- VCF South West - June 14 - 16, Davidson-Gundy Alumni Center at University of Texas at Dallas
- VCF West - Aug 2 - 3, Computer History Museum, Mountain View, CA
- VCF Midwest - Sept 7 - 8 2024, Schaumburg, IL
- VCF SoCal - Mid February 2025, Location TBD, Southern CA
- VCF East - April 2025, Infoage Museum, Wall NJ
-
Please review our updated Terms and Rules here
KIM-1 getting stuck in loop
- Thread starter TheSmJ
- Start date
nige the hippy
Veteran Member
More than likely then (I wasn't familiar enough with the 6530 to know if it has a uart on it)...
Seems strange (brain numbing) to bit bash transmit inverted, but receive true though!
Dwight Elvey
Veteran Member
Hi
When bit banging, polarity of the signal is trivial to handle.
As I recall in RS-232 a 1 is a negative signal while a
0 is a positive. It would make sense to invert once for
the simplest recieve method.
Dwight
When bit banging, polarity of the signal is trivial to handle.
As I recall in RS-232 a 1 is a negative signal while a
0 is a positive. It would make sense to invert once for
the simplest recieve method.
Dwight
nige the hippy
Veteran Member
Absolutely agree, but it seems daft to have one true and the other inverted, when a 7407 will buffer it in exactly the same way as the 7406 shown but without inverting. It keeps your ttl logic sense and your rs232 logic senses the same.
going O/T (but then we already are, and I'll start a new topic if necessary!) I was looking in the apple II users guide. on page 343 (game port) it refers to 5V as being TTL low and 0V as TTL high....
I noticed back years ago that enable lines and decoded outputs tend to be active low for TTL (and high for cmos), I was wondering whether at some point in the early days of ttl history, logic levels were inverted and that a 7400 was in fact originally a quad OR gate?
Dwight Elvey
Veteran Member
Hi
A quad NAND gate in TTL is faster than a Quad NOR gate. This is typical for
any particular logic family to favor one over the other.
Most early tansistor logic was 0v = false and -V was true.
The choice of which is the true level is really otherwise arbitrary. One type
of logic can be demorganed into the other.
I'd suspect that the conversion between RS232 and TTL was intended to be the simplest
buffer, that is invert. Even so, RS2323 really doesn't deal with 1's and 0's. It deals
with mark and space. There is nothing to say that after the start bit that the data
must be, RS232, true or false. It is just by convention that it is the way it turned
out.
My guess on TTL was that it was compatable with DTL. In DTL, it was common to create
the logic function of a AND by wiring the outputs to gether. This made the typical
logic NAND/AND based. This got carried over to TTL that no longer used the slower
wires AND structure.
Now, was there a time when DTL had logic levels reversed?? I don't think so.
I think that most RTL and DTL was always 1 was more positive and 0 was more negative.
Dwight
nige the hippy
Veteran Member
sorry, I think you may have missed my point, the 7400 reference was ... using De Morgan a NAND gate applied to negative logic becomes an OR.
And I know that looking at old schematics the "active" state was -ve (as a legacy of pnp transistors)
Just a musing, and a wonder if anyone here knows for certain why on earth the book should say as it does.
And I know that looking at old schematics the "active" state was -ve (as a legacy of pnp transistors)
Just a musing, and a wonder if anyone here knows for certain why on earth the book should say as it does.
I finally got back to this - it's been a busy few weeks.
I built the aforementioned RS232 adapter and after checking nearly everything with a VOM and rebuilding it a number of times, I can't seem to get the KIM to send a signal to any terminal program. No "KIM-1" message or anything.
Any ideas as to what I'm overlooking here? About the only thing I haven't tried is to connect it to a different motherboard running Win XP rather than Vista... but I don't see how this should make any difference.
I built the aforementioned RS232 adapter and after checking nearly everything with a VOM and rebuilding it a number of times, I can't seem to get the KIM to send a signal to any terminal program. No "KIM-1" message or anything.
Any ideas as to what I'm overlooking here? About the only thing I haven't tried is to connect it to a different motherboard running Win XP rather than Vista... but I don't see how this should make any difference.
MikeS
Veteran Member
You're thinking in terms of RS-232 to TTL conversion, but that's not the case here; the KIM (and AIM) interface is a current loop interface and the keyboard (input) is inverted in the KIM because it goes to +5 instead of ground. In practice, the circuitry on the XMT line can usually be omitted and pin U can be connected directly to the RX of the terminal or PC; only the KIM's RX on pin T has to be clamped and inverted (the AIM does that and has the proper polarity input on pin Y, and can also have the RX line connected directly to the TX of the terminal without any interface circuitry at all).
