• Please review our updated Terms and Rules here

Help with Olivetti M24 - AT&T 6300

Z

zinamo

Member
Joined
Aug 14, 2014
Messages
31
Location
Italy
Hello everyone,
I have recently acquired an old Olivetti M24 that was sold in the U.S. as the AT&T PC 6300 but the problem is that it cannot boot... I could not establish if it is working or not because I do not have the monitor and I am still waiting for the keyboard..
To have at least the video, I have built the adapter described in this page: http://www.olivettim24.hadesnet.org/doc.html but when powering I have nothing on the monitor, no beeps or floppy activity..

What the more, the computer came without the bus converter (this one: http://mastodonpc.tripod.com/personal/images/m24riser.jpg) but a small 90-degrees riser is plugged into the video-converter card to connect the mfm controller (never heared of this solution).

When testing I removed the hdd and controller to avoid damage and when powered up the psu makes a "capacitor noise" that disappears when connecting something (old useless ide hdd) to the molex where the hdd was connected; a part from that I checked with a volt-meter and the PSU produces the correct voltages when powered..

Coming to the motheroard firstly I checked the backup battery and it has leaked but not much and nothing looks corroded (I checked both sides of the mobo) except the pins of the power led that are close to the battery.. The only thing that I could assess is that the power led is dead..I mean it doesn't work but I have replaced it and the new led lights correctly when the system is powered..

I also have noticed that if the system stays powered for some minutes (5 to 10) the cpu and some chipson the video card become warm, not really hot, just warm..

So anyone has an idea how should I check / fix my system? I really hope to find a solution because I really wanted such a system in my collection (it's a kind of local computer as it was built in Scarmagno (Italy) less that 100 km from where I live) but nowdays are becoming rare and a little too expensive for my student budget..

oh, I also have this: https://www.youtube.com/watch?v=3iQUlO5Oocc&list=PL15B-32H5GlL2MaHAbSzq5p802TPj46qI&index=72 it is a video that shows how these machines were built in mid '80s! sorry but the speech is in italian..if someone is really interested I can provide the translation..
 
Suspiciously I have now owned two of the same machine (an M24 originally and later a PC6300) where even minimal leakage of the battery caused the machine to basically drop dead and never POST again. Poking at everything with a logic analyzer found the presence of a clock but no attempt to start the POST process. My assumption is that there is a component that is always damaged or the PCB sucks up the leakage and internally destroys itself. When you do get a keyboard the machine should initialize the keyboard controller before beeping and displaying anything on the screen. When running these machines make perfect Windows 1.01 boxes.
 
zinamo said:
What the more, the computer came without the bus converter (this one: http://mastodonpc.tripod.com/personal/images/m24riser.jpg) but a small 90-degrees riser is plugged into the video-converter card to connect the mfm controller (never heared of this solution).
Click to expand...

Neither have I, and I've had a lot of experience with these machines (expanding, repairing, etc.). Can you take some pictures of what the inside looks like? That video card is not standard to the 6300, does it have any specific markings on it?

So anyone has an idea how should I check / fix my system? I really hope to find a solution because I really wanted such a system in my collection (it's a kind of local computer as it was built in Scarmagno (Italy) less that 100 km from where I live) but nowdays are becoming rare and a little too expensive for my student budget..
Click to expand...

If this were a stock 6300 system, I would tell you to connect the monitor and the keyboard, and power it on. When the POST is in progress, the keyboard LEDs flash (as seen in this video), so even without a functioning monitor you can see if the BIOS and CPU are intact. But because the inside of your 6300 seems different/modified, I have no idea on what you should try next...

You may want to try the AT&T 6300 Service Manual which should be at the Downloads section of the 6300 Shrine. As others have warned you, however, the battery leakage is a bad sign. You might have to test many connections to see what else is damaged.

NeXT said:
When running these machines make perfect Windows 1.01 boxes
Click to expand...

If you run them with the 6300 640x400x2 driver, I agree.
 
