to switch between CV's preset with trimmer pots. But I want to use it do some audio routing as well (so I can switch the saw and sqr osc waves on/off).

Since I managed to blow up a 4052 yesterday trying to run audio through it (I guess they're strictly intended for DC voltages... cue laugh track from those more experienced than I), I'm pretty sure the correct way is to use a transistor, right? Or should I be using an op-amp? How might I implement that?

thanks....

mitchell

[celebutante]

I just found this:

http://www.uni-bonn.de/~uzs159/

then go to "misc/signal switch"

But it's a bit confusing, because it appears that there's a a resistor/transistor combo hooked to the 4053's address lines; can't see why you'd do that. I guess the A, B, C of the 4053 isn't meant to represent the address lines, but instead the signal outputs from the the three lines above (and the switch shown in the three identical diagrams is just a manual two pos input selector switch?)... anyone?

mitchell

[blue hell]

The transistors are just inverters on the address lines, probably for matching it to an active low control circuit ... but the switches above just represent the the ones in the IC, not manual ones anyway. No need to get confused

[blue hell]

[celebutante]

I'm basically using the Digisound 80 switcher circuit, except that I'm using a 4052 instead of a 4051 (4051 is a 1>8 mux, 4052 is a dual 1>4 mux). But my voltages are all the same as this.

I attempted to route a triangle wave from one my .com oscillators, which output 10v peak to peak, but I ran it through a passive attenuator. It sounded really weird and distorted, then when I played with the level on the attentuator, it just died.

Vcc is +15v, so I don't think that's a problem. But wouldn't an oscillator wave be an AC signal fluctuating between +5v and -5v (unattenuated)? i.e. below ground causing CMOS to fry? This my theory... if it's correct, how do I can a route a bipolar waveform through? (I think I'm missing something!)

mitchell

---

Picture 1.png
Description:
Filesize:	33.26 KB
Viewed:	4182 Time(s)

---

[bugbrand]

Heya,

You can run 4000 series chips off +/- 7.5v safely - you'll see this in some Buchla designs (eg. LoPassGate, I think -- yup, check Mark Verbos' site: http://www.simple-answer.com/DIY.html)

Also, may be worth checking a CGS design, the Sequential Switch - I ain't totally figured it, but he's usually got some great ideas:::
http://cgs.synth.net/modules/cgs28_seq_switch.html

[fonik]

or even from +8V to -8V, that's what doepfer does on his switch.
according datasheet it is rated for operation from 0.5V-18V...

here is a link to my redrawn schematic:
http://modular.fonik.de/pdf/seq_switch.pdf

[richardc64]

[celebutante]

thanks again all... I'm sure I can it going with all this info! I'll keep ya posted... right now I'm building A couple of Rene Schmitz 555-based envelopes which will be part of it (since I can't voltage control my .com envelopes, I realized I should just incorporate them, like the Sequential 700 programmer).

Mitchell

[françois]

Hello,

Just a few remarks. If you read carefully the datasheet for the 4051/4052/4053 switches, you'll notice that they have three supply voltages : VDD, VSS and VEE.

VDD and VEE are the analog signal rails, and VDD-VEE can be as high as 20V (but 15V or maybe 18V is safer).

VDD and VSS are the logic control rails and can also be as high as VDD-VSS=20V. This allows switching analog signals such as ±5V pk-pk under control of 0/+5V logic. Or ±5V logic as I like it (using a LFO as a clock for example).

These switches will happily accept AC inputs, within a limited bandwidth of course (they are CMOS, not known for its very high speed). At audio rates this is not a problem. But never ever have the analog or logic signal exceed the rail supply voltages ! Unless you mean to build an electronic fireworks...

-- françois

[richardc64]