Well, I'll just apologize now for being a newbie ... And for a long post ...

Would it be safe, in general, to try to use a salvaged PC power supply
to power a KIM-1 board ?

I rediscovered a KIM-1 recently (got it at a garage sale some time ago,
in a box with other stuff), stumbled onto some KIM site or another
probably Vern Garner's www.KIM-1.com, did't see an easy way to leave
a query there). While browsing around there, I realized all of the
sudden, that the board I had could (and was intended) to be actually
used, to some extent, just sitting alone on the desk, with an external
power supply! I guess I'd figured that with the edge connectors on it, it
was intended to be used with some sort of bus and/or chassis, which I
don't have, of course.

Which leads me to my question, basically, what's useful, comtemporarily
available, to power a KIM-1 board? I got a copy of the User's Manual,
used the A-Connector as stated to wire up +5V/Gnd from my DD-1
breadboard kit.

Unfortunately, I subsequently realized that although the DD-1 power
shows about +5.15V with no load, when connected to the KIM-1, it
drops to 4.3V.

Following the instructions in the "KIM-1 User Manual" and "KIM Hints"
pamphlet ... On pressing the RS key, the LED's flashed, but then when
attempting to enter the "simple program" (User Man 2.4), nothing
actually shows up on the LED's ...

Wondering if my power is just too low? Continuing to follow the
troubleshooting steps ... Between A-A and A-1, there's 4.11V, btw
E-21/E-22 is 3.97V, E-21/A-K is 4.11V ... Watching voltage btw
E-7/E-21 while pressing the RS button, goes from 0.01V to 3.96V ...

So, that's about where I'm at so far ... I'd like to try to use what I've
got lying around to make sure the KIM-1 board works at all, before I
even consider trying to breadboard up a power supply like the one
shown in the User Manual ...

Finally, direct questions ... How sensitive IS the KIM-1 to voltage
specs? It specifies +5V ±5%, obviously I'm below that ...

I have a number of old PC power supplies available (pack-rats of the
world unite!), but I'm cautious about those because they specify
around 22A (!) for the +5V line. The "KIM Hint" doc mentions specs
of 1.2A at +5V, and measurements between 700mA to 1A.

Also, I've read that PC supplies, while very well-regulated and stable
under normal use in a PC, get less so with a significantly smaller load
(like trying to use one for a hobby station, playing with gates and such
stuff) ... Only thing I've really done there, is put a VOM on one, and
watch it stay pretty stable, as far as I could tell, at 5.37V, and also with
and old drive attached, at 5.2V ... I've forgotten most of whatever I
once thought I knew about electronics, not really sure how to evaluate
this situation ...

Although I did search out one post which said that it was OK to use an
old PC power supply, as long as you put another load on it, such as an
old CD-ROM drive.

That one's here:
http://www.classiccmp.org/pipermail/cct ... 09486.html

But I didn't want to subscribe to a newslist just to ask the question...

This forum seemed somewhat recently active, so I thought I'd give it a
shot ... If anyone has any advice, please reply, or if I guess you could
email me at MinstrelMike@Yahoo.com if you wanted to go external.

I'll probably be impatient if I don't see a useful reply soon, and just try
out the PC power supply... Of course, then I'll reply here myself and
post the results, and of course, more questions if it doesn't work out.
Most likely, my attention-span will time-out on it within a few days
anyway ...

Well, if anyone's read the whole thing, THANKS!

Mike

PS: Maybe should have posted to the Hardware forum?

You'll probably be interested in the KIM-1 links at http://www.6502.org/oldmicro/trainers.htm

I don't know anything about KIM-1, but 65xx NMOS parts are supposed to have 5V +/-5%. Running them too low probably won't hurt anything, but they're not guaranteed to work at the lower voltages. (The CMOS ones will work much lower if the clock rate is lowered appropriately.) The 22-amp capacity of the power supply is not a problem as long as the voltage is right. It just means that if you short something, you'll get a bigger poof!

>You'll probably be interested in the KIM-1 links at >http://www.6502.org/oldmicro/trainers.htm

Yep, pretty much started there ...

The only part that really says anything about the power supply itself,
is the "User Manual", which gives a schematic to home-build one, but
they don't really explain it. I guess, back then, anyone who would
even think of playing with something like a KIM-1 board, would also
know enough about electronics to deal relatively easily with building
such a power supply...

Me, on the other hand, I've been doing a bit of reading up, and their
schematic looks like a variation on a basic capacitive-filter with a full-
wave bridge rectifier (whew!). On a seriously coincidental note, the
book that I'm mainly referring to here, it's chapter on
microprocessors features entirely the 6502 and it's support chips!

