HELP: CompactXVI troubles [FIXED]

[caius]

Hi all,
I have a CompactXVI with a SCSI to IDE adapter and a IDE to CF converter, it worked very well until some weeks ago when it has begun to show this error (not always) when he tries to boot from floppy disks (after  I insert a disk on prompt) or from  CF:





Translate to english it would mean more or less:"The system could not boot.  please reset." which is similar but not the same to the error in this article:

http://www.gamesx.com/wiki/doku.php?id=x68000:sram_issues

Obviously I tried to clear the SRAM without success.



So I thought it was matter of replacing all capacitors on motherboard and power supply.So, I made this whole recap but , sadly, after poweing on the system for two or three time, power supply has gone bad.I opened it and found the fuse (3.15A) blown and a two diodes literally ckacked in two.One diode is marked 1NGA and other is a zener diode which I was not able to identify.So, I replaced the fuse, the 1NGA diode 'D2'(there are many of them on the pcb) with a common 1N4007 and the zener 'ZD12' with a 6.3V one but  after powering on the fuse and the zener 'ZD12' has blown again.I post this picture for better understanding, bad components  are circled in gray, 'D2' diode(1NGA type ) is hidden under the R6 resistor:



Obviously, capacitors on ower supply were mounted correctly with right polarity and capacity, I made no mistakes.

So, now I'm using a pc ATX power supply but, also in this case, I obtain  (not  always because some time the system boots well) the same error above.
So, now I'm stucked and asking some help from you.
Thanks in advance.

[lydux]

Hi Caius,

What make you think ZD12 is a 6.3v zener ? Sound like this part is located after the rectifier stage. According to PSU schematics v1.0 and 1.5, it should be 15v. I' not sure through... Since I don't have compact psu schematics.
Please check with this one : http://nfggames.com/X68000/Schematics/PSU/X68000_EXPERT_PSU_Schematics_v1.5.pdf
It should be similar on area around mosfet Q1. And btw, Q1 is probably the one who mess your psu : a dead mosfet often act as a short. Located in this area, it make sense to me that it could burn the fuse and this zener.

As for your error message, did you try to press the Interrupt button ?
I don't remember if this message comes from IPL, disk bootloader or Human68k. Depending on the result of the interrupt button, it might give me more hint on where it hangs.
And one last thing : the next time you could reach human68k prompt, try a RAM test : http://nfggames.com/X68000/Misc/Memory%20Test%20Program%20v1.12/
Just in case of...

Good luck !

[caius]

Hi Caius,

What make you think ZD12 is a 6.3v zener ? Sound like this part is located after the rectifier stage. According to PSU schematics v1.0 and 1.5, it should be 15v. I' not sure through... Since I don't have compact psu schematics.
Please check with this one : http://nfggames.com/X68000/Schematics/PSU/X68000_EXPERT_PSU_Schematics_v1.5.pdf
It should be similar on area around mosfet Q1. And btw, Q1 is probably the one who mess your psu : a dead mosfet often act as a short. Located in this area, it make sense to me that it could burn the fuse and this zener.

Hi lydux and thanks for replying.I thought 'ZD12' zener diode was a 6.3V comparing my PSU with Expert schematics but, probably, I was wrong.
Yes, Q1 mosfet in Expert schematics is directly connected to the cathode of  'ZD51' which is a 15V zener diode.In the Expert schematics Q1 is a 2SK643 MOSFET type, in my PSU it's marked 'FEC K1101 1N' and it's always connected to the cathode of the 'ZD12' diode.
According to the MOSFET cross reference '2SK2601' is the recommended replacement for the '2SK643' mosfet of the Expert PSU.Is it the same also for the 'K1101' of the Compact?

By the way, D2' diode marked 1NGA was good and not bad, this is because I tested it in circuit and not desoldered (it gave me in circuit almost 20 ohms of continuity  on the multimeter).Anyway I replaced it with a 1N4007 which I presume to be an equivalent.Is it right?

As for your error message, did you try to press the Interrupt button ?
I don't remember if this message comes from IPL, disk bootloader or Human68k. Depending on the result of the interrupt button, it might give me more hint on where it hangs.
And one last thing : the next time you could reach human68k prompt, try a RAM test : http://nfggames.com/X68000/Misc/Memory%20Test%20Program%20v1.12/
Just in case of...

Good luck !

This is the screen I get after press a key on that error message:


And these are some values I obtain progressively  by pressing the  interrupt button:















Regarding the RAM test, I will do it for sure  if I  only could boot again in the Human O.S., now it seems that the error is permanent.

UPDATE:
I desoldered  'Q11 K1101 MOSFET and pin of gate and source are really  shortcircuited so it presumably bad.
So, since 'Q1' on Expert PSU is a '2SK643' MOSFET type (equivalent to 2SK2601) and  presuming that  'Q1' K1101 on Compact is the same as Expert   and since 2SK643 and  2SK2601 are discontinued,  I found these with similar  characteristics:

http://it.mouser.com/Semiconductors/Transistors/MOSFET/_/N-96mrk?P=1z0w94oZ1z0w951Z1z0w7iaZ1z0z63xZ1z0y3zrZ1z0w6o6

Are they good for my proposal?

