Best way to use 2 external writers - 1394 or USB2 ?

G

Guest

Hi.
I need some advice on how to make the best use of two external CD-R
burners with an IBM laptop (T23). The intention is to be able to burn
CD-R media in the least possible time, preferably while still being
able to use the computer for other applications.
I'm considering buying two units of an LG model which burns at 52x
speed. Of course I welcome any advice about the safest speed that I
could/should use in real world with such writer.

The laptop has two PC Cards to allow such setup (both cardbus):
- SIIG : one 1394a slot, two 1394b slots
- StartTech: two 1394a slots (one 6pin, one 4pin), two USB 2.0 slots
(the USB 2.0 slots aren't available yet, I'm still trying to make them
work)

I would like to know which method would allow me to burn faster and
with less load on the system, using the two burners simultaneously.
In other words, which approach would make the best possible use of
bandwidth and throughput allowed by the two cardbus slots.

Four possible methods come to mind:

a) In theory, I'd use two 1394b external enclosures, since I do have a
PC card for that (SIIG). But at this point, following my frustration
with an ADS enclosure, I regard Firewire 800 as pure mirage in the
desert. But I'd welcome some positive suggestions of a solution that
is proven to work with the SIIG card in Firewire 800. I've seen lots
of 1394b HD enclosures, but not 1394b enclosures for DVD/CD writers.
The ADS should do either, but does neither.

b) To the SIIG card's single 1394a slot, I could daisy-chain two
separate 1394a enclosures.

c) To the SIIG's single 1394a slot, I could connect a dual-drive
enclosure. I wonder how different this would perform from method "b".

d) To each of the StarTech's 1394a slot, I could connect separate
enclosures. (can't have both SIIG and StarTech cards inserted, as they
don't fit together physically).
I wonder if I could get better performance by using separate slots
from the same card.

And if I can ever make the StarTech card's USB 2.0 slots work, I'd
have two additional choices:

e) To each of the StartTech's USB2 slots, one separate enclosure.

f) One enclosure in a USB 2.0 slot, other in the 1394a slot.

Thanks for any word of advice.
If possible, please recomend a particular enclosure model that is
known to work well.
Best to all.
 
A

Andrew

In comp.sys.laptops (e-mail address removed) wrote:
: Hi.
: I need some advice on how to make the best use of two external CD-R
: burners with an IBM laptop (T23). The intention is to be able to burn
: CD-R media in the least possible time, preferably while still being
: able to use the computer for other applications.

You might want to consider Firewire (1394) instead of USB 2.0 because,
as I understand it, Firewire uses its own controllers whereas USB 2.0
uses the computer's CPU to shovel data. Transferring data between two
USB 2.0 devices might tie up the CPU more than two Firewire devices
that don't tax the CPU might. Just a thought.

Andrew
--
----> Portland, Oregon, USA <----
*******************************************************************
----> http://www.bizave.com <---- Photo Albums and Portland Info
----> To Email me remove "MYSHOES" from email address
*******************************************************************
 
A

Arny Krueger

I need some advice on how to make the best use of two external CD-R
burners with an IBM laptop (T23). The intention is to be able to burn
CD-R media in the least possible time, preferably while still being
able to use the computer for other applications.
I'm considering buying two units of an LG model which burns at 52x
speed. Of course I welcome any advice about the safest speed that I
could/should use in real world with such writer.
The laptop has two PC Cards to allow such setup (both cardbus):
- SIIG : one 1394a slot, two 1394b slots
- StartTech: two 1394a slots (one 6pin, one 4pin), two USB 2.0 slots
(the USB 2.0 slots aren't available yet, I'm still trying to make them
work)
I would like to know which method would allow me to burn faster and
with less load on the system, using the two burners simultaneously.
In other words, which approach would make the best possible use of
bandwidth and throughput allowed by the two cardbus slots.

Building up your own extenal drive enclosures is probably a good idea, but
why not avoid putting your eggs in one basket, and use combo USB2/Firewire
enclosures like these?

http://www.bixnet.com/firenkitfor5.html

Then armed with appropriate USB2 and 1394 PC cards, see what works best for
you? You might find your best operation with mixture or a match.

I have enough experience with actual use and know enough about how these
controllers are actually implemented to know that the actual results you
obtain, can't be reliably hypothesized from theoretical studies.

