smdev@csustan.UUCP (Scott Hazen Mueller) (12/20/86)
I'd like to throw out to the net an idea I had last night:
Is it possible to hook an PClone hard drive to the Amiga external
drive port? To my mind, all it would take is one drive w/controller
(~300-400$), a box with a power supply ($100?) and a custom cable to
go from the card-edge connector on the controller to the Amiga port.
Total system cost: about $500 for 10 or 20 megs. It sounds too good
to be true --- where did I go wrong? If it can be done, I volunteer
to be a test site :->.
\scott
--
Scott Hazen Mueller lll-crg!csustan!smdev
2020 Cheyenne Way #137 --------------
Modesto, CA 95356 | F. - I. W. |
(209) 527-1203 --------------candym@calgary.UUCP (Mike Candy) (12/22/86)
In article <346@csustan.UUCP>, smdev@csustan.UUCP (Scott Hazen Mueller) writes: > I'd like to throw out to the net an idea I had last night: > > Is it possible to hook an PClone hard drive to the Amiga external > drive port? To my mind, all it would take is one drive w/controller > (~300-400$), a box with a power supply ($100?) and a custom cable to > go from the card-edge connector on the controller to the Amiga port. > > \scott A nice idea but it won't work. If you could rig a cable to get from your hard drive to the floppy expansion port, the ultimate problem is the rotational speed of the hard drive. A floppy rotates at 300 RPM giving the controller 200 msec to write 11 sectors of data. The hard drive rotates at 3600 RPM or 16.7 msec per revolution, this means the maximum number of sectors on a track would be 11 / (200 / 16.7) = .91 !! Hard disk controllers write raw data at a much faster rate, and I know of no way to get the Amiga to write data faster. Does anybody know what controller chip the Amiga uses? If it is an FDC765A or equivalent, has anybody tried an FDC7265 which is supposed to get 20% more storage and be "completely software- and pin-compatible"?
daveh@cbmvax.cbm.UUCP (Dave Haynie) (12/23/86)
> Summary: Sorry won't work > Does anybody know what [floppy] controller chip the Amiga uses? If it is an > FDC765A or equivalent, has anybody tried an FDC7265 which is supposed to > get 20% more storage and be "completely software- and pin-compatible"? Sorry, you're out of luck. The Amiga floppy disk is controlled by a section of the Paula custom chip. This chip is told by the 68000 what track to read or write, then it goes about reading or writing during a specific DMA slot assigned to disk I/O. You always read full tracks with this controller. Once a track has been read in by Paula, it must be decoded (Paula reads raw MFM data). The blitter is used to decode this, usually a sector at a time, and it achieves this decoding in one blitter pass (though as I recall, it takes three passes to encode the data). Because of all this, floppy disk buffers are constrained to reside in Chip memory. -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Dave Haynie {caip,ihnp4,allegra,seismo}!cbmvax!daveh Why look here for inspiration, when its all around you anyway. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
sean@ukma.ms.uky.csnet (Sean Casey) (12/24/86)
In article <346@csustan.UUCP> smdev@csustan.UUCP (Scott Hazen Mueller) writes: > Is it possible to hook an PClone hard drive to the Amiga external >drive port? To my mind, all it would take is one drive w/controller >(~300-400$), a box with a power supply ($100?) and a custom cable to >go from the card-edge connector on the controller to the Amiga port. Yeah. Why do Amiga compatable hard disk drives cost THREE TIMES AS MUCH as the same capacity IBM PC drives (with controller). Is there three times as much magic inside? Sean -- =========================================================================== Sean Casey UUCP: cbosgd!ukma!sean CSNET: sean@ms.uky.csnet ARPA: ukma!sean@anl-mcs.arpa BITNET: sean@UKMA.BITNET
jmpiazza@sunybcs.UUCP (12/24/86)
In article <1164@cbmvax.cbmvax.cbm.UUCP> daveh@cbmvax.cbm.UUCP (Dave Haynie) writes: > >... The blitter is used to decode [data from the read track], >usually a sector at a time, and it achieves this decoding in one >blitter pass (though as I recall, it takes three passes to encode the data). Which explains why copying disk to RAM: is so much faster than RAM: to disk. (I have a single drive system -- could you tell?) Flip side, joe piazza --- Cogito ergo equus sum. CS Dept. SUNY at Buffalo 14260 (716) 636-3191, 3180 UU: ...{rocksvax|decvax}!sunybcs!jmpiazza CS: jmpiazza@buffalo-cs BI: jmpiazza@sunybcs GE: jmpiazza
