All RAM chips destroyed??

[Pr0f]

I had a thought on the 74LS257 passing the test and not working in the QL - I know that the Service Manual did qualify certain chips to a particular manufacturer - presumably down to timing tolerances...

I wonder if that's why your chip may have been suspect?

[Ruptor]

Why not use 74HC257 that is faster and just better all round? I haven't looked at the specs in 30 years but from memory the outputs are LSTTL compatible and the inputs are better level definition giving cleaner signals. No negatives making them a no-brainer.

[Technical-Chap]

Yes, I agree, it was possible down to timing, 74HCT seem to be the best as they have the right threshold voltages for the 1's and 0's. But 74HC work in most cases.

[Pr0f]

sometimes faster can be a problem - if someone has relied on a more latent signal that a particular manufacturer has, then using a chip with faster timings won't work. It's not really good design practice to do that anyway - should always design for worse case - but that wasn't always Sinclair's style

[Pr0f]

It was the 74LS257 (and the 74LS245) in the service manual that had a chip source / manufacture note:

IC19,20 74LS257 As NOT Nat. Semi
IC21 74LS245 Issue 5 NOT Nat. Semi

So not National Semincondutor version...

[Technical-Chap]

So looking back the two chips were TI and the replacement was Hitachi. But it does look as if I ringed the wrong chip in my first picture

[tofro]

Not sure if it matters, but NatSemi is part of TI today. (Would that mean the Service Manual warning extends to TI now?

Tobias

[Pr0f]

Not sure if it matters, but NatSemi is part of TI today. (Would that mean the Service Manual warning extends to TI now?

Tobias

It would be interesting to know the reason for that parituclar requirement, as there are so many more manufacturers of logic and also now new logic families, like HC, HCT etc - what are the exact timing requirements and current drive requirements for these devices, given their job is driving the RAM chips from the CPU for the most part (I said it that way because the 245 also acts as bridge to the Zx8302 for issue 5)

[Ruptor]

Hi Tobias

Not sure if it matters, but NatSemi is part of TI today. (Would that mean the Service Manual warning extends to TI now?

It is more like TI eliminated a competitor and scrapped Nat semi production of TTL since TI was one of the first making them and have enough plants so they have done us a favour.

[Ruptor]

Is it right that the 68008 address lines can only source 400uA because if so they should be buffered before going on to other things. An LSTTL input can take 100uA and I don't know what the two ROMs take then the lines leave the board via ROM and expansion connectors that is asking for trouble since there is no control over any external loads.