PDA

View Full Version : adding 3rd int drive to G3 B/W



RayS
01-31-2002, 06:35 PM
Anyone ever tried to use the power from zip bay to drive a third IDE drive in a mac B/W G3 rev 2??

ricks
01-31-2002, 07:40 PM
Ray,

Power should be power, doesn't matter where you get the four prong molex. You can also use a Y adapter available at any supply house incl. Radio Shack.

The main problem of course is the data cable. If you use the data cable that would normally go to the zip you'll be limiting the drive you install on it to ATA3 speeds. I think ATA3 is how they measure grass growth.
Rick

wingman
01-31-2002, 10:16 PM
No need for the splitter, there should be a free 4-pin Molex connector hanging there -- for the Zip drive after all.

Besides the pokey channel, there are also issue of heat. That area gets very little if any ventilation.

Although it costs more, a PCI ATA card (66, 100, 133mhz) will provide a performance boost over the motherboards ATA/33. Not to mention each of the 3 HDs will sit as single masters on separate channels, also a boost in performance.

ricks
01-31-2002, 11:07 PM
What he said!
Rick

RayS
02-02-2002, 12:44 PM
Well I went ahead and installed it directly above the CD rom drive ( You have to reroute the ribbon cable) So far so good .

Don`t know about the ATA33 bus speeds but I am playing back 24 tracks Pro tools sessions from it .( same as from my 2 other ATA HD and my SCSI ) Will let you know if heat is a prob.

How can I check the data transfer speeds for all the different disks?


Thanks

ricks
02-02-2002, 02:14 PM
Not ATA33. What I said was ATA3. The bus that the zip would normally be connected to is an ATAPI bus which moves data at a ATA3 specification. ATA3 has a max transfer rate of 3MB/sec. You will get performance about equal to a zip or a CD-Rom with a hard drive on that bus, if that's OK then your set.

You can check your drive performance using ATTOs,FWB or MacBenchs drive tests. Usually around here we compare the ATTO as the standard so everyones numbers match up at least from the test basis.

You can get ATTOs ExpressTools from ATTOtech or off the Gurus ftp site.

Rick

kaye
02-02-2002, 02:51 PM
I thought ATA-3 was 16MB/s? k

ricks
02-02-2002, 03:38 PM
Hey Kaye,
Good thing somebodys got brain hook up. Yessireee, 16.6 MB/s it is. I thought ATAPI was at 3 but the ATAPI bus on the G3 is ATA3 rated.

Thanks for the correct-a-mundo.

Does staring at a monitor cause brain cells to die faster?
Rick

RayS
02-04-2002, 10:01 AM
ran the Atto express tools benhmark and got the lowest transfer rates( read /write) from my scsi disk and the fastest from the internal G3 B/W ata drive , the one usin the zip conn. had slower read times but faster write and sustained write speeds( twice as fast).The surprise was certainly the scsi disk, that was a third slower than the ata ????

ricks
02-04-2002, 04:57 PM
Hey, that's better than I'd have ever guessed you could get on the ATAPI bus, great.

What SCSI card are you running and what model drive? Maybe your scsi bus isn't as fast a bus as the ATA?
Rick

Is wonderful to have a good test tool to find out what's happenin with those drives!

RayS
02-04-2002, 08:47 PM
Hi Ricks ,
the card is a Grappler F1.18 ,the disk is a Seagate Medalist
7200 rpm ,4.5 Gb.
Here`s more
ST-34520N


UNFORMATTED CAPACITY (GB) ________________
FORMATTED CAPACITY (GB) __________________4.55
AVERAGE SECTORS PER TRACK ________________246 rounded down
ACTUATOR TYPE ____________________________ROTARY VOICE COIL
TRACKS ___________________________________36,024
CYLINDERS ________________________________9,006 user
HEADS ______PHYSICAL______________________4
DISCS (3.5 in) ___________________________2
MEDIA TYPE _______________________________THIN FILM/MR
RECORDING METHOD _________________________ZBR EPRML (16/17)
INTERNAL TRANSFER RATE (mbits/sec)________110 to 193.88
INTERNAL FORMATTED TRANSFER RATE (MB/sec)_14.0
EXTERNAL TRANSFER RATE (mbyte/sec) _______20 Sync
SPINDLE SPEED (RPM) ______________________7,200
AVERAGE LATENCY (mSEC) ___________________4.17
BUFFER (/optional) _______________________512 KByte
Read Look-Ahead, Adaptive,
Multi-Segmented Cache
INTERFACE ________________________________SCSI-3 Ultra
ASA II, S.M.A.R.T., SCAM level 2
BYTES PER TRACK __________________________115,078 average
SECTORS PER DRIVE ________________________8,888,924
TPI (TRACKS PER INCH) ____________________9570
BPI (PEAK KBITS PER INCH) ________________179.2
AVERAGE ACCESS (ms read/write)____________9.5/10.5
Drive level with controller overhead
SINGLE TRACK SEEK (ms read/write) ________2.5/3.5
MAX FULL SEEK (ms) _______________________21.0/23.5
MTBF (power-on hours) ____________________800,000
SHOCK (G's):
operating (Read/Write) __________10
abnormal ________________________
nonoperating 11/2 msec __________75/200
ACOUSTICS (Bels typ/max) _________________3.9/4.2
POWER DISSIPATION (watts/BTUs) Active ____14.1/48.1
Idle ______12.3/42.0
Standby ___4.1/14.0
POWER REQUIREMENTS: +12V START-UP (amps) _2.5
+12V TYPICAL (amps) __0.81 idle
+5V START-UP (amps) __
+5V TYPICAL (amps) ___0.51 idle
IDLE (watts) _________11.0
LANDING ZONE (cyl) _______________________AUTO PARK
IBM AT DRIVE TYPE ________________________0 or NONE

