Sharp MZ-80K

[Gary C]

Might as well start a thread on my latest machine

Bought a Sharp MZ-80K 5 months ago, and because of this virus thing, only just managed to pick it up.



Love the details like the hinged lid with its support which even has a little screw to lock it in place



and the labelled monitor controls inside the case



Unfortunately it doesn't work. No output from the PSU. Strangely there is no mains fuse accessible from outside, its inside the PSU case and had blown big style. The input windings of the transformer were fine, so first suspect are the tantalum capacitors. Lo and behold one of the four is short circuit so they are all replaced.

Just waiting to get my hands on a slow blow 500mA 20mm fuse. Might as well use the correct one.

 

[Gary C]

New & correct spec fuse fitted and PSU now powers up. 12V is a little high but probably poor open cct regulation so plugged into the board, switched on and...

Greeted by "** MONITOR SP-1002" yay !

No tapes to boot from, but am very impressed that the drive belt looks brand new and all tape functions seem to work fine. Most of the chips have a date code of the end of 1980 so that makes it the second oldest machine I have after the PET.

I also got the expansion unit at the same time with a floppy interface card and a printer card. One issue with the unit is, its missing its door. Anyone got a spare, or is it time to get the tin snips out

 

[Atomtan]

I am in the process of restoring an 80k also. I thought I had reached the same point as you, but struggled to pull down a copy of 5025 Basic from the internet via a car tape converter. After some help from the people on the Sharp Facebook group I figured out that I I probably have dodgy RAM somewhere as the machine 'crashes' after a period. I managed to get an original 5510 basic tape, but that crashes at loading which again points to the RAM. Its a quirky machine, and doesn't have much of a following, but there is something about it really like and it makes my owned from new PET 4016 look pretty bloated. If you need a tape copy of 5510, let me know (I believe its partly backward compatible, though some features don't work).

 

[Gary C]

Cheers

I am going to try to get the floppy adapter working with a spare Gotek.

First though, I am overhauling the keyboard. Most of the keys have poor resistance when pressed, so I have disassembled it and removed each switch.



Dunked in IPA and operated several times seems to have cleaned up the contacts very well (the IPA was filthy afterwards)

One unusual problem, in the past the machine must have been used with a program that used the symbol keys as a numeric keypad and came with labels to go under the key tops. I have removed the labels, but the owner must have had more than one machine to change because two of the symbol key caps are exactly the same.



Will have to make a new one.

 

[Gary C]

Oh I love the smell of burning RIFA's in the morning.

Powered up the expansion box to the crackle and smoke of a gently combusting RIFA . I know I could have replaced them by rote but, ah well. Lots of spares in the drawer and they generally don't cause too much of a problem if you don't let them combust too much.

 

[Gary C]

Noo, no caps in the drawers, all used up.

RS to the rescue, glad I still get free next day delivery as it makes ordering things like this stupidly expensive from Farnell.
I seem to have ended up with a business account somehow and almost feel guilty when ordering

Kitchen (adjacent to garage) still has a faint lingering RIFA smell. Good job the wife's sense of smell is so poor.

 

[Atomtan]

Alway love the Robert Duvall quote, and once you have faced a smoking RIFA, you never forget the smell....my commodore pet one was the most unusual I have seen, its huge and slowly oozed goo as the smell gently increased, yet in continued to work until critical ooze mass, then pop and smoke...

I would be fascinated to hear how you get on with the Gotek as the Sharp seems to have the smallest following out there, and still seems stuck with the car tape adapter as the recommended way to get software across.

I need to get onto messing with the RAM on the 80k but I picked up and 80a for £10 from the local house clearance and this jumped the queue and is now fully working (and repainted which I guess some would see as a travesty).

Keep throwing the odd update on here and best of luck with the keyboard (I have only ever done one keyboard on an Apple II and vowed to avoid doing one again unless I have too).

 

[Gary C]

I/O unit reassembled and all powered up. The FD command responds with the BOOT DISK ? message so its reading the ROM on the disk interface board, but thats as far as it goes.

It has a T3444M microcontroller rather than a traditional disk interface and I cannot find any information on it at all, and given its a PROM type microcontroller, if its toast I dont see me getting a replacement.

So, the interface box itself works up to the ROM and I have not given up on the card yet but does anyone have anyone have any info on the card/microcontroller ?

 

[Gary C]

Found one

https://original.sharpmz.org/mz-80b/download/sm80fd.pdf

Looks like the motor on signal is produced by a 2 to 4 line decoder 74S139 then latched in a 74S174. What is really odd, the latch for the motor on and drive select functions is latched by a IO read of the port latching the address lines rather than a write, which explains the use of an IN command in the part of the code to turn it on, bizzare.

Anyway, I dont seem to have a motor on signal which may mean the problem is with other (replaceable) chips on the board. At least I have a starting point now.

 

[Gary C]

Ah ha !

The I/O request signal from the bus IORQB is present at the edge connector but not at IC1, corroded track under the IC. Chip removed and is fine, just need to remake the track. Rest of the tracks look ok, just this odd corrosion around the one chip.

 

[Gary C]

Next step

Repaired the corroded track at the pad of IC1 and the signal from the bus is now present at the chip, all looked good but still no response from the drive. Looking at the program in the ROM for the disk interface (the ASM file is available https://original.sharpmz.org/mz-80k/dldrom.htm) it applies motor on then waits for RDY before doing anything else. RDY is developed from the index pulse and that is missing !

