Slow speeds with Esata & Si3132 Expresscard

I'm wondering if anyone here uses a Si3132 chipset based Expresscard in order to run Esata. I'm getting incredibly underwhelming transfer rates of 45-50 mb/s when it is connected to my Dell E1705 laptop. I know there has to either be something wrong with the laptop chipset or the Expresscard itself because I get much higher speeds of around 70 mb/s when the same external hard drive is connected to a desktop.
The incredibly odd thing is, if I use a fresh installation of windows, with no drivers other than that for the si3132 expresscard installed, this 45-50 mb/s bottleneck doesn't seem to exist until I restart windows for the first time. From that point on, the bottleneck is pretty much there permanently.

I was just wondering if anyone has a similar setup working without any issues, or if anyone out there has a similar issue. It's driving me insane that I know I can somehow get my external hard drive working at full speed, but something is limiting it.

if it helps, my system specs:
core (first gen) duo 1.83ghz
Geforce 7900 go GS
2 gigs of ram @ 667 mghz
and yes, I am using the latest drivers.

 

I have a SIIG eSATA II ExpressCard and I experience rather low transfer rates as well around 50MB/s.

I have had alot of trial and error with this thing. See the problem was that with the drive plugged into the siig expresscard wouldn't format and would lag my computer, however it was only just recently that I discovered when I put the jumper on the HDD limitting the sata II hdd to sata I speeds, the drive then worked!

Good News right? Except that the transfer rate is much to slow.. I am looking to exchange my card for a different eSATA card. Which card do you have?

 

Under heavy file transfers peaking the limit, what is your CPU activity like? Is it skyrocketing, or is is pretty minimal as if it is being offloaded by the chipset?

 

I bench marked my HD connected via esata and got an average read speed of 55MB/s. I have a WD 250GB Caviar SE16, if you check tomshardware harddrive benchmark list this is about right.

http://www23.tomshardware.com/storage.html?modelx=33&model1=136&model2=124&chart=33

The 300MB/s speed refers to the SATA-II interface, which can transfer at 300MB/s, but that will ONLY occur when transferring data between the PC and the buffer on the disk. Once the buffer is full (or empty, depending on the direction of transfer) the transfer can't happen any faster than the platters on the disk can provide the data. That's why, in real life applications, there's little difference in the performance of SATA-I and SATA-II drives => in fact, WD has not bothered to update the interface on their Raptors beyond SATA-I, yet they are still the highest performing SATA drives you can buy (and they DO include NCQ support, which gives much more performance benefit than a switch to SATA-II would).

 

Dietcokefiend, to answer your question, my CPU activity is fairly minimal, usually staying under 10 to 15% usage.

I have found a strange temporary solution to the speed problem, which is to not plug in the expresscard + hard drive until after windows has finished booting up.

For whatever reason, if I boot the computer up with the actual expresscard plugged in, I will see the speed being limited. My only guess is that my laptop's chipset is doing something where the expresscard SATA controller's bandwidth becomes limited to that of my internal hard drive (which is also at around 45-50 mbs at its peak). I'm fairly certain that my laptop's internal hard drive is only SATA I, while my external is SATA II, so it is possible that my external hard drive is being "stepped down" even though it's on a totally different SATA controller.
Granted, I doubt stepping down to SATA I performance would actually limit my transfer rates to be a whole 30 mbs slower, because as far as I know, the difference between SATA I and SATA II on real world performance is practically negligible.