WAIT mod on an issue 3 board

[dinosaur]

Greetings,

Did anyone try the "WAIT mod" by Wilf Rigter (who apparently did it on an issue 1 board) on an issue 3 board ?

If yes, did you get a *stable* system afterwards and if yes to that second question, what PNP transistor did you use ?... I tried with a BC559B (i.e. a lower-noise-but-lower-max-Vce-BC557), and the result was catastrophic (totally unstable system with numerous random crashes). I think this is due to the bad switching qualities of the transistor (storage time >200ns, against 13ns for the ZTX313 it is coupled with). Had it been a NPN transistor, I had used a 2N2369A (same storage, rise and fall times as for the ZTX313), but I didn't find any PNP equivalent in my drawers and none either in my usual components suppliers' catalog: an ideal candidate would have been the 2N5771(20ns storage time), but it's impossible to find, nowadays...

I see that Wilf Rigter used a 2N2907A, but this transistor is just as bad a switching transistor as the BC55x (300ns switch off time, and probably a worst storage time), so I'm wondering it this mod is at all valid, at least on an issue 3 board...

[1024MAK]

On an issue 1 board, I used a BC307 transistor. It works fine

image.jpg

(1.07 MiB) Downloaded 245 times

image.jpg

(831.08 KiB) Downloaded 245 times

image.jpg

(364.77 KiB) Downloaded 245 times

Mark

[1024MAK]

image.jpg (33.16 KiB) Viewed 3566 times

image.jpg

(267.17 KiB) Downloaded 245 times

I have not yet tried on an issue 3 board, but as the circuit is the same, I don't see why it should not work.

And yes, the BC557 (along with the NPN versions) are slow in switching off once fully saturated. I tried a NPN BC547C to drive the CPU clock pin on a recreated Jupiter Ace board. I had to increase the value of the base resistor rather a lot to get it to turn off so that the clock signal would pass "through".

Mark

[dinosaur]

The BC307 datasheet doesn't even quote the Ton/Toff/Tstore figures but its bandwidth is even lower than the BC559's (130MHz against 300MHz)...
But I suppose that if you used a BC307 and not a BC307A/B/C, then its Hfe might be low enough that it doesn't saturate, unlike the BC559B I used (which had a measured Hfe of 300).
Using a general purpose transistor with the lowest Hfe (50 or so) to avoid saturation could be a solution, but I'd rather find a proper switching PNP transistor... I wonder if a Japanese transistor (2S*) could fill the bill...

[RetroTechie]

Perhaps it's possible to use a small "logic" FET for this application? Which would most likely avoid any saturation / switching speed issues. But tbh I have no idea whether that's possible, or if so which type.