TF330 Bus Switches SN74CB3T3245PW

[terriblefire]

I have removed the NMOS discussion. It got too far off topic. Please start another topic in hardware if you wish to debate NMOS vs CMOS etc.

[exxos]

supaduper kindly sent me some of these...

https://www.digikey.co.uk/product-detai ... -ND/770579

Though I still have issues

Its acting like ( I think ) the other texas part I tried out, like the chips not even plugged into the board.

I have VCC to 3.3V. GND to 0V, OE to 0V. Assuming bus pulls up on 5V CPU side, I placed in 10K to 5V on the left side of the IC.. On the right side, I put 10K to 3.3V..

I get 5V and 3.3V as expected each side of the IC. If either side goes to 0V, nothing happens the other side.

If I remove the 10K pull up on the 3.3V side, the voltage drops to 1V.

[ijor]

Though I still have issues
Its acting like ( I think ) the other texas part I tried out, like the chips not even plugged into the board.

It seems to me that you are probably making some obvious mistake and for some reason you don't see it. It happened to me more than once, it happens to all of us, I guess. May be you are sure you are connecting OE to GND, but you are not. I'm not saying it is precisely that, just an example. Sometimes the mind plays those kind of tricks with us.

If you want to find out what's going on, IMHO you perhaps need to share with us much more details, such as pins numbers, pictures, etc. Then may be, we might be able to spot the problem. Of course that not because we are smarter or know more. Just because we see this from the outside.

But I thought you tried the NXP switches and they work ok?

If you want to laugh, I can tell you something that happened to me years ago. Struggled days for an issue when the solution was obvious and was clear in front of my face all the time ...

[exxos]

The NXP ones work fine... They pretty much wire up the same..

I was just trying the Texas buffers as TF uses them and seem fine. So someone sent the same ones for me to try and I can't get those working either.

I may well be doing something dumb. But I don't see what. 3 solutions work just fine and are repeatable. 3 solutions do not work and are repeatable.

For things like OE, I always swap between 5v and 0v for the hell of it anyway. Even measure the voltages on the ic pins directly to rule out bad connections.

They do act like OE isn't working, but defiantly wired up and tested with 0v, 5v, 3v...

The only thing I haven't done much testing with is I've not connected unused gates to anything.. But I don't think that would matter for DC testing I am doing.

Its possible the chips are sensitive to heat, though I've done one in my own, with my gas iron and normal soldering iron. So seems unlikely.

Possible my little psu board is blowing them up for some reason, but its only really a 5v and 3.3v regulator on there.. Nothing fancy.

Maybe the chips don't like the flyleads I have for some reason and its killing the chips...

But while half my tests fail with Texas and idt parts, if they are failing then they are failing way to easily.

One of the old timers I used to work with always used to say the hardest faults to find are the self inflicted ones. Though my addition to that is there is always more than 1 fault which makes tracing them near impossible.

Even so, I've wired them as per data sheets. Plus tried everything I can think of to try and work it out.

I'm just going with TF theory that in have the touch of death... And NXP parts are indesctuctable..

[exxos]

BTW - I postaged images and findings in various places in my blitter thread already..

for example..
https://www.exxosforum.co.uk/forum/viewt ... 110#p15986

So may best best to discuss there as its not really on topic with this thread. Even so, whatever the problem, I will likely have to go with the NXP chips as they work for me, others don't.