X68000 Original Model - Caps? Leaks? Psu? Oscillator? Half working model!(FIXED)

[deathx]

Hi there,
I have an original x68k grey model but i'm having issues to giving him life again
I bought it as junk and when i receive it i turned it on: Fan spinning like crazy, The power button was static green and the timer button was static red. No highreso led and no leds on the drives. The Soft button wasn't working.
I thought was by the old psu, i changed it with an ATX using the 74LS04 inverter. All the wires were jointed togheter (es: all the red ones in the +5V, all the white ones in the +5VSB and so on).
I turned it on but i got the same exact result as before, meaning that the old PSU could be probably ok at first. Nevermind.

I was aware about the battery leak so i've opened it up and just as expected the nearby area had leak. Not that much, but pins and solder points were greenish.
I desoldered and cleaned the resistor array on the left of the battery, 74132 TTL, that little 8 pin TTL and changed all of the 5 caps nearby. I reflowed the oscillator but i did keep it in place.

After doing so, the soft button got back to work and i was able to turn it off, the red timer led is still on but i can turn on and off the green power led (fans always spin).
The hireso led and floppy drives still not turning on at all. I was able to pun in and pull out some floppy, drives seems active somehow.


I was thinking to change all of the other caps on the main board, and i have bought from digikey a new programmed oscillator at 69.55199mhz

Is anything else i'm missing here? (other than change the oscillator and full recap)
I'm afraid i'm not going to the right direction since i'm not seeing much improves here.
Can you give me some hints if i'm missing something?
Excuse my bad english
thanks

[kamiboy]

No, that is about as much as you can do. I did the same things myself, but in the end all I got was a fancy paperweight. The original model seems to be quite fragile, and the battery being on the mainboard, in the vicinity of many important components does not help things.

[deathx]

No, that is about as much as you can do. I did the same things myself, but in the end all I got was a fancy paperweight. The original model seems to be quite fragile, and the battery being on the mainboard, in the vicinity of many important components does not help things.

I see.. so after the full recap and oscillator, if is not going to work, there's nothing to do else? even if i complitely change all the components nearby the leak?
I'll wait for leon to let me know something, i know he succed in fixing a few of them, hopefully he can give me some advices here and there.
thanks!

[kamiboy]

Do share any repair tips that you get here. I might give my own one last try before giving it up for lost.

[vanpeebles]

I'd try the easy stuff first, try an ATX PSU, and change the caps on the mobo

[deathx]

Do share any repair tips that you get here. I might give my own one last try before giving it up for lost.

Don't give up on it! I'm pretty sure these can be repaired anyway, but it's all about individual skills and knowledge. Luckly for me i did many repairs in the past and i'm used to repair jamma arcade boards in a basic level.

I'd like to know, what are your led status? I've never saw a working one booting up so i don't know the led "sequence".. do you have the same as i have?

[kamiboy]

My machine is in pieces right now, so I cannot check. But I believe I get green power led, not sure about anything else. No sync signal from the video out either.

[leonk]

How do you show video out?  I made my own custom X68K to VGA cable which I plug into my PC.  Even if the video is not working, I know when the unit is on / sync as VGA monitor either turns off or stays on.

Back to your original issue.  I find the OG X68K to actually be must easier to "Save" then something like a compact.  All the capacitors on the motherboard are through hole and tend to still test good!  The PSU and SRAM battery are the failure points.  Unlike any other unit, the battery being that close to the crystal is the main failure point.  Even applying some heat to the original crystal kills it (to desolder it).

I strongly suggest you replace the crystal!  I assume you ordered CPPT1-HT0PT-ND (data sheet here: http://www.cardinalxtal.com/uploads/files/cpp.pdf)

As you can see, pin 1 is CONTROL / Output enabled.  On X68K, it's ground.  You need to clip this pin on the replacement part, and isolate the hole on the PCB with electrical tape.  Apart from that, just cleaning everything around the PCB and you should be good.

You can see a picture of I installed it, along with the ML2032 battery here:  https://twitter.com/leonkiriliuk/status/936385790317342721

Notice I also replaced the 2 IC's closest to the battery.  They are now in sockets.  The smaller IC is 5V sense.  If it detects voltage drop to 3V or less, it resets the system!

