Bad Part from TI and Avnet: 74LS368AN

In rec.games.video.arcade.collecting Robert Bullock wrote: : SN74LS368AN. Date code 43AD7HM : : In a Pac board at 8B, gives Matrix-like green lines and weird static on : screen. Putting the old part in made it go away. New socket. Put these in : loosely, pushed them in tight, etc. nothing worked to make it go away. Tin : double wipe sockets. Game play is unaffected. : : Old part worked fine under every circumstance. : : Ordered from Avnet Mar 10, quantity 11. I tried 5, all the same more or : less, amount of static varied. : : Shame on you TI. :) : : If you would like one, send me a self addressed stamped envelope. :) : : Don't ya just love bad parts right out of the box new?

They're probably not bad parts, just not fast enough to be used in the clock oscillator circuit. Was the original an LS? Did you pull the TI datasheet and see if that chip is spec'd at 18MHz?

The original was also an LS. The PCB component placement guide had an LS368 shown. Both were made by TI. I doubt TI lowered their spec. Feel free to do the math on the switching characteristics though. This shouldn't be that cosmic. I would suspect a batch of parts slightly out of spec perhaps. They worked, just not well.

The part might not be all that happy with the +5 being too low. The newer part might have less tolerance for voltage fluctuations.

-- Matt M and M Electronics

of

Speculations. The voltage was at 4.98 or so, and I still don't see where one part from one manufacturer that was old as the board would work where new ones would not. frankly, don't care.

If the chip couldn't take 4.98 and work with it, it's bad.

In rec.games.video.arcade.collecting Robert Bullock wrote: : Speculations. The voltage was at 4.98 or so, and I still don't see where one : part from one manufacturer that was old as the board would work where new : ones would not. frankly, don't care.

The simple fact is that that circuit doesn't drive the 74ls368 with valid TTL input levels. The resistors across the gates in the 368 result in the inputs being stuck at invalid TTL input levels, so the output levels aren't guaranteed. If the crystal/cap don't provide enough excitation to get enough feedback to start the circuit oscillating, it's the circuit's fault, not the chip's fault.

: If the chip couldn't take 4.98 and work with it, it's bad.

Most of my pacs work down to 4.3 (and many below).

I still don't see how the old chip would work where none of the new ones would. And the response about 4.98 was to previous speculation. I've seen them work at 4.3ish too.

I'll try to repost when my new batch of 368's come in.

Maybe I have met this one before, and maybe not. If you can put it in a socket, instead of directly on the board, and find that it works, I have met this one before. If not, no.

Jim

Robert Bullock writes:

: SN74LS368AN. Date code 43AD7HM

: In a Pac board at 8B, gives Matrix-like green lines and weird static on : screen. Putting the old part in made it go away. New socket. Put these in : loosely, pushed them in tight, etc. nothing worked to make it go away. Tin : double wipe sockets. Game play is unaffected.

: Old part worked fine under every circumstance.

: Ordered from Avnet Mar 10, quantity 11. I tried 5, all the same more or : less, amount of static varied.

: Shame on you TI. :)

: If you would like one, send me a self addressed stamped envelope. :)

: Don't ya just love bad parts right out of the box new?

: -- : "Quality is the enemy of production." : Homepage -