Getting instruments connected to a laptop

[AlanPound]

Nice post...

Maxsea though is a pretty powerful piece of software. It supports (in principle) as many serial ports as you have got on your PC, and for each piece of information you want to display or process, you can pick and choose which serial port to take it from, and (within limits) which NMEA sentence you take it from (different GPSs can use different sentences to output their data).

For the NMEA output port (to drive an autopilot for example), you can choose which serial port to send the relevant guidance data to (sorry, I'm too lazy to go and look up the NMEA sentence names).

But I too would be reluctanct to make a PC too vital a piece of my boat - however it shouldn't be too hard to arrange your autoilot to have an easily accessible 'plan B'... My CETREK will take it's guidance from one of three simultaneous NMEA inputs... you choose which one via the contol head.

I understand the poster's interest in collecting data to construct real-life polars for Maxsea. I thought about that myself - but you need to be more determined than I am in order to sort all of that stuff out, and actually use the result...

Alan

 

[extravert]

I agree, looking at the detail they are not the same. But just connecting NMEA-0183 to a PC's serial port always seems to work, so we can advise that they are compatible.

I haven't tried driving NMEA-0183 straight from a TTL signal, I have always used a MAX-232 line driver as I've got a few kicking around looking lonely. They are pretty cheap anyway. Going over a long distance like from the top of a mast may cause problems without a line driver.

I guess £40 for a PIC development environment means that you will be doing assembly language, sounds a bit cheap for a C compiler based system.

I agree that the cost of components will not be expensive, I can't see how the manufacturers of these multiplexers can charge hundreds of pounds. The processor I used (ST-10) is way over-specced for the job, I just happened to have some spare. But even using that processor the cost of components (buying them at one-off prices) would only be £20.

 

[RichardDunstan]

Re: Another product

Thanks,

I hadn't seen that one.

I also saw an NMEA amplifier/expander by a company called Vectra but couldn't find much in the way of info on it. It is also more expensive than the others. You can see it at http://www.iboats.com/products/401/888_nmea_amplifier_expander.html

Rich

 

[tome]

It's the pcb and packaging that costs, not the components. Also these days EMC testing and CE marking adds a packet to the unit price.

I'm in the happy position of developing nav products for an internal market which is performance driven and not cost restrained, so I can take these hits. Qualification testing of our latest nav system cost £22,000 and we only plan to build 100 or so!

 

[RichardDunstan]

Thanks for all the technical stuff.

I noticed on the ShipModul product that it advertised that it didn't need a PC to keep it working. One of their technical people dropped me a personal email and I asked him if he wouldn't mind me posting it. I've withheld his email address.

By the way, I have absolutely nothing to do with the company. I've been on a pocket maxi a couple of times, sat at the nav station and thought "why can't we do that?". So that's what we're trying to do.

Here's his reply....

Richard,

Being the manufacturer of the ShipModul MiniPlex-41(USB) multiplexers, I would like to add some comments.

All mentioned multiplexers do their job and they do it good. One reason for us to develop our multiplexers though is that the existing multiplexers in the market have a few shortcomings.

The CruzPro multiplexers have 4 or 8 inputs and one RS-232 output. That's it. You can connect the output to your laptop to get all instrument data in the laptop. But it depends on the used navigation software if the received data is echoed back out of the serial port to the autopilot. Sure the software itself will send steering information to the autopilot but that's mostly it. If you want to present the data of the other instruments to the autopilot, you would have to connect the data-out from the multiplexer to the data-in of the laptop AND the autopilot. But then the laptop cannot send data to the autopilot.

The Noland has the same problem, but has a separate NMEA output. This output can either send data received from the laptop through the RS-232 interface, or send data received from the NMEA inputs. In the last case, you have to interconnect the TX and the RX signal of the RS-232 interface on the multiplexer. Also, using the NMEA out makes one NMEA input inoperational.

The bottom line of the above is that you cannot create a configuration where the data from the attached instruments AND the computer is available to the autopilot.

Seeing these shortcomings of the existing product, we decided to develop a multiplexer that could do better. Our multiplexers have four NMEA inputs, two NMEA outputs and a computer port, being RS-232 or USB. NMEA output 1 always delivers data received from the four NMEA inputs. NMEA Out 2 can be configured to output all data from the NMEA inputs AND the computer or from the computer only. In all cases, the computer keeps receiving data from the NMEA input.
Another feature is that the baudrate of the serial port can be set to as high as 57600baud, preventing any possible overflow. Four channels of 4800 baud data into the multiplexer effectively needs at least 19200 baud at the output to get all the data through.