Physical:
Height (inches/mm): 1.007/26.9
Width (inches/mm): 4.000/101.6
Depth (inches/mm): 5.754/146.15
Weight (lbs/kg): 1.5/0.68

Already low-level formatted at the factory with 8033 spare sectors
per volume for both P and G lists.

ZBR = Zone Bit Recording = Variable sectors per track

Ray

ricks
02-04-2002, 11:22 PM
Hey Ray,
I just ran a benchtest on my rev2 B&W that has a ATA100 Barracuda on the ATA33 internal bus. I set the Max Transfer Size to 8mb which is what we've all kinda settled on as the most telling test and a standard size so we're all comparing apples to apples.

My BarracudaIII 7200 rpm drive got these numbers:

PR 23.96 MB/s
SR 23.45 MB/s

PW 22.02 MB/s
SR 20.50 MB/s

That's higher than the max capability of the Medalist drive. That drive was the internal factory drive that came in my 8600. I think in the 8600 on the narrow internal bus it moved 6 to 8 mb/s. I never wrote those down back then and am going by always faulty memory.

I believe that the Grapler is a 20MB bus max host.

It's really strange to me that your getting higher write speed with all I've heard about the slow write problems with the B&W. My ATA133 hardware raid setup gets in the mid 60s MB/s sustained reads and down to 49 in sustained writes. Same when I tested a scsi raid, the reads were way over 100MB/s sustained and the writes were still under 50MB/s. Someday I'll find a write fix and get rich if anyone still knows how to use a B&W by then.

I realized after posting that the IDE buses are not on the PCI bus and probably don't have the limitations of PCI connected devices. very interesting.

Sounds like you've found an appropriate way to add a drive and I got to learn something new from it.

Best of luck with the B&W, I love mine.
Rick

[This message has been edited by ricks (edited 04 February 2002).]

RayS
02-05-2002, 10:41 AM
Ricks,
this is where everything is :
internal ata0-id 0-cd-rom
internal ata0-id 1-maxtor 7200rpm ,30GB not partitionned
internal ata1-id 0-stock G3 drive 6GB
internal ata1-id 1-maxtor 40 GB 4 partitions
scsi pci slot 2- grappler card -seagate medalist 7200 rpm 4.5GB

also set the transfer rate to 8mb ran the benchmark :

scsi drive----------- G3 drive
r 8.14 ----------------17.43
w 8.16 ----------------7.56
sr 8.11 - ------------14.15
sw 8.14 --------------7.56
drive on Zip power
15.73
15.96
15.67
15.94
partitionned maxtor 40GB partition1
28.92
7.36
28.37
7.34
partitions2-3 &4
20.66
7.33
20.10
7.32


Ray





[This message has been edited by RayS (edited 05 February 2002).]

ricks
02-05-2002, 03:09 PM
Well Ray,
You've really given me something to think about. First, why the deplorable writes on the ATA33 bus yet the writes(and reads) are basically saturating the ATAPI bus with the new 30gig?

My Barracuda III is an ATA100 drive and I would expect similar performance on the ATA33 bus from most ATA66 and ATA100 drives. Your not getting the writes. Why?

I am surprised that the new Maxtor is saturating the bus so well on the ATAPI bus, that's great, I just want to know what the difference is between that and the ATA33 bus. Makes me wonder if the 30gig is faster than the 40gig and would perform even better on the ATA33 bus.

I see I'm going to have to try a couple of things in my B&W and see what happens.

I'll post back if there's anything interesting.

Rick