[lydux]

I desoldered  'Q11 K1101 MOSFET and pin of gate and source are really  shortcircuited so it presumably bad.
So, since 'Q1' on Expert PSU is a '2SK643' MOSFET type (equivalent to 2SK2601) and  presuming that  'Q1' K1101 on Compact is the same as Expert   and since 2SK643 and  2SK2601 are discontinued,  I found these with similar  characteristics:

http://it.mouser.com/Semiconductors/Transistors/MOSFET/_/N-96mrk?P=1z0w94oZ1z0w951Z1z0w7iaZ1z0z63xZ1z0y3zrZ1z0w6o6

Are they good for my proposal?

Ok so, ZD12 on compact is ZD51 on expert. That's definitively 15v.
And on MOSFET part, K1101 = 2SK1101, so yeah, basically almost the same as the EXPERT. This is a typical N-Channel power mosfet 30v 10a.
But seems to me any N Channel MOSFET that is at least 20v 8a could do the job here...
Just for insurance, I'll personnally go here for an IRF740 (cheap, good quality and widely used) : http://fr.mouser.com/Semiconductors/Transistors/MOSFET/_/N-96mrk?P=1z0z73rZ1yzrw77Z1yzrw76Z1z0y48cZ1z0y49eZ1z0xzzgZ1z0w7myZ1z0w7iaZ1z0z63xZ1z0y3zrZ1z0wa0b

For your D2 diode, I can't find anything about this 1NGA mark. Does its equivalent on expert is D3 ? But it doesn't really matter here... 1N4007 is ok.


Now, this write box message indicate that Human68k is correctly loaded, so that not a drive issue. But these values does not point me to the exact problem. So, I'm currently disassembling human.sys, and I were able to find the location of this error message. Unfortunatly, it seems to be a generic one which can be raised by many different failures, but generally occurs uppon human initialization. I'm now near to be able to hack this part, I would like to do some translations on human.sys (will serve me as well for ERSA driver) and mods this screen by something more informative than a single "I failed ! Move on..."

So please, be patient ! 

[caius]

Ok so, ZD12 on compact is ZD51 on expert. That's definitively 15v.
And on MOSFET part, K1101 = 2SK1101, so yeah, basically almost the same as the EXPERT. This is a typical N-Channel power mosfet 30v 10a.
But seems to me any N Channel MOSFET that is at least 20v 8a could do the job here...
Just for insurance, I'll personnally go here for an IRF740 (cheap, good quality and widely used) : http://fr.mouser.com/Semiconductors/Transistors/MOSFET/_/N-96mrk?P=1z0z73rZ1yzrw77Z1yzrw76Z1z0y48cZ1z0y49eZ1z0xzzgZ1z0w7myZ1z0w7iaZ1z0z63xZ1z0y3zrZ1z0wa0b

For your D2 diode, I can't find anything about this 1NGA mark. Does its equivalent on expert is D3 ? But it doesn't really matter here... 1N4007 is ok.

Thanks Lydux.I will replace these zener and mosfet and let you know.By the way, I think, that Compact PSU schematics are really needed since many people got this model!

Now, this write box message indicate that Human68k is correctly loaded, so that not a drive issue. But these values does not point me to the exact problem. So, I'm currently disassembling human.sys, and I were able to find the location of this error message. Unfortunatly, it seems to be a generic one which can be raised by many different failures, but generally occurs uppon human initialization. I'm now near to be able to hack this part, I would like to do some translations on human.sys (will serve me as well for ERSA driver) and mods this screen by something more informative than a single "I failed ! Move on..."

So please, be patient ! 

I was able to run the memory test and all goes fine.But now boot success is drastically low!I cannot go into Human O.S. anymore,
and when  I boot succesfully from hard disk (in my case a Compact Flash) , then I can see only one FDD (unit 0) and it doesn't  neither work (unit 1 is not listed after I run the drive.x command).I don't know what can be done, I'll start make a whole reflow, may it's due some dry joints or change some oscillators (I read that boot problems can be caused form these).Sadly I can not reprogram Eproms as they are all SMD, I presume.
We will be patient and take all the time you want, you are doing a fantastic work!

P.S.
I tested those resistors arrays on sub-board and some of them doesn't give me any value but I don't know if this depends from the fact that I have tested them in circuit.

[eidis]

 Caius,

I once had a similar problem. The system refused to boot and I pinpointed it to bad internal floppy cable. The X68000 needs to see both drives in order to boot. Maybe there is a bad connection or the FFC cable needs to be replaced. Try to flex it a little bit while reseting and see if anything changes. And once again about the capacitors If you haven't allready replaced them on floppy drives, it won't hurt to try that too.

