Jump to content
midibob

Tales from Midibob's workshop SYS85

Recommended Posts

Well, I've started on the SYS85 tech and it's all uphill at present due to the lack of tech info.

Still, even without it faults can be found and rectified which will no doubt save time later.

 

Two boards had suffered quite a bit of corrosion and it had got into the 64 pin ROM cart sockets. These are an absolute nightmare to remove but with a lot of patience it can be done. The first board I lost the top of one rivet which I thought was pretty good considering most of the joints had gone crusty. The second came off without any damage so that was a relief. Takes about 2-3 hours as even after it's removed all the holes and pads need cleaning or they don't take new solder.

 

Another board came out of reset once but never again. The reset cct on this tech (and Scorp 1) can be very annoying when it plays up as the watchdog trips the reset and the reset, resets the watchdog ....... ad infinitum.

 

Anyway, in this case the logic probe showed all the right signals going to all the right places but it wouldn't come out of reset?

So in come the big guns.... Sniffing round with the scope showed the waveform from the WG pin of the ULA chip (25) at only about 3v p2p and it should be about 5v. This is what feeds the watchdog timer and as the signal was low it wasn't enough to stop the watchdog triggering.

Checking the xtal waveform at 4MHz also only showed about 3v p2p?? All very odd. The 4Mhz signal is also generated on the ULA chip which is a custom one (TA11820).

Next thing was to re-check the 5v rail as it was good when I started. Yep, still 5v? So what's going on? That's when I saw this.

 

I love all this stuff. :D

 

post-6626-0-67307600-1467751092_thumb.jpg

 

I'm sure you'll all spot it straight away.

 

If anyone's got a game manual for SYS85 with the wiring diagrams, dip switch and test modes please upload it.  You wouldn't want to see a grown man cry!!

Share this post


Link to post
Share on other sites

Ah there s a leg out of bed id say bob !😀nice detective work there bob and full credit for the patience and skill I know it's not easy !great work as usual and.always look forward to.these threads .

Share this post


Link to post
Share on other sites

Great pic Bob for once even I spotted that after scanning everything else first lol

looks a bit bow legged :-)

Share this post


Link to post
Share on other sites

Great stuff Bob, you have got lots of patience, that's for sure.

 

Keep up the great work, very impressive! :)

  • Like 1

Share this post


Link to post
Share on other sites

midigod never stops

 

happy days :)

Share this post


Link to post
Share on other sites

Thanks for the feedback guys, it makes it all worthwhile.

 

Yeh, with that leg adrift I'm not sure if it was touching slightly hence the dropped voltage or what. Pin 40 is the 5v rail but I didn't bother to investigate further once I saw it.

 

More to follow soon......

Share this post


Link to post
Share on other sites

Thanks for the feedback guys, it makes it all worthwhile.

 

Yeh, with that leg adrift I'm not sure if it was touching slightly hence the dropped voltage or what. Pin 40 is the 5v rail but I didn't bother to investigate further once I saw it.

 

More to follow soon......

Keep them coming Bob! Love it :)

Share this post


Link to post
Share on other sites

Yeah impressive to say the least. Excellent work and it's great you take the time to log all this in posts. I have some CPU's if you need any. Let me know....

Share this post


Link to post
Share on other sites

Thanks for that. Luckily I do have a few stashed away but as yet not found any boards with faulty ones. Long may that continue :)

Share this post


Link to post
Share on other sites

First bad CPU. LOL

 

I've got quite a few of these boards to write up with various faults, some stinkers and others just good old fault finding. Before I do though I want to make sure they are fully tested and this is where I'm a bit stuck.

 

I'm in the process of building another Heath Robinson test rig and so far I'm able to emulate the reels, meters and have an alpha so I can get into attract mode. From here things get tricky. There's no information on which strobe and data inputs/outputs go to which switches or lamps, also nothing on how to get into test mode and what tests there are.

 

So far I've managed to locate the start button but not it's lamp (there's 127 to choose from!), I've also found the DEMO mode switch so can actually kick off some games although I haven't got a clue what it's doing. :lol: 

 

As no docs appear to be available I'm hoping somebody might be able to tell me what wires go to what pins on the motherboard even if it's just the buttons for now. A big ask I know but fingers crossed. :pray:

 

Cheers

  • Like 1

Share this post


Link to post
Share on other sites

Things have been moving along nicely and I've now completed 6 boards.

 

Most of the general repairs are components around the battery area, no surprise there. 5V regs going out of spec, the 2R7 dropper resistor, IC7 (74HC00), R171 (4k7) and the odd track corroded or burnt especially the 50v AC feed. I've also had a few plug Z's with corrosion in them (the 64 pin ROM cart socket) and swapping these out is a horrible job especially with the solder contaminated.

One board had some previous repairs and after removing IC1 (556) I found some of the pads were missing? With a couple of new rivets fitted it booted OK.

