CompuServe Thread

#Maxtor Problems

19 messages in this thread
#158130From: Dave CinegeAug 1, 1994 1:54 AM
I am having problems with a Maxtor MXT-540S. (S not SL, this is the IBM OEM drive) I am using it in an A3000. The problems I am having concern the transfer of large files. Some times it will take a long time (maybe 10-30 seconds as opposed to a maximum of 2) when transfering a file over 200k. When this happens the copied file, has correct size, but the data is corrupt. Someone on BIX has the same drive, and placed a single terminator on the A3000 motherboard to correct this problem. I have tried this but it has helped very little. Dave Elite MicroComputers – Flying high on AutoPilot
#158136From: daniel wolfAug 1, 1994 3:08 AM
Suggest you try reducing the MAXTRANSFER number in HDTOOLBOX to 1FFFF or so from 7FFFFF. It is possible the large file is outstripping the Amiga/drive ability to keep up and reducing MAXTRANSFER may help the system keep up more easily.
#158145From: Dave CinegeAug 1, 1994 6:21 AM
I have played around with the MAXTRANSFER. It didn't help any, just slowed the entire drive down. But I will try what you gave me. Dave Elite MicroComputers – Flying high on AutoPilot
#158193From: Dave CinegeAug 1, 1994 7:12 PM
Nope. MaxTransfer 1FFFF – 7FFFFF didn't help any….. Dave Elite MicroComputers – Flying high on AutoPilot
#158272From: Jim SteeleAug 2, 1994 11:16 AM
You could try swapping out the WD SCSI controller chip with a newer version. I've seen some problems with 3000s and some drives that I felt were firmware problems between the "proto" chip in the 3000 and the drive. -Jim MOS via Autopilot & Amiga
#158277From: Dave CinegeAug 2, 1994 12:13 PM
Do you know what the current REV of the WD chip is? I have a WD33C93A-PL 00-04 9114 WDC PROTO Dave Elite MicroComputers – Flying high on AutoPilot
#158286From: Dale LarsonAug 2, 1994 1:39 PM
-08 is a current part. Dale L. Larson Intangible Assets Manufacturing (info@iam.com) "That's the best part about being crazy. You see things no one else can see."
#158353From: Dave CinegeAug 2, 1994 10:02 PM
UGHHH……I'm 4 revs back? Time to search for a new SCSI chip. Anyone know any sources? Is GrapeVine still around? Dave Elite MicroComputers – Flying high on AutoPilot
#158385From: Dale LarsonAug 3, 1994 4:31 AM
Rev -04 works fine for most people using AmigaOS. Under alternate OS's, the -08 is nicer because it fixes bugs (AmigaOS works around the known bugs in each of several revs). It is not an Amiga-specific part, so availability isn't much of a problem. Dale L. Larson Intangible Assets Manufacturing (info@iam.com) "That's the best part about being crazy. You see things no one else can see."
#158323From: Jim SteeleAug 2, 1994 7:13 PM
-8 is the last version I've heard of. I recently saw someone in Amigaworld advertising them, but a quick look thru the last couple issues was fruitless. It's a standard part though, so it should be available thru the usual electronics outfits. -Jim MOS via Autopilot & Amiga
#158363From: daniel wolfAug 2, 1994 11:37 PM
excellent suggestion. A friend had a 'similar' problem recently with a maxtor 120 on a A1200 which we solved by formatting the drive from scratch and reinstalling 3.0. He copied some software onto it and whenever he went to use it it crashed. I think a virus crept in.
