• Updated 2023-07-12: Hello, Guest! Welcome back, and be sure to check out this follow-up post about our outage a week or so ago.

Challenging SE/30 logic repair.

redruM69

Active member
I have an SE/30 board in for repair that I just cant figure out.  It's been recapped, and corrosion was minimal.  It shows the attached pattern.  No chime, although I do hear 2 distinct "clicks" from the speaker at time of chime.  All address and data line and other traces have been checked between ROM, RAM, VROM, VRAM, and UA8-UG8.  Known good RAM/ROM/VROM has been swapped in.  All UA8-UG8 chips have been replaced.

Where would you go next on this one?

87037551_10215704638549703_2517529427403866112_o.jpg

 

IlikeTech

Well-known member
Make sure the rom is seated properly.  That looks like the screen you get with no rom.  Maybe corrosion on the slot contacts?

 
Last edited by a moderator:

IlikeTech

Well-known member
Also, is it possible something is pulling down the bus?  Maybe scope the address and data lines?

 

dochilli

Well-known member
Did you check the voltages at the floppy port? Must be 12 and 5 V. Perhaps they are to low.

Do you have another working SE/30. Then you could change the logic boards and see if it works with another good known AB.

 

LaPorta

Well-known member
I got that screen once when I just plum forgot to have the requisite number of RAM slots occupied. Do you have at least four sticks in there?

 

redruM69

Active member
Make sure the rom is seated properly.  That looks like the screen you get with no rom.  Maybe corrosion on the slot contacts?
ROM has been swapped numerous times with known good.  Contacts are clean.

Also, is it possible something is pulling down the bus?  Maybe scope the address and data lines?
I can give this a shot.  I haven't logic probed anything yet.

Did you check the voltages at the floppy port? Must be 12 and 5 V. Perhaps they are to low.

Do you have another working SE/30. Then you could change the logic boards and see if it works with another good known AB.
This is my SE test rig.  Analog is in great shape.

I got that screen once when I just plum forgot to have the requisite number of RAM slots occupied. Do you have at least four sticks in there?
Yup, 4 known good SIMMs installed.
 

 

techknight

Well-known member
Check the traces between the ROM and back to the ADC/SWIM/SCSI. As well as check the ASC between the GLU. 

Also, Check D-line to D-line for shorts, as well as A-line to A-line for shorts. as well as to both VCC and GND

If that fails, grab an oscilloscope and probe the data lines for activity, and more importantly, bus contention. Also if the SWIM goes bad, it will hold up the boot process. Same with the ASC if it cant be communicated to properly.

SCC wont matter, it would just freeze system 7 before it reaches the desktop. 

Last but not least, check the interrupt lines on the VIA as well as all 3 IPL lines into the CPU to make sure they arnt being held in a stuck interrupt state. 

 
Last edited by a moderator:

redruM69

Active member
No shorts between traces, or vcc/gnd found yet.

I did poke at the address and data lines with a logic probe.  All data lines show activity.  All address lines show activity, except A0, and A15-A19 are held low while powered.  They are not shorted to ground though.
Is that typical behavior?

 

techknight

Well-known member
Hard to say, it will all determine where its at in ROM. the CPU is attempting to execute code. But its dying off somewhere. Probably trying to initialize a hardware peripheral and keeps running into the address error or bus error trap handler exception. 

if you remove the ROM, it should attempt to walk the bus. so monitor the address lines and make sure. It should treat it as an OR.I instruction so itll just keep going. However if not, then see below:

You can also make a nop-inserter ROM Simm, (basically hardwiring $4E71 on the data bus, but driving with only pull-up resistors to prevent bus contention when others are selected) and watch the address lines like a counter. they should all be div/2 from each other. you would check these at all peripherals. 

Also watch these lines with an oscilloscope. you need to make sure each data bit and address lines are able to drive to the full bus voltage without contention issues. 

 
Last edited by a moderator:

redruM69

Active member
I threw in the towel and shipped the board back to the customer.  There's a point at which the time/money spent on diagnosis and repair exceeds the value of the board, and this is WELL past that.  I was hoping to follow through to the end, but I have a stack of other repairs growing.  

Thanks for the info guys!

 
Top