wagner@utcs.UUCP (12/25/86)
In article <1164@cbmvax.cbmvax.cbm.UUCP> daveh@cbmvax.cbm.UUCP (Dave Haynie) writes: >The Amiga floppy disk is controlled by a >section of the Paula custom chip. This chip is told by the 68000 >what track to read or write, then it goes about reading or writing >during a specific DMA slot assigned to disk I/O. Actually, there are three DMA slots (hardware manual, pg. 187). Why so many? If I understand the disk section of the hardware manual, each time slot transfers two bytes to/from DSKDAT(R) (pg. 235). Each slot should be good for 30K of data transfer (15K+ scans per second*2). The disk can only produce 28K/second of data (5 RPS*.5Ksectors*11sectors/rev). Are the second and third DMA slot always unused? Are they used for interleaving other disk work? Can any other disk work be interleaved? Can one disk seek while another transfers data? Can two transfer data at the same time? Can it walk and chew bubble-gum at the same time? :-) >You always read full >tracks with this controller. Once a track has been read in by Paula, >it must be decoded (Paula reads raw MFM data). The blitter is used to >decode this, usually a sector at a time, and it achieves this decoding >in one blitter pass (though as I recall, it takes three passes to encode >the data). Can the blitting to decode a sector overlap with reading another track? (I think that Matt first suggested this). In current practice (V1.2), does overlap occur? Is it envisioned for a future release? Presumably, it would require double-buffering for the tracks (not a bad idea, in and of itself). Perhaps it should be an option, so that storage-constrained systems wouldn't be hampered. >Because of all this, floppy disk buffers are constrained to >reside in Chip memory. I presume this is true only for the track buffers, and not for the cache? (although I notice that, in the current release, the memory comes out of chip and not fast memory). Michael
grr@cbmvax.cbm.UUCP (George Robbins) (12/25/86)
In article <5396@ukma.ms.uky.csnet> sean@ukma.ms.uky.csnet (Sean Casey) writes: > >Yeah. Why do Amiga compatable hard disk drives cost THREE TIMES AS MUCH >as the same capacity IBM PC drives (with controller). Is there three times >as much magic inside? Mostly because there are more PC compatible drive sets sold, by a factor of 30-300 thousand. There are many vendors of controller boards and chip sets, both domestic and asian. The market is very competitive and the controllers, and more so the drives are sold at a very small profit margin. The people selling Amiga drives are trying to find a way to get some return on their investment on initial quantites of a few hundred units. Obviously, the pricetag is going to be much higher. Also, PC drives require no casework, power supplies, or FCC approval. -- George Robbins - now working for, uucp: {ihnp4|seismo|rutgers}!cbmvax!grr but no way officially representing arpa: cbmvax!grr@seismo.css.GOV Commodore, Engineering Department fone: 215-431-9255 (only by moonlite)
lachac@topaz.RUTGERS.EDU (Gerard Lachac) (12/26/86)
In article <1172@cbmvax.cbmvax.cbm.UUCP> grr@cbmvax.UUCP (George Robbins) writes: >pricetag is going to be much higher. Also, PC drives require no casework, power >supplies, or FCC approval. ??????? Boy this one has got me totally confused!! Is IBM that powerful that it tells the FCC to "pass us by"??? Could you elaborate on this? Thanks -- ---------------------------------------------------------------------------- "Isn't fun the best thing to have?" lachac@topaz.rutgers.edu
wagner@utcs.UUCP (12/26/86)
In article <1768@sunybcs.UUCP> jmpiazza@gort.UUCP (Joseph M. Piazza) writes: >In article <1164@cbmvax.cbmvax.cbm.UUCP> daveh@cbmvax.cbm.UUCP (Dave Haynie) writes: >> >>... The blitter is used to decode [data from the read track], >>usually a sector at a time, and it achieves this decoding in one >>blitter pass (though as I recall, it takes three passes to encode the data). > > Which explains why copying disk to RAM: is so much faster than RAM: >to disk. (I have a single drive system -- could you tell?) I don't think so. Those blitter passes should be fast. But the fact remains that writing with AmigaDOS is about half the speed of reading (according to the benchmark program posted recently). Adequate caching of the disk does not improve this significantly. Nor does large blocksize. Still remains a mystery. Michael