>I don't know anything about KIM-1, but 65xx NMOS parts are
>supposed to >have 5V +/-5%. Running them too low probably
>won't hurt anything, but >they're not guaranteed to work at the
>lower voltages. (The CMOS ones

OK! Sounds as if it should work, then, if there's nothing ELSE wrong
with the KIM-1 ... I think it basically sat in a box in a dusty garage
for at least the past couple of decades, until I got it accidentally in
with a box of miscellaneous stuff at an estate sale. I was mainly
trying to score the DD1 Digi-Designer; it's old, but for $10...

I'll give it a try with the PC supply, and double-check the
voltages on the other pins, etc. ...

It's been a LONG time since I ever studied this sort of stuff, basically
just the required electronics for my CS degree, and since I haven't
really used any of those skills for about 15 years, I've forgotten ...

I don't know if it would mean anything, but I was also thinking of
checking the resistance across some of the +5V/Gnd pins on the
KIM-1, then putting an equivalent resistor across the +5V/Grd pins
on the PC supply, see if the voltage appears to drop, and see if
the resistor gets very warm ...


Thanks again!

(I know, I'm getting annoyingly chatty, I can't get into the habit of
writing more terse emails, I write as if I'm writing a letter)

Mike

It won't be a linear resistance, so what you measure with an ohmmeter probably won't be anywhere near what it would be operating with 5V. If they say 700mA though, 5V/700mA is just over 7 ohms, and 5V*700mA is 3.5 watts; so if you can get a 6.8- or 7.5-ohm 5-watt resistor to artificially load the power supply, whatever voltage you measure with that load should be pretty close to what you'll get with the computer connected to that power supply. (You could get away with a 2W resistor and maybe even a 1W if you leave it on only long enough to get a reading and then remove it or turn off the power.)

It needs to have a regulator too. Just diodes and capacitors coming off a transformer will definitely not be accurate enough. There are power supplies available all over the place though, including a load from Jameco-- workbench supplies, wall-warts, and ones made to mount inside computer cabinets.

Yes, that's how I power mine.

_________________

What I did was buy a surplus switching power supply from one of my usual dealers (I can't remember if it was Marlin P. Jones or All Electronics) for about $5. It's a brick-style fully-enclosed supply that provides 5V and +/-12v. I forgot the current ratings exactly, but they were more than sufficient to power the board. It had a DE9 on the end so I just whipped-up a corresponding cable for the KIM. Works great.

The thing with PC supplies is that depending on the model, it may have a minimum load requirement in order to guarantee regulation. I don't know how the minimum load is calculated for each supply (usually available in the spec sheet which might be available if bought surplus; otherwise usually not available), but I would think that the KIM draws enough on the 5v rail (1A or so) that it should be OK.

Rich

_________________
Rich Cini
Build Master and maintainer of the Altair32 Emulation project
http://cini.classiccmp.org
http://altair32.classiccmp.org

Ok, guys, thanks a lot for the quick responses! (not to
mention reading my long posts :)

Here's where I'm up to so far ... Power diagnostics seem
better, but still can't get past Reset...

Again, following the instructions in the "User Manual" and
"Hints" ... Checking various pins: Vcc (E-21/E-22) = 5.08v,
Reset (E-7/E-21) = 5.06v to 0.06 w/RS pressed, and pin
E-V at least seems to have a "pulse" (don't have a scope,
so checking with my logic probe shows a pulsed signal, but
no idea about the actual frequency).

So far, this seems OK ... But when I continue per UserMan,
and attempt to begin the "Simple Program" (2.4), pressing
[AD] then [0] [0] [0] [2] keys, NOTHING shows up on the
display LEDs.

One further note: a couple of things that I've noticed
DIFFERENT about my KIM-1 verus the various documentation
that I've gathered from the net... Mine has, instead of the
eight MOS 6102 IC's, has eight NEC P77826-730 D2102AL-4.

I'm assuming that this has NO impact on functional docs
and instructions. Must have just used different, compatible
memory for some reason.

This unit has both "Commodore" and "MOS" logos, "Rev G",
latest IC date code on the MOS MPS6530 003 is 1078 (Oct
1978?).

Also, the keypad is a little different: the SST switch is at
the upper left, rather than upper right, as in the UserMan,
and is labelled "On" to the left, instead of right. The
UserMan says to keep it to the LEFT (SST Mode Off), while I'm
assuming that the keypad is labelled appropriately, and am
keeping it to the RIGHT for SST Mode Off.

So, that's it for the moment. Any further advice?

I may eventually get brave enough to fry my brain by
looking at a couple of schematics that I've found, maybe try
to trace a few signals with my probe and/or DMM.

Also, although I don't have an oscilloscope, I've downloaded
several programs which attempt to use the PC sound card
inputs to emulate scopes and/or frequency counters. I've
tried a few, still not too sure how well they actually work.

