Super HD : Soft Power on/off problem

[costa]

Hi there,
Some more troubleshooting on X68000... this time, one Super HD is not powering off using the soft power on/off button.
It tries to switch the computer and disk off, and starts slowing down the drives, but then restart spinning faster again... we can listen to this, and it continues forever.

I was looking at the Soft-power troubleshooting and repair http://www.gamesx.com/wiki/doku.php?id=x68000:soft_power_switch_troubleshooting_and_repair on wiki, and need to check some parts, but could not find all caps indicated there.

Before I try replacing them, I would like to ask some experienced guys about why you think of these other caps in the pictures ...
Are they only rusty, or is that signs of leak? They are not "fat", the upper side looks plain...
Should I replace them? Can it be responsible for the power on/off problem?

PS: battery is dead, I measure with voltmeter and got 0v on its terminals... any relation with the problem? I will replace it with a CR2032.





They are located here in the sub board:



And this one (the smaller) is on the main board:


What do you think?

[eidis]

 Hi Costa !

The health of capacitors can not be judged by the looks alone. It can have no signs of being bulged or leaked but due to age its electrolyte can dry up, its ESR can increase and its capacity can decrease. This happens more faster with systems which generate heat and have more demanding hardware which stresses the capacitor. That is why old 8bit (Speccy Rulez ) and 16bit systems (except for their psu) last much longer than more modern ones. A super high quality capacitor can last up to 20 years and super cheap ones for up to 5 years. It all depends on how much the hardware is being used, how much heat is near the capacitor and how much current the hardware draws. During trial and error I have come to a conclusion that 100uf and larger capacity capacitors can cause overcurrent and kill the system. This is a true story about Atari 130XE which was resurrected. C64 power supplies tend to kill the computer as well. And all because of one filter capacitor in the PSU.

The SUPER is more than 20 years old so there is no way around it. The capacitors must be replaced or there is a high probability that something could go wrong.

Keep the scene alive !
Eidis

[H68k]

Following eidis's advice.. I'd highly recommend you use the highest quality, longest life rated capacitors you can get, that way. once you've gone to the trouble of recapping every board in the machine. you wont have to bother with it again for some time. (except.. take it apart to clean it now and then of course)





VIVA LE SPECTRUM!

Now if you'll excuse me.. I have to run before the commodore 64 snobs turn up looking for my blood.

[costa]

1. Have you guys seen and counted how many caps are there on that sub-board? Replace all of them?
2. Do you think the spotted caps have anything to do with the power off problem?
3. Can I easily damage the main board / sub-board if I am not really, really good with an iron solder ?

[eidis]

 Hi Costa !

1. It is highly advisable to replace all of them.
2. Please download and run the X68k Power Test program which can be found in the following link. It will give you a much better understanding of what is wrong with your X68000.

Soft-power troubleshooting and repair
http://gamesx.com/wiki/doku.php?id=x68000:soft_power_switch_troubleshooting_and_repair

3. Use 16w soldering iron and do not heat one leg for more than 4 seconds and you will be fine. Pre compact models are very service friendly.

Keep the scene alive !
Eidis

[costa]

Hi.

My Super can now boot from CF card (using the AztecMonster), and can also boot from the floppy ONLY when the AztecMonster is connected.

When I remove the AztecMonster, it won't boot anymore.
SRAM boot device is set to STD.

Another symptom is that it is not displaying the usual message with the floppy on the screen... that message has simply gone! Even without a floppy, without the AztecMonster... not boot message asking for floppy.

[eidis]

 Hi Costa !

Try reseating the riser and using a different floppy cable.

Keep the scene alive !
Eidis

[costa]

Hi, thanks for the quick reply...

I thought the riser was that small board with the external floppy and scsi connector to the outside... but that will not need re-seating, so I must be wrong.

Sorry for this dumb question, but can you point out was is the Riser?

By the way,this is the second floppy cable I try, so I am assuming it is Ok at this point.

Hi Costa !

Try reseating the riser and using a different floppy cable.

Keep the scene alive !
Eidis

[eidis]

 Hi Costa !

The riser is a printed circuit board which extends connectors away from another board. This is the board where we can plug in the I/O expansion boards. I had similar problems with floppy due to poorly seated riser card.

Keep the scene alive !
Eidis

[costa]

Re-seated the raiser, still the same.

I could do another test, and look how interesting (or should I say worrying?):

1. It booted the Memtest68k boot disk. It is really a quick to boot, then it ejects the floppy and starts testing memory.
2. It boots my HumanOS disks from both drive 0 or 1 when AztecMonster is connected.
3. The "insert bootable disk" message (with the floppy) is never displayed.