[deathx]

How do you show video out?  I made my own custom X68K to VGA cable which I plug into my PC.  Even if the video is not working, I know when the unit is on / sync as VGA monitor either turns off or stays on.

Back to your original issue.  I find the OG X68K to actually be must easier to "Save" then something like a compact.  All the capacitors on the motherboard are through hole and tend to still test good!  The PSU and SRAM battery are the failure points.  Unlike any other unit, the battery being that close to the crystal is the main failure point.  Even applying some heat to the original crystal kills it (to desolder it).

I strongly suggest you replace the crystal!  I assume you ordered CPPT1-HT0PT-ND (data sheet here: http://www.cardinalxtal.com/uploads/files/cpp.pdf)

As you can see, pin 1 is CONTROL / Output enabled.  On X68K, it's ground.  You need to clip this pin on the replacement part, and isolate the hole on the PCB with electrical tape.  Apart from that, just cleaning everything around the PCB and you should be good.

You can see a picture of I installed it, along with the ML2032 battery here:  https://twitter.com/leonkiriliuk/status/936385790317342721

Notice I also replaced the 2 IC's closest to the battery.  They are now in sockets.  The smaller IC is 5V sense.  If it detects voltage drop to 3V or less, it resets the system!

Leon, thank you for your thoughts
I have a built cable that i bought from a friend, but since this was supposed to be used with a FmTowns II machine, i will defenetly made a new one for the x68. The monitor turns on and off when i power it but there's nothing to be showed on screen, i just used once since is not pin-out precise, anyway the sharp is not ok so i quitted using it and just compare the leds if something change after all the works i'm doing.

About the crystal, yes, i just bought 2 of them, exact part number with 69.55199Mhz requested (they told me i'll be 69.552, just like yours in the picture). Thanks for the datasheet, i was infact going to ask you what was the pin 1 since i've barely used them in my arcade repairs. I will cut the leg and isolate it!

I bought on ebay the ttl replacements since here in italy i can't even find shops that selles those things (digikey was out of stock). i noticed that my 74HC132 is a 74HC132N and not 74HC132AP, i bought the same to be sure.

So, it looks like i'm the right direction. May i ask you about the 5V molex that you mentioned in another thread? i've read that you directly put 5v on it because the system was not powering up correclty. What molex were you talking about? With my ATX i connected all the cables togheter, i took all the red ones and connected all togheter to the 5v, did the same for all the colors (blue 12V, white 5VSB, grey -12V and so on).. is this ok? Also, are amp specs matters? Psu wattage? (mine is 500W)

[leonk]

I'm not a power supply expert so don't quote me but I suggest you look inside your ATX PSU; do all the common voltages originate from the same source? I remember reading a lot time ago that in some power supplies (especially high watts ones) not all the voltages come from the same rail!!  So combining 5V lines together (for example) might be a bad thing to do.

In my case, I was trying to get all the voltages from the ATX connector. This turned out to be a bad thing as the 5V lines on my ATX did not have sufficient current to start the system. I ended getting 5V from IDE molex power connector and system started.

For the final power design, I purchased a 1U ATX power supply; did a cap kit on it to quality capacitors and then wires the X68K original power wires into spots on this ATX PCB. I then transplanted the ATX PCB into the old power supply case.

The last thing I did was replace the 2 wire power with 3 wire (ground added!) and replaced the fuse on the power input board with a fuse holder and better fuse value for the power supply (I think 6A fuse)

But all this was cleanup work AFTER I got the system to post. No point in doing all this work if system doesn't post.

The HIRESO led must turn on at correct bootup. "bios" screen and HumanOS are all 31k/highreso. That light turns off only for 15k games.

[kamiboy]

I replaced my crystal with a programmable drop in replacement from ebay. I do not think that I need to clip any legs, the pinout is the same as the one in the x68000 if I remember correctly. I verified the crystal to be working using an oscillator. I have replaced a couple of IC's that were broken, and one several times since it mysteriously kept failing. I have not looked at the reset IC that you speak of yet. How did you test yours? Which pin will be high, or low if it is resetting the machine? It should be an easy check I think.

Edit:

Had look at your repair thread and found the datasheet. I guess I have to look at pins 5 and 6 which are reset and inverted reset respectively. I'll check tonight and report back.

