So I've acquired the famous IBM Blue Lightning board. I can boot into MR-BIOS and RTC charge is fine.
But for some reason I can only see black screen after the initial memory count. Is it a normal behaviour for that board? If not, where could be the problem?
I must be some kind of standard: the anonymous gangbanger of the 21st century.
Congratz 🤣
Here is the manual
https://exeyesoftware.com/wp-content/uploads/ … ystem-Board.pdf
try disabling the L2 cache in the bios, to rule out bad cache chips. I will have to get mine plugged in, since I do not remember any special diagnostic screens after post. I just remember it booting to A: or C: whatever I set up could be wrong.
Check to see that you are not overclocking the CPU. Mine would only run at 75mhz, since it was a BL3 I heard BL4s will run at 100mhz but mine would just hang at memory count if I tried to overclock it,
Ok, after tweaking L2 cache read/write wait states to 1 the board has finally showed me the boot options.
I must be some kind of standard: the anonymous gangbanger of the 21st century.
here is the manual
https://exeyesoftware.com/wp-content/uploads/ … ystem-Board.pdf
My board runs with zero wait states on everything. I used 60ns 16mb with parity on my ram. The cache worked at zero wait state also.
So far my board has successfully booted into WIndows 95 OSR2. 256kb 20ns L2 cache had to be disabled completely though.
I must be some kind of standard: the anonymous gangbanger of the 21st century.
Other thing useful, the IDE on board will work with the updated driver from here its superfast.
http://vogonsdrivers.com/getfile.php?fileid=9 … &menustate=35,0
sounds like you have bad cache chip, or chips. mine ran windows just fine with the cache enabled.
I used this ontrack disk manager with that IDE driver I posted and it worked good. The newer driver came with a older version of the ontrack, but it works great with the fat32 version.
https://www.philscomputerlab.com/ontrack-disk-manager.html
My board has different Local Bus IDE. It's Appian ADI2. Also it doesn't have any factory repairs.
I must be some kind of standard: the anonymous gangbanger of the 21st century.
It's the same thing.
One thing is that the onboard IDE hooks IRQ14 and there is no way to disable it. So if you use a VLB controller need to make sure it can use IRQ 15 or somthing else. Currently I use the onboard IDE and my CD is pluged into ESS sound card on IRQ 15.
Theres no reason not to use the onboard IDE though its pretty fast.
Look here should be the same thing as AIC-25VL01
http://www.ryston.cz/petr/vlb/adi2.html
this is CF card on mine.
So far my attempts to fix the L2 cache were a total failure. The original cache can work with relaxed wait states set to 1 and 2. But other 20ns cache at my disposal won't boot Windows even at such relaxed settings. 15ns ISSI cache was even worse, but it's somewhat rusty.
So either this board is VERY picky or has some problems.
I must be some kind of standard: the anonymous gangbanger of the 21st century.
I don't know what else to suggest. you seem like a smart enough guy. All the symptoms lead to a cache issue. I am not sure what else to auggest.
wrote:I did notice though, that the board insists on good quality SRAM chips. This board kept crashing and glitching until I bought a bunch of AT&T 12ns chips.
386/486 PC with Alaris Cougar 486BL MoBo
Though I don't think 12ns is necessary, My board runs fine with 20ns chips.
Another Thread about cache chips
A tale of 3 Blue Lightnings... tale 1
All the symptoms lead to a cache issue.
On the contrary, I've already had similar experience. But looks like the mentioned problem with SRAM quality is right on the money. So I'll have to try more cache modules.
Another Thread about cache chips
What's even more funny, the board came with the exact same 20ns SRAM chips.
I must be some kind of standard: the anonymous gangbanger of the 21st century.
After rechecking everything, I've noticed that my board was missing one jumper. Without that jumper the cache is very unstable. Right now I can push the original cache to 0/1 WS (Read/Write).
EDIT: No, it's still not stable, but at least that's a progress.
EDIT2:
Another batch of perfectly working 15ns SRAM and another failure. I give up. It's just a sh*tty board through and through. It's incredibly picky with SIMMs, SRAM and the CPU doesn't work stable at 100Mhz. Doesn't worth all that hassle to make it work properly. Good thing it didn't cost me much.
EDIT3:
And even more surpisingly, all my problems with cache disappeared after I've installed 486DX. Heck, I can even overclock it to 40Mghz with 0 WS. What is wrong with that board?!
I must be some kind of standard: the anonymous gangbanger of the 21st century.
Ok, this board is just cursed. After installing some CPUs into the Socket 2, it refused to boot. Apparently some socket pins just broke or bent after few installations.
Performance of the integrated 486BL is also very low. SpeedSys reported only 8.56 points, which is roughly on par with 486DX 25Mhz. Doom performance was slightly better: ~4050 ticks, which is somewhere around DX33.
I must be some kind of standard: the anonymous gangbanger of the 21st century.
Are you using a different motherboard PCB revision compared to others? Sounds like your board is optimised for i486DX. Were there different BIOS revisions you could try? Did you try running the BL3 at 2x25 MHz or 2x20 MHz?
Plan your life wisely, you'll be dead before you know it.
B revision. Production date: November 1994. PCB doesn't have factory added wire.
I am not aware of BIOS revisions to this board. Not sure how to force multipliers or 20Mhz.
Anyway, for now the board is drying up after water session. Maybe it will fix the problem with the Socket 2. Or it will die completely.
I must be some kind of standard: the anonymous gangbanger of the 21st century.
Any benefit in copying the wire mod?
Plan your life wisely, you'll be dead before you know it.
How would I know? Judging from late production date, the wire mod won't do anything or could harm it.
I must be some kind of standard: the anonymous gangbanger of the 21st century.
After VERY thorough cleaning I have finally managed to get stable L2 cache with 0 WS and 22.77 points in SpeedSys. It didn't helped with the Socket 2 though, so most likely it will require resoldering or replacement. Perhaps it's fixable by removing the upper part of the socket, but I have some doubts how to remove it without damaging in the process.
Doom results didn't changed much, scoring only 3946 ticks. But that's understandable, with 25Mhz VL-Bus and cheap CL-5428 card.
EDIT: After switching to Chips F6400 card the result was improved to 3810 ticks. Which is probably would be akin to DX2-50 or maybe DX2-66 without a fancy VLB card.
I must be some kind of standard: the anonymous gangbanger of the 21st century.
SIMM's or DIP's were dirty/corroded then 'eh? Did you brush them with a toothbrush and some kind of solution?
Plan your life wisely, you'll be dead before you know it.