I have tried all my boot disks, no success... only memtest68k will boot.

Run out of ideias...

[eidis]

 Hi Costa !

What happens when you try booting Motos from floppy drive ?

Keep the scene alive !
Eidis

[costa]

Hi Costa !

What happens when you try booting Motos from floppy drive ?

Keep the scene alive !
Eidis

I don't have that game in Floppy right now, but I can make one.

Does it have any special characteristics that makes it good for diagnosis's ?

[BlueBMW]

One more thing to check....  this is kind of a long shot but the rear HDD/FDD connector block... check and see if has two blue resistor blocks (I think they're resistor blocks) installed.  I know I had a system that would not boot disks properly when those were missing.  They must terminate a bus of some sort.  Another system worked fine when they were missing but only when an aztec monster + HD cable was installed.  With a non HD cable connecting it the system wouldnt boot from floppies.

I think the number on them was:

BI 18Z331J221J

I found some replacements on ebay but I cant seem to find any more right now.

[costa]

Motos did boot.

In the mood, I tried another game - Xevious, and it booted also... and on this one I could see the autoexec.bat being processed and everything.
Odd.

Is it something in the autoxec of config.sys of the HumanOS?

Hi Costa !

What happens when you try booting Motos from floppy drive ?

Keep the scene alive !
Eidis

I don't have that game in Floppy right now, but I can make one.

Does it have any special characteristics that makes it good for diagnosis's ?

[costa]

Hi BlueBMW

Yes, there are two resistor networks next to the external SCSI connector - I mean, the resistors are there, soldered to the PCB.
There are also two sockets (for another resistor network, I think, because I saw that in a IO another board I have here - not related  to this issue).

So, are you saying that I should also put two resistor networks on those sockets?

I made a system disk from the Xevious floppy... it has an old version of HumanOS - command 1.0.
I can boot from this floppy, but not from HumanOS 3.0.2.

One more thing to check....  this is kind of a long shot but the rear HDD/FDD connector block... check and see if has two blue resistor blocks (I think they're resistor blocks) installed.  I know I had a system that would not boot disks properly when those were missing.  They must terminate a bus of some sort.  Another system worked fine when they were missing but only when an aztec monster + HD cable was installed.  With a non HD cable connecting it the system wouldnt boot from floppies.

I think the number on them was:

BI 18Z331J221J

I found some replacements on ebay but I cant seem to find any more right now.

[BlueBMW]

Yeah, the system I had issues with I had to install resistors into those sockets.  So total I ended up with four resistor networks installed.   I'm not sure if it has to do with the hard drive bus or the floppy drive bus but for whatever reason their absence caused an issue with certain configurations.

I guess the sockets are labeled RM601 and 602



And the parts I installed looked like these:

[costa]

This is a resistor network, I would try any if I knew the resistance... I think that event one of different resistance would not hurt.
I will see when I can buy one of these for example that I found at Farnell:
BOURNS - 4612X-101-103LF - RESISTOR, BUS RES N/W 11, 10KOHM, 2%, SIP

Edit: I just checked the data sheet, and one application of these resistor networks is guess what: provide SCSI termination!
https://www.bourns.com/data/global/pdfs/scsi.pdf

[BlueBMW]

I'd guess then that the Aztec monster provides termination to the bus when installed but if its termination is turned off or if its not present at all then the SCSI bus would not be terminated at all.  Maybe this causes issues with the way the system boots in the same way that a poorly seated riser card causes boot issues even though IO slot cards / connections may not seem relevant to floppy drive booting etc.

Might be worth a shot at least...

[costa]

Talking about terminators...

Man, I put a ScSI terminator on the external SCSI connector, and it all got back as usual!!

Mystery solved... I did not know a Super HD without HD needed a SCSI terminator!
I have one terminator here, that I used plugged to my CDROM, so I plugged on the back of the Super after removing the AztecMosnter, and voile... got back that well know message with the floppy asking for a boot disk.
Inserted HumanOS 3.02 and it booted.

Still a mystery for me why it boots an old version of the HumanOS and games... but at least I have now confirmation that there is nothing wrong with my Super.
Thanks all for all ideias and suggestions.
I will still try the resistor's network as soon as I can - but only when I have list of components I need so I can put an order online.
cheers!!!

[BlueBMW]

It must have something to do with the way certain versions of human68k address hardware.  Maybe earlier versions dont look for an intact hard drive bus before booting....  One of the really smart guys should know