Keep the scene alive !
Eidis

Update: Just ran some tests on my ACE HD and I get the same error when either one or both floppy drives are disconnected. My BOOT parameter is set to ED100.

Update #2: This could be linked to bad FFC cable. The original two have contacts on the same side, 1.25mm pitch, maximum 34 circuits, 8cm and 6cm. The closest compatible one which can be bought in small quantities is available here:

http://avnetexpress.avnet.com/store/em/EMController/_/A-13966017/An-0?action=part&catalogId=500201&langId=-1&storeId=500201&CMP=KNC-Supplyframe_VSE-T11

7.6 cm, contacts on the same side, 1.25mm pitch, maximum 35 circuits. One needs to be cut off, but that should be a simple task.

[BlueBMW]

Ahh, good to know a source of replacement compact floppy cables.  I know one of mine is a little flaky.

[eidis]

 BlueBMW,

Just a hint. If you ever need an uncommon cable or connector it is more than likely that www.molex.com manufactures them. After identifying the part number and confirming that it is compatible from the manufacturers provided datasheet, go to www.findchips.com to find a shop that sells them at reasonable price.

Keep the scene alive !
Eidis

[caius]

Just to inform you that I fixed the problems with my CompactXVI:

These are the repair logs , I hope they could help other people:

- CompactXVI PSU 
Symptom:
PSU stop to working after a spark come out from it.

Fix:
At a closer inspection fuse (3.15S 125VAC) was blown and 'ZD12' 15V zener diode were cracked in half. Desoldered the near the 'Q1? MOSFET (2SK1101) and found that GATE and SOURCE pins was shortcircuit.Replaced it with a common IRF740 MOSFET along with a new 3.15A fuse and 'ZD12' 15V zener diode fix the problem.

- Compact boot error message:
Symptom:
Compacts stop to boot (either from SCSI or FDDs) and show a message error (sse above for picture)

Fix:
X68000 need to see both floppy drives in order to properly boot.After I changed capacitors on PCBs floppy and check connection between the FFC cables and respective connector, it boots again normally

I want to thank Lydux and Eidis for pointing me in the right direction, I couldn't have done it without you guys!

[eidis]

 Hi Caius !

A fine work of art. Thank you for resurrecting the original PSU and not going for the "quick fix"

Keep the scene alive !
Eidis

[lydux]

Yeah ! Good job Caius ! 

I still don't get why human load correctly from storage, but can't boot because it can't find a drive... I've probably miss something in bootloader. But anyway, that's good for you !

I can now confirm by reverse engineering that most of time, this error message raise because of a drive issue (or no drive at all).
I'll continue to translate human kernel, that might help...
Screenshots !

[caius]

Yeah ! Good job Caius !  

I still don't get why human load correctly from storage, but can't boot because it can't find a drive... I've probably miss something in bootloader. But anyway, that's good for you !

I can now confirm by reverse engineering that most of time, this error message raise because of a drive issue (or no drive at all).
I'll continue to translate human kernel, that might help...

Thanks Lydux for your help in troubleshooting.You are doing a great!From my own (and Eidis know this) I'm carry on my project about connecting ordinary FDDs to X68000, I succesully hooked up a Samsung SFD321B (and more models are on the list) , it reads and writes fine, the only problem is the autoeject feature, I'm using a software driver called expfd.x  that can simulate this feature but most of the time it doesn't work.The ideal would be an hardware circuit capable of handle this proprietary signal (INTERRUPT) of X68000 FDDs, there are some circuits floating on the net, I would to try them but I need some help to interpretate them (they are all in japanese).I think you are the right person (if you only had time to dedicate to  ...)..

[caius]

Yeah ! Good job Caius !  

I still don't get why human load correctly from storage, but can't boot because it can't find a drive... I've probably miss something in bootloader. But anyway, that's good for you !

I can now confirm by reverse engineering that most of time, this error message raise because of a drive issue (or no drive at all).
I'll continue to translate human kernel, that might help...

Thanks Lydux for your help in troubleshooting.You are doing a great!From my own (and Eidis know this) I'm carrying on my project about connecting ordinary FDDs to X68000, I succesully hooked up a Samsung SFD321B (and more models are on the list) , it reads and writes fine, the only problem is the autoeject feature, I'm using a software driver called expfd.x  that can simulate this feature but most of the time it doesn't work.The ideal would be an hardware circuit capable of handle this proprietary signal (INTERRUPT) of X68000 FDDs, there are some circuits floating on the net, I would to try them but I need some help to interpretate them (they are all in japanese).I think you are the right person (if you only had time to dedicate to  ...)..

[lydux]

Sure ! Might be interesting, start a new thread or pm me. 

[gypsie]

Hi,

I ve got an error message after i ve changed all the capacitor on my Compact.

The reason was : rear switch (internal fdd order) set to 2-3.