Starting to think that maybe I'm over my head, but not
drowning just yet :)

Thanks again for any and all past and/or future assistance!

Mike

Oh, and Garth:

>It needs to have a regulator too. Just diodes and capacitors
>coming off a transformer will definitely not be accurate enough.

Yep, you're right, of course! I just wastn't paying attention!
The schematic in Appendix D "Suggested Power Supply" DOES
in fact, specify two regulators, one for each, +5V 1.2A and 12V,
respectively: LM309 and LM78?12 ...

>There are power supplies available all over the place though,
>including a load from Jameco-- workbench supplies, wall-warts,
>and ones made to mount inside computer cabinets.

Just trying to make some use out of what junk I have piled
and scattered about, at the same time being "correct" enough
to try not to damage the KIM-1 before I have a chance to
play with it some :) I have lots of wall-warts around, but
NONE at +5V, mostly 3, 6, 9, or 12, most common, I guess,
and none of them seem nearly as stable as the PC PS ...

Thanks again,
Mike

Hi Mike,
I am not sure if I understood right, but certainly I could give you a few hints:
Is the display working continuosly? (What I mean is does it show the information of a random adress and data). I think there must be a wiring of one connector to enable internal RAM.
I believe PC- power-supply sometimes need a "power good" signal to work right.
Hope this can help you.
I have neither a KIM (but I would like to have one!), only a hobbyist and learning english was long time ago...
best regards
Charly DL4DBY

No, no, no, no, no. Power good is ALWAYS an output from the PC power supply. One Amiga magazine gets this wrong twenty years ago and people are still quoting it as true.

The purpose of power good is to tell the motherboard that it can come out of reset as the power supply outputs are now stabilized.

Lee.

You certainly are right, Lee. What I had in mind is: Some PC-style power-supply do not work when power consumption is very low. And ATX- Supplies work when one wire (the green one?) is connected to ground.
Seems I mixed this up. Thanks for the info.
Greetings from germany
Charly

I'm trying to be more concise, not really working :)

Just typed a nice long message in the forum reply text box, only to have
it all disappear after a time-out.

MinstrelMike>> So far, this seems OK ... But when I continue per UserMan,
MinstrelMike>> and attempt to begin the "Simple Program" (2.4), pressing
MinstrelMike>> [AD] then [0] [0] [0] [2] keys, NOTHING shows up on the
MinstrelMike>> display LEDs.

DL4DBY> a few hints: Is the display working continuosly? (What I mean is
DL4DBY> does it show the information of a random adress and data). I

No. Always, just flashes (at least a few segments on all size LED's)
when the RS key is released, but when attempting to enter addresses
and/or data, etc., nothing at all...

I've also attempted to press gently on the keyboard and displays at
various points, thinking that if there's a loose connection, maybe
a momentary flex might allow a brief display; nothing so far.

Only ONE exception to "prove" this "rule": once (only ONCE!),
while pressing each key in order and (more or less) randomly, to
try to "feel" if the keys were "pressing" properly, the following
appeared on the display: "3800 88" ... I've tried this again along
with Resets, dozens of times, but not knowing if I might have
accidentally triggered some specific sequence, no changes so far.

DL4DBY> think there must be a wiring of one connector to enable internal
DL4DBY> RAM.

Well, there's nothing like that mentioned in the User Manual, unless
I've missed or mis-read something.

The only other setups required in the Manual, in addition to the
main power 5v supply, is to ground pin A-K (Decoder Enable).

I have tried a few further diagnostics in the meantime ...

Trying to check the keyboard pins on the Application Connector,
the Keyboard Column pins (A-18 thru A-21, A-W, A-Y) appear to
ALWAY be "HI" (4.85v), while each the Keyboard Row pins (A-17,
A-V, A-X, A-Z) appear to go from "LO" to "HI" when an associated
key is pressed. Seems that the signals are at least getting
off the keyboard correctly?

Looking at the schematics in the User Manual, it seems that
the Keyboard Column pins should be "pulsed" instead of just
always "HI". Not at all sure how to interpret this; how
could one tell from these pins which specific key was
pressed? Maybe I'm misunderstanding completely; if these
pins are used only for connecting an external keyboard,
should they make any sense as "output" anyway?

The only other thing to mention, checking pin E-V for 1Mhz
(p.C-1, Step 6); since I don't have a scope, I can't really
try this, but my logic probe does indicate a "pulsed"
signal. I'm planning to buy a better DMM, with a
frequency counter, soon (will justify as a Christmas
present to myself :), then I'll try again ...

Well, that about does it for now (not exactly concise :).
If anyone with sufficient detailed knowledge of the
KIM-1 board is willing to advise me further, I'd
greatly appreciate it!

Thanks again, everyone who has replied so far!

