Al, I've done some tentative receive measurements on a Flex 5000. With a digital scope I monitor WWV symultaneously on an analog radio and on the Flex, and measure the tick delay.
At 48k/256, with 4096 FFT, I get around 230ms latency. CW turnaround takes 50-100ms additional. CPU utilization is about 15% I also disclaim authority. Anyone else have 5000 numbers? 73 Ed W2RF On 5 Oct 2007 at 16:20, a.groff wrote: > Out of concern and curiosity, I decided to measure my TX Latency from > mic input or VAC input through PowerSDR. to the TX I & Q signals. > The hardware: SDR-1K, AMD 3300+ CPU 1gb ram (CPU utilization < 10%) , > on board sound card, D44 sound card & Ozy/Janus sound card. > The software: PSDR 1.10.3, Audacity 1.2.5, VAC 4.08 > With external cables and 'Y's, I setup Audacity to capture the mic > input to PSDR on Left stereo channel and the PSDR TX I output on the > Right stereo channel. I then recorded brief transmissions with Audacity > and measured the delay between the left channel (mic/VAC) the right > channel (TX I or Q). > > In the first test series I measured delay for the O/J mic input to O/J > TX I output. For the second series I measured delay for D44 mic input > to D44 TX I output. > Mode LSB,VAC off, 48000 BPS sample rate > Sound Card SC buffer DSP buffer delay > O/J 2048 4096 270 msec > O/J 2048 256 120 msec > O/J 256 256 40 msec > > D44 2048 4096 310 msec > D44 2048 256 160 msec > D44 256 256 Failed. (no signal > through put ) > > For the VAC port latency tests, in the PSDR VAC setup form, I selected > the D44 3/4 sound card as the input ( no VAC cable ), the measured delay > was therefore from the D44 mic input, through the AC channel of PSDR and > out the O/J TXI/Q. > Mode DIGL,VAC on, 48000 BPS sample rate > VAC-O/J 2048 4096 320 msec > VAC-O/J 2048 256 155 msec > VAC-O/J 256 256 190 msec > VAC buffer size seemed to have little effect on latency for this test. > > Immediately AFTER toggling TUNE ... > VAC-OJ 2048 4096 2450msec ( yes 2450 > msec ! ) > (toggling VAC restored original measurement.) > > Next I included the actual VAC software (v 4.08) in the latency test. In > this test I routed the test signal through the D44 3/4 sound card, an > audio repeater, and a VAC channel to the VAC input of PSDR and out to > the O/J TXI/Q. > VAC-O/J 2048 4096 900 msec > VAC-O/J 2048 256 450 msec > VAC-O/J 256 256 390 msec > Again the VAC buffer size seemed to have little effect on the TX latency. > > Disclaimer... This is a sample of 1.. Different CPU's will have > different latency... Small buffer sizes have other (undesirable ? ) effects. > > I would be interested in hearing other measurements. > > YMMV > > AL, K0VM > > > > > > > _______________________________________________ > FlexRadio mailing list > FlexRadio@flex-radio.biz > http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz > Archive Link: http://www.mail-archive.com/flexradio%40flex-radio.biz/ > FlexRadio Knowledge Base: http://kb.flex-radio.com/ > FlexRadio Homepage: http://www.flex-radio.com/ > _______________________________________________ FlexRadio mailing list FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archive Link: http://www.mail-archive.com/flexradio%40flex-radio.biz/ FlexRadio Knowledge Base: http://kb.flex-radio.com/ FlexRadio Homepage: http://www.flex-radio.com/