NeXT said:
Suspiciously I have now owned two of the same machine (an M24 originally and later a PC6300) where even minimal leakage of the battery caused the machine to basically drop dead and never POST again... My assumption is that there is a component that is always damaged or the PCB sucks up the leakage and internally destroys itself...
Click to expand...
I've had similar experiences with this machine as well.
 
Hello everyone and thanks for your replies... I will wait for the keyboard hoping it arrives soon (by the way they are quite rare so it was a real pain to find one), in the mean time I have tested with the voltmeter and the keyboard connector outputs the 12 V needed so hope dies last :)
Anyway it is possible that the leackage has killed the motherboard internally because from outside looks absolutely perfect and no component looks damaged (exception for the above described led).. is there any way to asses if it's the mobo dead or the video card? As anyone tried building the adapter mentioned above?

By the way, I have also found this: http://issuu.com/adpware/docs/mc063?e=1322916/3135827 from page 79 it's about Olivetti's HDD, it's in Italian but again if someone is interested..
 
zinamo said:
is there any way to asses if it's the mobo dead or the video card?
Click to expand...

The service manual, section 5, is about troubleshooting when you don't have a functional POST. Download that (URL in my last post) and see if anything there helps.

ChuckG on this forum has built the adapter and verified it works. But don't worry about that yet; get the keyboard and see if the lights blink when it turns on. If so, that will be hopeful.
 
well actually the video card looks like the ones commonly used in these machines (as far as I know) but there is something like this: http://www.ebay.it/itm/AAEON-PCM-10...780?pt=LH_DefaultDomain_0&hash=item4aca5a714c (but with the 8 bit isa slot) plugged into it; then the hdd controller is plugged on the adapter.. when I get back home anyway I will take a picture and post it! Probably someone for unknown reasons removed the original bus converter..

What the more I have also found a spare motherboard that is sold for about 45€ shipping from USA included but I don't know if it is wotrh getting it because maybe also the video card is dead an I would also need the above mentioned bus converter to expand the computer with some basic features but those are very rare and costly so maybe in case the modo is dead I should get a whole new machine.

Anyway this evening I will have some spare time so I could look at the manual you mentioned to find if there is something usefull..
 
I have 2 spare bus converters and 2 spare video cards that will fit this machine. I hope that I will never need all of them. See photos below. They are a little old and dirty, but were pulled from working machines around 20 years ago. I have both a working AT&T 6300 and XEROX 6060 system. I could easily test the video cards in my computers and ensure that they still work, but am not sure if the machine would give an error if a bus converter was bad as it will obviously POST without one. I know that you said these are rare and costly, but perhaps we can make a deal that works within your budget? a trade for some trivial souvenir from Italy perhaps? Feel free to PM me.

PA150008.jpgPA150005.jpgPA150001.jpgPA150002.jpg
 
The M24 without bus converter is quite common as a kind of low cost version. I have such one and it uses an angle riser for the HDD as you describe. I have another one with bus converter. But I only have one working power supply, the 2nd is dead. But both of my M24 share one same problem, I can't read/write any floppy disk, that is where I have to investigate in the next time. I am in contact with a nice guy from Italy which collects just M24, AT$T 6300 and Xerox 6060, he seems to have plenty of them.

Once one of my M24 is fully operational I will investigate into the Olivetti Keyboard 2 which came with some of the M24. I have two of them, but different versions. The older one is having the 9 pin connector for Olivetti keyboard mouse, but I don't have such a mouse, they are quite rare. But from service manual I couls see the pin assignments and if I look the signal names, that reminds me on Commodore Amiga, Atari ST and Schneider Euro-PC mouse! It should be possible to connecgt such a Atari/Amiga mouse to the M24 keyboard! I will discuss this in detail as soon as I was able to test this.
 
While the actual AT&T 6300-branded mouse is rare, the Logitech model it was OEM'd from is slightly more common; you should look for that. I demonstrate the logitech model in this video and it works the same as my 6300-branded mouse. I believe the video is sharp enough to read the model number, but if not, let me know and I can dig it out to look it up.
 
