Tuesday, April 24, 2007
It works well! Great rainbow ribbon cable!
To Peter Alfke, Symon and John Larkin: Today (April, 24th), we get back the PCB board to implement the transmission of signal using ribbon cable. We use a 47Ω serial-termination resistor and a 1.5 meter long rainbow ribbon cable. It works well. However, one of the waveforms is a little bit strange. It is actually not as good as the waveform we get using a normal single cable in the simple experiment previously. The follow figure is the strange waveform which is on the resistor, can you tell me why?
Subscribe to:
Post Comments (Atom)
1 comment:
From a comp.arch.fpga posting on 4/25/2007, 12:16 pm Pacific Daylight Time:
Thanks for sharing the pictures.
Please note that for source-series terminated lines, the signal at the load
(destination) will be different and closer to ideal than the signal at the
driver (source). If your blog image is from the transmit side of the cable,
look at the other end.
It may also be that the series resistors need to be adjusted. From the
looks of your signal, I'd estimate you're using about 2 meters of ribbon
cable rather than 1, showing the time from transition until the reflected
signal as about 18 ns. I'd estimate the speed of light across ribbon cable
as about 2/3 C or about 4.5 ns per meter (but I'm not certain). The return
trip for the reflection doubles the time from the driver transition to the
observed reflection. You should see a big difference on the other side of
the cable.
Try slightly greater resistor values (60 ohm to 100 ohm) and see if your
signal at the receive side improves further.
- John_H
Post a Comment