3-27-2006

I'm starting this new section to share my experiences selling, repairing, and designing electronic devices from the early 1970's forward.  I was a field sales engineer for some early electronic companies including CTS Microelectronics based in West Lafayette, IN, Magic Dot, in the Minneapolis, MN area, Signetics of Sunnyvale, CA, Monolythic Memories, Rockwell represented by the John G. Twist Co, and EXCERT, INC, a mail order computer company I started and folded after joining Intel for a brief time in the early 1980's.  Most of my initial attempts will be to place online some information about the Rockwell AIM-65 that might be in sparse supply.  I am willing to act as a broker for people wishing to sell or acquire 6502 apparatus. 

After searching the web for specific threads about the AIM-65, I noticed that some people are attempting to resurrect these early development systems.  I functioned as a distributor, repair and parts center and system integrator for many years while these were in production and therefore have retained some knowledge, documentation, development boards from the varied vendors, and some spare parts.  I am willing to try to respond to any questions one might have regarding these items.

An initial topic that I picked up from the Dead AIM 65?? thread was that these boards were known for their inability to start up reliably.  I only noticed that specific problem on a batch of boards that came through my business in about the middle of the production run.  I found that the CPU oscillator wouldn't start up reliably unless it was slightly off-balanced in one direction.  If I remember correctly, the Rockwell engineers designed this circuit using 7404 invertors with 1.8k resistors to both sides of the crystal which was probably okay normally because the tolerance of the values usually meant one side was off value by 10% or more from the other.  But possibly they ran into a batch of resistors that was in very close tolerance to the others and those oscillators would not start up reliably as they went into a sort of quasi linear modes requiring several resets to get going.  I think I soldered a low value cap (10pf?) across R3 (3.3- a pull down resistor) or one of the 1.8k resistors to provide the necessary unbalance upon start up.  I'll try to dig back into my notes and find out exactly how I accomplished this.

Some have complained about the 2114 RAM chips going dead even while sitting on the boards in storage.  I have not run through my stash of chips and tested them so I'm curious to see if I have the same conclusion.  I have chips from several manufacturers so it might be localized to just  a vendor problem.  I have noted that some of my sample chips that were protected with black static foam rubber have had their leads rot off while sitting around in my basement over the years.  These were normally the chips with gold plated kovar leads on ceramic packages.  Again I haven't tested these to see if they still function.  I will have to mount them onto a socket and carefully solder the remaining stub to the socket pin to be able to test them!

I have many samples of MMI PROMS, Signetics & Intel chips, and CTS hybrids,   Plus I have an assortment of Rockwell chips including parts from the PPS 4 and PPS 8 families including a Universal Assemulator  & TTY using an RS 232 current loop and paper / Mylar tapes.

Some have complained about the lack of application notes AN's surrounding these families.  I will try to scan and post them as time permits or by necessity in answering inquiries!  Should I use the PDF format or just DOC files?

David E. Colglazier (DaveC)

651.429-2272  USA

4-03-2006

Another subject that I read about shared tecniques for trouble shooting the AIM-65.  An oscilloscope surely helps in determining if the crystal is doing its job by feeding the CPU and I'd encourage anyone to learn how to use one.  It is only necessary to have the CPU, the lower RAM Chips (Z2,3),  the  monitor ROM set, for the machine to come alive...you won't know it but it will still be looking for its peripherals that are coded into ROM.  At this point the 555 timer reset can be checked with a VOM or O-scope.  Once you're sure that the reset line is not stuck, you can go on to the crystal circuit.  I pull all the other socketed devices from the board at this point to get through these checks.  Once you're sure you've got a live system, you can start adding back in the I/O devices to get the bad device located.  The display would be a good start and then the keyboard so you have some interaction with the system.  Most of the time we assume an electrical component failure but more often than not, I also found mechanical/electrical failures at the interfaces.  One particularly hard troubleshooting job was an RAM error intermittently!  I would check out all the chips in different sockets and they'd all work just fine except when inserted into one address space.  All the socket pins looked fine but on careful inspection of the solder side, I noted one pin not protruding above the bubble of solder.  This pin had bent back on the underside and while making contact, it was not firmly attached to the solder or had come loose from the twisting of the board or the insertion of the RAM chips.  I repaired that socket and everything worked fine from then on.  Physical inspection of all electrical/mechanical interface can reveal the oddest problems.  No doubt, components go bad, but sometimes it can be just that a simple resoldering will fix the problem.

4-04-2006

I've added some comments to the section above and will now go on to a new tool I've been using since I returned to PC repairs.  A Techie friend, Jim Justin, turned me on to a quick way to remove RAM chips that are soldered directly to a through hole PC board.  What he said to do was to get a heat gun and if not trying to save the board, just heat up the solder side and insert a small screwdriver or similar tool under the chip and just pry it off.  Most chips have just a few pins bent to retain them while going through the solder wave machine and this worked just fine as long as I didn't want to save the PC board.  I've refined that method since and am able to do surface mount chip removal and replacement by adjusting the heat carefully to prevent trace lifting.  I've used it to replace 486 CPU's and smaller devices.  I even repaired a Pentium level CPU area when I discovered one of the inductor chips had been sheared off during handling or shipping.  I never found the component but salvaged one from a power supply area on a defective PC board.  The old solder sucker still has it's place for me but when I need to heat up a whole chip for replacement, I normally use this new method.  Careful regulation of the distance of the heat gun from the board is the critical technique.  I also acquired a heat shrink gun with a focusing tip and found it to be sufficiently hot enough to do very precise component repair work without disturbing many surrounding devices.

Another tip Jim shared with me concerned multi layer PC board troubleshooting to find hidden defective traces.  This one works but is very hard to do unless one has access to a known good board.  If you know you have a broken or intermittent internal trace, you need to locate it and reroute the conductor externally assuming this is not an overly critical high speed date line and can tolerate another and possibly longer or shorter route.  With no power applied to either PC board, you take a pad of steel wool and connect it to a VOM, hopefully one with a audible signal indicating continuity.  Then by dabbing the pad to the boards alternately, and changing the other lead to single traces, you can locate a general area where the broken trace is located.  Hopefully you won't find more than one but this can be used successively even in that event to find additional broken traces since you have a known good board.  Once the general area is located, you can lose the steel wool pad and revert to point by point continuity checks.  It is time consuming but it does work!

4-18-2006

I have been corresponding with a fellow AIM user and he had never heard of the Seawell group up in Seattle in the late 1970's so I put up a couple of scans of their offering on this site at

http://www.originalwoodworks.com/aim65/

One of their more humerous product titles was their Little Buffered Motherboard.  It is a great product and I still have two of them for using their Prommer II and their RAM/ROM expansion.  Unfortunately, I didn't retain their disk system as it was quite a luxury to have one of those back then!

12-21-2006

You will also find a narrative about the various AIM-65 schematics and what revisions were made to the PC main board at

http://www.originalwoodworks.com/aim65/aim65s~1.htm

I will try and scan the large schematics and post them when I find out how to do it and also the time!  I finally found my Commodore C64 1541 disk drive interface card for the AIM with schematic, manual, and software in PROM.  I had tucked it away along with some Auto Vector keyboard cards, a Micro Interface tape card, and some Cubit interface cards and PROM programmer.  I also plan on putting up a small area to sell and exchange AIM 65 surplus parts.


DaveC