[Solved] Slow SPI (Half Speed) BlueSCSI DB25

[68kPlus]

(Also posted on 68kMLA)

Hi everyone,
I received my BlueSCSI recently from Bruce at RecapAMac in Australia, and I found that no matter what SD card I use, it always seems to run in half SPI speed (25MHz). While the speed difference isn't noticeable according to the creator of the BlueSCSI, it seems like my Mac Plus runs noticeably slow even with SCSI Accelerator 7.0, with only around 240KB/s Read and 210KB/s write.
Can anyone help?

Thanks!

 

[eric]

Hey @68kPlus, thanks for posting over here - unfortunately cant do support there.

SPI speed is how fast it is communicating with the SD card, we try 50mhz first, and if it cant communicate with the SD card at that speed we try half of that. The causes are usually poor/old/low quality SD cards or bad contact with the card and the card holder.

I'll admit this happens so infrequently I don't have much more to add for debugging than what is already at https://github.com/erichelgeson/BlueSCSI/wiki/Slow-SPI - though even at 25mhz on a Plus you shouldn't notice (I can test later today manually setting the speed to half) - IIRC there was maybe 10% difference on the top end in speed.

I'll try to replicate your setup and speeds. Make sure you go through the basic troubleshooting steps as well - different card, full overwrite format with SD card fomatter, etc

 

[68kPlus]

Hey @68kPlus, thanks for posting over here - unfortunately cant do support there.

SPI speed is how fast it is communicating with the SD card, we try 50mhz first, and if it cant communicate with the SD card at that speed we try half of that. The causes are usually poor/old/low quality SD cards or bad contact with the card and the card holder.

I'll admit this happens so infrequently I don't have much more to add for debugging than what is already at https://github.com/erichelgeson/BlueSCSI/wiki/Slow-SPI - though even at 25mhz on a Plus you shouldn't notice (I can test later today manually setting the speed to half) - IIRC there was maybe 10% difference on the top end in speed.

I'll try to replicate your setup and speeds. Make sure you go through the basic troubleshooting steps as well - different card, full overwrite format with SD card fomatter, etc

Hey Eric,
I would have thought that if the card holder had issues with contacting the SD card it would have shown up during testing conducted by Bruce from RecapAMac and then resolved. Perhaps it's on my end, but the fact that I have tried 3x cards with no fix makes me think it's the BlueSCSI.
I'll try some more cards later on (I have stacks), and I'll see if I can get it to reach the 50MHz SPI.

My setup:
Macintosh Plus (4MB RAM, Platinum 1990)
DB25 BlueSCSI (powered via USB)
32GB Kingston Canvas Select Plus (with MacPack .hda file and several 2GB partitions - all working)

 

[68kPlus]

*updated profile picture*
Just letting you know to avoid confusion.

 

[eric]

Did some testing on the Plus and found something that changes an assumption we make. First 25mhz SPI is 5% slower in scsi director on raw transfer speed.

The assumption we made was that the SPI was either full or half speed (and nothing in between) - we actually used to check every mhz from 50 to 25 and only warn if under 40mhz - https://github.com/erichelgeson/BlueSCSI/commit/1cfaf005943342fdeddb2c5fe9626953a0b502a1

But! In the tests I just did I see that is incorrect and we should go back to the old way of testing values between 50 and 25. I have a card that fails at 50, but is successful at 49, and there is no difference in the scsi director tests.

I've attached a USB bin you can flash via QMK Toolbox. It will try 50mhz down to 25. Let me know if it resolves the 25mhz spi for you.

Thanks for reporting this!

 

[68kPlus]

Did some testing on the Plus and found something that changes an assumption we make. First 25mhz SPI is 5% slower in scsi director on raw transfer speed.

The assumption we made was that the SPI was either full or half speed (and nothing in between) - we actually used to check every mhz from 50 to 25 and only warn if under 40mhz - https://github.com/erichelgeson/BlueSCSI/commit/1cfaf005943342fdeddb2c5fe9626953a0b502a1

But! In the tests I just did I see that is incorrect and we should go back to the old way of testing values between 50 and 25. I have a card that fails at 50, but is successful at 49, and there is no difference in the scsi director tests.

I've attached a USB bin you can flash via QMK Toolbox. It will try 50mhz down to 25. Let me know if it resolves the 25mhz spi for you.

Thanks for reporting this!

Hi Eric,
Thank you very much for making a custom firmware! It's so cool to be working with you.
I flashed the BlueSCSI, and it appeared to work.
It booted up fine but in the log.txt file it still didn't seem to change.
Do I have to delete the old log.txt and then start it again?
I'm unsure what to do now.
No problem - I love to help these sorts of projects! It's so cool to think that I might be helping with the BlueSCSI and it's future development!
Thanks again,
Bailey

 

[eric]

The log should be overwritten - is the version in the log 1.1-20220918-SNAPSHOT? If so I'll add some logging and we'll try another

 

[68kPlus]

The log should be overwritten - is the version in the log 1.1-20220918-SNAPSHOT? If so I'll add some logging and we'll try another

Unfortunately no, I think the version date was around July.

 

[68kPlus]

Also some notes worth mentioning:
My STM32 BluePill appears to be 96MHz (I think there is a faster version?)
Version I'm running: 1.1-20220626-USB-96MHz

Also here is my log file:

 

[68kPlus]

I'm going to now update my BlueSCSI to the latest firmware and see if that helps.

 

[68kPlus]

Okay update:
Despite "updating" the BlueSCSI, it still shows up as the older June update.
Going to try using BlueSCSI Updater now.

 

[68kPlus]

Okay another update:

I needed to change a jumper to another spot to make the update work.

Strangely it gave me an error code upon booting up the Plus:
0F0003 (and sometimes 0F0002 and 0F0004).
After changing the HDA file it works again, which is odd.
Either way, it's now running at 50MHz SPI, which solves my issue.
Thank you for all the help Eric!

P.S:
The BlueSCSI is slightly faster now.
Sustained Read: 415KB/s
Sustained Write: 224KB/s
Access: 5MS
Seek: 0.0MS (?)

Overall index (HDT Benchtest) 7.5 (up from 2.1)

 

[eric]

Awesome! Glad it's working as expected now! These changes will be in the next release. Thanks again for taking the time to report the bug and work with me through it.

 

[68kPlus]

Awesome! Glad it's working as expected now! These changes will be in the next release. Thanks again for taking the time to report the bug and work with me through it.

No problem! I loved helping this project!