[leonk]

It was thanks to JulBS0 suggestion of replacing the crystal that I managed to get my OG X68K working.

Pin 1 of the CPPT1-HT0PT-ND is the option pin.  On the X68K motherboard, this is ground.  Which will actually disable the oscillator!

As he pointed out in my repair thread, "The internal pull-up is rather weak, but since we leave the pin unconnected, it should not be a problem."  As per data-sheet, option pin needs just 2V min to start working; leaving it floating seems to be OK.

[leonk]

I replaced my crystal with a programmable drop in replacement from ebay. I do not think that I need to clip any legs, the pinout is the same as the one in the x68000 if I remember correctly. I verified the crystal to be working using an oscillator.

Did you verify in circuit or out of circuit?? 

Maybe you are using a different programmable crystal than I am.

[deathx]

So, it's now late evening and i'm out of force for today, but, only god knows how, i have already received the oscillators from Digikey, USA->ITALY in a day O_O (take that amazon).
I have also received a good cap kit in tantallium (those are great but little expensive) i'll fill what i'm missing with these (specially the drives after some clean here and there).
I will post my result tomorrow morning, hopefully for the best... cross fingers!

Btw, there's something i've never understood even if i personally do not find differences in what i was using for, but, put higher Voltage capacitors with same uF, does matter? I mean, i'm used to change the capacitors from, lets say, 100uF 16V to 100uF 25V (of course, if i'm out of 16V at that moment). I've never found differences in working conditions or replaces. I'm probably wrong here, but where's the point on this?

[leonk]

The uF is more important to get exactly the same than voltage.  Voltage needs to be rated or higher.

Think of uF as how much power will be put back on the circuit when it drops; voltage is the maximum amount of voltage the capacitor can withstand before it pops.

I strongly suggest to NOT use tantalum capacitors.  That X68K was designed to only work with Electrolytic capacitors!!!  Not only do they exhibit different electrical behaviour .. Electrolytic capacitors leak when they fail.  Tantalum tend to blow up / catch on fire.  REALLY!!!

[caius]

Btw, there's something i've never understood even if i personally do not find differences in what i was using for, but, put higher Voltage capacitors with same uF, does matter? I mean, i'm used to change the capacitors from, lets say, 100uF 16V to 100uF 25V (of course, if i'm out of 16V at that moment). I've never found differences in working conditions or replaces. I'm probably wrong here, but where's the point on this?

The nominal voltage rating indicated on a capacitor is the maximum voltage that may be applied to its terminals  You should start from the beginning and study the physical quantities : voltage, current, resistance and their correlation (Ohm's law) 

[caius]

  Tantalum tend to blow up / catch on fire.  REALLY!!!

Sure but if you polarize them inversely   

[kamiboy]

This is where I got my crystal: https://www.ebay.com/itm/TCXO-0-1ppm-Crystals-Oscillators-Clock-Frequency-can-be-customized/182144121804?hash=item2a68a2bbcc:g:xiMAAOSwAvJW8rTq

Indeed it seems that pin 1 is labelled as Vcon/NC, which I am not sure how I should interpret. Is the pin not connected? If not then it should not matter whether I leave it floating or connected to ground.

I should mention that upon initially swapping the original crystal with this one the machine sprang to life and ran normally. But it did die later on when I was diagnosing a soft power on/off problem.

Oh, and green led lights up when I push in front power button, no led lights up when it is pushed out. No other leds light up. And some IC's get burning hot, like IC43. I've socketed them and swapped them, but they still get hot, so the cause is elsewhere.

In other news, if I am not mistaken it seems my reset IC might indeed be ok. I am reading 0V from reset pin and 3.8V from inverted reset pin. I was expecting 5V from that pin, though I guess 3.8V is high enough to be logical high. The ref input voltage is in the 3V range as well which might be the cause, or maybe not.

[deathx]

This is where I got my crystal: https://www.ebay.com/itm/TCXO-0-1ppm-Crystals-Oscillators-Clock-Frequency-can-be-customized/182144121804?hash=item2a68a2bbcc:g:xiMAAOSwAvJW8rTq

Indeed it seems that pin 1 is labelled as Vcon/NC, which I am not sure how I should interpret. Is the pin not connected? If not then it should not matter whether I leave it floating or connected to ground.

