Robust RF Links

[enauman]

I need some RF advice. Which of the inexpensive transmitters/receivers out there perform the best in an extreme EMI environment such as near a TIG welder with high frequency start operating. Is there one modulation scheme more resistant to EMI than the rest? I want to transmit a single control command to a receiver to set a digital pot. I have everything working on the bench, but haven't been able to find an RF link that will withstand the EMI. As they say "...RF is FM (F'n Magic)"

[dhouston]

The most important thing is to use FSK modulation. This is analogous to FM (Frequency Modulation) and is inherently more noise immune than ASK (analogous to AM). FSK hardware tens to cost 4-5 times ASK.

I have no idea what noise frequencies you will encounter but imagine that it's very broadband. Still, you might be well served by transmitters/receivers that can be tuned over a very broad range. This might give you a chance to find a band that has less noise than others.


http://www.hoperf.com/rf_fsk/rf_transmitter.htm (http://www.e-madeinchn.com/Product.htm)
http://www.hoperf.com/rf_fsk/rf_receiver.htm

You should also use highly directional antennas on both ends. You can purchase printed circuit yagis but they tend to be available only for certain frequencies. Here are the results of a search:


http://www.bing.com/search?q=printed+yagi&form=OSDSRC

One other thing to consider, if you have line-of-sight, is using either IR or a modulated laser pointer.

http://www.instructables.com/id/Laser-Transceiver/

[enauman]

Thanks for the reply and the advice. I will look into FSK. The reason I can't use IR or directional antennas is that the device will not always remain in place and most likely will not have line of sight.

Side note:
I worked as a flight test instrumentation engineer for the Skunk Works for 30 years and had a lot of unique telemetry problems/fixes on aircraft. The difference with my experience and what I'm trying to do now is that I don't have $80K+ to spend on telemetry equipment like I did in those days :)

[dhouston]

Well, you could just use a more powerful laser and make your own line-of-sight - that's the Skunk Works approach. :rolleyes:

Seriously, there is one technique that uses inexpensive ASK hardware but double modulates the signal. Philips uses it with the RF on their top of the line Pronto touch screen universal remotes. It is explained briefly at the bottom of this web page...


http://davehouston.org/ir-rf_fundamentals.htm

If you know anyone with an RF signal strength meter, you might be able to get an idea of the noise frequencies you are dealing with.

[enauman]

Thanks Dave,

I have a signal strength meter, but it doesn't have a spectrum display. I have a feeling that the welders EMI is pretty broad spectrum.

[dhouston]

I have a feeling that the welders EMI is pretty broad spectrum.I suspect you're right. I did a search but did not find a definitive answer - some said it only affected AM, some said it also affected FM, some said it even affected TV. All of the manufacturers seemed to have instructions on proper grounding, etc. to minimize RFI but none said what frequencies were involved.

I used to run a company that sold precision machinery to the sheetmetal industry so I visited a lot of aerospace manufacturing plants but I retired at about the time that all but a few leading edge companies were just starting to wire their facilities for automated data collection so problems like this were still to come.

We had electronic controls on most machines but they were self-contained and did not communicate. I cannot recall an instance where EMI/RFI caused us problems.

[cncmachineguy]

Back when we had hardwired phones, we could hear the welder during alum TIG. On the old phones (the kind that had hold and line selection functions on the bottom), We just had to have the annoyance of listening to it. When we upgraded to a phone system, it would drop the calls.

We tried everything we could think of like moving the welder to the other side of the shop. Un rolling the ground lead (~50 ft) so as not to have the big electro magnet hanging on the wall. Nothing helped.

It may be nice to note - this was only on Aluminium (A/C setting). TIG of steel or stick welding never was noticed. Also, this was clearly radiated as the original phone system was not connected to line power. Also the fact that we could listen to it may give an idea to frequency?

[dhouston]

I have a signal strength meter, but it doesn't have a spectrum display. I have a feeling that the welders EMI is pretty broad spectrum.How is this coming along?

I just came across the XRF (http://www.hotsolder.co.uk/xrf-radio-data-module-406-p.asp)which, while a bit costly, might work in your environment given that it's tunable over such a wide range (and probably has power output higher than legal for the US). The developer asked me for help doing an XRF to X-10 RF Gateway which only took a few minutes to explain. I can ask him if they've had any experience with TIG. Maybe they could even test it if there's someone nearby with TIG.

[enauman]

Thanks Dave. The project has been on hold for the last week. I got sidetracked (pleasantly) by a friend of mine at Freescale Semiconductor. I got their codewarrior development system in exchange for a commitment to do my next project using their hardware and software tools so I have been studying their IDE and brushing up on my C++ skills. I plan on getting back to it the welder project tomorrow. I will check the XRF out. I have a Miller Dynasty 300 TIG (inverter type) so I can do the testing myself. I'll keep you informed of my progress on this thread.

Thanks again for your help.

Ed