 

A couple of boards had track corrosion under the chips, which is a real pain. Here's one which I didn't spot originally as the chip was bad anyway. After replacing with a socket the board would stop if it was flexed even under the slightest pressure. This was IC7 (74HC00) and it was only when I removed the socket again and checked under the microscope I saw it. I don't know whether anyone can spot it but here it is. Which pin?

 

post-6626-0-29689000-1472674440_thumb.jpg

 

One board had me going a bit. The Alpha went potty when the board had been on a few minutes. A lucky guess led to squirting some freezer on IC15 (74LS241) which restored normal operation. The chip wasn't really running warm either?

I then found the Alpha went funny from cold with an associated ROM error. This proved to be IC9 (6850) as it went tits up with the freezer. 2 heat related faults on the same board, one hot one cold, you couldn't make it up.

 

The next one was a right stinker. It started off with power applied and a hot smell coming from the other side of the board from the 5V reg heatsink. IC15 (74LS241) and 16 (74LS273) were running really hot, too hot to touch in fact so they were quickly whipped out. After removal, the resistance across the 5V line was 80 Ohms and it should be around 850!

As you can imagine the 5V goes to almost every chip so I wasn't looking forward to this one.

The vbatt voltage was low (3.9v) and IC7 (74HC00) was slightly warm which is not good for an HC series! Yet another blown chip.

 

Before removing chips willy nilly a quick check round revealed the CPU was cold also there was no output from pin 25 IC6 (WG). This is the watchdog pulse and if it's not there it usually means IC6 has had it. The chip was slightly warm too and it should run cold. These are custom ULA chips so the only replacements available usually have to come off of donor boards. (Scorp 1's or Sys85) More on this later.

 

So, back at the ranch I'm tracking down this low resistance. Next suspect was IC19 and once removed I measured 350 Ohms so going in the right direction. I'm now wondering if this board has had a lightning strike or something as the silicon count is getting rather high!

The board still doesn't boot so there's obviously more chips playing up. I'm now seeing no activity on the address lines and although the data lines appear to be active they all seem to be the same low waveforms. All very odd.

 

Moving on to IC5 (AY-3-8192) and this was low resistance across the 5V.

Next chip was IC9 (6850) and this was stone cold and was found to be shunting the data lines.

Lastly IC8 (8279) was removed and bingo the resistance went up to about 1K.

 

After all parts were replaced I was greeted with nothing on switch on. To cut a long story short when the board was flexed it burst into life. Turning it over I noticed I hadn't soldered pins 1+2 of the CPU socket. That's what comes of working too many hours.

 

So into test mode to check all the functions and one button doesn't work. T36 was s/cct.

 

Two more boards are still giving me sleepless nights but I will crack them eventually so look forward to some more tales soon.

 

Back to the custom chip IC6. This has appeared in a number of guises  mostly labelled TA11820. I have also seen one with the original SYS84 numbering ie. LA05-170. Another number is BFM1071 and I have seen these listed in the large semiconductor distributors sites. Whether they are the same chip I don't know but I dread to think how much they might be and how many you'd have to buy.

Another number I came across was NEC 9036 KP but I can't find any info on that one.

 

 

Finally one last tease.

This had a reel index error, I'm sure you'll all spot it.....

 

post-6626-0-37035200-1472677903_thumb.jpg

Share this post


Link to post
Share on other sites

Brilliant work bob you know you love a challenge really 😉the dedication is outstanding the two freeze repairs were a gem.

T40 and 41 look a bit suspect are the legs touching due to being crushed ?

Share this post


Link to post
Share on other sites

I think you have taken repairs to a new level Bob!

 

Amazing :)

Share this post


Link to post
Share on other sites

The suspect pin was #1 and the break was barely noticeable (even under the microscope) so I'm not surprised nobody has commented.

 

Moving swiftly on to the last two stinkers!!

 

The first one was failing with an 'Error 11' which is the first reel not indexing.

3 transistors were found to have low output T34, 35 and 36 ZTX550. These provide the driving pulses for the reel and switch inputs. Despite changing these it still failed? I then noted one of the input pins of IC8 (8279) had gone low. Once swapped out all was fine and it booted into attract mode.

Checking the memory retention showed that credits were retained and other info after a power cycle but every time  it rebooted I had a 'check battery' error message? Further scratching of the head didn't produce any thoughts as to why this should be so I left it for a couple of hours and when I tried it again it was fine??

How does it know it hasn't got a battery (or it's flat)? There's no circuitry to measure the Vbatt voltage so I can only assume it must write something to RAM but if so why didn't it remember it when it remembered everything else? Answers on a postcard please.

 

Finally on to the last one which I almost gave up on.

This wouldn't boot farther than a 'meter error 30' and this was due to no CS output to the driver chip IC19 (74LS273).

A load of components had already been changed due to corrosion including the 64 pin ROM socket. All the solder joints were contaminated so this was a nightmare in itself to get off.

First port of call was to replace IC6 (TA11820) but this proved to be exactly the same?

