I need some help with this

[dbminter]

Well, opening up the case was a goddamn waste of time. The PCI card won't even start! Code 10. Cannot start. Fuckers! I guess the card is too old.

[cornholio7]

using hdd tools on my usb2 i get this if it helps you at all

 

I 06:03:54 Average I/O Rate: 25.755 KB/s - Maximum I/O Rate: 25.755 KB/s

[dbminter]

Actually, yes, I think it does.

 

 

That I/O rate is in the same mean as my own. Between 25 and 31. How long did it take for that to run? Approximate time will do. And, what is the size of the drive?

 

 

Thanks!

[cornholio7]

i ran it for about 1 min on a 160gb drive

[dbminter]

EXCELLENT!

 

 

That's basically how long I let my values run. Until the I/O Rate returns a solid value, or a minute, or when the values level off... or I get pissed at waiting.

[dbminter]

Now, more importantly... how confident are you that your USB 2 HD is running at Hi-Speed?

[cornholio7]

totally

[dbminter]

Good!

 

 

I am still puzzled by at least one thing, though.

 

 

What is causing that Hi-Speed device connected to a slow speed port message?

[dbminter]

Okay, I think I finally have some "SENSE" of this damn problem. There's something happening that I cannot explain its reason, but, it's the cause. There ARE two different speeds being reported by the 2nd USB drive. Whenever the 2nd drive is connected/powered on for the first time, it behaves "normally" writing an image file in Build mode at about 11x. Whenever this drive is the ONLY one connected, it can be powered off/disconnected and reconnected without problem. BUT, if the drive is connected WITH the other USB 2 HD, THEN something interesting happens. Disconnecting the 2nd drive or powering it off and reconnecting it/powering it on causes the slow speed thing. The write rate drops to about 4x instead, or about 1/3. WHY this happens...

 

 

Did a quick test with this in HDD Tools. The I/O Rate plunges from 27,xxx KB/s to 9xx KB/s. Need to reboot and see if this process repeats on the first drive. IF it doesn't, then, it seems to backup, FINALLY a conclusion of mine through all this silly process!

[dbminter]

Okay, rebooted, checked the I/O's on drives 1 and 2, powered off drive 1, powered it back on, rechecked the I/O, and got the same value as before. Powered off drive 2, powered it back on, checked the I/O, and, as expected, the I/O plumetted. Checked the I/O on drive 1, and, strangely, as I expected drive 1's I/O did NOT change. Without rebooting, turned off drive 1, turned it back on, rechecked the I/O, and it had not changed! Checked the I/O on drive 2 again, and, it was the expected still lesser value.

 

 

So, the conclusion is that something in drive 2 causes this effect ONLY if it is NOT the only attached drive and ONLY if it has been powered off/powered back on or disconnected/reconnected. Which means for drive 2, I need to remember to reboot IF I ever have to turn it off/disconnect it and turn it on/reconnect it.

 

 

I wonder if the same process repeats if I connect one drive to the front and another to the back? I wonder if it repeats if I use Safely Remove Hardware first... time to get to work! It DOES repeat if I Safely Remove first AND power off and power on. It does NOT repeat if I Safely Remove and disconnect and reconnect the drive. Time to test just disconnecting and reconnecting. It's only with the drive powered off and on, it seems. Even without Safely Remove, disconnecting and reconnecting the data cable did not cause the problem. And, the last test. If the drive is powered off at any point and powered on, the system MUST be rebooted to restore the full write speed. Just logging off and back on did NOT help.

 

 

I also wonder what it is about the 2nd drive that causes this phenomenon... something in the bridge? ALI?! Or the drive in the enclosure itself? DAMN! I had the drive opened up the other night! I should have checked the chip when I was able to look at it. Oh, well, at least, once I got rid of the screw that WOULDN'T come out without pliers... the drive is easily reopened for future examination.

[chewy]

using hdd tools on my usb2 i get this if it helps you at all

 

I 06:03:54 Average I/O Rate: 25.755 KB/s - Maximum I/O Rate: 25.755 KB/s

 

 

26KB/s - 26KB/s

??????????????????????????????????????????????????????????

 

or 26,000 KB/s

[dbminter]

Yeah, 26,xxx instead of 26. Typed in anger.

[chewy]

you said 50 minutes for say over 4 gigs

 

4,360 MB / 3000 seconds

 

1.4MB/s too slow

 

you should be getting at least 10MB/s 436 seconds

 

436/60 about 7 minutes

 

I was doing this a few years back over in the msnewsgroups with W98

when somebody jumped my shite over my methods, a senior w98

performance msmvp verified it was the most accurate real world test.

[cornholio7]

using hdd tools on my usb2 i get this if it helps you at all

 

I 06:03:54 Average I/O Rate: 25.755 KB/s - Maximum I/O Rate: 25.755 KB/s

 

 

26KB/s - 26KB/s

??????????????????????????????????????????????????????????

