paralleling up a cmos chip soldering one chip on top of another

[dazza000]

HI

can I
parallel up a cmos cpip's input , output and suppliy pins - soldering one chip on top of another - or connecting them together on a pcb in order to gain more current output?

thanks
david

 

[KlausST]

Hi,

generally i think this is not a good idea but for sure it depends on the circumstances.

Maybe with digital ICs... but usually there are logic families with higher output capabilities.
And the question is: is the device the best choice if you need to parallel some of them? Is´nt there a better chooice for a single IC?

Klaus

 

[Audioguru]

I DOOD it about 44 years ago.
There was a new office computer that had a problem. It used DTL logic and I fixed the problem by adding another IC on top of an existing IC to reverse a paper-handling motor when a paper jam occurred.
It did not parallel the outputs, instead it changed the logic.
It worked so well that I was awarded an electric shaver.

You can probably parallel logic outputs since they will all switch high or low together at the same time. But you cannot parallel slowly moving analog outputs unless each one has a series resistor to a common output.

 

[dazza000]

HI

can I
parallel up a cmos cpip's input , output and suppliy pins - soldering one chip on top of another - or connecting them together on a pcb in order to gain more current output?

thanks
david

Actually it is a 4013 flip flop. I'm just using one half and I have tied all the inputs to the other flipflop to vcc. ( I think that is the right way to treat unused inputs.)

Any way I think I'll try it and see what happens.
thanks for your repliess guys.

over and out

 

[ard]

I DOOD it about 44 years ago.
There was a new office computer that had a problem. It used DTL logic and I fixed the problem by adding another IC on top of an existing IC to reverse a paper-handling motor when a paper jam occurred.
It did not parallel the outputs, instead it changed the logic.

Me too! Back in the Spectrum / Commodore days I could identify the individual DRAM chips gone bad by merely piggybacking a good chip on each of the 8 in the bank. Worked a treat.

 

[betwixt]

Back in the earliest days of PCs, IBM uses a similar trick for a different purpose. When large dynamic RAMs were scarce and expensive, they piggy-backed two DIL ICs of half the needed capacity each (I think they were 2 x 128K to make 256K) but one had it's lower half disabled and the other had it's upper half disabled. I guess they were using rejected 256K devices, selected for ones with low and ones with high addressing errors. In those days the manufacturing yield of DRAMs was very low so making use of 'half select' devices seemed a good idea. They were shipped from the factory already soldered on on top of the other. The ones they used were made by MOSTEK but something similar looking is here: https://eda360insider.wordpress.com...d-ics-six-decades-of-3d-electronic-packaging/

I still have some ceramic 28-pin DIL processors from Signetics with built in sockets for a piggy back chip to be plugged into them!

Not quite the same trick as boosting output current but interesting all the same.

Brian.

 

[Audioguru]

You do not need to activate the unused half of the 4013. Simply connect to ground the Set, Reset, Clock and D input.