I should mention that upon initially swapping the original crystal with this one the machine sprang to life and ran normally. But it did die later on when I was diagnosing a soft power on/off problem.

Oh, and green led lights up when I push in front power button, no led lights up when it is pushed out. No other leds light up. And some IC's get burning hot, like IC43. I've socketed them and swapped them, but they still get hot, so the cause is elsewhere.
In other news, if I am not mistaken it seems my reset IC might indeed be ok. I am reading 0V from reset pin and 3.8V from inverted reset pin. I was expecting 5V from that pin though, though I guess 3.8 is high enough to be logical high. The ref input voltage is in the 3V range as well which might be the cause, or maybe not.

NO lucky with me today... but that's EXACTLY what i get now. Only Green led when soft-pushed, when i turn it off no other leds are on (got a red power when turning off with the switch on the back, just for a second). I have put a floppy on both drives and the light turn on on both. No video at all. I suppose we do have the same problem here..
At this point i don't know what to do. Was thinking about the other nearby oscillator (38mhz) and i've read at nfg website talking about soft problems related to 74ls244/74ls04-08 something like that but i'm not sure if this is the case. I might just pull them out and test them.. i've been working 3 hours this morning for absolutely no joy in the end
This morning i have changed the oscillator with the new one, cutted pin1 and isolated. I have put the new 74132N and 7705CP, also changed the remaning caps.
I also noticed, since i have 2 1Mb pcb ram expansions, one of them (the one that came with the system) does light up the red timer, the other does not. They are the same board from I/O.

[deathx]

Btw, there's something i've never understood even if i personally do not find differences in what i was using for, but, put higher Voltage capacitors with same uF, does matter? I mean, i'm used to change the capacitors from, lets say, 100uF 16V to 100uF 25V (of course, if i'm out of 16V at that moment). I've never found differences in working conditions or replaces. I'm probably wrong here, but where's the point on this?

The nominal voltage rating indicated on a capacitor is the maximum voltage that may be applied to its terminals  You should start from the beginning and study the physical quantities : voltage, current, resistance and their correlation (Ohm's law) 

Oh ok, so let's say that the voltage nominated is the actual limit for that capacitor, but in terms of funcions is no big deal, i'm i right?

[leonk]

How are you connecting your video display device? are you sure if the system was working fine you will have video?

How confident are you in your power supply? I swapped mine with a 1U ATX. Try to minimize what's plugged in when you turn on. No extra ram. no floppies. I had mine running out of the case!

I did replace the 2 ICs next to the sram battery with modern equivalents.

first one was 74HC132AP. Next was a TL7705ACP

[deathx]

How are you connecting your video display device? are you sure if the system was working fine you will have video?

How confident are you in your power supply? I swapped mine with a 1U ATX. Try to minimize what's plugged in when you turn on. No extra ram. no floppies. I had mine running out of the case!

I did replace the 2 ICs next to the sram battery with modern equivalents.

first one was 74HC132AP. Next was a TL7705ACP

For the video part i have 2 options: a vga adapter and a scart cable adapter, first one has been maded by me, the scart cable was purchased on ebay (i'm actually testing it on a scart tv).
I think the power supply is ok, it's a near-brand new model that i've barely used in years. It's a low brand but should do the job. I'm confident in the supply since all of the voltages are joined together in groups and they are marked on the pcb by squares, i have cut many extra wires and i'm currently using one cable for individual purpose. (i've used a continuity multimeter and all the cables were joining the same the place, ground included).

I might be wrong with the two TTL chips. Mine has a 74132N but i'm currently using a 74Ls132N (brand new) i saw the datasheet descripion and they were in the same family (i'm using it since the one i bought from ebay is not coming home yet) and i've found many 7705CP on my junk arcade board, i picked up one that was looking fresh on a late pcb (same reason, waiting for the one on ebay).
Since, if i'm not wrong, those 2 chips are "fine" i might be thinking when i will receive the ebay ones won't change my situation.

Btw, now is doing something different, all the sudden. The soft button is not turning it off again and the power led goes from green (on) to a blinking green (off). Other Leds are always off.
I noticed that if i unplug the video board from the top and power it on, the HIGHRESO led turns on! =_= (still, no video of course)

Should i try to give him onother power supply meanwhile i'll wait for the ttls coming?

