slipring ethernet test

That is unusual

economical

Signal generator on one side, o'scope on the other; rotate and watch for dropouts.

Less so

Signal generator on one side, signal analyzer on the other, synch the signals outside the sliprings (direct from generator to analyzer) and compare with output across sliprings, bit for bit.

OR if you want to do it digitally

Put a source with a tremendously large file (or series of files) on one side wrapped in a CRC. Transmit to a receiver on the other side and do a CRC check. Faster, but won't give you bit-by-bit failures. But use a non-autocorrecting CRC.

Determine in advance what is your allowable packet loss, since you will have some. Otherwise spend your time trying to decide what is good enough.

And since TCP/IP will retransmit lost packets, maybe the good enough is something below the "failed" criteria for a given packet?

Hello,

We do have signal generators, what frequency would you suggest?

In terms of the signal analyzer which one should I get? I like the idea of

doing it digitally. What's a CRC?

Is there any way to simply hook up a PC and to monitor its

digital throughput with either an analyzer or just a digital storage scope?

thanks for your help

Russ

Rather than lead you astray

http://en.wikipedia.org/wiki/Ethernet_over_twisted_pair

100BASE-TX follows the same wiring patterns as 10BASE-T but is more sensitive to wire quality and length, due to the higher bit rates.

A 100BASE-TX transmitter sends 3 differential voltages, +1 V, 0 V, or -1 V^[1].

http://en.wikipedia.org/wiki/Cyclic_redundancy_check

I'd be tempted to set up the null modem communication communication between two computers using a communications package like the old Crosstalk and a cross-over cable, because it flags drop outs. Once you have a package that supports error flags and CRC you have everything you need to start spinning your slip rings and do a comparison.

And don't make CRC too hard, it is simply a semi-accurate bit count. It is not infallible, but a good indicator.

Then you'll eventually want to do bit for bit comparisons.

you might do better by converting to coaxial cable, and then finding a slip ring radar antenna feeder as used in the cheap ship board radar scanners for pleasure boats. This has the advantage of needing only 2 wires and has less complex slip rings.

most radar feeds will be 50 ohms, so make your twisted pair drivers 50 ohms or you will need to match impedances.

I am sure you can google these

once setup you can use it in series with a test site called www.dslreports.com to access the internet

test with a standard wire and get bandwidth, then test with the slip ring connecetd to the internet.

The DSL reports gives you many tests to run

Gentlemen:

Thanks for the input. I'll look into this further,

and let you know what's up.

I appreciate your comments.

Russ

I hate to be negative, but why even try to do it this way at all????

I would use a standard high quality WLAN to bridge between moving and fixed parts. You have a HUGE amount of equipment available today......it should be very, very easy. Just use the sliprings to get the supply voltage needed onto the moving parts.

Anything else is just wasting time and money.......do not forget to use at least WAP2 as part of your security.....

Have a great day.

Excellent Answer!

Don't forget a large capacitor to bridge any gaps in the power over the sliprings.

...that only works with DC as far as I am aware!

True! I was assuming the DC power to energize the remote.

Actually AC power over slip rings can be more useful than you think......but it really doesn't matter.

What I was saying is that one should qualify the statement fully, or someone will/might try Caps on an AC slipring!!!

I have tried the RF version a few years ago without success. Even with a spread spectrum modem. Finally, I ran the RS-485 (19200bps) signal through slip rings and got much better results. I also had 208V powering switched heaters on the other rings. The machine has been running this way for many years!

Try it! Just wire the Ethernet cable between your computer and the switch through a slip ring. Download a movie while rotating the rings and see if you notice a slow down. Ethernet should have better error recovery. You might not be able to see any difference.

That was going to be my initial approach, as I've

enjoyed success with the same idea with USB signals.

I was just wondering if maybe there was an inexpensive

piece of equipment, tailored to this type of testing, capable

of yielding a nice picture of the system's performance.

Thanks gentlemen.

Russ

You could rent an Ethernet protocol analyser and look at the error rate if you want to quantify the results.

I would stick to a 10baseT and short cables to reduce the effect of impedance mismatch through the rings.

Good luck.

Would you consider converting signal to optic and transmitting via an optic rotor. We use this system up to 100Mb and dont seem to encounter any problems.