Always worked for me anyway...
mike
MikeS
Veteran Member
-----
Did you connect V to 21 on connector A? That enables the TTY.
And type Rubout (whatever that is in your terminal emulation) after a reset?
Like I said, you can probably do without the IC, transistor etc. on the KIM pin U and connect it directly to RX (pin 2 if it's a DE9, pin 3 if DB25) on the PC (*NOT* TX or you may be sorry ;-)
mike
mike
No I had not known about connecting V to 21 on A... after I do that, the KIM's display stays dark after pressing RS so it's definetly a step in the right direction. Thanks for pointing it out to me!
Now, the KIM still does not display anything on the terminal. I'm using 1200bps, 2 stop bits, and no parity or flow control. I tried the backspace key, the right arrow and the Delete key as rubout (since I don't know what the terminal defaults to) and still no luck. I tried forcing the protocol to TTY rather than "Auto" and no dice there either.
I'm starting to wonder of the 2N2222 transistor I have is bad. I might go out to Radioshack tomorrow and pick up a pack of them and try again.
Now, the KIM still does not display anything on the terminal. I'm using 1200bps, 2 stop bits, and no parity or flow control. I tried the backspace key, the right arrow and the Delete key as rubout (since I don't know what the terminal defaults to) and still no luck. I tried forcing the protocol to TTY rather than "Auto" and no dice there either.
I'm starting to wonder of the 2N2222 transistor I have is bad. I might go out to Radioshack tomorrow and pick up a pack of them and try again.
MikeS
Veteran Member
Some thoughts:
Make sure you're using the right pins at the PC end. The schematic in User Notes#0 is for a DB25; if your PC has a 9-pin RS232 connector the pin numbers are reversed and ground is on pin 5 instead of pin 7. This is *especially* important if you decide to do without the stuff on pin U and connect directly to the PC; if they're reversed, you'll damage the KIM.
If you have a voltmeter, you can check to make sure:
With the PC or terminal turned on but disconnected from the KIM, you should see abt. 2.5V on the RX line (the one that goes to the 200 ohm resistor and ultimately to pin U) and abt. -9V on the TX line that goes to the 4.7K resistor and ultimately pins R & T (referenced to ground of course).
Assuming that's OK, you can check the RX 2N2222 by grounding the base (the junction of the 6.8K and 4.7K resistors; when you ground it, pins R & T should go from ~0V to +5V or so.
Good luck!
mike
Make sure you're using the right pins at the PC end. The schematic in User Notes#0 is for a DB25; if your PC has a 9-pin RS232 connector the pin numbers are reversed and ground is on pin 5 instead of pin 7. This is *especially* important if you decide to do without the stuff on pin U and connect directly to the PC; if they're reversed, you'll damage the KIM.
If you have a voltmeter, you can check to make sure:
With the PC or terminal turned on but disconnected from the KIM, you should see abt. 2.5V on the RX line (the one that goes to the 200 ohm resistor and ultimately to pin U) and abt. -9V on the TX line that goes to the 4.7K resistor and ultimately pins R & T (referenced to ground of course).
Assuming that's OK, you can check the RX 2N2222 by grounding the base (the junction of the 6.8K and 4.7K resistors; when you ground it, pins R & T should go from ~0V to +5V or so.
Good luck!
mike
Well that would have been good to know beforehand... I haven't seen a DB25 serial port on a PC in 15 years.... I assumed this was DB9 but then again these schematics must be at least 30 years old themselves.
I'll swap things around and let you know....
Still not working. Now I have to ask what component would have been damaged on the KIM if U had been directly connected to the PC while I was using DB25 schematics, since I had already tried that before I knew these pins weren't for DB9. Outside of the RS232 stuff the KIM works fine, but now I may or may not have damaged RS232 related components to begin with, so I won't bother to continue until I can test that, hopefully with a VOM since I am lacking an osilloscope.
MikeS
Veteran Member
------
Well, if you have damaged anything I'm certainly very sorry, but I did caution you:
<quote>
...you can probably do without the IC, transistor etc. on the KIM pin U and connect it directly to RX (pin 2 if it's a DE9, pin 3 if DB25) on the PC (*NOT* TX or you may be sorry)
</quote>
That's why I say RX and TX (receive & transmit), to avoid confusion because of IBM's non-standard usage of the RS-232 pins; maybe I should have been more emphatic.
Hopefully there's no damage; if there is, it will probably only be U26. Check the voltages on pins 8, 9 and 10 and see if they make sense for a NAND gate with output on pin 8.
Fingers crossed,
mike