[kamiboy]

How do you show video out?  I made my own custom X68K to VGA cable which I plug into my PC.  Even if the video is not working, I know when the unit is on / sync as VGA monitor either turns off or stays on.
---
Notice I also replaced the 2 IC's closest to the battery.  They are now in sockets.  The smaller IC is 5V sense.  If it detects voltage drop to 3V or less, it resets the system!

I have mine connected to an X68000 monitor, and I also test the sync pins using an oscilloscope to verify the presence of both a V and H sync. One of which is currently missing.

When you replaced your 5V sense IC, did you measure voltages on the pins? I get 3V reading on the inverse reset, and ref voltage pins and wonder whether that is normal.

EDIT:
It seems the voltages on those pins are normal, so my sense IC is working ok.

[deathx]

One of my friend is going to send me his motherboard in exange of me installing him the new oscillator. That board is working, so, i will do the best i can to figure out differences from his working board to mine that is not. He also had bios eeprom on that board, i will dump them and compare to mines (wich are sharp proms, not eeproms) i'll share them in case of need.
I hope i will get the curplit of this and also help you to solve the problem... if not, at least there's some record for future purposes.
This is probably a good chance to figure out the problem, but is also the last one i can do... if you have suggestion on what i should check let me know about, i have the probe and a HP logic comparator (but i haven't started to use it properly)

[leonk]

I have a concern with HIRES turning on/off depending on vga cable plugged or not. How did you make your cable??  Sounds like you might have a short to ground there.

[deathx]

All right! I got the  motherboard from my friend, same exact model. After fixing some minor issues and replace him the oscillator (he wanted a new one from me since i've ordered 2 of them), board is working really great. May sound unusefull, but if someone wants to know, the boot sequence is:

Turn Power switch ON (soft button off): RED POWER LED
Push Soft Button ON: POWER LED goes green, HIRESO power red turn on shortly after, green POWER LED blinks twice and then go static, FDD 0 led turns on blinking and video is displayed.

I was able to test my upper video board, drives and also sub board, they all work good as should, at this point my problem is defenetly in the motherboard.

The problem is, what should i have to check and compare? That's a real task. I could not imagin to probe the WHOLE board.
I'll do some "simple" steps before trying the probe. I'll try to swap CPUs and BIOS proms, just to be sure that mines are working

(EDIT: after doing so, they are ok)

[kamiboy]

Testing a working board against a broken one is no easy task. Measuring voltages will not really get you far as far as bus buffers are concerned, and are of limited utility with logic chips, since one cannot be sure whether a difference of logic state on an IC pin is significant or just a artefact of on board being in a different state.

If you had a logic probe you could follow the CPU state from a cold boot and see where the two diverge. But then you would have to be privy to the workings of bios boot sequence code to see why the boot sequence fails.

Honestly my knowledge level falls short of the necessary for this sort of task by several orders of magnitude. I also have two ACE motherboards that do not function for unknown reasons, and it is quite frustrating not knowing why, and not being skilled enough to be able to figure it out.

[deathx]

I have finally fixed it. Now is working with no problems.
Turns out that my leak went that far enought to kill the 38Mhz oscillator (among with the 69mhz). When i pulled it out the first time i wasn't sure if was ok or no due to the fact that i do not have any oscilloscope to test these, there were minor leak underneath but wasn't look that bad. Apparently it was not as i supposed.
Before find out was the oscillator, i did replace also all TTLS nearby the battery (74Ls11, 74ls244, ecc ecc) and oscillator for last. So it might be a ttls failure and oscillator coulded be ok, i don't know.

At this point, if changing the 69Mhz won't do much, i suggest to change everything nearby (resistors, capacitors, ttls, diodes, oscillators)

Thanks to all helping me out with this

[vanpeebles]

Excellent work! Well done! 

[mnht2]

I had an X68000 original model that worked back in 2010. When I tried to boot it up 1 month ago, I only got a black screen.

I was quite desperate until I found this thread. I checked my system, found a leaking battery and followed the guide (replacing some diodes and the 69Mhz oscillator).
Yesterday I finished, started the X68k and it booted fine again.

Thank you all for this great thread!