Only with Linux to date. I have just installed rtlinux and have not yet tried it with the board.
At 11:30 PM 6/19/02 -0400, you wrote: >At 11:27 AM 6/19/2002 -0400, Harland Alpaugh wrote: >>I am using a PD-MFS board on a Pentium 3 machine and have not been able >>to get the card to PCI host bus rate above approx. 100 samples per >>second. The input to the board from the outside world is on the order of >>1 million samples per second. This is with Linux. > >With Linux or with RTLinux? As Loren Frank describes below, under RTLinux >the data transfer rate for a PD2-MF64 is much slower than the theoretical >PCI bus maximum (33 MHz x 32 bits, or 133 MB/second). We have found it to >vary between 6.6 MB/second (fastest, and acceptable since you can actually >pass all the data to the computer at or near the board's top data rate) >and 2.2 MB/s (slowest, and unacceptable, since it is too slow to transfer >all the data collected at a high rate). This appears to depend not so much >on the processor (Athlon vs. Pentium) as on the motherboard: the Intel >i850 chipset works poorly, and the new nVidia nForce chipset seems to work >even worse in spite of its theoretical speed. > >These results, by the way, assume you use the faster pd_ain_xfer_samples >function, which theoretically use DMA. If you are taking samples one at a >time, it will be much slower. > >The reasons for this problem are not clear: we have tried to get UEI to >explore this problem but they have not been very helpful so far. We would >like to use their cards for a commercial product, so we are going to >continue to try to get to the bottom of this. > >To get back to your original question: if you want to transfer data pretty >fast under RTLinux, I can send you some GPL code derived from Loren's as >an example. I have not tried the Linux drivers, but since you are on the >rtl list, I would suggest if the linux drivers don't work for you you try >collecting the data under RTL and passing it to your program via shared memory. > >Good luck, > >Pablo Alvarez > >P.S. Does anybody else have experience with this card at high data >transfer rates? Any ideas on why the chipsets would have that kind of effect? > > > > >At 10:02 PM 6/18/2002 -0400, Loren Frank wrote: >>On Tue, 18 Jun 2002, Khoa Nguyen wrote: >> >> > > >> > > Anyone have comments about or experience with UEI DAQ boards? >> > > >> > I made a couple of modifications for my PD2-DIO-128 boards. Don't know >> if this >> > is the one you have. >> >>I'm using three PD2-MF64 boards, with two of them taking continuous data >>at 30 >>KHz / channel. They are currently redesigning these boards to correct a small >>timing problem, but for the most part I've been happy with them. >> >>The only serious problem I've run into is that, for reasons that still >>are not >>clear to us, high rate data collection works well on Athlon machines, but >>the Pentium 4 bus seems to be too slow (or the driver is poorly written). >>That issue only comes up at very high speeds, though, so for more normal data >>acquisition applications, they seem to work well. >> >>Loren >> >> >>-- [rtl] --- >>To unsubscribe: >>echo "unsubscribe rtl" | mail [EMAIL PROTECTED] OR >>echo "unsubscribe rtl <Your_email>" | mail [EMAIL PROTECTED] >>-- >>For more information on Real-Time Linux see: >>http://www.rtlinux.org/ > > >------------------------------------------------------------ >Pablo Alvarez, Ph.D. Phone: (617) 867-0244 >Software Engineer email: [EMAIL PROTECTED] >Neural Arts >21 Bay State Rd. >Boston, MA 02215, USA >------------------------------------------------------------ > > >-- [rtl] --- >To unsubscribe: >echo "unsubscribe rtl" | mail [EMAIL PROTECTED] OR >echo "unsubscribe rtl <Your_email>" | mail [EMAIL PROTECTED] >-- >For more information on Real-Time Linux see: >http://www.rtlinux.org/ -- [rtl] --- To unsubscribe: echo "unsubscribe rtl" | mail [EMAIL PROTECTED] OR echo "unsubscribe rtl <Your_email>" | mail [EMAIL PROTECTED] -- For more information on Real-Time Linux see: http://www.rtlinux.org/
