R61 transfer speeds with PCMCIA cards

[charlienyc]

hi,
my R61 8932CTO running Win7/32 has an unusually slow file transfer rate when an external drive is connected using a PCMCIA card, either eSATA or Firewire 400. i've used the computer in this configuration quite often in the past, yet have never run into this problem before. the one thing that's possibly different is that i have the machine connected to a dock. i am using the laptop's PCMCIA slot though. what would cause this? i try to run ThinkVantage and Windows updates pretty frequently. did Windows updates cause this? old BIOS?
cheers,
-charlie

[precip9]

I am surprised that you ever had a different experience.

There is a problem that seems common to the Windows drivers for all the chipsets of external ESATA cards, limiting the bandwidth to a fraction of the slot capacity.

For these reasons, I abandoned the use of ESATA slot for peripheral expansion. USB 2.0 is actually much faster in practice.

[charlienyc]

thanks for your message. for the work i'm doing, i need synchronous data transfer. Firewire and eSATA are synchronous whereas USB is asynchronous. i did try daisy-chaining a hard drive through a firewire audio interface. this was connected to the on-board 4-pin 1394 port and worked fine. that was an interim solution though, as interfaces and drives should be on different buses.

[axur-delmeria]

Why not use the Expresscard slot (assuming that your R61 has one).

[Hans Gruber]

I have one esta external hard drive. Essentially the speed is equal to a standard 3.5" drive. A bit faster than USB 2.0 but hardly a fast connection.

[precip9]

thanks for your message. for the work i'm doing, i need synchronous data transfer. Firewire and eSATA are synchronous whereas USB is asynchronous. i did try daisy-chaining a hard drive through a firewire audio interface. this was connected to the on-board 4-pin 1394 port and worked fine. that was an interim solution though, as interfaces and drives should be on different buses.

It sounds like you might be overdubbing. But allow a questioning of the synchronous requirement. Hard drives have latency much larger than the usb interface. Windows also buffers writes, unless you have turned off write-buffering. If you haven't turned off write buffering, and it works, then usb 2.0 is worth a try.

The synchronous advantage of Firewire is essential with a dubbing loop that does not depend on hard drive writes. Then the full-duplex guaranteed latency of Firewire really does matter.

Back around 2003, I did a lot of on-site recording with a 2.1 gHz Centrino single core. It was connected to the interface with T.I. Firewire, but the write drive was just USB 2.0. Steinberg captured 16 channels/96kbs with no glitches, because the Firewire and USB had separate buses to the chipset.

[charlienyc]

thanks for all the responses.

actually i'm trying to sort this issue because i need to mix down 30-40 tracks at 24bit / 96k. i've been able to do it before via eSATA. you're exactly - separate buses are needed. that's why i'm using the onboard Firewire port for my interface.

btw, this machine has a PCMCIA slot, not expresscard. weren't they mutually exclusive on ThinkPads?

i have not had time to test, but now i really do think the dock is at fault. i'll post an update once i can spend more time with this.

It sounds like you might be overdubbing. But allow a questioning of the synchronous requirement. Hard drives have latency much larger than the usb interface. Windows also buffers writes, unless you have turned off write-buffering. If you haven't turned off write buffering, and it works, then usb 2.0 is worth a try.

The synchronous advantage of Firewire is essential with a dubbing loop that does not depend on hard drive writes. Then the full-duplex guaranteed latency of Firewire really does matter.

Back around 2003, I did a lot of on-site recording with a 2.1 gHz Centrino single core. It was connected to the interface with T.I. Firewire, but the write drive was just USB 2.0. Steinberg captured 16 channels/96kbs with no glitches, because the Firewire and USB had separate buses to the chipset.

[precip9]

On Thinkpads with an internal pci-e bus, which includes at least the T400/500 going forwards, the old Cardbus/PCI slot had to be "faked" by an interface chip supplied by Yamaha. The same chip does the Firewire on such machines, with a really bad audio reputation. It also makes using a SATA adapter in the PCI slot (as opposed to the Expresscard slot, which is 1X PCI-e) a bad idea, because the "faked" PCI slot is on the same bus as the Firewire.

[charlienyc]

On Thinkpads with an internal pci-e bus, which includes at least the T400/500 going forwards, the old Cardbus/PCI slot had to be "faked" by an interface chip supplied by Yamaha. The same chip does the Firewire on such machines, with a really bad audio reputation. It also makes using a SATA adapter in the PCI slot (as opposed to the Expresscard slot, which is 1X PCI-e) a bad idea, because the "faked" PCI slot is on the same bus as the Firewire.

This is excellent information - thank you. Time for an upgraded (i.e. new) machine!