Purpose of D1 in PAL-1

[Magnus Olsson]

I'm studying the schematic of PAL-1 and I can't figure out what's the purpose of the reverse-biased diode D1 in the RS-232 output circuit. It connects an open-collector output of one of the 74LS38 buffers to ground, so it will only conduct if that output goes below -0.7 V, but how could it do that? 

That output is pulled up to VCC by R33 and can pull down to logical zero (which is > 0 V) when the output becomes active, and it's connected to the inputs of a NAND gate which can't go negative either, so why is this diode needed? And it's an 1N4001 which can take a whole amp of current. What could happen to get that kind of current at that point in the circuit?

[GN L]

The schematic is simplified for KiCAD. The real connection of these two diode you can refer to CR5 and CR6 on page 30 of the Kim-1 user manual .

The PAL-2 will optimize for more readable schematic.

> On Apr 1, 2022, at 5:02 PM, Magnus Olsson <mne...@gmail.com> wrote:
>
> I'm studying the schematic of PAL-1 and I can't figure out what's the purpose of the reverse-biased diode D1 in the RS-232 output circuit. It connects an open-collector output of one of the 74LS38 buffers to ground, so it will only conduct if that output goes below -0.7 V, but how could it do that?
>
> That output is pulled up to VCC by R33 and can pull down to logical zero (which is > 0 V) when the output becomes active, and it's connected to the inputs of a NAND gate which can't go negative either, so why is this diode needed? And it's an 1N4001 which can take a whole amp of current. What could happen to get that kind of current at that point in the circuit?
>
>
>
>

> --
> You received this message because you are subscribed to the Google Groups "PAL 6502 computer" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to pal6502+u...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/pal6502/56bc7c5b-dd92-4378-b407-2c358042fd26n%40googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.

Best,
Liu

[Magnus Olsson]

Thanks for the reference, but that doesn't quite explain it.

The schematic in the KIM user's manual is for the 20 mA current loop interface. As far as I can see, Q7, CR5 and CR6 are connected to each other in exactly the same way as Q8, D1 and D2 in the PAL-1 schematic, but then they are connected directly to the current loop. In this case, the diodes make sense since they are connected directly to the outside world (I don't know very much about current loop interfaces so I may be missing some details here).

But in the PAL-1, it seems as if one gate and two transistors were added after the current loop interface to translate the signals to the RS-232 levels, and the diodes should never become forward-biased. The best explanation I can think of is that the first part of the PAL circuit is simply copied from the old current loop interface and that the diodes are there because they were there on the KIM. But then I don't know much about RS-232 either so I may have missed something obvious :)

[Jim McClanahan]

I think you're probably right that it is an artifact left from the original design that doesn't do anything now. I can't see any set of realistic conditions where it would not be reverse biased (or at ground on both sides) and non-conducting.

Thanks,

Jim W4JBM

[GN L]

Very useful investigation! I also think this part of circuit is simply copied from the original KIM-1 design. The TTY RS232 interface and expansion interface on PAL-1 are copied from the design of Rich Dreher and Vince Briel on the Micro-KIM, and the Micro-KIM is almost keeps the design from KIM-1, which I think is a good design principle for a KIM-1 replica.

This part will change on PAL-2, I want to put the ttl interface onboard, then add an rs232 add-on card to the mainboard for backward compatible.

Thank you for the study!
Liu

> To view this discussion on the web visit https://groups.google.com/d/msgid/pal6502/c7ae15ba-27c1-430e-81be-f99d3f8c78afn%40googlegroups.com.