ARPA on recreational radars

[twistedtree]

There is a good discussion over here on Panbo about recreational radar, and MARPA/ARPA performance in particular. Be sure to read the comments, not just the article.

 

[tpbrady]

Wow. A mini radar tutorial not to mention real world stories.

 

[HenryD]

TwistedTree - It would be helpful if Navico / Simrad would answer some of the questions posted.

 

[timjet]

The MARPA function on my Ray 18" HD dome unit is unreliable at best. It's probably a lot to ask of a small antenna and perhaps the manufactures put this feature on these small units for competitive reasons. Pretty much worthless IMHO.

Thanks twistedtree for the post. Now I know it's not just a problem with my unit.

 

[twistedtree]

TwistedTree - It would be helpful if Navico / Simrad would answer some of the questions posted.

I agree, but don't expect it will happen. No company wants to talk about shortcomings in their products. I expect all we will see is a note in some new software release about improved MARPA performance. Hopefully this increased exposure will get the improvements sooner rather than later. Lots of Simrad radar owners would benefit from improvements.

 

[twistedtree]

It's probably a lot to ask of a small antenna.....

This was Ben Ellison's argument too, but if you read through the comments after the article you will see there is no technical merit to that argument, and ample proof that small radars are just as capable of high-quality MARPA as large radars. The small Furuno radars, for example, give excellent MARPA performance with no discernible difference from their large radars.

So it was proven that a small radar is no excuse for poor MARPA.

 

[JDCAVE]

Hey Peter thanks for this--great debate, that's for sure. It's going to take me a while to go through all the discussion, at which point I might post. I do have a couple of points that I will post here.

First the water was getting muddy by the interchangeable use of the terms: accuracy and precision. It strikes me that precision might be improved with a narrower beam width, but the devil might be in the details. Systematic error (accuracy) might be more likely due to alignment or other installation issues.

Second, frame rate or sampling frequency will also have an impact on resolution and precison, I suspect, but don't know for sure.

I'm meeting today with my former research scientist, who is a physicist and hydroacoustician. Sonar and radar have many similarities. If you have a specific question for him, let me know and I will raise it.


Jim
Sent from my iPad using Trawler Forum

 

[twistedtree]

Hey Peter thanks for this--great debate, that's for sure. It's going to take me a while to go through all the discussion, at which point I might post. I do have a couple of points that I will post here.

First the water was getting muddy by the interchangeable use of the terms: accuracy and precision. It strikes me that precision might be improved with a narrower beam width, but the devil might be in the details. Systematic error (accuracy) might be more likely due to alignment or other installation issues.

Second, frame rate or sampling frequency will also have an impact on resolution and precison, I suspect, but don't know for sure.

I'm meeting today with my former research scientist, who is a physicist and hydroacoustician. Sonar and radar have many similarities. If you have a specific question for him, let me know and I will raise it.


Jim
Sent from my iPad using Trawler Forum

I'd love to hear what your friend thinks.

Yes, terminology can clarify or it can further confuse a discussion. Sometimes it even does a little bit of both.

At issue in the discussion are two incorrect assertions in the article that were claimed to make it inherently harder for a small radar to accurately locate and track a target. Neither is true, and in reality a small radar has no inherent disadvantage over a larger radar. This can easily be demonstrated in practice as well by certain small radars that perform ARPA very well, and comparably to much larger radars.

The first issue was whether a long pulse length makes it harder to measure a target's distance. It does not. A radar's ability to accurately measure the distance to a target is determined by it's ability to measure the time between the pulse start and the echo return and is typically accurate to 20-25m. It has nothing to do with pulse length. What a longer pulse width DOES do is obscure objects directly behind the the first one the pulse encounters. The radar is basically blind to anything less than the pulse length's distance beyond the first target. So if there are two boats, one behind the other, it will only see one target. Once they are spaced more than the pulse length further apart, they will be detected as two targets. But the distance to the first encountered target is measured very accurately regardless of pulse length.