The Olivetti/AT&T keyboard mouse does not talk RS232 serial or PS/2 protocol. Like on Commodore or Atari it has two x- and y-signals directly comming from the light sensors at the stroboscope wheels and two/three mouse button signals. The Keyboars 2's mouse connector also supplies 5V. That is very similiar to Atari and Commodore mice. Just the pin assignment is different, so a simple adapter should do the job. But if someone would offer me the original Olivetti mouse I also wouldn't be unhappy.
 
Hello guys,
I have received the keyboard (Olivetti keyboard 2) and when starting the computer, both leds stay still for about 1 sec. then they starts blinking fast (like it is shown in the video linked previously) but they never stop blinkng and still no video or beep sound... I think the motherboard may have suffered corrosion..

I have also found this link: http://www.win.tue.nl/~aeb/linux/kbd/scancodes-2.html if you scroll down a bit you will find a lot of infos about the M24's keyboard and mouse port; what the more, about this I have to say that on my keyboard the mouse port was hidden behind a removable plastic cover, so I think all the keyboard may have such interface.. and another thing, will it still be possible to connect a serial mouse right?

For those who asked, I have taken some pictures of my actual machine and of the angular riser, see below..

@ Trixter, so the video is yours? let me say it is very nice and complete!

@ 1ST1 may i ask you to PM me the contact of the person you mentioned?

2014-10-24 15.26.35.jpg 2014-10-24 15.27.17.jpg
 
Video was mine, yes. Glad you enjoyed it.

Thanks for the pictures -- I have never seen a 6300 without a bus converter board before! That seriously limits the usefulness of the system; only one "slot"...
 
I totally agree with your point.. luckly I have found PCFreek that has two bus coverters so we are trading for one :)
 
I have received the keyboard (Olivetti keyboard 2) and when starting the computer, both leds stay still for about 1 sec. then they starts blinking fast (like it is shown in the video linked previously) but they never stop blinkng and still no video or beep sound... I think the motherboard may have suffered corrosion..
Click to expand...
Not looking good.... The blinking should stop when the keyboard controller is programmed and the system is initialized, immediately followed by a beep and POST on a display. :(
 
Here you can sownload the complete service manual for the M24. http://bitsavers.informatik.uni-stu...livetti_M21_M24_Theory_of_Operation_Nov84.pdf It contains complete functional description, also for POST diagnostics, it might help to fix the problem. First check if the dip switches are set according to your current configuration, specially for memory size and FPU presence.

Aditionally I found somewhere this:

Olivetti M21/M24 (AT&T 6300)
The M24 went to the US as the AT&T 6300. It had an 8086, so was faster than the PC, albeit
difficult to work on.
POST codes are sent to 378 (LPT1). If a fatal error occurs, it performs more initialization of DMA
and interrupt controller circuits, tries to display an error message, complements the lower 6 bits of
the POST code, sends the result to port 378, and halts the CPU, so numbers will flicker on the
POST display with bit 6 on and the lower bits running from 0 upward. The codes start at 40 because
a black box was used to monitor POST status at the parallel port. Bit 6 was set true (to a 1) to alert
the box that the POST was starting.
Code Meaning
40 CPU flags and register test failed (fatal)
41 BIOS ROM checksum test failed (fatal)
42 Disable pdma controller command and test 8253 timer channel 1, mode 2, refresh counter
(fatal); display sub-error code of 1 if interval is below window, 2 if above, and 3 if timer
does not respond.
43 8237 DMA controller test failed (fatal)?amaster clear the controller, set the mask register,
read the control registers, test all 8 read/writeable channel registers. Test registers 0-3
DMA address and count with FFFF then 0000.Set up channel 0 for 64K RAM address
refresh. Set up memory-to-I/O transfer, unmask the RAM refresh, and let refresh begin for
the first time. Set up the 8253 for proper refresh count. Test for unexpected DMA request
(suberror 3), and init DMA channel 1 (not used), 2 (floppy), 3 (display), and init nibble
latches. Check for proper DMA transfer into lowest 64K bank of RAM (suberror 4 if parity
error).
44 8259 PIC test failed (halt)?ainitialize stack to lower 64K RAM area just tested, init and
disable 8259A, set up interrupt vectors in RAM, set up software then hardware diagnostic
interrupt vectors, test software interrupts, then hardware interrupts. Disable interrupts via
8259 mask register, check for hot interrupts, convert hot mask to IRQ number, save any
error code, install interrupt vectors, initialize video, and display any error messages (H:#,
where # is the hot IRQ#).
45 Install real interrupt vectors, determine system configuration from switches, and initialize
video mono and color. Set video mode 3, clear the screen, and display any passing error
messages for CPU, ROM, DMA, or interrupt controller. Size and clear RAM at every 64K
bank past the lowest 64K, displaying the tested RAM as test progresses. Display any error
in form cc:y000:zzz:wwww:rrrr, where cc is the config number, y the failing segment, z the
offset, w the written data and r the read data. Test the MM58174 clock calendar, and
display message if fails Test 8253 real time clock count capability, and tone generator.
Display any error, and halt if failure.
48 Send beep to display and initialize all basic hardware. Init 8041 keyboard controller,
determine parallel port configurations and test their registers, determine serial 8250 and
Z8530 configurations, check for game card, set up interrupt controller, set all 4 Z8530
serial controllers to 9600 baud, no parity, 1 stop and 8 data. Set up interrupt vectors,
initialize RAM variables, clear the screen, initialize the hard disk controller, test for and
initialize option ROMs, verify ROM checksums okay, initialize floppy disk controller, allow
user to select alternate Z8000 processor if installed and perform INT 19 cold boot.
Click to expand...

So you need a device connected to the printer port which decodes the 8 bits to BCD (74ls47, 74ls48 or 74ls247) and then displays them on two 7 segemnt LED display. Or just 8 LEDs and convert binary to decimal (or hex?) manually.
 
Hello guys, I have some updates.. As the original mobo was doomed I managed to get another one. I also had the original bus converter; for this particular part I want to say thanks to the user PCFreek that not only provided me the part, but he was also very kind in forwarding me the mobo (that came from U.S.A. too) togheter with the bus converter so I could save a lot on the shipping cost.. Really thanks PCFreek!!

Coming to the machine, replacing the mobo and intalling the bus converter brought the machine back to life so I could finally hear the boot beep! I had some issues with the video card: I built the famous VGA adapter but even if the monitor led turned green (meaning signal received) I could not see anything on it.. the lcd didn't even lit.. so I tried messing with the mobo switches and the video card jumpers; at this point as I was quite tired and even a little desperate about the video card beeing dead too, so I wrongly shorted pins 1-4 (see picture below)..

gggggg.jpg

After this move, the monitor led blinked orange and red (meaning troubles).. Only then I realized that the problem could have been the monitor compatibility.. so I was forced to disable the integrated video card and use an external VGA.. to do so, since I am on 1.43 bios, I set SW1-5 and SW1-6 to "ON" and using an 8-bit paradise VGA, I eventually had video signal!

Now the machine appears to be working I even managed to low level format the Olivetti made Lexikon HD 670 20Mb hard drive and install PC-Dos 3.3 on it! There still are many things to fix like the 5.25" floppy that doesn't read anything (hopefully a good clean will give the fix) and the computer case that is not in bad shape but shows it's age.. One more thing.. I was thinking in using a 3.5" 1.44mb floppy unit on this machine (still thinking on a smart location to place it) with the method described here: http://www.minuszerodegrees.net/transfer/35_inch/2m-xbios.htm but does anyone knows if it is possible to disable the integrated floppy controller? I have looked on the service manual but I couldn't find anything.. Otherwise the machine will stay with a 3.5" 720kb unit!
 
Well, from the schematics, it seems that if you defeat pin 17 (~765CS) on IC 6E on the motherboard, that will defeat the floppy function. If 6E is in a socket, that's easy--just bend pin 17 slightly so that it doesn't enter its socket opening.

I haven't tried it myself, but it should work.
 
You must log in or register to reply here.
Back
Top