grr@cbmvax.cbm.UUCP (George Robbins) (12/26/86)
In article <8118@topaz.RUTGERS.EDU> lachac@topaz.UUCP (Gerard Lachac) writes: >In article <1172@cbmvax.cbmvax.cbm.UUCP> grr@cbmvax.UUCP (George Robbins) writes: >>pricetag is going to be much higher. Also, PC drives require no casework, >>power supplies, or FCC approval. > > Boy this one has got me totally confused!! Is IBM that powerful that >it tells the FCC to "pass us by"??? Could you elaborate on this? As I understand it, the current FCC rationale is that expansion cards and devices that plug into an FCC approved system, and *do not* have any external connections do not *currently* require FCC approval. Also some devices the seem primarily passive, including interrconnection cables, adapters, joysticks and *mice* do not *currently* require FCC approval. Of course, the FCC is a government agency, and is not required to be sensible. These distinctions probably represent a compromise between a bureaucratic desire to control everything, and the amount of work required to do so... The ultimate irony is that the rules only apply to devices that generate and *use* radio-frequency energy. Appliances that don't use RF, but squirt it out anyway, like your bug zapper, electric drill, neon sign or vacuum cleaner are not covered by these regulations. -- George Robbins - now working for, uucp: {ihnp4|seismo|rutgers}!cbmvax!grr but no way officially representing arpa: cbmvax!grr@seismo.css.GOV Commodore, Engineering Department fone: 215-431-9255 (only by moonlite)
kdd@well.UUCP (Keith David Doyle) (12/27/86)
In article <5396@ukma.ms.uky.csnet> sean@ukma.ms.uky.csnet (Sean Casey) writes: >Yeah. Why do Amiga compatable hard disk drives cost THREE TIMES AS MUCH >as the same capacity IBM PC drives (with controller). Is there three times >as much magic inside? > No, we're just a good 3 times as desperate to get ahold of hard disks as the average IBM owner. The law of supply and demand. Keith Doyle ihnp4!ptsfa!well!kdd
lachac@topaz.RUTGERS.EDU (Gerard Lachac) (12/27/86)
In article <1180@cbmvax.cbmvax.cbm.UUCP> grr@cbmvax.UUCP (George Robbins) writes: >As I understand it, the current FCC rationale is that expansion cards and >devices that plug into an FCC approved system, and *do not* have any external >connections do not *currently* require FCC approval. So what your saying here is that when people start producing the ZORRO expansion boxes, and make things to go in them (eg SidecarD, clock, hardcard, etc) then these cards won't have to be approved. And if I understand right, then end bus cards like the Allegra don't have (or need) FCC approval... -- ---------------------------------------------------------------------------- "Isn't fun the best thing to have?" lachac@topaz.rutgers.edu
chapman@eris.BERKELEY.EDU (Brent Chapman) (12/28/86)
In article <5396@ukma.ms.uky.csnet> sean@ukma.ms.uky.csnet (Sean Casey) writes: >Yeah. Why do Amiga compatable hard disk drives cost THREE TIMES AS MUCH >as the same capacity IBM PC drives (with controller). Is there three times >as much magic inside? Unfortunately, it's not magic; it's just economics. Companies selling PClone hard drives can buy them from manufacturers at much lower prices than companies selling drives for Amigas, simply because there are a lot more PClones out there, and the PClone-serving companies can buy in much larger quantities. Also, there's a _lot_ more competition in the PClone market than in the Amiga market, which leads to still lower prices.. Brent -- Brent Chapman chapman@eris.berkeley.edu or ucbvax!eris!chapman
cmcmanis@sun.uucp (Chuck McManis) (12/29/86)
. = kdd@well (Keith David Doyle)
.. = sean@ukma.ms.uky (Sean Casey)
.. Yeah. Why do Amiga compatable hard disk drives cost THREE TIMES AS MUCH
.. as the same capacity IBM PC drives (with controller). Is there three times
.. as much magic inside?
..
. No, we're just a good 3 times as desperate to get ahold of hard disks
. as the average IBM owner. The law of supply and demand.
Argh! Please consider this, if you want to sell hard disks, and stay in
business, you have to sell them above your cost.