Mike

> Just typed a nice long message in the forum reply text box, only
> to have it all disappear after a time-out.

If your logged-in time times out when you're writing a long post so that clicking "Submit" asks for your name and password again, just enter them and keep going. It will remember your book and post it after you've logged in again. Of course there are always options like writing in another program (often offline) and copy-and-pasting too once you log back in too.

About the keyboard-- I have no KIM-1 knowledge or info here, but there are various ways to read a keyboard. I'll mention one example here. From your post, I'll hypothesize that "true" a logic high state. When no key seems to be pressed, it makes sense for the software to quit scanning for which key is pressed-- after all, no key is. They might be keeping all the columns high until a row is sensed to have gone high. That would tell the software now to just start doing one column at a time, so that with both column and row information, it can tell which key is pressed. After it has found out which key, it wll keep watching that key. The strobing is suspended as long as that key it is still being pressed. After the key has been determined to be released for the full de-bouce period, then the software will look to see if any other key on that column is being pressed. If not, it goes back to the first state and starts over. This lowers the software overhead, does the software key-debouncing, and allows two-key roll-over.

Hi,
I think you're meaning this (mentioned in the manual):
Note 2: The jumper from pin A-K to Vss (Pin A-1) is essential
for system operation. If you expand your system later,
this jumper will be removed and we'll tell you what to
do to pin A-K. Figure 2.2 Page 6.
Means Application- connector Pin "1" and "K" to ground, same connector pin "A"= 5v, N=12 V (only for cassette operation).
Also be shure that NO connection is established between V - 21 (TTY enable).
GO- and ST- Key need to be open.
Good luck,
Charly

Sorry about not replying for a while ... I was off-line after running over my ISP dial-up limit and getting my account disabled until enough time had passed...

GARTHWILSON> If your logged-in time times out when you're writing a
GARTHWILSON> long post so that clicking "Submit" asks for your name
GARTHWILSON> and password again, just enter them and keep going.
GARTHWILSON> It will remember your book and post it after you've
GARTHWILSON> logged in again. Of course there are always options
GARTHWILSON> like writing in another program (often offline) and
GARTHWILSON> copy-and-pasting too once you log

Well, it didn't happen quite that way: after the time-out, submit, and username/password prompts, it did NOT appear to remember the "book" ...
DID wind up back in the reply/edit form, but it was empty ... Sometimes I
forget that a form textbox does NOT actually send data until you submit
the whole thing, therefore the time actually spent typing counts toward the time-out ...

GARTHWILSON> About the keyboard-- I have no KIM-1 knowledge or info
GARTHWILSON> here, but there are various ways to read a keyboard. I'll
GARTHWILSON> mention one example here. From your post, I'll
GARTHWILSON> hypothesize that "true" a logic high state. When no key
GARTHWILSON> seems to be pressed, it makes sense for the software
GARTHWILSON> to quit scanning for which key is pressed-- after all, no
GARTHWILSON> key is. They might be keeping all the columns high until
GARTHWILSON> a row is sensed to have gone high. That would tell the

Thanks, Garth ... I'm familiar with the general theoretical keyboard scanning, but have never had the opportunity to deal with one in the "real" world. Your hypothesis seems to explain my observations, but still leaves no way for a "human" in real-time (much slower than the "strobing" frequecies of course) to determine which key is being pressed.

I'm going to go back to reading and see if there's any clues as to exactly how the KIM-1 board handles things.

DL4DBY> I think you're meaning this (mentioned in the manual):
DL4DBY> Note 2: The jumper from pin A-K to Vss (Pin A-1) is essential
DL4DBY> Also be shure that NO connection is established between V - 21

I think I've understood the above and have it setup as instructed. I only meant that there was nothing ELSE specified in the manual to do for the minimal setup (w/out casettete or TTY) ...

DL4DBY> GO- and ST- Key need to be open.

Yeah, I can imagine how having either of those stuck would freak things out. But, they seem Ok, as far as watching the Keyboard Row pin goes from "LO" to "HI" when they're pressed. If they were stuck closed,
seems like the row pin would also be stuck "HI" ...

MinstrelMike> I'm planning to buy a better DMM, with a frequency counter

Well, got that, seems to work much better than my old cheap one!

Trying again to measure the frequency at pin E-V, instead of a pulse, I get a 3.74v (and the logic probe just gets a "HI" now) ... Not sure what's happened to it in the meantime, but it seems that I have a problem, at least, with the clock signal.

So, I guess that probably does it for this KIM-1 for now ... I'm getting ready to head out of town to do the Christmas bit with the family; maybe I'll take it along, play with it if I have time, try tracing some of the PCB lines around.


Thanks, guys! I'll reply again after New Year's, maybe sooner if I get a chance, have anything new to say ...

Mike