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/

Reply via email to