I hope I have been able to clarify things a bit. As for practical experience: we have been able to get approximately 120 customers in a few months' time, quite a few of them having troubles with the Noland multiplexer. Of course you can take this as marketing talk, but I sincerely think that we developed a good product.

If you have any questions, feel free to ask.

Kind regards,

Meindert Sprang
-----------------------------------------------------------
www.shipmodul.com
-----------------------------------------------------------

 

[bedouin]

I think the spec for NMEA is TTL levels - which is why it's so easy to implement but may also explain why you hear problems with interference. In theory you could damage the receiving equipment by using RS232 levels - but I'm sure most equipment is suitably protected.

The beauty of PIC is that there is a vast array of development stuff in the public domain; so you can get a range of compilers for nothing. However the PIC instruction set is pretty simple, so I would tend to use an assembly style of programming rather than C.

The reason they charge that amount is that they are low volume items. Elsewhere in this thread there is mention of USB/RS232 conversion. That is MUCH more complicated than NMEA muxing - but yet you can buy on for < $30, because they make and sell them by the thousand.

I did talk to someone recently about designing and selling a Mux like this - but even with a component cost of £10-£20 (and you might need to add a bit for a professional-looking box &c) you would have to sell them for about £100, and still not make enough money to offset the development and support costs.

 

[bedouin]

A slightly odd posting.

To re-iterate what has been said above - there is no real problem connecting two listeners to the same NMEA output - if you are worried than a few pence would buy a buffer to make it safe.

At the risk of being shot down in flames, I would also suggest that in many cases you'll get away with connecting multiple outputs to a single input, but I won't necessarily recommend it.

Hidden in the posting is the really (potential) problem with MUXes - you cannot actually mux 4 NMEA into 1 and GUARANTEE results., you may end up with just too much data and have to drop some of it - it then becomes an issue of what data to drop, and whether there is a problem with some sentences being delayed by a few seconds.

 

[RichardDunstan]

I must admit that is the way we look to be headed.

There is another post that talks about some of the technical incompatabilities of the NMEA and RS232. It looks as though you can work round them but that isn't fixing the problem. Hence, my concentration on the multiplexer route.

I was also worried about the PC failing. One big wave and we're bound to find something new and unexpected on the cabin floor. Last time it was one of our female crew members.

Anyway, one of the products that I looked at said this:

"Stand-alone

It is not necessary to use the multiplexer with a computer. As soon as power is applied, the Miniplex-41 starts reading data from the NMEA inputs, queues it and sends it to the outputs."

Now, this would only really be of use if I wanted to set up an autopilot to sail on wind angle for instance. Without a PC, our A to B functionality isn't compomised. We don't rely on charting software (we hook a charting program to the GPS currently) for our only solution. We still backup the positions onto the paper based charts and write up a log. So if the computer goes down, we don't lose any information.

For racing, the computer is central to our data processing capability. Hence, what we would lose is the competitive advantage that the rapid calculations and course optimisations would give us. If it takes a terminal dive, we have to go back to the nav's rather slow brain. Either that or take a spare laptop.

What is worrying me a bit is the capability of the charting software to accept multiple inputs from the one serial port. I imagine that these multiplexers cobine all the wind speed/angle, log, depth, sea temp and all that good stuff into a single RS232 stream that the PC gets. The nav software then reads this. I'll ignore the "talker" bit for now.

Is the software capable of recognising that a single input is "responsible" for "sending" all these NMEA sentences.

I asked MaxSea Tech Support what they though and this is what they said...

"The NMEA multiplexer can be connected to Maxsea to receive several sentences on a same COM port.
But, we did not try out this kind of connectivity.

In fact, you need to check if you receive correctly NMEA sentences in the "check
data" box into Maxsea."

The bottom line is that, although I think I'm on the right line and most people seem to think I am, there are not all that many people out there who have done this. It appears as though most people have gone through the Raymarine SeaTalk, Silva Nexus or B&G FastNet route and neatly sidestepped the problem.