Take as a given that you cannot sell a hard disk to *everyone* that has
an Amiga, and if you are lucky you will sell to 10% of them. Now the
latest hearsay on Amiga sales puts them at 150K in the U.S. That is
15K hard disks. Assuming you are a conservative person and want to hedge
your bets, you try to break even at 10,000 units sold, and I will also
assume you can sell 10K in one year. The first thing to do is add up your
cost of the parts and labor to produce a hard disk subsystem .
Part I. Calculating the Costs
[See the notes below for some assumptions]
Hard disk drive, 20Meg 1/2 height $250
SCSI controller $75
Cable (50 pin, shielded) $5
Power Supply (12V@4A, 5V@3A, switcher) $30
Allegra sized PC board (4 layers) stuffed & tested $100
Enclosure for Amiga Interface $30
Enclosure for Hard Drive $50
Fan $1
Power Cord, Decals $1
Manual, Unit packaging (shipping Box etc) $3
-----
Per Unit Recurring Costs $545
FCC Approval $10000
Driver Software (by contracted consultant) $2500
Maintanence Utilities (backup, chkdsk, also cons.) $4000
Engineer (1yr, initial design, debug) $40,000
Technician(1yr, support, documentation) $18,000
Advertising (1yr, 4 large ads in two publications) $36,000
Office/Phone (could be a garage, 1yr) $12,000 min
-------
Minumum NonRecurring costs $122,500
or as a cost/unit ($122,500 / 10K Units) $12.25
=======
Your cost/drive shipped $557.25
Dealers Suggested List Price (50% margin) $1114.50
Discounted price (mail order) $900.00
Notes and Assumptions:
1.) You are buying 1000 drives at a time from the drive manufacturer,
as well as 1000 units of the other stuff.
2.) You can find a cheap contractor to write your software for you
and still have it be reliable.
3.) You advertise in specialize periodicals like AmigaWorld and
Compute! and stay away from BYTE.
4.) You do one mailing to the Commodore registered owners list -
150,000 17 cent stamps is $25,500.
5.) You have either no office or one of those one room jobs. Or
you run this thing out of your house which cuts down on how
much the consumers trust you to stay in business.
6.) All you are doing is selling hard disks for the Amiga, if you
do other things like sell software you can share office and
advertising costs, and maybe even your technician but that
only affects the enduser cost by about $10.
Part II. Reaping the rewards :
As the reader should note there is no mention of "your" salary in the
costs above. That is because you "own" the business and you get all
of the profits! So using the figures above lets assume you make your
market forecast and actually sell 15,000 units. Now the first 10,000
went to paying off the operating expenses of the company and the
FCC approval and such like. Now you get 5000 units of pure profit for
the year, lets see how that adds up ...
5000 units @ ($557.25 - $545.00 materials cost) = $61,250 for the
year.
Wow! You made $61,000 dollars. But what about next year? Will another
150K amigas be sold? How do you deal with all those people bitching
to buy $500 hard disks? Even if you never sell another disk in your
life your customers will be expecting support. What it boils down to
is now you have to sell at least another 5,400 drives to pay for the
Tech, the Advertising, and the Office, 8,660 drives if you want to
keep the Engineer around (more if you give him/her a raise) And another
5,000 on top of that to pay your own salary (unless you take a pay cut).
Thats another 10,000 drives minimum the second year.
Part III What is the Point?
The point of this entire article is to point out three things :
First, the Unit Recurring Cost to the manufacturer is biggest
contributor to price, since a reduction of $100 will pull $200 of
the suggested list price. When you sell to the IBM PC market, 10%
of the market is 60,000 drives and you don't need either a case or a
power supply since the PC provides them and you don't need any custom
software since IBM supplies that also. With a volume of 4000 drives a
month and the only parts being a $200 drive, $50 interface, and a
$.50 ribbon cable the suggested list is now around $500. With mail
order going for about $400.
Second, the amount of money the manufacturer is making is usually
so low that the difference between selling "at cost" and at their
regular price is at most $50. They make there money by selling 60,000
drives (thats $3,000,000 for the PC Drive seller) The dealers haul
in a bundle but one dealer may only sell 50 drives the whole year
so although he makes $25,000 he also has a lot of overhead like store
space and sales clerks etc.
Third and last, until the volumes go up, and the cost gets reduced
the Amiga drives will continue to be more expensive than PC drives.
Look at the Macintosh Plus drives, the only difference is the built
in SCSI interface on the Mac (a $100 charge to the Amiga drive maker)
and their disks are almost exactly $200 lower than the Amiga drives.
