NMEA interference on navtex

[JerryHawkins]

I have installed a Simrad IS15 instrument system and when it is transmitting NMEA data, the interference (which can be heard on a world-band radio tuned to 500KHz) blots out my navtex radio reception. With the instruments off completely (or the NMEA output temporarily turned off) everything is fine.

Any ideas on how I can get rid of this interference. I've tried using screened twisted pair cable for all NMEA connections with the screen connected to ground (battery negative), but this hasn't helped. I've also tried using 0.1uF capacitors to 'de-couple' the data lines - no help. I'm now thinking of buying (over £200!!) an NMEA data combiner (see http://www.actisense.com) which 'opto-isolates' NMEA in and out.

Cheers,

Jerry

 

[ccscott49]

Surely this is a problem for Simrad to solve, after all they can't sell equipment that interferes with radio sgnals, I thought that was illegal or not within certain regulations at least! CE and all that!

 

[rogerm]

As an Electronics Engineer I would suggest that if the measures you have taken so far have not helped then the NMEA 'data combiner' won't help either since all it does is to provide 'ground' loop isolation between the different legs of the circuit.

It would help quite a bit if you knew just which part of the setup is creating the interference. Leave the NMEA enabled but disconnect all the inter-instrument connections. Assuming the Navtex then works OK connect each instrument in turn until the problem comes back. The NMEA speed is usually 4.8Kbaud. Depending on just what data is being sent the highest possible frequency if 4.8Khz but is typically less than half that. You might like to try a couple of very small coils such as is used to suppress electric drills in series with the two NMEA signals where they come out of the 'master' instrument ie the one sending the NMEA signal VERY close to the o/p pins. If the coil is much more than quite small it will upset the NMEA signals.
This all assumes that the Navtex aerial lead itself is screened and the aerial is just as far outboard as possible ie on a stern aerial 'christmas tree'.
Roger

 

[ParaHandy]

Assume Jerry's problem is likely to be a ground loop? These are v. nasty problems to cure. The ground of the device attached to the NMEA is using the -ve to go to ground through the Simrad IS15 rather than it's own ground (which assumes that the device is not opto-isolated which is not v. good).

Try just removing the NMEA signal ground.

You will, of course, lose the signal to any other device which is connected to the NMEA and which is correctly isolated.

 

[JerryHawkins]

Thanks for the advice and tips. I believe the rules were changed a while back - the onus is now on the receiving equipment to reject noise rather than the transmitting equipment not to produce it!

I've tried leaving the NMEA 'transmit' from the instruments turned on, but with the cable unplugged - the noise goes away completely, so at least the instrument head itself is not 'radiating'! This is why I thought the NMEA 'combiner' would help if I place it right next to the instrument and only have a couple of inches of cable from instrument to combiner acting as the 'radiating aerial'. Once the signal comes out it should be 'clean'.

Yes - my navtex antenna is on a transom mounted pole and the cable is screened, and grounded. Should I disconnect this ground? This was recommended by ICS (I have a Nav6Plus). As an aside - ICS have given me a free replacement Nav6Plus with the latest circuitry mods which improve interference rejection! It has certainly helped as I now get messages through although they are scrambled to some degree or other.

Cheers, Jerry

 

[HaraldS]

First of all I think something must be actually wrong in your system to have the radiated noise you describe. I have my Navtex and other receivers quite near a lot of NMEA cabling and I don't hear a bit of a noise.
I don't have an IS15 system, but from what I know about it I have some ideas of what you could check and try.

I presume that when you say NMEA transmit, you mean the actual NMEA port, I think it's the right most looking at the rear of the instrument.

I also presume that your instruments are wired to each other through the robnet daisy-chain and that you don't have noise using just this instrument interconnect.

The actual NMEA ports: From what I know about the IS15, these can be selected to be either receivers or transmitters.

Have you tried using them as receivers from some other NMEA source and was that ok?

I assume you have the troible when you configure them as transmitters and actually have something connected. (Since you said with the cable off, you don't have noise). What are you connecting to?

Could it be that you somehow ended up with two transmitters on the same cable?

The other suspicion would be that the NMEA outputs aren't as ground free as they should be, or at least with too small a common mode range. Or one of the sides gets shorted to ground on the receiving end and it doesn't like it.

The NMEA drivers are quite powerful, so that a short on the output could feed back noise through your power connection.

See if you have noise when you leave the cable on the output of the instrument, but leave the far end open. If you have no noise in that case, which is what I suspect, then we know it is ok in a no load case.
You should also be able to terminate the far end with a 120 ohm resistor without yet getting a problem. If that test passes too, I would suspect a problem with the input side.

Not sure what you have tried with the 0.1 µF capacitor, did you put it accross the two NMEA poles, or in series with one or both wires?

 

[JerryHawkins]

Harald, thanks for your very detailed reply!

