Vintage Signetics IC help needed

[valley001]

Hi all,

I have on my bench two HP 3450A DVMs.  I am working on making one functional unit between the two.  The problem I have run into is differing ICs installed in the same locations on these two units (on various boards), at issue currently is the data counter board A10.  The manual available online indicates these should be the HP part number 1820-0094 (A10 IC 4-5, 7, and 10) however the unit I am working to make functional has these vintage Signetics ICs installed at IC 4,5,7 and 10 locations and the two are not compatible (pic of thr Signetics IC attached).  The traces on the board lead to different pins so I know the at the very least the pinout is different.    I would simply use the donor board with the more common HP parts but that board is damaged from prior poor repair attempts. 

There is no reference in the HP manual to this Signetics IC at all.   I have had no luck searching the internet for information.  I wonder if both are the referenced DTL NAND QUAD 2-Inp with differing pinouts?

Before I resort to buying vintage Signetics product catalogs from the late 60s I thought I would see if any of you have info on this IC, hopefully a cross reference could be determined. 

Thanks all for looking.   

[fzabkar]

Have you looked here?

http://www.bitsavers.org/components/signetics/_dataBooks/

[David Hess]

I have several old Signetics databooks but that part number is not enough to go on and the logo is not any Signetics marking that I am familiar with.

[Andy Watson]

Are you sure it's Signetics? I think the logo looks more like General Instruments.

[valley001]

I was not sure at first on the logo but now it seems confirmed based on the post from fzabkar.  The logo from 1968 seems to match.  I glanced through and did not see reference to a 0303, but a lot of part numbers in the 300s. 

[David Hess]

That answers that but the parts in the two available UtilLogic databooks skip 303.

[valley001]

That answers that but the parts in the two available UtilLogic databooks skip 303.

Right, so probably a custom IC for HP?  I did notice in the 3450A manual, page 4-8 figure 4-9,  under the pinout diagram for the 1820-0094 (2 input gates) there is an additional part number 1820-0349.  I thought this may be the HP number for this Signetics IC but the vendor for 1820-0349 is listed as Fairchild. 

[floobydust]

With the HP part number 1820-0094 I got alternates, DTL Quad 2-input NAND gates:
SN15846N  "with 6k pullup {output} resistors"
MC846P
DM946N
SN4501
SC6903PK (OP has 6934? might be the Signetics part number)

edit: maybe a SN7400 (wrong pinout) would work, SN7401 open-collector matches SN15846P pinout.

[valley001]

That's the trouble the 1820-0094 is not compatible with this Signetics IC, so I assume those cross references for the 1820-0094 IC will not work in this case.  I suppose I should now study the pinout on the PCB and see where the differences lie. 

[floobydust]

The other quad NAND gate IC is 1820-0303 which is Signetics/Philips SP680N or SP680A (from IC10 A5) 5V DTL. They all have a wacky pinout 
source: Signetics 1972 Databook http://www.bitsavers.org/components/signetics/_dataBooks/1972_Signetics_Utilogic_2.pdf

[valley001]

Excellent, thank you for this information.  This 1820-0303 may be it.  I notice the vcc and ground are on 1-8 rather than 1820-0094 on 7-14.  Ill probe around tonight and see if this matches the ICs installed in my A10 board. 

[fzabkar]

The SP680A looks like a 7400 lying on its back. Can a 7400 be installed on the other side of the PCB?

[valley001]

The SP680A looks like a 7400 lying on its back. Can a 7400 be installed on the other side of the PCB?

There is some room back there, maybe 1/4 inch or so.  Ill keep this in mind.   Could the pins be bent 180 degrees?  Then it could be installed in the socket. 

[valley001]

Quick shot of the unit working (finally), after calibrating it to my Fluke 332D.  This one showed overload constantly.   Digital section mostly checked out.  Running through the troubleshooting manual led me to replacing the cooler that contains some input fets and the reference.  That did not fix it.  Turned out to be the pulse transformer that relays signals between the digital and analog sections, somehow one of the windings went open at some point.  Luckily I had some good parts from a unit I received a while ago that was damaged in shipment. 

This is by far my favorite piece of test gear. 

[floobydust]

That's great work    it's a complex beast, state of the art in its day. I was surprised to see the 'inguard' metaphor back to 1971.

I had an idea, you could take an SOIC-14 to DIP adapter, solder in a SMT SN7400D, and flip the board to solder in the header pins. This would give you (flipped die) pinout matching SP680A. I don't think the nixie gods would be offended    If you bend DIP pins 180 they just break off.

[magic]

If you bend DIP pins 180 they just break off.

IME it can be done. But only once

I would give it a try on a cheap current production part.

[CaptDon]

I have seen some commercial production equipment where
the board layout guy must have screwed up and indeed I
have seen I.C.'s 'under' the board and the 180 degree pin
bend remedy also!!! This was back in the mid-late 70's when
RTL, DTL and even ECL were common!!
My favorite was the 'memory expansion' by soldering 2102A's
on top of each other and bending out the CS pin and wiring it
separately.