stuartw
Member
When I bought my old tub some 4 years ago now, it came with a host of 1980's navaids, ie Loran, Decca plus the addition of a 1990's GPS. The Autohelm 6000 it was also fitted with, also came with a Navigation Interface, which allowed it to be controlled by the Loran or Decca.
Now this old interface was a seemingly crude affair which took the cross track error info from Loran or Decca, and indexed the steering computer 1deg left or right, depending on the cross track error direction, thus reducing the error to zero. I never used it, as I never really bothered with the Decca or Loran. However last year on a trip to Guernsey & France, the constant re-adjustment of the Authohelm to keep the cross track error small, was a pain
However earlier this year, I did some cable re-routing and fed the GPS into the Navigation Interface. It didn't work, which was not altogether surprising, but there were signs of life. I managed to find the original Nautec design engineer - still with Raymarine who took them over, and after lengthy discussions, we both came to the conclusion, that the more modern GPS was giving out too many NMEA sentences for the Nav. Interface to handle, as it only head a crude micro and little memory. All it was programmed to see, was the XTE sentence. My one clearly was seeing it, but getting drowned with all the rest. It was also suggested the checksum at the end of the XTE sentence would also be a problem
Undaunted, I set about trying to filter out all the other sentences, and just presenting the XTE sentence. During an internet search, I found a Swedish company- Marin Electronik AB, who make such filters call "Ferrets". They agreed to do me a PCB version, just extracting the XTE sentence, removing the checksum and putting that back out. Sadly when I received it, it did not work, so it was back to the drawing board. After further discussions with Raymarine, who, I must admit, were brilliant, I decided to use a laptop to feed in synthetic data, and see what the interface really wanted to see.
Initially by using the laptop configured as a hyper terminal, and creating sentences in notepad, I managed to find the sentence which the interface responded to. The real problem was that the original system used a bastardised version of NMEA 0180, which was not quite NMEA 0183. It was only looking for a crosstrack error of two significant digits and two trailing ones. ie 09.99 to 00.00. Unfortunately the modern GPS format, is for cross track errors of 9999.99 miles, why that big? beats me.
There was one final twist however. The old system turned belly up when it received zero cross track error, configured as "00.00". After further playing about, "0.0" worked, only for the zero XTE situation.
I contacted Marin, who were prepared to do another version and send me a new processor. However the Gulf war got in the way, as they were very busy. So I decided in the interim to write a program for my laptop, which would take in the GPS data, extract the XTE sentence, strip out the excess leading zero's, do a test for zero, and substitute the new data and send it out to the Interface. It worked. Initially I set up a waypoint with a small error, and checked that the rudder position changed in the right direction to try and correct the error.
I didn't really have an opportunity to try it in a live situation until recently. The new Marin version, for one reason or another I haven't got yet. So it was on my recent trip to Cherbourg and St. Vaast, I was able to put the system to the test. So with my laptop "plumbed" in we set off.
It was fantastic, exceeding my wildest dreams. Although the Nav. Interface is in the inner helm and I drive from the flybridge, I was able to leave the interface switched on, but would put the Autohelm to standby or Auto from the flybridge I could see from the display on the Autohelm control, the cross track error and the direction. From Bembridge to Cherbourg I had a maximum error of 0.08nm.
From Cherbourg to St Vaast, which is a tortuous route, much the same, but in much worst conditions. On that bit of the trip, I did take it out of auto on each waypoint turn as they were significant turns, and re-established a new bearing before putting back into auto.
On the way back to Gosport, in a horrendous lumpy sea last Thursday, it performed well. At one stage the XTE was 0.5 miles, but the system brought it back to around 0.02nm. I decided at the Bembridge ledge, which was my 3rd last waypoint, to see if the system would steer itself round the corner. To my delight it did, bit by bit, doing a shallow turn to port. Just as it was about to straighten up, a tanker coming out made me take over and steer round it.
You may ask, why have I bothered? Well when you look at the price of an all singing & dancing replacement, which really isn't that much better, you can see why. Besides, being retired, it's a challenge.
There must be lots of old systems like mine out there, which also could be rejuvenated.