 

From here on things went rapidly downhill as I couldn't fathom out why there was no output from IC6 pin 23 (meter CS pulse). Eventually the only course of action was to remove all the chips that had access to the address and data buses to see if anything was amiss. I don't usually like this scatter gun approach but I was getting desperate.

I'd already tapped out all the connections to the ROM cart plug which were good, all the data and address buses and so far nothing untoward was found.

After removing all the support chips I came across this underneath IC9 (6850).

 

post-6626-0-33904500-1473451147_thumb.jpg

 

That doesn't look good, also pin 1 fell off. Still, after a bit of cleaning.....

 

post-6626-0-69662700-1473451146_thumb.jpg

 

Thinking this was it, another 6850 made no difference.  At this point I was about to give up.

 

After this my fault finding technique went right out the window as I was replacing the chips one at a time but not keeping all the new ones in. After a few more days of blood, sweat and tears sanity was restored and with a complete new set in it booted. Then came the faulting process but this time working backwards.

IC6 was indeed faulty so was IC8 (8279) and IC9 (6850), yep the spare was knackered too!

 

After all that I was hoping a quick function test and wrap it up. Nope.

Some of the lamps weren't working and one Data line was missing and a couple of others were low as well as one strobe line.

Four of the power transistors all had fractured legs T28, 30, 31 and 32 (BF64) also TR52 ZTX550 was o/cct.

Now all the lamps were working so the final check was the payslides and guess what, one didn't fire. Give me strength!

 

This proved to be T9 which was getting rather hot and it's associated 22R resistor which was open circuit. Now T9 is a FST254 and luckily I had a donor board with one on as I haven't been able to find these. The later Scorp 1 used a ZTX653 so I reckon that's a good substitute.

 

Sys83 coming soon.

Share this post


Link to post
Share on other sites

Brilliant work bob and that last board sounded like it nearly beat you .they say sys85 is a unreliable tech but I'm. Thinking this board will now be nearly new with what you have had to replace .sys83 sounds good to me .

Share this post


Link to post
Share on other sites

Another Sys85, still on the operating table!

709777635_P1030895(Custom).jpeg.5b0dd7a2e00f74e7009519e82260f7d3.jpeg

 

This one has been fun and games so I thought it was worth a mention.
It came in as a no boot.
The 7805 regulator legs were a nice green colour and the power dropper resistor had lost all it's body due to rot! The 74HC00 was also rotten so these were all replaced.

On fire up I was greeted with nothing, which wasn't surprising as it wouldn't come out of reset. I have a particular hate for the reset/watchdog circuit in this tech (same as Scorp 1) as it uses the custom ULA chip.

The watchdog is reset by an output pulse from pin 25 and sometimes I would see an output and other times nothing so the reset would then operate. Forcing the reset high still didn't produce the watchdog pulse so things weren't well somewhere?

A quick check round showed the DATA lines were being pulled down by about 1.5 volts, not a lot but I suspected enough to cause a problem. After putting the CPU in a socket so I could isolate one of the data lines full voltage was noted so I knew I was on the right track. Now it was a case of isolating D0 on every connected device until I found the one responsible. I eventually got round to the RAM and, bingo. With a new RAM chip in all the waveforms looked good but there was still no boot as it wouldn't come out of reset. Luckily I had one of the custom ULA chips, which I knew was good, so that went in and lo and behold I got a constant o/p from pin 25. It then came out of reset.... and then did absolutely nothing!:headache:
After wasting time swapping out the 6850 I just happened to look at the ROM cart (my home made one) and noticed that two of the tracks on it were damaged! After fixing these it burst into life.

As I'd already swapped out the CPU, Custom ULA and the 6850 I thought I'd put the originals back one by one. Blow me, they all worked! You couldn't make it up.

Now I had a working system the reels wouldn't index (my Arduino reel faker that is) and I kept getting error 11. This turned out to be T40 (ZTX550) which had gone s/cct.

So far I've checked all the triacs fire and all the lamps run so I'm almost there.

Still got some funnies to sort out. Firstly it always greets me with a 'Check battery' error message on boot up. I know the RAM is being powered by the battery OK and being as there's no voltage sense circuitry this error can only be caused by something not being written to RAM or not being stored. As I know the RAM is OK it has to be the first option.

The other issue is the user buttons and coin switches aren't behaving correctly, haven't got my head around what is occuring as yet but more head scratching will surely follow.

Will I get it sorted before 2019?:lol:

Share this post


Link to post
Share on other sites

Wow ..thats some work grate to see thank you for shareing 

Share this post


Link to post
Share on other sites

Andrew, yes but you'll need a slight mod.  I'll send you a PM.:)

Well it's amazing what a good nights sleep does for you.
Back on the case this morning I found there were some dodgy joints on my scratch test rig, one of them was the door switch and of course the software needs to know the door is closed before it will write to RAM. As soon as that was sorted, no more 'Check battery' errors.:)

The next one on the bench is all lamps going out after a short while. It hasn't made an appearance yet but I'm hoping it will. Watch this space.

  • Like 1

Share this post


Link to post
Share on other sites

×