Here's an example. A friend wanted to use an audio interface that is
attached by firewire with a certain PC chipset. The vendor, presumably
having been burned with other customers, recommended against relying on the
motherboard chipset's firewire controller, and instead recommended using a
certain firewire controller on a PCI card. I assure you that an a priori
theroetical study would indicate this is so much garbage.

My friend attempted to use the on-board firewire controller and the machine
spontaneously rebooted every time he tried to record with the audio
interface. He then installed the recommended PCI card and everything ran
well.

Probable-cause was a defective implementation of the firewire chip on the
motherboard chipset. It was a Nvidia chipset- 'nuff said.

A lot of the issues you are strugging with can easily involve factors that
aren't part of the standard theoretical description of computer equipment
operation.
 
V

visitor2

Transferring data between two USB 2.0 devices might tie up the CPU
more than two Firewire devices that don't tax the CPU might. Just a thought.

I tend to agree with that, from what I've read elsewhere.
And unless I can fix the issue with the USB2 installation, the fact is
that, in practice, I will be left with either of two adapters for my
laptop, both of which only granting Firewire (1394a) operation.

Another aspect which puzzles me is the actual rate of data throughput
in PC cards.
There is one manufacturer from Taiwan (ViPower) which makes several
models of external enclosures, with several connection types:
Firewire, USB 2.0, Cardbus, pcmcia.

In their site, there's a little chart stating that CD operation would
perform as below:
(It states the maximum transfer rate achieved by each connection type,
and which CD speed allows for it)

SPEED USB 1.1 pcmcia Cardbus
Firewire USB 2.0
6x 900k
24x 2000k 3600k
40x
(?)
6000k 6000k

So, the best that could be gotten with USB 1.1 is 6x operation. 24x
would be the best out of either pcmcia or Cardbus, with the latter
having higher transfer rate at such speed. USB 2.0 and Firewire, in
theory, would happily use speeds hiher than 40x.

But then, the inevitable question to my situation: if the above chart
is accurate, then no matter if USB 2.0 or Firewire is used, the
transfer rate will be cut down to the limits of cardbus and I wouldn't
be able to burn at speeds higher than 24x.
Which brings me again to the availability of cardbus connecting
enclosures. If I used them with each cardbus slot, instead of plugging
two enclosures to a Firewire card in a single Cardbus slot, wouldn't
it be a more agile method ?
I know, all theory, but that's all I've got.
Thanks for helping, Andrew.
 
V

visitor2

Speed USB 1.1 pcmcia Cardbus Firewire USB 2.0
6x 900k
24x 2000k 3600k
40x
(?) 6000k 6000k
 
V

visitor2

Arny Krueger said:
why not avoid putting your eggs in one basket, and use combo USB2/Firewire
enclosures like these? http://www.bixnet.com/firenkitfor5.html

Thanks Arny, it seems like a nice and cheap suggestion, I've made a
note of it.
Then armed with appropriate USB2 and 1394 PC cards, see what works best for
you? You might find your best operation with mixture or a match.

Yes. The only thing is I've already spent preciously saved money
attempting to find a working solution, so I already have two different
cards as I mentioned in the original post, and so fat USB 2.0 is not
an option since the card that offers it didn't install successfully.
Whereas the 1394 entries appear sound in Device Manager.
I have enough experience with actual use and know enough about how these
controllers are actually implemented to know that the actual results you
obtain, can't be reliably hypothesized from theoretical studies.

Excellent point.Your friend probably found that while trying to make
his stuff work.
It seems we're still shooting in the dark whenever we attempt a
different hardware solution. And we have no guarantee that one
equipment will work with other.
Probable-cause was a defective implementation of the firewire chip on the
motherboard chipset. It was a Nvidia chipset- 'nuff said.

Before comitting to buy the equipment, I made note of remarks which
pointed at best manufacturers and best chipsets. Things to look for
would be TI 1394 chipset, Oxford 911 implementation (Oxford 922 for
1394b), and SIIG and ADS were among top respected brands. I have a
feeling the SIIG card is alright, but the ADS enclosure doesn't
inspire trust.
USB 2.0 was a desperate attempt to have SOMETHING to count on. But I
just read that with USB 2.0, the chipset to look for is NEC, while
this Startech uses other one, I guess ALi.
So I'm trying to spot some kind of solution from whatever tools that I
already have. Either cards appear OK in Firewire mode, I just don't
have a Firewire enclosure, must be special-order where I live (outside
the US), like I did for the ADS, which sadly didn't work. So it's more
than one issue to deal with.
I just found a local shop that carries a double-drive enclosure that
is both USB2/Firewire capable. Very, very expensive, but it could be a
last resource, even though I'd prefer separate enclosures (not all
eggs in one basket, and easier to carry one around).
Thanks very much for your advice. Please, have a look at my response
to Andrew, and the interesting chart that I mentioned. It got
word-wrapped in the first post, but I got it right in a second
attempt.
 
