K178LK1 soviet logic gate

[sahko123]

I'm repairing an old 1979 nixie tube calculator from the ussr and as part of that I want to RE the schematics to some extent. I have datasheets for all the IC's but im confused about one in particular. The logic configuration says 2AND-2OR/2AND-2OR-NOT but im not exactly sure how its configure because of the diagram below. My assumption atm is that is two 2 input AND gates going into a 2 input OR gate. But the IEC based diagram is kind of confusing because it implies a 4 input and gate followed by a buffer. Can anyone confirm how the inputs are actually configured?

[coppercone2]

I heard someone I know had trouble passing signals through it, the 'papers?' pin kept going high every time. (small carbon contact in the middle of the underside of the package)

[ataradov]

The function of each element is this: y2 = (x1 & x2) | (x3 & x4); y1 = ~y2; EDIT: typo is corrected, thanks dobsonr741.

I think the symbol should have horizontal divider and "&" sign for the second half. But this is the way it is drawn in all the documents.

[dobsonr741]

Best original specs I found: http://electro-tehnyk.narod.ru/docs/Datasheet/K172.pdf

Small typo in ataradov’s formula - should read: y2 = (x1 & x2) | (x3 & x4)

[DBecker]

I heard someone I know had trouble passing signals through it, the 'papers?' pin kept going high every time. (small carbon contact in the middle of the underside of the package)

Is there a bias contact on the package itself?  Is there a standard bias voltage, or does it vary based on long-lost per-batch documentation?

My understanding is that soviet chips were about 9 years behind western ones.  They were a mixture of copied masks, hand retouched and adjusted for the local fabs, and original soviet layouts.  The latter were often basic RTL or TTL logic gates , fabbed as a common layout that was customized in the final metal layer.

RTL and TTL don't typically use a substrate bias voltage.  A bias is typically need with NMOS and CMOS.  Starting in the mid-1970s that bias was generated by an on-chip charge pump, saving a pin or two at the ever-decreasing expense of a little area and tighter process control.

[ataradov]

K178 is a PMOS series, but I don't think it had active biasing.

Bigger devices, like CPUs did have internal bias circuitry and the bias voltage was measurable outside of the device because they were often bonded to a metal heatsink (?), which often did not get covered well by the epoxy molding. And it would also be available on the sides of the package if you remove a bit of plastic.

That bias voltage was a great way to diagnose the devices on complicated boards. All you had to do is power on the board and go over all the ICs measuring the voltage.

[Vovk_Z]

I edited a picture of its logic. As I undestand, it should be as here:

[dobsonr741]

A little backgrounder on and-or-invert: https://en.wikipedia.org/wiki/AND-OR-invert

[Njk]

I edited a picture of its logic. As I undestand, it should be as here:

I think that's the best answer to the original question.
For those interested in the vintage stuff, abundant sovietology information is available at http://www.155la3.ru

For that particular part: http://www.155la3.ru/datafiles/k172_k178_pasp.pdf

Notice, the factory data sheet does not provide the function drawing, only textual description of the logic. So the latter takes precedence in case of contradiction, I presume. As for the drawing, it's typical product of the kibbutz-style society (that's what they're trying to re-build now). Digital electronics was a new technology back then and not every drawing lady in the committee was understanding what's that all about. BTW, Russians are not alone in that regard...