• Please review our updated Terms and Rules here

KIM-1 shorted?

C

ClassicHasClass

Veteran Member
Joined
Mar 22, 2013
Messages
2,276
Location
(not so) sunny (No) So Cal
Working on my pre-ROR KIM Rev A tonight and when I reconnected the power, I smelled something bad and it went crazy. I immediately powered it off and plugged in Dwight's debug board. The lowest level test will flash the LEDs, but weakly, and I'm noticing the power supply LED also pulsing at the same time. Holding RS will stop the flashes, so it seems like the CPU is sort of working, but I can't get it to do more than that.

I connected the debug board and the same power supply to a known good KIM and it all worked.

I think I've blown a component but I'm not sure how, and it seems very low level. I still smell a "burnt" smell running it. Any recommendations for places to start looking?
 
After I calmed down and hooked it back up again a couple hours later, it runs, passes all of Dwight's tests, and the smell is gone. I might burn it in for a little bit just for confidence but I (fortunately) can't replicate what happened earlier. I really am wondering if there's an intermittent short on the board, though.
 
probably a capacitor went boom. You should check them (visually), and if you can't see anything, consider checking them properly (even if it means desoldering them).
 
Maybe. I ran it this morning for an hour and got no repeats. The TTY is bad, but that was known (the LED and keypad work and are serviced by the same RRIOT, so I don't think it's that, probably one of the downstream components).
 
The I/O ports are relatively easy to talk to thru the monitor and a minimal of code. Debugging anything there should be relatively easy. There are not that many things that could have shorted other than the wires at the connectors.
Dwight
 
The schematic suggests that a failure in either of the 7400s at U15/U16, or the 7488 at U26 or the transistor at Q7 (which doesn't seem to be labeled on my board), could cause the TTY to fail. I'm assuming Q7 is the transistor physically labeled between U15 and U26.

However, it seems to me that I could just sample from the RRIOT's PA7 and PB0 to get TTL-level input and output and work with that ... or did I misread it? That would give me serial and also prove the RRIOT is working, right?
 
ClassicHasClass said:
The schematic suggests that a failure in either of the 7400s at U15/U16, or the 7488 at U26 or the transistor at Q7 (which doesn't seem to be labeled on my board), could cause the TTY to fail. I'm assuming Q7 is the transistor physically labeled between U15 and U26.

However, it seems to me that I could just sample from the RRIOT's PA7 and PB0 to get TTL-level input and output and work with that ... or did I misread it? That would give me serial and also prove the RRIOT is working, right?
Click to expand...
Sounds like you are on the way. You might get one of the Maxim (sp?) chipsto convert 5Vlogic to RS232. That might be more convenient than a TTY connection.Also, on the TTY, you likely need a current source, unless what ever you are connecting it to supplies that.
 
I got this working, but it turns out to be a bit more complicated.

Tapping pin 25 (PB0) on RRIOT 002 works fine for output when connected to a TTL-USB converter for serial output.

For input, however, I could not get anything on pin 34 (PA7) initially. After staring at the schematics for a couple days and having some false starts, I found that the NAND gate coming from U26 was always low because its inputs, including from the inverter at U16, were always high, presumably suppressing the input. I wired pin 9 on the 7400 at U16 to +5V to make the output on pin 8 going to the NAND gate low, and serial input now worked. The tipoff was looking at $1740 and seeing that the high bit was never set (i.e., $7f) until I forced pin 9 high (now $ff).

On this busted KIM, however, the connection didn't work ... until I took +5V off U16 pin 9. Indeed, $1740 was always $ff without that, so I'm guessing this means that U16 is bad at minimum. As a consequence there is no echo when connected to this KIM, though that's something I can live with. Otherwise it checks out just fine. Having a little trouble loading paper tape into it but that could be due to these irregularities. At least this means the RRIOT is almost certainly working, which is a relief.

Now I need to decide if I want to desolder and replace U16 or not.
 
Actually, I'm now reconsidering whether it's U16, because it seems to properly invert on that line. Pin 12 to U15 (not U26, I was tired last night), on the other hand, has an abnormally low voltage (~1V) compared to a working KIM (~4.5V). This goes to what looks like a pull-up resistor at R17, which I checked and seemed to have the wrong resistance value, so I replaced it. I didn't break anything, but it didn't fix it. The voltage across the resistor on the working KIM is nearly 5V, but R17 here (both the old and new resistor) drops down to that same one volt. I checked the resistance of the old resistor out of circuit and this time it's right (~1K), and the new resistor in circuit is now showing the same wrong resistance. On the working KIM these resistors all test properly in circuit.

R17's other connection is to the collector terminal of transistor Q7. Could that be causing what I'm seeing? Or is U15 doing something?
 
You must log in or register to reply here.
Back
Top