--
--Chuck McManis
uucp: {anywhere}!sun!cmcmanis BIX: cmcmanis ARPAnet: cmcmanis@sun.com
These opinions are my own and no one elses, but you knew that didn't you.gnu@hoptoad.uucp (John Gilmore) (01/03/87)
In article <1172@cbmvax.cbmvax.cbm.UUCP>, grr@cbmvax.cbm.UUCP (George Robbins) writes: > >Yeah. Why do Amiga compatable hard disk drives cost THREE TIMES AS MUCH > >as the same capacity IBM PC drives (with controller). > Mostly because there are more PC compatible drive sets sold, by a factor of > 30-300 thousand. There are many vendors of controller boards and chip sets, > both domestic and asian. The market is very competitive and the controllers, > and more so the drives are sold at a very small profit margin. > > The people selling Amiga drives are trying to find a way to get some return on > their investment on initial quantites of a few hundred units... The question is, why didn't the Amiga designers build it so you could plug in all those cheap IBM compatible hard disk drives? Like by including a standard disk controller chip on the motherboard? (The chips, connectors and all that would be cheap, since they are commercial parts used in all the IBM clones.) At least they could have done a SCSI port like Mac Plus. I suspect the answer is: because they thought they were building a "rock-shooter" (game machine) rather than a real computer. -- John Gilmore {sun,ptsfa,lll-crg,ihnp4}!hoptoad!gnu gnu@ingres.berkeley.edu I forsee a day when there are two kinds of C compilers: standard ones and useful ones ... just like Pascal and Fortran. Are we making progress yet? -- ASC:GUTHERY%slb-test.csnet
lachac@topaz.RUTGERS.EDU (Gerard Lachac) (01/04/87)
In article <1610@hoptoad.uucp> gnu@hoptoad.uucp (John Gilmore) writes: >all the IBM clones.) At least they could have done a SCSI port like Mac Plus. > I was curious recently about Hard Drives for the Mac + (my roomate has one) and was perusing Computer Shopper for prices. I couldn't find a 20 meg for under a grand. My question: What good is the SCSI port if the drives are this expensive anyway?? I was under the impression that with a SCSI one could plug in just about any Hard Drive and expect results. What gives?? -- ---------------------------------------------------------------------------- "Isn't fun the best thing to have?" lachac@topaz.rutgers.edu
fnf@mcdsun.UUCP (Fred Fish) (01/04/87)
In article <1610@hoptoad.uucp> gnu@hoptoad.uucp (John Gilmore) writes: >I suspect the answer is: because they thought they were building >a "rock-shooter" (game machine) rather than a real computer. Actually, my understanding of the situation is that is exactly what they were building. Turning it into a "real computer" was an afterthought. It is also the last computer without some sort of (even rudimentary) MMU that I am going to buy. I'm getting awfully tired of errant processes trashing the system and other processes... -Fred -- =========================================================================== Fred Fish Motorola Computer Division, 3013 S 52nd St, Tempe, Az 85282 USA {seismo!noao!mcdsun,hplabs!well}!fnf (602) 438-5976 ===========================================================================
farren@hoptoad.uucp (Mike Farren) (01/05/87)
In article <1610@hoptoad.uucp> gnu@hoptoad.uucp (John Gilmore) writes: >The question is, why didn't the Amiga designers build it so you could >plug in all those cheap IBM compatible hard disk drives? Like by including a >standard disk controller chip on the motherboard? (The chips, connectors >and all that would be cheap, since they are commercial parts used in >all the IBM clones.) At least they could have done a SCSI port like Mac Plus. > >I suspect the answer is: because they thought they were building >a "rock-shooter" (game machine) rather than a real computer. There's probably a bit of truth in that - after all, the Amiga DID start out as a game machine. However, I think that the bigger reason is simply that at the time the machine was designed, there were no "cheap IBM compatible hard disk drives" - a 10M drive cost approx. $1K. The development of a very large market was what got the drives for IBMs down to a reasonable price in the first place. Also, the idea of using IBM compatible controllers falls down when you consider that all of them are designed to the IBM I/O bus, which I would HATE to see in a 68000 machine. As far as SCSI, it's worth noting that Amiga hard drives are approaching SCSI drives pricing pretty rapidly. -- ---------------- "... if the church put in half the time on covetousness Mike Farren that it does on lust, this would be a better world ..." hoptoad!farren Garrison Keillor, "Lake Wobegon Days"