Sorry this is so long, but it was difficult to summarise.
<hr width=100% size=1>
Now this old interface was a seemingly crude affair which took the cross track error info from Loran or Decca, and indexed the steering computer 1deg left or right, depending on the cross track error direction, thus reducing the error to zero. I never used it, as I never really bothered with the Decca or Loran. However last year on a trip to Guernsey & France, the constant re-adjustment of the Authohelm to keep the cross track error small, was a pain
However earlier this year, I did some cable re-routing and fed the GPS into the Navigation Interface. It didn't work, which was not altogether surprising, but there were signs of life. I managed to find the original Nautec design engineer - still with Raymarine who took them over, and after lengthy discussions, we both came to the conclusion, that the more modern GPS was giving out too many NMEA sentences for the Nav. Interface to handle, as it only head a crude micro and little memory. All it was programmed to see, was the XTE sentence. My one clearly was seeing it, but getting drowned with all the rest. It was also suggested the checksum at the end of the XTE sentence would also be a problem
Undaunted, I set about trying to filter out all the other sentences, and just presenting the XTE sentence. During an internet search, I found a Swedish company- Marin Electronik AB, who make such filters call "Ferrets". They agreed to do me a PCB version, just extracting the XTE sentence, removing the checksum and putting that back out. Sadly when I received it, it did not work, so it was back to the drawing board. After further discussions with Raymarine, who, I must admit, were brilliant, I decided to use a laptop to feed in synthetic data, and see what the interface really wanted to see.
Initially by using the laptop configured as a hyper terminal, and creating sentences in notepad, I managed to find the sentence which the interface responded to. The real problem was that the original system used a bastardised version of NMEA 0180, which was not quite NMEA 0183. It was only looking for a crosstrack error of two significant digits and two trailing ones. ie 09.99 to 00.00. Unfortunately the modern GPS format, is for cross track errors of 9999.99 miles, why that big? beats me.
There was one final twist however. The old system turned belly up when it received zero cross track error, configured as "00.00". After further playing about, "0.0" worked, only for the zero XTE situation.
I contacted Marin, who were prepared to do another version and send me a new processor. However the Gulf war got in the way, as they were very busy. So I decided in the interim to write a program for my laptop, which would take in the GPS data, extract the XTE sentence, strip out the excess leading zero's, do a test for zero, and substitute the new data and send it out to the Interface. It worked. Initially I set up a waypoint with a small error, and checked that the rudder position changed in the right direction to try and correct the error.
I didn't really have an opportunity to try it in a live situation until recently. The new Marin version, for one reason or another I haven't got yet. So it was on my recent trip to Cherbourg and St. Vaast, I was able to put the system to the test. So with my laptop "plumbed" in we set off.
It was fantastic, exceeding my wildest dreams. Although the Nav. Interface is in the inner helm and I drive from the flybridge, I was able to leave the interface switched on, but would put the Autohelm to standby or Auto from the flybridge I could see from the display on the Autohelm control, the cross track error and the direction. From Bembridge to Cherbourg I had a maximum error of 0.08nm.
From Cherbourg to St Vaast, which is a tortuous route, much the same, but in much worst conditions. On that bit of the trip, I did take it out of auto on each waypoint turn as they were significant turns, and re-established a new bearing before putting back into auto.
On the way back to Gosport, in a horrendous lumpy sea last Thursday, it performed well. At one stage the XTE was 0.5 miles, but the system brought it back to around 0.02nm. I decided at the Bembridge ledge, which was my 3rd last waypoint, to see if the system would steer itself round the corner. To my delight it did, bit by bit, doing a shallow turn to port. Just as it was about to straighten up, a tanker coming out made me take over and steer round it.
You may ask, why have I bothered? Well when you look at the price of an all singing & dancing replacement, which really isn't that much better, you can see why. Besides, being retired, it's a challenge.
There must be lots of old systems like mine out there, which also could be rejuvenated.
Sorry this is so long, but it was difficult to summarise.
<hr width=100% size=1>
) <A target="_blank" HREF=http://www.jersey-harbours.com>http://www.jersey-harbours.com</A>
