• Please review our updated Terms and Rules here

KIM-1 no phone home

GanjaTron

GanjaTron

Experienced Member
Joined
Jan 11, 2007
Messages
201
Hi folks,

recently got a KIM-1 (Revision G) on eBay which was guaranteed working -- there were photos showing it in action accompanying the auction, running the simple add-two-numbers test program described in the manual.

Lo and behold, when I received it and hooked it up to my 5V bench supply, I got *nothing*. Zilch. The LEDs don't even flicker. Hitting reset does nothing either. I checked the voltages on the chips, and they're fine. There also appears to be a clock signal, although I only checked it with my VOM. I did notice the RES signal is *low* and doesn't change when I hit reset (and no keys are stuck), which is definitely abnormal.

I called the seller and he *swore* it worked (and I have no reason to doubt him given the photographic proof). Am I doing something wrong or do we have an obscure case of dopant migration and catastrophic chip failure here?

Note that this is my second attempt at getting my hands on a working KIM-1 (which doesn't exactly help my frustration). My first KIM (Revision F) worked for *one* day and then died, which seems uncannily familiar now! This particular unit flashes briefly on powerup but also remains unresponsive. Is this a systematic failure mode with the later KIM-1 revisions?

Loads of thanks to anyone who can enlighten me!

--Roland
 
I'm not going to be much help trouble-shooting these since I've never attempted to power mine on, however at Vince's wondeful MicroKim workshop someone had said that the MOS 6502's are notorious for being over sensitive to ESD. I could see how easy it would be for anyone not being static concious to possibly handle it wrong resulting in blowing the CPU or damaging some resistors or capacitors.

I'm guessing you also tested the power supply for voltage and amperage not just the chips for voltage? I would be reluctant to swap the CPUs in the event that either is more damaged, however you could certainly try reseating anything or looking for cold solder joints. My MicroKim is covered in my shoddy soldering and colorful solder burns and cold solder here and there however it has enough newer components to forgive me. The older equipment may not.

- John
 
Well, without going into a lot of detail, if the RESET line is low, it is in a RESET state and not running. You have to get that RESET line to at least 2.5V for the system to work. Now, what is causing the problem? Not sure totally but there is a debounce circuit connected to the step switch and RS (reset) using a 556 IC. Start by reseating that chip. Grab a set of schematics online and see if you can see the chip I'm talking about. It maybe directly soldered to the board, but it could have a socket. You never know with these boards.

Also, there is a signal on the CPU that tells you if it is fetching instructions. Check pin 7 the sync line. It goes high when fetching and instruction from memory. So, it should be pulsing too fast to see and therefore the voltage will be around 2.1V.

As long as the 6530's aren't damaged, they are fixable machines.

Sorry, I have to ask... Are you sure, 5V and GND is hooked up correctly and not pulling more than 5V? A 5V wallwart power supply can actually have 6-9V initially and get pulled down under load. Too high voltage will cause problems.

Vince
 
vbriel said:
Well, without going into a lot of detail, if the RESET line is low, it is in a RESET state and not running. You have to get that RESET line to at least 2.5V for the system to work. Now, what is causing the problem? Not sure totally but there is a debounce circuit connected to the step switch and RS (reset) using a 556 IC. Start by reseating that chip. Grab a set of schematics online and see if you can see the chip I'm talking about. It maybe directly soldered to the board, but it could have a socket. You never know with these boards.
Click to expand...

Hi Vince,

yeah, the stuck RES line is pretty strange. I've already spotted the 556 (which is soldered in) and could readily replace it. Which still leaves me wondering: why did it die a few days after it was tested?

I reseated the 6502 and 6530's, which are the only socketed chips. I also grounded myself with a wrist strap, so I'm ruling out static discharge.

Also, there is a signal on the CPU that tells you if it is fetching instructions. Check pin 7 the sync line. It goes high when fetching and instruction from memory. So, it should be pulsing too fast to see and therefore the voltage will be around 2.1V.
Click to expand...

Thanks for the hint, will (hopefully) check that this weekend.

As long as the 6530's aren't damaged, they are fixable machines.
Click to expand...

Yeah, the 6530's worry me. Apparently that's what died in my other (Rev. F) KIM. These are of course irreplaceable...

Sorry, I have to ask... Are you sure, 5V and GND is hooked up correctly and not pulling more than 5V? A 5V wallwart power supply can actually have 6-9V initially and get pulled down under load. Too high voltage will cause problems.
Click to expand...

It's an adjustable lab power supply, so it's regulated and should be ok. I've checked the output with a VOM and found it to agree with the voltage indicator. I've also checked it with a scope for excessive ripple, which looked ok too. When I set it to 5V the KIM draws around 0.5A, which seems reasonable. I don't think the PSU is a problem, but I'll recheck it. I assume moderate overvoltage doesn't immediately destroy the circuit, but increases heat dissipation. The 6502 and 6530's are warm but not hot. But then again it's CMOS...

