Unhappy AIM-65

[SiriusHardware]

We await developments with interest.

I paid a visit to the North West Computer Museum (in Leigh Spinner's Mill, Leigh, a few miles WNW of Manchester) earlier this weekend.

Among the many interesting items on display is an AIM65, which I would not have recognised if this thread had not drawn my interest to it.

[River25]

Do you also supply 24V for the printer? On power-up, if the 24V is available, the AIM will print out a message. Does this happen?

[Pamphonica]

When originally received, I did connect the 24V as well and got no printout on startup. Just a quick burp from the printer on power off.

Unfortunately, I have now managed to fry the 74LS155 decode chip and have no spares! I removed the decode chips and associated ICs to test them, then installed nice sockets. But I was tired late last night and put the 74LS155 back in the wrong way round. It got v hot and is now dead! The 74LS138 tested fine as did the 74LS33 and the 7404.

I have meanwhile tested my small stock of 11 2114 RAMs on a small RAM tester from ebay. Two failed but the rest are fine so that potential fault area is probably eliminated.

I will see if I can do bus testing later this week, using the Polar B2000, using a 74155 that does not really seem to work in place of the LS155.

So until I get a new LS155, fault finding may have to wait a bit. And I may also try to get spares of the LS33 and 7412, neither of which I have "in stock"

-Jeremy

[Pamphonica]

Quck update.
My £5 (Audiojumble) Polar B2000 uP Bus Tester now "speaks" 6502. It is amazingly useful! All this fault-finding was done with that device.

• No shorts on any data or memory lines to rails or adjacent lines
• RAM all fine - full 4K
• Monitor ROM disasssembles fine
• Memory map test showed a faulty 74LS155 chip selector - replaced
• Also the 6532 RIOT chip was selecting itself onto the bus when not required - replaced
• Unfortunately the replacement RIOT chip has D7 and D6 data bits stuck in its RAM section (but continuity to the bus is fine)
• I can write characters to any display position fine so all that bit works.

The RIOT fault is the key bit. From the Monitor listing I realised that the 128 byte RAM in that chip is used all over the Monitor, to hold status words including for display character-writing and printout, so is essential. The top bits stuck low explains why the display data is never written (display write bit is bit 7, so never happens).
So I will have to find another spare RIOT (££!) and try again. At least I understand why the display is not happening.
Off to raid my piggy bank...
-Jeremy

[SiriusHardware]

Having read a few PET disc drive repair threads where the RIOTs also seem to be used I think someone has made a substitute which consists of a small daughter board which plugs in to the RIOT socket. May be worth searching for if you can't find an original working RIOT.

[Pamphonica]

I have ordered a few each from a couple of Chinese companies. Even if only a few work I will be grateful! Curiously, one RIOT had no RAM working at all, and hogged the bus when not asked to, the other seems better behaved but has lost a couple of bits in RAM. Shame. BTW I did not fry the 'LS155 - I fried a poor innocent LS138 that I had put in the wrong socket, and backwards! But the original LS155 was faulty in the end.
I feel that I am eliminating a lot of potential fault areas, and homing in on the problem. It's just a matter of time before I get this thing going!
Picture of my £5 tester attached (unplugged from the CPU socket in the photo), and the test rig for display modules.
-Jeremy

[Phil__G]

oooh the MK14E should have a couple of those displays!

[Pamphonica]

Expensive, at £20 a shot! Luckily only one of mine was faulty.

[ortek_service]

Originally Posted by SiriusHardware

Having read a few PET disc drive repair threads where the RIOTs also seem to be used I think someone has made a substitute which consists of a small daughter board which plugs in to the RIOT socket. May be worth searching for if you can't find an original working RIOT.

Dave Curran at Tynemouth Software mentions he did have some NOS 6532 RIOT IC's, a few years ago in this thread: (Where he made some Commodore RRIOT replacement adaptor boards):

http://blog.tynemouthsoftware.co.uk/...lacements.html

But it seems much of this is no longer available - although the 6530 adaptorr PCB's are still listed here: https://www.sellmyretro.com/offer/de...ment-pcb-61997
However, I can't see any 6532's listed in current shops that that required.

I did also see there that there's an in-house number for these on ones inside Atari 2600's: http://blog.tynemouthsoftware.co.uk/2018/07/
- which used to be rather-common (I've picked up several for a few pounds in total many years ago) / may be some scrap ones going cheap (would now be a shame to strip down a working one).

[ortek_service]

Originally Posted by Pamphonica

Quck update.
My £5 (Audiojumble) Polar B2000 uP Bus Tester now "speaks" 6502. It is amazingly useful! All this fault-finding was done with that device.

• No shorts on any data or memory lines to rails or adjacent lines
• RAM all fine - full 4K
• Monitor ROM disasssembles fine
• Memory map test showed a faulty 74LS155 chip selector - replaced
• Also the 6532 RIOT chip was selecting itself onto the bus when not required - replaced
• Unfortunately the replacement RIOT chip has D7 and D6 data bits stuck in its RAM section (but continuity to the bus is fine)
• I can write characters to any display position fine so all that bit works.