I think we will try it anyway.

I'll let you know how we get on.

Rich

 

[extravert]

Yes, testing and tooling is expensive for small production runs, but I would expect these manufacturers to sell more than a few hundred. Maybe they don't, which would explain the price.

My old NMEA-0183 multiplexer which was part of my old Danaplus system was in a box that looked like it had come straight out of the RS catalogue with a label stuck on the front. It was not meant to live outside and looked like a low IP rating. Perfectly adequate and fit for purpose. I expect that is not acceptable now for a commercial product to be in a bog standard box without swoops and curves, even if it is mounted out of sight.

> ... not cost restrained ...

must be military!

 

[kdf]

Heres what I have - it may help.

I have a B&G Network system. This is not a fancy bus but multi-drop NMEA. Each display on the network acts like a mux in that it takes whatever comes in , adds its own data, and resends the lot to the output. That way the next display gets all the data and so on.

I break this network, feed it into my mux, also feed in the GPS Nmea. The mux adds these together and the nmea output from the mux is fed back onto the b&G network so that I can use a GPS repeater to display position etc. This is also fed into a chartplotter.

The RS323 from the mux goes to the laptop. I have connected NMEA straight into the PC via the com port and this works fine as well. THe PC is running Transas Navigator and sees all the NMEA data on the net plus the gps data. THe output from the PC goes to the Autopilot via a switch which allows me to failover to the chartplotter if the PC fails.

 

[tome]

Can't really comment on how many they sell as I've never been involved in the leisure market.

I have a small pcb which my team developed sitting on my shelf which would do exactly what is required, including timetagging of all data to an absolute accuracy of around 10 micro-secs for replay. It can route NMEA and various other data formats in and out of 8 bidirectional ports, filtering messages as required either by type or update rates, and driving both RS232 and RS422 levels. It will also log raw data to a hard drive, and can respond to or initiate external events. It's a smart piece of 3rd generation kit but to put it on the market, we'd have to sell it for close to £2,000 assuming we could sell a good few hundred.

To go with it, there's another smaller board which digitises almost any analogue signal and converts it into NMEA transducer sentences (XDR).

The good old days of die cast RS boxes with a label stuck on the front have, sadly, disappeared and everything now has to be CE marked etc. Some see this as progress...

BTW, not military but same or sometimes higher standards has to work desert and arctic.

 

[AlanPound]

Hang on...

"I asked MaxSea Tech Support what they though and this is what they said...

"The NMEA multiplexer can be connected to Maxsea to receive several sentences on a same COM port. But, we did not try out this kind of connectivity."

Hmmm

IIRC (and I've got the Maxsea software with me, but not my bl**dy dongle), on the appropriate setup page, you go one by one through every piece of information that you want to display or log, and determine the instrument (and hence the serial port), and the message that you wish to derive it from.

If all you have is a GPS, then everything it knows about, it gets from that. And again, IIRC, if you are getting position from the GPS, and you want to display SOG, you can tell Maxsea to get it from "Maxsea" rather than the GPS, and it will work it out for itself from the ongoing *position* information it is already getting from the GPS (smart - but just one of the little tricks it does...)

Anyway, my point is, provided you have enough serial inputs - and you can get any number using a bunch of serial->USB adaptors - then you can collect data from as many instruments as you wish ... and the potential overflow problem that you would get by MUXing NMEAs together doesn't exist, as USB is fast enough to bring all of it into the PC (ie. there really isn't any MUXing going on as such...)

... so I can't see you have a problem...

Alan
PS I don't work for Maxsea
PPS I got the Setsail/Maxsea edition - same thing but cheaper (at the time)...

 

[stefan_r]

email sales@kmelectronics.co.uk and ask him for some help, though the NMEA > SeaTalk box is a RayMarine bit about £100 I think because everything from RayMarine is >£100 of course!!

mailto: stefan@athito.com

 

[stefan_r]

Re: PCMCIA

they are errrr specialist....can be got though - PM me if anyone needs some pointers for retailers

good manufacturer is 'Socket'

mailto: stefan@athito.com

 

[johna]

Richard

Thank you for your reply. Not being a racing man, well I sometimes am but only when there is another boat on the water, I thought polar diagrams were only used by the mags to compare boats. A new view on these matters from now on.

Thanks again.

John