Black Box and Happy Drive

[+Larry]

At various times over the years, I've noticed that sometimes my Happy 1050 is a little "flaky" when running with a Black Box connected. IIRC, typically producing an Error 139. Today I was working on a 130XE MB and decided to run the Happy diagnostics. Of course, C304 of the 130XE had to be clipped to get the Happy to pass the high-speed transfer test. But after clipping it, I ran the Enhancement diagnostic and it failed on Ram Test 2. I repeated this several times with th same results. It consistently produced "Disk Error 1764 8C" So I disconnected the BB from the ECI, and immediately the Happy passed the diagnostics -- multiple times. Has anyone run into this before or has some clue what the error message might mean? If anyone has the Happy Tech Docs (very rare) it might be mentioned in that, although I think that Happy was no longer being made/sold by the time the the BB came out. (?)

Thanks,

Larry

[Gunstar]

Well, I don't know anything about either product, but it seems obvious to me that there is a direct conflict with both devices. They both seem to be trying to use the same memory location or something.

[MEtalGuy66]

There was a firmware update for the BlackBox called the BlackBox enhancer. Perhapse it adressed/corrected these problems. I dont know because I have no happy drives atm to test with it. You can alwayse email bob puff and hope for an answer.

[Mathy]

Hello guys

 

There was a firmware update for the BlackBox called the BlackBox enhancer. Perhapse it adressed/corrected these problems. I dont know because I have no happy drives atm to test with it. You can alwayse email bob puff and hope for an answer.

 

The BB Enhancer did other things then fix bugs. Bugs where fixed with new versions. IIRC the latest version of the BlackBox ROM is 2.16. This includes the Enhancer and IIRC the FloppyBoard code. This ROM will work on all BlackBoxes but if there is no FloppyBoard, you can't use FloppyBoard features. I was under the impression that version 2.16 was on my BlackBox page (Bob Puff told me I can publish this version. Saves him a lot of time, I guess). But a quick check tells me it isn't. I'll have to fix that later today.

 

Greetings

 

Mathy

[1050]

Larry, IIRC the error number might be significant?

"Disk Error 1764 8C"

 

On my BB with Happy as #1 drive setup, the byte 8C winds up in the third byte position on all sectors written to if trying to use Disk Based SpartaDOS versions 2x thru 3x. 8C could be error 140 which is supposed to be "Serial Bus input framing error" what ever that really means. My experience with SIO error numbers is that relevant routines usually just grab a number and send it without really trying to send something discriptive about the actual failure at hand. In other words, it could be just about anything actually causing it.

 

My system has to warm up before the Happy wants work right, until then there are a lot of misreads, restarts and total failures to write. But once warmed up it works fine. And mine fails the test as well with the BB in place and passes without the BB. I don't think the Happy software was written very well especially when it comes to working with SpartaDOS. The BB is throwing hi-speed SIO stuff into the mix as well and the Happy will NOT accept custom sector skews of which the BB has it's very own flavors of that it tries to get capable drives to use. The trouble may be that that the BB isn't treating the Happy in the right manner?

 

It's only been the past few years that my Happy has really laid down on me as described above, it used to be rock solid and ready freddy, good to go at all times. I can't imagine what needs tweaking, but something sure does. I have the alignment disk, I have a passable scope and I still don't want to risk losing the only Happy I gots by attempting a fix. A spare/second Happy is so far down in the budget list, it just isn't going to happen either.

[ijor]

Hi 1050,

 

I don't know much about the BB or the FloppyBoard, but there is a known documented issue with the Happy and some SpartaDos versions: High-Speed xfer (as used by SpartaDos) is incompatible with non-buffered writes. That is, you must enable fast writes to use hi-speed, or you will get errors.

 

Fast writes are disabled by default on the Happy. Some later versions of SpartaDos, including SpartaDosX, detect Happy drives and enable fast writes. Older versions just detect the happy as an USD compatible drive, it then uses USD hi-speed, but doesn't enable fast-write.

 

I don't remember which versions have a Happy driver and which don't. But just in case use the Happy software to enable fast writes.

 

Regarding fast or custom sector skew. The Happy doesn't support the 'f' command to set a custom skew. It doesn't because it doesn't need it. The Happy as most other buffered drives don't care about the skew/interleave, they read the whole track in one revolution disregarding the skew. If you still want to use fast skew with the Happy, Steve Carden made a small utility for this. He did it because he wanted to format disks in the most optimized way for his customers (if you told him you have an USD, he would format the disk with the Happy using USD fast skew).

 

Btw, your drive needing warm up hints about a mechanical issue.

[+Larry]

Larry, IIRC the error number might be significant?

(snip...)

 

Well, that would be a likely error code -- good thought! From the Happy 7.1 Docs (and don't I wish I'd bought the Tech Docs from Happy!) it would lead me to the conclusion that the Happy is only testing itself *internally*. But if that were true, then it wouldn't make any sense that the error only occurs with the BB attached. The BB and PBI/ECI devices in general are know to be associated with timing errors, so I'm going to try this with a couple of other XE's and see if this might somehow being a timing problem that is cured by the "Stable-XL" fix from Bob Puff. But as best I can tell, my Happy is rock-solid (when used alone). I guess while I'm at it, I'll dig out another Happy and see if it does the same thing. I will report back with the testing results.

 

OTOH, I remember some years ago talking to Bob about Happy Drives and he mentioned problems with R/W errors, but I don't remember any real details -- just something in passing.

 

And thanks for all the replies. I do have the latest Enhancer ROM (Floppy Board system).

-Larry

[ijor]

Larry, IIRC the error number might be significant?

(snip...)

 

Well, that would be a likely error code -- good thought! From the Happy 7.1 Docs (and don't I wish I'd bought the Tech Docs from Happy!) it would lead me to the conclusion that the Happy is only testing itself *internally*.

 

The test is internal, but the computer must first upload the test program to the drive on each test step.

 

What failed in your case is the transmission of the test code, not the test itself. 8C is the SIO error code, serial framing error. 1764 is the PC of the code (in the computer, not in the drive) when the error was produced.

[+Larry]

Larry, IIRC the error number might be significant?

(snip...)

 

Well, that would be a likely error code -- good thought! From the Happy 7.1 Docs (and don't I wish I'd bought the Tech Docs from Happy!) it would lead me to the conclusion that the Happy is only testing itself *internally*.

 

The test is internal, but the computer must first upload the test program to the drive on each test step.

 

What failed in your case is the transmission of the test code, not the test itself. 8C is the SIO error code, serial framing error. 1764 is the PC of the code (in the computer, not in the drive) when the error was produced.

 

Hi ijor-

 

Thanks! I think I've solved this little mystery, and you are correct about the actual cause of the error.

 

The cause of the problem turns out to be the high-speed SIO transfers in some OS, or supplied by the Black Box drivers. (It can be disabled in the BB menu.) In the original tests, I was using APE Warp, but the CSS UltraSpeed Plus OS also produces the same errors, as well as the stock XE OS with the BB attached and using its high speed drivers.

 

To make a long story short, if high-speed SIO is used, the Happy will normally fail the diagnostic test when doing "Ram Test #2." If regular SIO speed is used, it will pass 100% of the time. So long as the normal SIO speed is enabled, the BB had nothing to do with the error.

 

-Larry