>>
>>
-Jeremy

Co-incidentally, I also picked-up one of these Polar Instruments B2000A uP BUS Tester with printer for a few pounds nearly 30 yrs ago. But I never used it, as it was missing any probes, and I think that was before eBay / the Internet was common. So I'll have to look into getting some for it.
I did also pick-up a Dataman MicroDoc a while later, that looks like it has similar capabilities and luckily did come with a few common CPU plug-in probe adaptor boards. But I'm still trying to recall where that went, to look at documenting what's in it, as there doesn't really seem to be any info on it on the 'net.

[Pamphonica]

Originally Posted by ortek_service

Co-incidentally, I also picked-up one of these Polar Instruments B2000A uP BUS Tester with printer for a few pounds nearly 30 yrs ago. But I never used it, as it was missing any probes, and I think that was before eBay / the Internet was common. So I'll have to look into getting some for it.
I did also pick-up a Dataman MicroDoc a while later, that looks like it has similar capabilities and luckily did come with a few common CPU plug-in probe adaptor boards. But I'm still trying to recall where that went, to look at documenting what's in it, as there doesn't really seem to be any info on it on the 'net.

If you drop me a PM with your email address I will send you all the details I have on the B2000/A. I should also be able to provide some personality modules and photos of the cables.

-Jeremy

[ortek_service]

Originally Posted by Pamphonica

If you drop me a PM with your email address I will send you all the details I have on the B2000/A. I should also be able to provide some personality modules and photos of the cables.

-Jeremy

Thanks - PM now sent.

I've just found my Polar B2000A (Not sure yet what the A-suffix version gave you extra), which I'd got at Telford 1999 Radio Rally for £1 (without any pods) and opened it up (possibly for first time), discovering a Z80 (ironic you've used it to repair a 6502 system !) + SRM2016 SRAM, 27128 EPROM (readout and attached my B2000A's version marked NP27 with 8543- date-code?), and 8255 PIA + loads of 74HC/LS373's.
I also found an infamous 3.6V 100mAh PCB-mounted NiCad - with the usual legs on some adjacent parts starting to go green. So will whip that out and maybe replace with an off-board 3xAAA holder and NiMH Low Self-Discharge cells for better memory retention if that's important.

A quick Google also found it discussed on Arcade-machine repair forums / a copy of the manual - as well as your recent post of EEVBlog about replacement pods: https://www.eevblog.com/forum/testge...up-bus-tester/

Surprisingly, Polar Instruments (Most well known for their famous 'Toneohm' PCB short-locator repair aid, I've also got some of - and first I got was also missing its Kelvin-connection wired? probes) are still around.
But they now mostly do PCB-design aids software products, and only test equipment still listed as current is an updated Toneohm + a rather-more expensive / professional-use Flying probe tester. It seems the B2000A was rather short-lived, by 1987 being discontinued (their first product to be): https://www.polarinstruments.com/sup..._warranty.html

I did also find I had a rather more obscure CIL Microsystems Ltd PCI-1281 16 channel Logic Analyser ? in the same-size case. This was also had a Z80(A), along with four! Z80A-PIO's and a Z80A-SIO for its RS232 Interface. I can't seem to find much info on this on the'net, and wonder if those v.cheap USB LogiC Analysers from China are more powerful these days.
But I've also readout the CIL PCI-1281's 2732 4K EPROM - finding Z80 Mnemonics in this! (So looks like it may have produced Disassembly output).

See attached zip of the readout-firmware for both of these, for interest / archiving for possible future repairs.
With the Dataman Microdoctor also long-discontinued. maybe there might be some interest in re-creating (an open-source, for maintainability, unlike some Arcade-machine IC testers) one, possibly using an Arduino (Mega?) / RPi controller, to assist in repairing 6502 / z80 etc. 8bit systems that won't boot-up and maybe easier to diagnose via the processor's bus to check memory map etc.

[TonyDuell]

Firstly a word of warning about 'whipping out the NiCd'. I've met many devices that charge said backup battery through a resistor from a 12V or so rail. They use the battery effectively as a shunt regualtor to keep the backup voltage at around 3.6V. This is then fed via a diode to the RAM +5V line (with another diode to supply the RAM from the normal 5V line when the thing is powered up). The result is that if the battery goes open-circuit or is removed the RAM is effectively powered from 12V which it tends not to like.

Do check the circuit and if necessary disable the charger or stick a 3V9 zener diode across the battery connections in place of the battery.

Anyway...

I have a strange logic analyser made by Zircon Insturments. It's quite slow, but has a lot of channels and I suspect was designed for educational use. The pod I have for it has 4 off 40 pin headers on it so that a straight-through ribbon cable ending in one of those DIP-Clips can be put onto a Z80, 6502, 6800 or 6809 processor and it will route the address and data pins, etc, to sensible inputs of the analyser.