Thanks for the hints, I'll let you know what I find out!

--Roland
 
yeah, the stuck RES line is pretty strange. I've already spotted the 556 (which is soldered in) and could readily replace it. Which still leaves me wondering: why did it die a few days after it was tested?

Hi
I'd check the capacitor used for timing with the 556. It is more likely
to fail than the 556.
Dwight
 
Dwight Elvey said:
I'd check the capacitor used for timing with the 556. It is more likely
to fail than the 556.
Dwight
Click to expand...

Great idea! Those caps tend to cause a lot more headaches than silicon. Thanks for the hint! I'll check it out...

--Roland
 
Yes, good idea. It would make sense that a cap would fail. This could cause the circuit to be in a constant reset state.

Vince
 
Okay, that'd be C2 and C3 around the 556 in the lower right corner next to the keypad. Actually, are these electrolytics (which are notoriously prone to failure)? They appear to be something else, at least from the casing.

--Roland
 
Hi
Actually, ceramics that have low current potential across them,
such as those in timing circuits, fail quite often with high leakage.
When used as bypass capacitors on supplies, these self
repair themselves by blowing the shorting silver fingers that
form over time.
When used for something like a timing circuit or signal filter,
they can be bad as well.
These are often hard to locate as well because most don't realize
that they fail in this way. One rarely sees them fail as power supply
bypass capacitors.
Dwight
 
Dwight Elvey said:
Hi
Actually, ceramics that have low current potential across them,
such as those in timing circuits, fail quite often with high leakage.
Click to expand...

But could they (theoretically) work for a while after years of disuse, and then suddenly die?

--Roland
 
GanjaTron said:
But could they (theoretically) work for a while after years of disuse, and then suddenly die?

--Roland
Click to expand...


Hi
It is not likely to have a sudden failure but should be considered.
The 556 reset circuit can be checked for the most part with a voltmeter.
Dwight
 
Hi
One other note. If the output of the 556 was held in the wrong
state, by a short on the output, it would get noticeably hot.
These were designed to drive quite a bit of current. This
leads me to suspect something on the input side.
Dwight
 
Hi again folks,

I finally got round to testing the KIM this weekend -- and it paid off!

I checked the input side of the 556 with a scope and found ST and RST working, so the capacitors are fine. Then I checked the output side, and found that only ST responds, so apparently the 556 is shot (though it's not running hot).

I then boldly proceeded to briefly short RST (pin E-7 on the expansion connector) to GND... lo and behold, the KIM came alive!!! I punched in the add-two-numbers test proggie in the user manual, and all was well! :)

So there you have it. I'll replace the 556 next weekend (got a spare lying around). Why it died right after the seller tested the KIM is anyone's guess...

Thanks again for your help!

--Roland
 
The 556 is a debounce circuit so it can be bypassed with no problems. Sweet, good to see a KIM come back to life. Do you still have the other KIM-1 that wasn't working?

Vince
 
vbriel said:
The 556 is a debounce circuit so it can be bypassed with no problems. Sweet, good to see a KIM come back to life. Do you still have the other KIM-1 that wasn't working?
Click to expand...

Hi Vince,

yeah, still got it sitting around at home, gathering dust. Apparently the one or both 6530s died, so it's definitely beyond repair. I guess it'd still make a fine display piece, although I did replace and socket two or three chips (including the 6502 and 556) in an attempt to get it going, but to no avail. :(

--Roland
 
GanjaTron said:
Hi again folks,

I finally got round to testing the KIM this weekend -- and it paid off!

I checked the input side of the 556 with a scope and found ST and RST working, so the capacitors are fine. Then I checked the output side, and found that only ST responds, so apparently the 556 is shot (though it's not running hot).

I then boldly proceeded to briefly short RST (pin E-7 on the expansion connector) to GND... lo and behold, the KIM came alive!!! I punched in the add-two-numbers test proggie in the user manual, and all was well! :)

So there you have it. I'll replace the 556 next weekend (got a spare lying around). Why it died right after the seller tested the KIM is anyone's guess...

Thanks again for your help!

--Roland
Click to expand...


Hi
It could still be a problem with other circuits. I've seen many cases where
the reset circuit was faster than the oscillator start up. This can be cause
by old silicon or old crystals.
One can try using a different crystal or in some cases, a little addition
phase shift by adding a 10-20 pf cap on one of the crystal leads to ground
helps.
Crystals take time to start up because they have such high "Q".
Another thing to try is to incease the size of the capacitor used to time
the reset.
It looks like you otherwise have a working machine.
Dwight
 
You must log in or register to reply here.
Back
Top