Radar Detector CPU

[Swift101]

Hi all,

Would anyone know what kind of CPU radar detectors have? And would the speed of the processor have an effect on the reaction time? My theory has always been that if an RD was equipped with something like a 600mhz-1gz CPU it would have a blistering fast reaction time & would be able to sweep Ka band very quickly. If they don't have CPU's with that much juice I was thinking, what's holding back manufactures from creating an RD that would boast a Smartphone like CPU. Forgive me and please correct me if I am assuming the wrong thing here. I would love to find out how these devices really work and if anyone knows anything about this topic, please share.

(Also I do realize dash mounted units are subjected to a lot of heat, so cooling a powerful CPU could be problematic, but perhaps cooling fins or a heat sink, maybe with a small fan, could cool the unit down.)

-Thanks

[ghz1]

I'm sure the professor can answer this one.

[SHOtime2511]

Sweeping Ka band isn't a problem.
It is the falsing problem that comes from running RDR/Ka guard off.
It is much better solved by segmenting the Ka band and turning off the segments not used by police radar.

[jimbonzzz]

I haven't really looked at this lately, but I know:

- The Belscort S7's used a Motorola/Freescale MC68HC908GP32 (older models used similar)
- Older V1 models used a Motorola MC68HC705C9ACFN
- The BEL STi has an ATMEGA32 (not sure if it has others)
- The 9500ix and similar GPS-enabled detectors supposedly have multiple processors controlling different aspects of operation.

You might be able to find more by searching the internal photos people post on here every once in a while....

[nine_c1]

They all appear to use some variation of an 8-bit RISC (Reduced Instruction Set) Processor.

Here is a detailed internal of the MCUs and various other chips inside the 9500ix.



Speeding up the sweep rate by virtue of using faster processors is not really what you want, as the sweep rate can be too fast reducing the dwell time on each frequency which would reduce the sensitivity of the RF section itself.

Valentine uses a different detection scheme than Escort/Bel (able to detect multiple responses.....or frequencies) for any given LO frequency which is what makes the V1 inherently faster than the Belscorts.

Escort/Bel have a design in the skunkworks that will use a different method of receiving raw signals and feed them directly into multiple digital analysis engines (processors) that will effectively speed up their detectors.

Might see that sometime next year.

[Swift101]

Thanks everyone, really intresting stuff, these are some serious devices.

[North Alabama]

With the faster V1 (still about a 2 second alarm delay with the Ka Guard on, non pop Ka), and even the slower Belscorts, I also want more speed from a RD to let me know at the instant that I am being hit with I/O or around lidar to give me a better chance to react. To put doubt in the Leo's mind on which is the fastest auto in the fastest display (certain models), try to beat the finger on the lock pad (certain models), and not give the Leo a long fast reading so that he will question what he is seeing. With traffic worked around a speeder, a non pin point radar can not nail a certain speeder except with the MPH Ranger EZ, thus there is hope of beating I/O with a fast RD like the V1 with the Ka Guard off (just don't be in the target display)....

[nine_c1]

With the faster V1 (still about a 2 second alarm delay with the Ka Guard on, non pop Ka), and even the slower Belscorts, I also want more speed from a RD to let me know at the instant that I am being hit with I/O or around lidar to give me a better chance to react. To put doubt in the Leo's mind on which is the fastest auto in the fastest display (certain models), try to beat the finger on the lock pad (certain models), and not give the Leo a long fast reading so that he will question what he is seeing. With traffic worked around a speeder, a non pin point radar can not nail a certain speeder except with the MPH Ranger EZ, thus there is hope of beating I/O with a fast RD like the V1 with the Ka Guard off (just don't be in the target display)....

Good points, but the V1 with Guard On and the Belscorts (disregarding BS/RDR/USA on certain models) are really not THAT slow today. The V1 will catch more than it misses above .3 seconds with the exception of 33.8 which can take up to .5 seconds.......the Bels up to .7 - .8 seconds on all bands (unless you put them in POP mode which can extend over 1 full second).

The problem is, they all sweep one band at a time using the same basic design.......a swept LO in the 1st stage fed into IF detector and on into another mixer circuit and more IF detectors.........all part of an FM demodulation technique. Again, sweep rate cannot be too fast or sensitivity goes in the crapper.

So in order to break the limitations of conventional detection, a completely new technique is required to produce the kind of response times we would like to see. It looks like the new design will use a multiple of fixed LOs in the 1st stage and feed the resulting IF signals into an analog to digital converter. From there this digitized data goes off to a gang of processors running in parallel for analysis..........a true "digital detector".

Another benefit of all this signal analysis should be a reduction in false alerts, something that normally does not go hand in hand with speeding up the response time of a detector.

We'll see. But to get what we want somebody had to think outside of the box and it looks like Belscort is doing just that.

[jimbonzzz]

The problem is, they all sweep one band at a time using the same basic design.......

Really? I was always of the understanding that the V1 sweeps K and Ka simultaneously... unless that is old technology now

[nine_c1]

The problem is, they all sweep one band at a time using the same basic design.......

Really? I was always of the understanding that the V1 sweeps K and Ka simultaneously... unless that is old technology now

You caught me Jim!

Yes, there are multiple conversion paths that permit K Band and a portion of Ka to be swept during a single LO sweep cycle on the V1. The Higher portion of Ka is completed during a 2nd sweep of the LO.

So this comment was not entirely correct:

Valentine uses a different detection scheme than Escort/Bel (able to detect multiple responses.....or frequencies) for any given LO frequency which is what makes the V1 inherently faster than the Belscorts.

Rather I should have said (using multiple conversion paths) during a single sweep of the 1st LO which is what makes the V1 inherently faster than the Belscorts.

This comment as well needs to be amended:

The problem is, they all sweep one band at a time using the same basic design

As you indicated (correctly) the V1 sweeps K-Band and a large portion of Ka during one sweep . Obviously the Belscorts are substantially different in this respect as they perform discreet sweeps of K and Ka.

So I did a poor job of accurately explaining the the issue, which is, they are all sweeping a single narrow radar channel at any given point in their respective LO sweeps.

The key to Escort's new design is effectively scanning a plurality of channels at the same time.