BlueSCSI not playing nicely with Powerbook 170

[penguin86]

Hello, complete SCSI noob here... I'm in the process of restoring a Powerbook 170. Two years ago I've got a BlueSCSI v1.1 (the STM BluePill based one) from Joe's Computer museum but I never managed to make it work. The bluescsi is the internal one, with the 2,5" SCSI connector, this one: https://github.com/dehij/powerbook-bluescsi

Here what I've done:

• I've followed the hardware and software troubleshooting instructions on https://github.com/erichelgeson/BlueSCSI/wiki/Troubleshooting
• I confirmed (reading the logfile created on the sdcard) the bluescsi reads the sdcard, mounts the volume at ID0 LUN0 and is happy
• I tried both supplying power to the STM via USB or have it powered by the computer SCSI bus

What I've got:

• If I place a bootable image on the bluescsi, I see the happy mac icon for a few seconds, then the boot fails
• If I boot the computer from physical floppy disk I created with system 6 (a real floppy in the floppy drive), I can see the bluescsi disk image appear on the desktop
• If I try to format the bluescsi disk, it formats ok
• If I try to copy data from the floppy to the bluescsi disk, it copies a burst of data, then hangs: I hear it stops reading the floppy and after a while it trows an IO error. It seems like it waits for a timeout before showing the error. When it stops writing the BlueSCSI green status led turns on and stay on until system6 give up and show the error.

I read a lot around the Internet and it seems something like this may be caused by wrong termination on the SCSI bus. I read on BlueSCSI troubleshooting page:

Make sure your termination is set correctly. If it’s the last device on the SCSI chain, you must have termination enabled by having the 2 jumpers on the BlueSCSI board near the SCSI port and resistor networks.

In this case this is the only device, so I think it should be terminated? BTW, I have two jumpers: one has only two pins (so on or off), but the other has 3 pins and the documentation doesn't explain in what position am I supposed to place the jumper.

But, looking at the schematics (attached to this post), i can see that placing both jumpers as indicated should result in shorting 5v to ground or shorting TERM_PWR to ground, and it doesn't seem correct to me. I tried, but the BlueSCSI in that case doesn't turn on.

I'm sure I'm doing something really stupid, but at this point I have no clue... Anyone can help me?

Thank you!

 

[penguin86]

Forgot to attach an image of the board. I have this one on hand, sorry if it's not very clear. The arrow shows the jumpers: as shown also in the schematics, there's a two pin one (JP1) and a three pin one (JP2), both labeled as "Termination".

 

[eric]

I've followed the hardware and software troubleshooting instructions on https://github.com/erichelgeson/BlueSCSI/wiki/Troubleshooting

Did you do a full overwrite format with the SD Card Formatter utility linked? (just need to validate) Also make sure you are not using a SanDisk and that you are on the latest firmware.

Measure the voltage on the termpower line and the 5v at the BluePill.

Yes you would want it terminated (it is at the end, and only device).

But, looking at the schematics (attached to this post), i can see that placing both jumpers as indicated should result in shorting 5v to ground or shorting TERM_PWR to ground, and it doesn't seem correct to me. I tried, but the BlueSCSI in that case doesn't turn on.

This these are pull up and pull down resistors (termination) so when it's "close to ground" it get pulled down to ground and vice versa. The choice is either if you are powering from TERM POWER for the termination or from the 5v source on the BluePill.

V2 has it's own 2.85v regulator so is more tolerant of weak termination.

 

[penguin86]

Did you do a full overwrite format with the SD Card Formatter utility linked? (just need to validate) Also make sure you are not using a SanDisk and that you are on the latest firmware.

Measure the voltage on the termpower line and the 5v at the BluePill.

Yes you would want it terminated (it is at the end, and only device).

This these are pull up and pull down resistors (termination) so when it's "close to ground" it get pulled down to ground and vice versa. The choice is either if you are powering from TERM POWER for the termination or from the 5v source on the BluePill.

V2 has it's own 2.85v regulator so is more tolerant of weak termination.

Hello Eric!
I made a miriad of tries in these weeks, so today I started from scratch.

I can confirm the voltages are ok *if i use external power* via the bluepill usb port. In that case I have 5v on the 5v bluepill pin, 3.3 on the 3.3 and 4.9 on the TERM_PWR.
If I try to power it from the SCSI bus, I have very low voltages ( less than 4v on the 5v, about 2v on the 3.3) and the bluepill bootloops or don't turn on at all.
So the following tests are made with external power and setting the termination jumper to 5v (similar results setting it to TERM_PWR). So JP2 to VCC and JP1 open (is it correct?)

I can confirm I'm using the latest firmware (I flashed it about a week ago, before there was an older one), v1.1-20231116.

I was using two sdcards and one was a Sandisk extreme 64gb, so i replaced it. Now I have a samsung and a microcenter, both 16gb class 10.
I did a full overwrite with that utility. That leaves the sdcard formatted as fat32, so I tried it both leaving as is or reformatting as extfat.

Leaving it as fat32 I had the best result until now: I had System 7.5.5 freezing about half way in the boot process. Never went so far.
Once rebooted, the disk with the question mark appears. If I boot from floppy, it says the bluescsi volume is corrupted, but I can initialize it and copy files to it.
If I copy the system folder from the booted floppy it copies without errors, but can't boot.

In the logs generated on the sd the only notable thing is it says that fat32 is slow and I should use exfat.

I _think_ the computer is ok because I can boot from floppy with system 6.0.8 or 7.1 and use it without errors.

Thank you for your help!

 

[eric]

So the following tests are made with external power and setting the termination jumper to 5v (similar results setting it to TERM_PWR). So JP2 to VCC and JP1 open (is it correct?)

I believe so - I don't have one of those handy at the moment to validate - but if you're getting ~5v to the jumper when jumped then that's correct.

If your internal power is that low you unfortunately wont be able to run this without addressing that or running a USB cable (which I'm not sure how you'd do, or even want to do)

Unfortunately with v1 it's going to be a bit of a guessing game here - bluepill quality varies widely. You might try flashing one of the OverClocked 96mhz or 128mhz firmware - some clone bluepills work better that way.

There was an issue with mouse bites (the perforations on the side of the board) on the initial batch - could you just validate that the traces near the side have continuity?

 

[penguin86]

Yes, I' getting 5v at the jumper when the bluepill is powered externally. The internal power is that low only when there's the bluescsi connected, so I thought my machine wasn't able to provide enough power for the bluescsi (I read in some machines is needed to power it externally). But I seem to understand that a Powerbook 170 in a good state should be able to do it, is it correct?
If that's the case, I should do some more troubleshooting. The machine is otherwise working normally and I can measure a stable 5v with almost no ripple around the board, even with the screen at maximum brightness.
Does the traces issue apply also to my board? I'm not using the standard BlueSCSI with the 2,5" adapter, but a fork (https://github.com/dehij/powerbook-bluescsi) meant for powerbooks. Anyway, I'll check, and I'll try also the different firmwares.
Thank you!