For first testing, I am using a TEAC FD55A as that should remove any Gotek problems. Checked at the floppy itself, and still no index pulse, then Doh, the disk is in the drive upside down. Put in the right way, index pulse at the chip and the head now loads and steps to track 0.

Looks like the interface card now works, so the next challenge is getting disk images onto either an actual floppy or the Gotek.

 

[Gary C]

And another step forward.

Got a tape adapter of Amazon (https://www.amazon.co.uk/gp/product/B001AZKRQ6/ref=ppx_yo_dt_b_asin_title_o01_s00?ie=UTF8&psc=1). The cable runs nicely down inside the machine from the tape player and has plenty of lead to connect to the PC.

Disk basic loads up from a WAV file nicely and the DIR command accesses the connected disk (but obviously gives an error on the none MZ80K disk).

No format command ?, what was it with machines of this era that they couldn't put the format command in the system ? Anyway, loaded another WAV file called masterboot and that then read the disk basic WAV file from the tape and attempted to write it to the disk. Everything looked good until the end when it threw up an error.

Loading back disk basic, a DIR command now gives a disk header so it has written to the disk, but I suspect its because my FD55A is a single sided drive. I thought the MZ80K were double but treated each side as an individual disk (similar to the BBC) but I'm not sure.

Anyway, moving forward

 

[Gary C]

If anyone with an MZ-80K could photograph this key

without any glare if possible.

I can't find a decent quality one online and with a good photo I might be then able to recreate the missing ke ytop

 

[Gary C]

Connected up a spare Mitsui drive (80 track but fine for testing) and managed to write a small program to a bootable master disks from a WAV file, but a bigger file fails, think I am getting write errors, but it works.

Connected up the Gotek using an IMG.CFG to define the disk geometry, with a dummy file of big enough size to write over and it works !!!

Next plan is to make up a PCB that fits in the interface chassis that I can mount the Gotek on and pull power from the interface unit.

Sill need a hires image of the missing key top if anyone can help.

File I used for IMG.CFG

[default]
cyls = 35
heads = 2
secs = 16
bps = 128
id = 1
mode = fm
interleave = 1
cskew = 0
hskew = 0
rpm = 300
gap3 = 0
iam = yes
rate = 125
file-layout = sequential

File used for FF.CFG

interface = shugart
host = unspecified
pin02 = nc
pin34 = nc
write-protect = no
side-select-glitch-filter = 0
track-change = realtime
index-suppression = no
head-settle-ms = 50
Followed by the normal stuff for image selection and display control.

 

[inotarobot]

Hi Gary, your doing great. Interesting read thank you for posting so far. Those caps can be a pain !!! Seems like you got on top of it.

And yes the smell does linger for quite a while.

Will be interesting to hear next step.

 

[Gary C]

Hi

Cheers,

I tend to write stuff as it helps the thinking process, but nice to know someone is reading.

Just ordered a 8mm pushbutton that I am going to hide in an existing hole just under the front of the machine and use it as a reset button to reduce the stress on the PSU and the mains switch. There is a connection on the board for a switch already. Might use push in connectors to leave the system totally unmodified.

I really need to decide on how to mount the Gotek. I really would like an original drive case but I am realising how lucky I was to get the I/O unit so thats probably not going to happen anytime soon. My I/O is missing the side panel, so I might knock up one with the Gotek mounted in it as an all in one solution.

 

[Gary C]

New door for interface unit crafted



And fitted



Just need to find out how to match the paint.

Also fitted a neat low profile reset switch in an existing hole just under the front edge to save the PSU being cycled too often. Cant be seen and didn't need anything drilling.



And thats the next issue. I noted that the +12V no load voltage was a bit high at ~15v but when attached it falls to ~10v. Also when I press play on the tape player, the screen momentarily shrinks suggesting that the 12v regulation is a bit poor. It uses a SG3524 to drive a darlington pair to regulate the supply and as far as I can see this is well outside of the spec of the device but as it uses external components to set various parameters, it could be anything (though probably a capacitor)

Anyone trouble shot one of these ?

 

[Gary C]

Still looking for a missing high quality photo of the missing key top if anyone has an MZ80K

But a new question.

The Argo emblem and power light plate on the case seems to be a separate bit, has anyone removed it to allow the case to be painted ?

My case is a bit battered and needs a spray, but that bit is fine and it would be a pain to mask it off so before I start prying, has anyone else removed one ?

And if you have the paint codes too

 

[Gary C]

Ah HA !

The regulator is a simple affair. So much easier than the Apple implementation with its flyback gubbins and high voltages. There is a simple step down transformer then a bridge + smoothing to give a nice ~30v DC to the rest of the board with three DC regulators.

The 12V and +5V regulators simply compare a desired voltage against what is being delivered to the machine (suitably ranged by a resistor ladder) then turns the DC supply on and off, varying the mark/space ratio to control the voltage.

Key to all this is the output capacitor and inductors, when 30V is applied, the LC network soaks this up and when the supply is off, it continues to deliver, smoothing the cycling DC to an average DC of the correct voltage and just as importantly, stops the voltage changes being faster than the regulator can cope with.

Looking at my supply, as soon as the IC fires the darlington pair, the voltage climbs too fast then falls back until fired again. Rather than a steady DC with a few blips, I'm getting a big step up and a die back, then a big step. The IC cant control it. Whipped out the electrolytic and put it in my tester and nothing. Not even a pF.

Looks a bit manky too, slightly bulged at the bottom. Have a set of electrolytics on order so, fingers crossed that should sort it.

 

[Atomtan]

Gary, really enjoying watching your work here. If you have not received it already, I will dig my 80k out in the next day or so and try and shoot you a photo of the key.