L

Logan Shaw

SPEED USB 1.1 pcmcia Cardbus
Firewire USB 2.0
6x 900k
24x 2000k 3600k
40x
(?)
6000k 6000k

If you're going to use spacing to make a chart like that, you
probably want to use a fixed-width font instead of whatever
proportional one you used. There is no way for anyone to know
what font you used, so the spacing will be all off for more people.
Even if there were a way to specify the font, the spacing might
not be exactly the same and/or not everyone will necessarily even
have that font.

- Logan
 
H

Hal Laurent

Logan Shaw said:
If you're going to use spacing to make a chart like that, you
probably want to use a fixed-width font instead of whatever
proportional one you used. There is no way for anyone to know
what font you used, so the spacing will be all off for more people.
Even if there were a way to specify the font, the spacing might
not be exactly the same and/or not everyone will necessarily even
have that font.

This is a news group, Logan. There aren't any fonts associated
with the posts. What font is used for display is the purvey of
your newsreader. Your newsreader may very well have a way
of displaying a post in fixed-width font (and it doesn't matter
which one, all fixed-width fonts will line up the same way).
If it doesn't, you can always cut-and-paste the chart into
Notepad or something and choose a fixed-width font there.

Hal Laurent
Baltimore
 
S

Scott Dorsey

Hal Laurent said:
This is a news group, Logan. There aren't any fonts associated
with the posts. What font is used for display is the purvey of
your newsreader. Your newsreader may very well have a way
of displaying a post in fixed-width font (and it doesn't matter
which one, all fixed-width fonts will line up the same way).
If it doesn't, you can always cut-and-paste the chart into
Notepad or something and choose a fixed-width font there.

The problem is that the original poster did not compose it with a
fixed-width font. Therefore it is unreadable if you are using anything
other than whatever variable-width font it was composed with. Since nobody
knows what font it was composed with, it is therefore unreadable by anyone.
This is not useful.

It is also crossposted all over the place, which is also a bad thing.
--scott
 
L

Logan Shaw

Hal said:
This is a news group, Logan. There aren't any fonts associated
with the posts. What font is used for display is the purvey of
your newsreader. Your newsreader may very well have a way
of displaying a post in fixed-width font

It does, and I am using it. Which is the why the improperly-formatted
table above comes out wrong.

- Logan
 
H

Hal Laurent

Logan Shaw said:
It does, and I am using it. Which is the why the improperly-formatted
table above comes out wrong.

Ah, sorry! I misunderstood what you were saying. You're quite
correct.

Hal Laurent
Baltimore
 
V

visitor2

I'm sorry for the inconvenience. Please have a look again, I re-posted
the chart and this time it appears to be readable.
 
V

visitor2

I already apologized, and re-did the chart, maybe it's now
understandable. If not, I'm sorry again not being able to fix it.

As to cross-posting being "a bad thing", it's something that is hard
to understand. I remember once posting separate copies of a same
message to each of several groups that the subject was related to. At
that time, I DID think that I shouldn't just lazily shoot it to
several groups. Well then I got a reply from a very furious group
user, stating that the right thing to do was exactly the contrary so
people and software would have an easier time managing a message that
was already read.
So again I'm sorry, Scott, I hate the idea that I have caused
inconvenience to anyone. (altgough this does sound like an old joke of
the soviet-Jew coming back from a concentration camp in Siberia: tried
to do right, did "wrong" again, sent back).

So how about sharing some knowledge and comment the mentioned
manufacturer's transfer rates, which look pretty odd to me (the
cardbus at max 24x for a 3600k rate, while 1394 and USB2 at above for
loose 6000k). It suggests that Cardbus won't allow either 1394 or USB2
full capacity. Do you know something about this ?
Thanks for your help.

Speed USB 1.1 pcmcia Cardbus Firewire USB 2.0
6x 900k
24x 2000k 3600k
40x
(?) 6000k
6000k
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Top