The second incorrect assertion was that the wider beam width found in smaller radars makes it harder to measure a target's bearing. It does not. The center of a target is the center of the echo return, and it's accuracy is determined by the radar mechanism's ability to track where it's aimed and is not influenced by the beam width. Reading through a bunch of product specs, 1 deg appears to be the norm for how accurately a radar can measure a target's bearing. What the beam width DOES impact is how wide the target echo appears. The echo will be the target's actual width plus the beam width. But the center will remain accurate to within the radar's specs of 1 deg. The other effect of a wide beam is that if there is a second target within the stretched echo, it will be lost in the echo. It won't be until the targets are spaced further apart than the beam width that the radar will be able to tell them apart and show two echos instead of one.

The bottom line is that a small radar scanner is not a disadvantage when it comes to ARPA. However there are significant differences in how well this function performs from manufacturer to manufacturer. Based on the feedback in the article's discussion, it seems confirmed that Simrad/B&G/Lowrence/Navico is quite poor across all their products, it appears that Raymarine is about as bad but there are fewer reports, and Furuno is quite good across all their products.

 

[tpbrady]

Jim

Do you know the sampling rate used for converting the raw radar return to a digital output for display on the MFD? There is probably an error associated with the sampling rate but the question is how much and is it possible that the perceived problem with erratic MARPA is processor limitations not experienced in larger commercial units?

Tom

 

[MYTraveler]

FWIW, the MARPA (or maybe it is ARPA, since it will automatically acquire targets) of my NN3D radars (a 6kw and a 25kw) work really well. I overlay AIS on the display, and rarely is there any material difference in either position or course/speed/CPA/TCPA info for the target vessel. And, although I have not focused on a side-by-side comparison, I do believe the two units perform their ARPA functions equally well.

 

[twistedtree]

FWIW, the MARPA (or maybe it is ARPA, since it will automatically acquire targets) of my NN3D radars (a 6kw and a 25kw) work really well. I overlay AIS on the display, and rarely is there any material difference in either position or course/speed/CPA/TCPA info for the target vessel. And, although I have not focused on a side-by-side comparison, I do believe the two units perform their ARPA functions equally well.

This is my experience as well. I've owned and operates a NN3D 4kw radar, a Furuno 1835 4kw, and a Furuno FAR-2117. ARPA works very well on all of them. I have yet to encounter anyone with different experience on Furuno radars.

 

[N4712]

This is my experience as well. I've owned and operates a NN3D 4kw radar, a Furuno 1835 4kw, and a Furuno FAR-2117. ARPA works very well on all of them. I have yet to encounter anyone with different experience on Furuno radars.

Add me to the list, My TZT with a DRS6A works great with ARPA.

 

[twistedtree]

Jim

Do you know the sampling rate used for converting the raw radar return to a digital output for display on the MFD? There is probably an error associated with the sampling rate but the question is how much and is it possible that the perceived problem with erratic MARPA is processor limitations not experienced in larger commercial units?

Tom

Clearly something is different, but it's really hard to say what. I did extensive testing on my Simrad system in conjunction with Simrad's radar engineering team trying to figure out what was going on with MARPA. From that we (not just me) concluded it was NOT:

- Poor heading source. I got the same results with an RC42 rate compass and an expensive Sat compass.

- Multiple conflicting heading sources. I stripped the network down to just the radar, MFD, heading, and GPS. It made no difference.

- 4G vs Pulse radar. I tested with both 4G and Pulse and both showed the same problems.

- Small vs big radar. The 4G is a 24" dome and my pulse radar was a 10kw 6' open array. Same results with both.

- Sea state related. I could reproduce the problem is dead calm conditions.

- Just me. Simrad reproduced the problem, as have many other people.

- NMEA 2000 configuration or electrical problem. My network was inspected and reviewed by numerous people and passed with flying colors every time. And we even tried with a second small network with minimal devices on it. Same problems.

I don't know what's different inside, but Furuno's small recreational radars have great ARPA, so it clearly can be done and does not require a big, or expensive, or commercial radar.

 

[MYTraveler]

This is fun!