Mac SE + BlueSCSI - Won't Bus-Power

[displaced]

Hi!

I've just built my BlueSCSI and connected it up to my SE. To my surprise, it won't power via the bus. I've got the jumpers on the two termination headers (it's the only drive) and I've checked my soldering job. All looks OK.

With termination enabled (jumpers in-place), there's no sign of life from the BlueSCSI board whatsoever. If I disable termination, the board's power LED illuminates, but of course, the machine doesn't detect the unterminated drive.

If I power the board via USB, it works perfectly.

Now, I've got a Molex > FDD adapter on the way, so I can power the board from the SE's hard drive power cable. But I'm interested in finding out why my Mac can't power the board via the bus.

The PSU and analog boards have been recapped. With the original 20MB HD fitted, I was measuring a solid 12.5V and 5.3V on the hard drive's Molex connector. So I *think* power's good.

Is it likely to be just 'one of those things'? Could perhaps a recap of the logic board help? Or maybe there's something on the BlueSCSI board I should check?

Any help appreciated!

Chris

 

[eric]

That is very odd! Could you measure the power on BlueSCSI itself? A safer area to attach it would be on the actual termination jumper. Regardless be careful to not short anything.

The 5v rail is shared between the TERM power and 5v molex connector on the SE so unfortunately I dont think the Moelx to berg will make a difference in this case.

If you could try to attach a clear photo of the back and front to see if there's anything obvious issues.

 

[displaced]

Thanks, eric -- I’ve taken the photos (attached) and will get the meter out to test the voltages this evening after work. Just to check - should I test across the two termination jumpers (with them closed)?

The only deviation I made from the instructions was to mount the diodes on the front of the board, since with the socketed BluePill there’s room for them there.

Thanks for helping!

 

[displaced]

Well, thought I might as well give the Molex-to-berg adapter a try on my lunch break.

Amazon sent me the wrong item (just a regular Molex splitter), but I found a Berg adapter in my box of stuff.

And it works perfectly. Lovely to see the machine boot so quickly!

Might this suggest something’s amiss on the logic board? I’ll still check the voltages this evening with and without the molex connected just out of curiosity.

 

[eric]

Looks clean, and yes the diodes can sit like that if there's enough clearance. Only thing I cant see is the resistor nets - just make sure they're pin 1 is correct.

The safest way to check would be to have them jumpered and put the lead on the one closest to R1 - that should read 5v, the other jumper should be ground.

5.3v is slightly high, but the diodes should bring it down to right around 5v. I dont think that is the issue but with 30 year old machines, who knows! If you had another vintage mac to try it in it might rule that in or out (or an excuse to get another )

 

[Branchus]

Do you have a multimeter to test the diode at CR1 on the SE logic board?

 

[displaced]

Ok, whilst my mounting bracket for the BlueSCSI was printing, I measured the voltage between ground and the closed TERM jumper.

0.77V!

With the board powered from the Mac PSU via the molex and berg connector, the same point shows 5.1V.

Thanks for the tip, @Branchus - I’ll check that diode tomorrow. I do have a multimeter - am I right in thinking I’ll need to remove the diode to test it properly? No problem if so - just want to make sure I do it the right way.

Thanks!

(edit: just a quick aside…. Would the Mac’s activity LED work if I very delicately desoldered the PC13 LED from the BluePill, soldered a couple of wires to the pads, and ran them to the case LED’s connector? I’m guessing ‘yes, unless I break something along the way’, but worth checking!)

 

[Branchus]

Thanks for the tip, @Branchus - I’ll check that diode tomorrow. I do have a multimeter - am I right in thinking I’ll need to remove the diode to test it properly? No problem if so - just want to make sure I do it the right way.

You can test it in situ. Just put the multimeter in diode mode, then put a probe on each side of the diode. With the red probe on the striped side of the diode, you should get an open reading. Then swap the leads around so that the negative lead is on the striped side, and you should get a reading of around 0.6V.

 

[displaced]

Looks like we’re headed in the right direction, @Branchus !

Red lead to the striped side gives 0.75V. In the direction where it should be an open reading, I see 0.52V

Gonna clean up the area and re-solder, just on the off-chance it helps, but I imagine I’ll be needing to replace that diode. If that’s the case, what part would I need? Or do I need to start poking around elsewhere first?

Cheers,
Chris

 

[displaced]

Heh. Never had this before - sprayed a little IPA around the diode and gave it the slightest knock with a soft toothbrush and CR1 literally disintegrated!

So I’ll definitely need to replace that

(edit: looks like it’s a 1N4001 and I can’t find my diode kit. But I’ve just scraped through Amazon’s same-day delivery, so I’ll have a replacement this evening. I’ll remove the dead guts of the old one and clean up the area so it’s all ready to go)

 

[displaced]

All sorted! Replaced CR1 with a new 1N4001 and the BlueSCSI’s running perfectly on bus power.

Thanks very much indeed for your help, both!

 

[nottomhanks]

This thread is interesting. I have 3 BlueSCSI, but I’m having trouble with termination and random Folders with Question Marks that prevent it from booting the machine, even though it did previously. I’m wondering about getting Molex to Berg adapters as well.
Any advice?

 

[eric]

Any advice?

Low voltage could mean weak termination and cause issues. It will dpeend on other factors too. Maybe you could start a thread with your specifics and we can work through some things to try there?