Anyone know how to use a Phillips part PSSI2021SAY 1

That's rather clever!

Pulling 3 and 2 low at the same time, yes you will get a negative AND function there. And, yes. Current is set the way you think. There are two diodes, but one is to compensate for the B-E drop in the transistor and the remaining 0.7 V is for the current source function. The internal resistor (from pin 5 to transistor emitter) is around 14 kohms (.7V/50uA). Choose a 69 ohms external resistor for around 10 mA in the LED.

>> The LED should be placed in the output, where it belongs. The device is not guaranteed to behave like a constant current device at the power supply connection.

>> An LED current greater than 10 mA may pose a problem to the signal trying to sink that current. You may need to buffer the signal with a high current driver, in which case, you have the opportunity to put the logic function elsewhere

>> TTL outputs are usually not guaranteed to be higher than about 2.7 to 3.0V, so you might not actually shut down the current source completely. In fact, given the scenario of the next item, it will be impossible to shut off the current source using TTL, since the outputs will clamp to about 5.7V

>> The internal base resistance of the device is 35 K, ( which means that if you need 20 mA and assume a VERY generous beta of 100, you'll need 200 uA of base current, which means you'll need a minimum of 7V across the base resistor. This means that the power supply needs to be a minimum of 8.5 V, which exceeds TTL maximum ratings.

TTFN

I wrote the above without looking at the transistor model. The spec beta is only 120, which is the PEAK value, so 100 is EXTREMELY generous.

TTFN

IR,

Sometimes one has to "think outside the box". There is nothing said about TTL or supply voltage levels in the original question. If the application already involved gates or micros, I am sure that this "trick NAND" would never had surfaced.

The typical supply voltage for this part is 12 V DC and it can take 75 V maximum. So it is obviously not a 5 V part. I think that huntleyb's solution will work very well. He might have to do some tweaking. But it will work.

I'm merely pointing out that IF he is using standard TTL, there are additional considerations to be taken.

That's part of the engineering process, whether you're in or out of the box.


TTFN

Thanks for the input guys.
I will now know when both of my 15V relays are on at the same time. When I get the samples in, and built this circuit I will let you know how it turned out.

huntleyb