The IS15 system has a single NMEA port per instrument head (far right as you say) which can be either a talker or a listener (these auto-configure depending on what they 'see' connected. I have two heads connected via the Roblink. I've switched connections so that they have both been talker and listener at one time or another. Everything works fine. All NMEA data is 'seen' where it should be. I have a a Simrad CP32 GPS chart plotter which talks (to the IS15 system) and listens, ICS Nav6Plus which listens, Simrad WP30 autopilot which listens, and the IS15 which listens to the chartplotter and talks to everything else. So all the listeners see only a single talker - the IS15 system. I haven't tried with the cable connected at the IS15 talker connection but disconnected from the listeners yet - I'll have to try that next weekend when I'm back on the boat. I'll let you know. I tried the capacitor across the NMEA talker poles to try and 'sink' the high frequencies.

Thanks again for your response. Regards, Jerry

 

[HaraldS]

Hi Jerry,

have been thinking some more about what could cause your interference problems. Normally NMEA should not be doing this, but there has been another thread here a few weeks ago with similar problems.

I have quite a few things NMEA connected on my boat, but I don't pick up any noise in any of the receivers. I too have a Simrad plotter CA40, but a Raytheon instrument system talking to each other and several listeners like Epirb and DSC controller.

My first guess was that there might be noise fed back via the DC power connection, but than the other chap with similar problems noted that he could pick up the noise with a independant battery radio, which rules out the noise on the power line theory.

As you know, NMEA is in iteslf very low frequency, so for a radio to pick it up you are wondering where the RF is coming from.

A possible explanation could be the square wave shape of the NMEA signal. A true square wave signal, with an infinitely short rise time, in theory contains every harmonic frequency there is. Zero rise time is impossible, but NMEA drivers are quite powerful and fast (20ns rise time), as they are designed for much higher frequencies (> 1 mega bit), so they certainly have the potential.

They are also quite low impedance, which means they can drive a substantial current at these frequencies. Higher currents would only flow though if one or more of the receivers is low impedance, or maybe clipping at some voltage level.

So it certainly makes sense to try to connect one receiver at a time and see what happens.

A high speed RS422 or RS485 (multipoint) network is usually terminated with 120 Ohms on either far end, to avoid reflections, but then people also match the cable to that impedance.

With NMEA all this doesn't matter because it's so slow. So in fact you could take out it's "zip" by putting two resistors in series at the transmit + and -, before going into your cable. Also, after those two resistors, I'd say try 330 or 470 Ohms, you could put your .1µF capacitor and here it would have some sensible effect flattening the edges.

The receivers usually have an input resistance that is well above 10k Ohm. So even with three listeners you'd be above 3k, so that the serial resistors wouldn't give you much signal loss, but they would limit the current if we get some clipping. And if that was the source of the radiation, it should disappear.

Also I think that these combined NMEA inputs and outputs of the IS15 are suspect, at least with respect to common mode immunity relative to ground. And there seems to be some asymetry in them, as Simrad notes in one of their updated manuals (Rev. H):

"Note ! Even if the NMEA input is automatically polarized on the IS15
instruments, it may be necessary to interchange the wires from
XX32/40/50 series if depth sounding problems are observed in
models that contain a depth sounder."

In clear this means the software automatically finds the polarity of a NMEA input signal, but the hardware is obviously not entirely symmetric. So try to flip the NMEA input side and see if that helps. (That's from the plotter as talker to the IS15 as listener.) This may affect the output side.

In your other post you note that you also have problems with the transceiver box in conjunktion with the sounder. So that might be related.

If all else fails to cure your problem, note that there is also an enhanced transceiver box, called expander, which in addition to depth and speed transducer connections, offers two entirely separate NMEA ports, with separated input and output lugs. Maybe they'll trade you up to that box if you can't get rid of the problem any other way.

Cheers,
Harald

 

[JerryHawkins]

Thanks Harald!

Harald,

Thanks again for all your advice - I can't wait to get back down to the boat to test your theories and ideas. I will let you know how I get on as soon as I have done the tests.

On the depth problem front; I was talking to Geoff Sargent today (Simrad after sales service manager) and he told me that my depth problem is not just me - others have reported it. It has come to light with their new IS12 system that uses (I guess) the same transceiver. It seems that anything but a hard surface on the seabed confuses the depth circuitry with the multiple echoes. They are working on a fix.

I too saw that note in the IS15 manual about swapping the data lines, but assumed that as it was working, all was OK - but I'll try it. I think the relationship of this to depth problems are if you're using a CE32 (combined chart plotter and echo sounder) that has the depth transducer connected directly to it, rather than my CP32 which is plotter only but can display a depth graph from NMEA data input.

Thanks, Jerry

 

[HaraldS]

Re: Thanks Harald!

Jerry,

good luck, and please let me know what you find out, I'm really curious now.

Cheers,
Harald