The Sinclair QL Forum

That is absolutely correct.

-80 or -100 components would be ideal. Faster chips are, in fact, more of a problem than slower chips - don't go faster than -70 and -55 would be very problematic. This is because of the noise caused by the fast rise and fall times of the faster components.

Alright, I'll give it a try. I found some -08's on eBay.

Wim

Most interesting! How much would it bump the speed of a trump card by changing the chips? Are any other changes needed?

As soon as I get the faster chips, I'll post a comparison. Hopefully early next week.

Wim

Well, I received the -08 chips today. Eagerly swapped the 150ns versions for the speedy 80ns ones. Plugged the expanderam back into the QL, booted, and ran the ram speed test. It took exactly the same amount of time (22.5 seconds) to finish as before! Bummer. I guess there's something else that's holding it back.

Any thoughts?

Wim

The timing is not set by the DRAM chips, they just limit the response time that will work reliably.

Without looking at the schematic, I believe the circuit you need to adjust is whatever controls the timing of the DTACK signal.

Mark

Don't worry. SuperRAM will show up soonish.

Also, endless supplies of new 40LF220s.

I wrote:The timing is not set by the DRAM chips, they just limit the response time that will work reliably.

Without looking at the schematic, I believe the circuit you need to adjust is whatever controls the timing of the DTACK signal.

I've had a quick look at the schematic and there are at least three, maybe four resistor/capacitor timing circuits that may need changing. Without having a working board and some suitable test gear (4 channel 'scope at least), it's not possible to work out which components need replacing, or what values the new parts should be. Sorry.

Mark

Oh well. Thanks for checking though. I guess we have to wait for some new hardware developments - superram comes to mind.

Wim