#158775From: Gary Vander MeiAug 6, 1994 8:40 PM
Daniel, The problem is the drive is made by Maxtor.. Amiga owners must enabe one of the reserved jumpers for some reason. Contact Maxtor (18002Maxtor) as to your particular drive. // Gary Vander Mei on Auto-Pilot \X/
#158893From: daniel wolfAug 7, 1994 7:05 PM
Just ran into a WEIRD problem with a Maxtor 25128 120mb IDE 2.5 inch drive on a A1200. Formats perfect, shows no bad blocks, but when certain programs are 'called up' from the drive they act weird. Virus checker says all is well. Other drive (Seagate) does not experience same weirdness. Interestingly QBTools says BOTH partitions on the MXT25128 have 'problem' on the root and says it will fix it but it doesn't. And that's IMMEDIATELY after a completely fresh 'format' operation. Absolutely stumped – never seen such a weird problem. Drive acts clean, but affects programs.
#158362From: daniel wolfAug 2, 1994 11:35 PM
If MAXTRANSFER doesn't help, I'm at a bit of a loss unless the data transfer is somehow screwed up or you've got a virus.
#158372From: Dave CinegeAug 3, 1994 12:23 AM
It's not a virus, I saw this problem as soon as I set up the drive. I also never had any problems like this with the old drive. Do you know anything about how the MASK works? I don't really know what it does, but in playing around with it, I've really increased the noticable speed of the drive. (When it decides to copy large files correctly :< ) Dave Elite MicroComputers – Flying high on AutoPilot
#158474From: daniel wolfAug 4, 1994 1:56 AM
I believe MASK is a number which MUST be at a certain value for each of the different file systems the Amiga has. Good luck. I hope it turns out to be the old style SCSI chip in your A3000 and you can replace it. I was quite surprised the first time I saw an early A3000 SCSI chip with BETA written right on it! But the SCSI drivers are written to compensate for it as much as possible – perhaps this is a case where it just kills your throughput to have the older chip.
#158485From: Ronald B. RomineAug 4, 1994 3:24 AM
DC >>Do you know anything about how the MASK works?<< I believe it deals with "where" within the Amiga's memory, the data can be DMA'd. And whether the transfers are 32-bit, 16-bit, or 8-bit in alignment. — The MASK is the highest area of memory (in BITS) that is accepted as a viable DMA area. The last digit of the MASK (c, e, or f) marks the alignment as 32, 16, or 8 bit. (If I got my math & facts right…) A mask of 0xfffffe = 16,777,215 bits (2-megs). This allows DMA transfers to the first 2-megs of address space, but doesn't allow for DMA transfers to other memory areas.. This would be the CHIP memory area. A mask of 0x7fffffe = 134,217,726 bits (16-megs). This labels the first 16-megs of address space as an approved DMA area. This would allow DMA transfers from HD to CHIP RAM, FAST RAM (Zorro-II), and Slow-Fast (if it is DMA'able). A mask of 0x7ffffffe = 2,147,483,647 bits (256-megs). This is first 256-megs of address space, which would include CHIP, FAST (Zorro-II), Slow-Fast, and the A3000/4000 motherboard memory, & A3000/4000 co-processor slot RAM. — The last digit of the mask affects whether the transfer is to be an 8-bit, 16-bit, or 32-bit transfer. 0x7fffffff = 8 bit alignment 0x7ffffffe = 16 bit alignment 0x7ffffffc = 32 bit alignment Since my A1200 has 32-bit memory, I use "0x7ffffffc" for EACH partition. This should allow me to have HD transfers aligned with my 32-bit FAST RAM. Also, the MAXTRANSFER rate should be '0x1fe00' for EACH partition of an IDE hard drive. Ron. – ("I'm not stupid, I'm not expendable, and I'm not going!" – quote made by any low level RPG player.)
#158543From: Eulogio (DJ) GarciaAug 5, 1994 12:02 AM
Ron, Thanks for the encyclopedic explanation of the Mask (a movie I want to see soon B-). I filed it for reference. DJ
#158560From: Brian BartlettAug 5, 1994 1:10 AM
Ronald, thanks for the information. Even though I rarely have to deal with such, it gives one the warm fuzzies to have such detailed information around. The Four Seasons of Southern California: Fire, Flood, Earthquake, and Riot.