Robas, Teodor wrote: > Hello Joerg, > > > The analog front end is not built right now. I just feed the signal > thorough some resistors into the analog input. In fact most of the > components in the analog part have poorly calculated values. The analog > circuit shall have a detailed review after the digital part will do that > 12MHz sampling. >
Ok, then put the feedback resistors for the differential amps on the review list as well. 10K is way too high, a few hundred ohms is more customary. > I tried to measure the number of samples in a bit from a serial line but > the length of a bit is not very stable. Anyway it seems that the number > of samples correspond to about the same sampling frequency of 300KHz. > There is another way to test the sample rate, doesn't even require you to fire up the soldering iron: Put in 50Hz (divided down from a doorbell transformer etc.), take your 8k samples and then look how many samples per 50Hz cycle you've got. It's pretty easy if you feed the data into Excel or OpenOffice Calc where you should see a sine wave if you graph it. If it turns out that your clock rate is higher than you thought you can also use a test signal from your sound card output, just set it to a few kHz. > The divider ideea is great ! My multimeter can measure up to 20KHz so > with a divider I can easily reach 20MHz. Seems that I will go this way > for now. > If you can find a discarded TV set with a PLL tuner it gets even better. Most TV sets from the last 20 years have that and when the flyback transformer blows people often throw them away. Many of their tuners contain a divider that can take in frequencies up to 900MHz. -- Regards, Joerg http://www.analogconsultants.com/ _______________________________________________ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user