Topic: Chirp for 898?

[sonarbear]

I heard that a chirp function will be added to the 898 in a coming update. Is this correct? How could it possibly work?

[Bob B]

I don't know where you heard that, but it seems pretty unlikely since the 898 is 2 Humminbird generations older than the Helix models .... None of them have chirp except the 12.

I have seen SPECULATION that Humminbird could add it to other Helix models, but I doubt would be really surprised if the xx8 or xx9 series ever get chirp......but you never know what they will do with a software update.

[Rüdiger]

Hi sonarbear

How Rickie in topic http://forums.sideimagingsoft.com/index.php?topic=9139.msg54592;topicseen#new  writes is CHIRP not only software but also hardware.

mvh
Rüdiger

[sonarbear]

Exactly. That's why I was wondering. But to create a sweep in the 898 should be possible via firmware update. The xducer would still be single F but it should be good enough. However I was wondering how the Chirp receiver works. Shouldn't the sent signal be correlated to the received waveform, peaking for each return? It seems to me that what is sold as chirp doesn't do that but rather just sends a sweep and still just plus receive strength. Does somebody know?

[rnvinc]

I have read (from a reliable source in Sonar acoustic engineering -non fishing Sonar) that it can be possible to add CHIRP to a non CHIRP system ... If there is enough processing power ...

HB did add CHIRP to ONIX in a software update with 2.700 ... HB will not claim whether they may had already foreseen this by adding any possible CHIRP hardware from the beginning release of ONIX ...

I highly doubt a series 2 generations old (and already Legacy status) would be considered for this possibility with the processing needed ...

Rickie

[sonarbear]

Interesting. So chirp can in principle be added! Maybe one of the Humminbird people could comment?

[rnvinc]

Interesting. So chirp can in principle be added! Maybe one of the Humminbird people could comment?

As I mentioned ... HB added CHIRP to ONIX with a software release ...

So in theory, CHIRP could be added to any sounder ...

If there is enough processing power available to make it work ...

Rickie

[rnvinc]

Interesting. So chirp can in principle be added! Maybe one of the Humminbird people could comment?

BTW ...there are no HB employees that regularly post here since Greg has been too busy to attend ...

Deb at HB occasionally will post when I send her direct communication that we need a question answered ...

I even recall a time or two where Joby at HB came to answer a question when he was summoned directly ...

We are all merely HB users sharing information as best we can ...

Rickie

[sonarbear]

So in theory, CHIRP could be added to any sounder ...

If there is enough processing power available to make it work ...

Sorry to keep insisting, but I find that that's exactly the question. If we assume that the head unit can create a sweep (some processing power involved) and the transducer has at least some bandwith, any unit can send out a chirp. Yet the return echo is, what's interesting. In a real chirp system the received waveform is basically correlated with the sent waveform, generating a peak of the product if they match. But that requires a digitized received waveform at sufficient resolution. I have some doubts that such a signal is present in the units. Does anybody know? Basically one could just make a unit that sends a broadband signal and receives it traditionally and call it chirp. It does chirp, yes, but can't do much with the information.

[sonarbear]

Ok. I did some patent digging and math book reading. Here are my results:
Basically, the chirp is a frequency sweep, which is later detected through a sliding correlation, i.e. cross correlation of received signal and template (sent signal, possibly adjusted for transducer or system non-linearities).
The older analog units, which detect the received signal based on the power in a small band via an abused FM intermediate frequency system (the legendary SA604SA) cannot be upgraded. Here the digital part has no time domain information of the received signal. Only of the power over time.
The 898 should be a more modern system however. If I read correctly, Humminbird has a (slightly ridicuolous because it just takes what's done for radar and replaces the word radar unit with fish finder) patent on sub-sampling the received sonar signal and digitally processing the result.
To quickly explain, the sonar wave itself oscillates to quickly to be resolved with any affordable A/D converter. However, if one samples the signal slower, a distortion (or "mixing") occurs, similar to what a radio does with analog parts. What is present in the unit then is a time domain signal, which contains the sonar signal received shifted to a lower frequency. What the 898 seems to be doing is to just calculate the power of that signal and some filtering.
However, since there is a true time domain representation of the received signbal, it can also be correlated with a sent chirp waveform also sub-sampled. The correlation peaks will in this case be way more resistant to noise and have higher temporal resolution. In result, one would get better lateral resolution for the sidescan (not vertical resolution though, that's determined by the transducer and boat speed) as well significantly more sidescan range.
If the transducer is no real chirp transducer, the resolution improvement will be not as good, but it will gain bigger range.
What would be necessary to get it on the 898 are two things:
1. somebody at HB to sit down and port the code that they probably already have to the 898
2. the 898 having sufficient processing power for the correlation to avoid slowing the unit down. For example, my Lowrance Hook does get noticeably slower when using dual chirp plus downscan.

I hope my explanation isn't too hard to understand and helpful!

[Bob B]

The Onix was the first unit Humminbird started advertising all DSP (digital signal processing) on the 2D.  The word at that time was that the previous models were all analog 2D.

Of course it is often difficult to differentiate marketing hype from reality.

[ckd022]

Sonarbear - what was the patent number that describes that invention?  It would save me some searching time and I'd like to check out the first claim to see what is really novel there as mixing up or down via aliasing caused by either up-sampling or down-sampling is well known in digital signal processing.  I would find it hard to believe that got through the examiner's office, but I've also found that often the examiners in some of these more specific knowledge areas don't have a real good understanding of the theory behind what is being claimed.

Thanks for sharing your findings!

Kevin

[sonarbear]

http://www.google.com/patents/US20060013067

Sorry, Raymarine is the culprit not HB. my bad. but still stupid.