 

or 26,000 KB/s

we use dot instead of comma over here 4,35gb takes just over 4 mins

[dontasciime]

Should be getting anything from 22 meg a second to about 28meg max over usb2.

 

Db if you buy a device and it says usb2.0 (full speed) it means it will work on usb2 and 1 and give you the full 12mbps, (dodgy wording and has conned loads of people)

 

If you buy something that says usb2 high speed *with certified logo then it should signal at 480mhz and has max theoretical transfer of 480/8... though 480/8/1.125/and maybe further divided by another 1.125 to get a more accurate max theo transfer speed.. which would be 47,407k a second..

 

To further complicate it Not all logo'd certified high speed usb2 devices are that high speed , as previously mentioned i have a topfield 5800 Pvr that has this LOGO , due to the nature of the machine though it can only transfer just over 12mbps @ about 16mbps and as such can be certified HIGH SPEED, Annoyed the hell outta me as i was expecting like all my other high speed usb2 devices to be able to sustain at least 22 meg a second.

 

So as long as your getting 22meg a second say whilst builing iso , shrink etc then your good to assume your using usb2 high speed.

Edited August 11, 2006 by dontasciime

[chewy]

using hdd tools on my usb2 i get this if it helps you at all

 

I 06:03:54 Average I/O Rate: 25.755 KB/s - Maximum I/O Rate: 25.755 KB/s

 

 

26KB/s - 26KB/s

??????????????????????????????????????????????????????????

 

or 26,000 KB/s

we use dot instead of comma over here 4,35gb takes just over 4 mins

 

whoever said science and math were the universal language lied

 

26 x 10 to the third KB/s

 

but good chipset in your usb2 corn

 

let the dogs out:

http://www.newegg.com/Product/Product.asp?...N82E16817106097

live in the fast lane

Edited August 11, 2006 by chewy

[dbminter]

If I'm getting 28,xxx KB/s, then, that should be roughly equivalent to 28 MB/s. CONSISTENCY, PEOPLE! So, that seems to be on par with what USB 2 should be.

 

 

The slow speeds appear to be caused by that one weird case that happens with the 2nd drive, but, not with the first. Either because of the design, construction, or configuration of the HD in the enclosure or the enclosure itself. Because it only happens when the drive is powered off and then back on.

[Defenestration]

Approx 28MB/s for an external enclosure is VERY good (but within specs)! Does the 2nd problematic enclosure have the same chipset as the first ?

Edited August 11, 2006 by Defenestration

[dbminter]

I don't know the chip set in the first drive. Never opened it. The 2nd one I opened but didn't think to check. It may be NEC, BUT I could just be transposing the chip that was on the PCI card that didn't work. However, opening the drive to check, should I ever need to, is relatively easy.

[Defenestration]

I mean the interface chipset on the enclosure rather than the drive ? I would not recommend opening up the drive itself since this is sealed in a dust free lab, and is never supposed to be opened.

[dbminter]

Yeah, I know what you meant. The bridge chip. We've had more than a fair share of bridge posts with those bloody ALI's and semaphore time outs.

[GOL]

As it may have already been said in here usb1.1 is 11MB/s i.e...1mb/s

USB 2.0 is 480MB/s i.e. 40mb/s

[Defenestration]

As it may have already been said in here usb1.1 is 11MB/s i.e...1mb/s

USB 2.0 is 480MB/s i.e. 40mb/s

Sorry to be picky, but......

 

It's a generally accepted notation that bytes (being larger) is denoted by an uppercase 'B' and bits is denoted using a lowercase 'b'.

 

Your calculation is also a little off. There are 8 bits in a byte, so....

 

USB 2.0 -> 480 Mb/s (ie. Megabits/second) = 480 / 8 = 60 MB/s (Megabytes)

 

USB 1.1 Hi-speed -> 12 Mb/s = 12 / 8 = 1.5 MB/s

 

USB 1.1 -> 1.5 Mb/s = 1.5 / 8 = 0.1875 MB/s = 192 KB/s (Kilobytes)

 

These are only maximum theoretical bandwidth - As a (rough) general rule of thumb you can half it to get the actual bandwidth. Also, USB has a higher overhead than Firewire so although USB 2.0 has a higher theoretical bandwidth, Firewire often has higher actual bandwidth.

 

Edited August 16, 2006 by Defenestration

[chewy]

Anyone getting over 20MB/s sustained transfer from usb2 please post results.

[dbminter]

Sorry to be picky, but......

 

 

Be picky all you like, because, in the end, the notation of the units, etc. doesn't matter a flying fuck. The manufacturers NEVER stick to anything consistent. They never explain if they're using megabyte, megabit, or mebibyte. They often times use all captial letters as a shouting advertising technique, so, it doesn't matter.