As I mentioned in the Osborne restoration message, I have a full in-circuit emulator for the Z80 (only). Although it will run Z80 code, let you set breakpoints, etc, most of the time I just use it as a bus tester to access RAM memory and I/O ports in whatever I am repairing. So a more general bus tester would be of interest. I would prefer it if it was a stand-alone device with its own keypad/display or perhaps a connection for a serial terminal rather than something to hang off a USB port. There isn't room for a PC on my bench. The processors I would want to be able to handle are the Z80, 6502, 6809 and 8085.

[ortek_service]

Yes, I was intending on replacing the obsolete NiCad with a same-voltage equivalent. But have read elsewhere on here that the replacement NiMH PCB-Mounted batteries can also leak a bit and that mounting one off-board may be safer to prevent any more corrosion of the PCB and nearby components.

So I'd like to remove original NiCad fairly-soon, to prevent more damage occurring - Thanks for info about some equipment charging it from a 12V rail / using it as a regulator, which seems a rather dodgy thing to do.. I've only ever seen these being charged from the main +5V rail.

I've checked the circuit in the back of the manual, and confirmed the Polar B2000 does also do this (and has a 200R resistor in series with battery, so wouldn't shunt that much). They have actually also included a C106D thyristor over-voltage crowbar on the PSU main output - Although that's a bit over 5V, as there's a diode-drop feeding the +5Vcc rail - rather than having the crowbar directly across +5Vcc (Not sure if this could be lifted by mis-connection of the pods to a higher voltage on target system etc, which could be wise to guard against).

The Polar B2000 supported the usual Z80, 8085, 6502 & 6800 (but might be able to extend that to a 6809). I do vaguely recall the Zircon make - There were quite a few companies back then (Inc. Philips-Fluke) who made this type of thing, but these now seem to have all gone, so probably some requirement for a suitable replacement being available.

[Pamphonica]

Now if only we could get someone to make up a replica PCB! I reckon its a fine little device. The only odd component (apart from a till-roll printer) is the 14-segment 8-digit LCD display. It seems to be fed by a 4-bit data bus and a strobe, plus 5V and ground. I tried to find an alternative (in case mine collapses) but no luck so far. Does the B2000A use the same display? Maybe we could get some photos (inside and out).
Another happy coincidence is that the AIM65 and the B2000 bus tester both use the same width thermal paper. Perfect!
We may need to split this thread. It's turned into a bus-tester fanzine!

[SiriusHardware]

Just to pull back onto topic, where are you at the moment with the AIM65? Still waiting for a RIOT from somewhere?

[Pamphonica]

I’m away for a bit so will have to pick this up again in a fortnight. I have ordered some RIOT chips and they should be waiting for me on my return. Thanks all for the very helpful advice. I will post once I get back to the workbench.
-Jeremy

[ortek_service]

Originally Posted by Pamphonica

Now if only we could get someone to make up a replica PCB! I reckon its a fine little device. The only odd component (apart from a till-roll printer) is the 14-segment 8-digit LCD display. It seems to be fed by a 4-bit data bus and a strobe, plus 5V and ground. I tried to find an alternative (in case mine collapses) but no luck so far. Does the B2000A use the same display? Maybe we could get some photos (inside and out).
Another happy coincidence is that the AIM65 and the B2000 bus tester both use the same width thermal paper. Perfect!
We may need to split this thread. It's turned into a bus-tester fanzine!

Last week I did try using the 'Report' function (Being as it said this could also be used, to request re-opening of closed threads, as well as report issues with content of an actual message), to request to the Moderators that posts #30 to #35 of this thread were split to a new 'Microprocessor Bus Testers' post in this area - In order to further discuss / upload pictures of the Polar B2000(A) and other similar products, without causing this thread to diverge from what some readers might be expecting in it.

- As if I just asked via a reply here, I wasn't sure a moderator would necessarily see it (unless there's a special directive to alert them?). But I'm not sure this has worked as expected, as I can't see move of the requested messages to as new thread (unless mods are all busy atm). And I can't see a way of seeing the 'report' I did / doesn't appear in the list of all my recent postings in profile / the private messages.

So any suggestions on how to achieve this, would be welcome.

[Pamphonica]

I can finally report success!
The RIOT 6532 chips from China turned up while I was away. Listed as “pulls” they actually look new, and the legs were characteristically splayed like NOS. Anyway, I did a quick check of the RIOT 128byte ram in three of the chips, and all was fine so I put back the original processor and display driver and booted up. Perfect display and printing now! Phew!
It’s been a long journey, and if I hadn’t been able to find out from the monitor listing that the small RAM scratchpad in the RIOT is used for lots of variables, and prove that the RIOT RAM was faulty in my original chip, I would not have guessed that the RIOT was the problem all along.
I’ve probably gone a long way round to get here, but have learned a huge amount along the way.
I can’t thank you all enough for your suggestions and encouragement. Great team effort!
Off to play some more with the inbuilt Basic.
-Jeremy

[Mark1960]

I have a Rockwell 1984 Data Book that has quite a lot of Aim65 info. I found the link below to archive.org copy that you might be interested in.

https://archive.org/details/1984rdb