btb DISK II pcb rev1 success!

This is great!

Thanks for the feedback.  I might build one of these just for the fun of it.

What souce code did you use for the EPROMs?

Thanks.

PSA: 2N4355 is obsolete, and it's direct replacement, PN4355 and similar MPS4355 are also obsolete.

The closest substitute I could find from a reputable source is the MPS751.  https://www.digikey.ca/en/products/detail/onsemi/MPS751/1049444

In post #1, 'Aurbo' wrote:

" Happy to report that my "btb DISK II" card is alive and working. "" Q2-Q3  are 2n4355 transistors "

On post #4, 'baldrick' wrote:

"PSA: 2N4355 is obsolete, and it's direct replacement, PN4355 and similar MPS4355 are also obsolete."

Uncle Bernie comments:

Beware of the timing issue with the 2716/2732 in UB3 position, for the "Woz Machine". I wrote about this issue seen in these btb cards here on Applefritter and showed the timing calculations. It won't work unless the EPROM is much faster than its typical spec of 350...450ns. But if you pick a 2732 which was made in the 1990s or later (i.e. ST did that) it is very likely to be much faster than its datasheet says, and then it may be fast enough to make this card work. In case of doubt, look through the window and compare the die size to the early 2732 made in 1978 ... over the decades they did several mask shrinks and consequently these parts got faster and faster. For typical 2716 there is not much hope ... I just measured a few early 2716 specimen made by Intel (c) 1977 for my Apple-1 color graphics card and these "450ns max" parts have actual access times of around 300ns ... 350ns if their capacitive and resistive loads are reduced to one LSTTL load, far less than the huge loads allowed in the datasheet. Which were required for the typical applications of these EPROMs back in the 1970s - 2 kByte is not exactly much memory, even back then, and so they made large PCBs with lots of these EPROMs, all outputs tied together, and this is why the specified load capacitance had to be that large.

As for the transistors, the schematic on github calls for Q2,Q3 = MPSU51 which is overkill. Even the 2N4355 is overkill in this circuit. You could use the 2N4354 available from Anchor Electronics in Santa Clara. The only difference between 2N4355 and 2N4354 is that the latter has less guaranteed current gain (beta) at 100mA and it's just a 2nd grade selection of the 2N4355, but this does not matter for this application, just look at those base resistor values ... enough base current in any case.

From Anchor, according to their Y2025 price list, you get 4 of the 2N4354 for $1. No need to pay $1 each for the MPS751 from Digikey.

Oh, and IMHO you could use almost any "switch grade" or "general purpose" silicon based PNP as long as it can sustain the 100mA...150mA drawn by the circuit. If you use a CMOS EPROM, i.e. TMS27C32, then the current drain will be a quarter of the NMOS part, almost any silicon PNP in your lab drawers can handle that, and you can get speed grades down to 150ns and even 100ns which will help to make this card work.

- Uncle Bernie

(just in case you wonder, I am in receipt of two BTB cards, a DISK II and a SmartDisk II, which don't work, so their builder got disappointed, so you were lucky that yours did work --- I'm not a fan of BTB's circuits but his layouts look very nice).