9500i food for thought

[nineballrakm]

I assume everyone knows how the false alarms are eliminated with the new 9500i. It marks the spot of location with a GPS system with the specific band down to at least 0.001 gigahertz (plus or minus 0.001). Then when you pass by this same location, 9500i stores this information and does not alert you the next time. I have noticed with the RX65 there is consistancy with false alarms and LEO within a fraction of 0.001 but still will alert you to the threat. Question, what would happen if LEO engineers develop radar guns that would allow LEO's to "tweak" the gigahertz numbers so that the radar gun would match the Shell Gas Station false alarms as Radar Roy mentioned with that specific location?

[AirMoore]

They do' tune' their radar guns, so that they are within correct frequency specs, and legal for use...

With that said:


It would be so hard (nearly impossible) to get two radar units frequencies tuned exactly identical to one another.... or so close it blends as exactly the same frequency as the door opener.

(Thats my .02, Jim will probabaly chip in here).

[nineballrakm]

Thanks . I geuss I don't all the physics involved but it was scary thought for the 9500i.

[jimbonzzz]

I assume everyone knows how the false alarms are eliminated with the new 9500i. It marks the spot of location with a GPS system with the specific band down to at least 0.001 gigahertz (plus or minus 0.001). Then when you pass by this same location, 9500i stores this information and does not alert you the next time. I have noticed with the RX65 there is consistancy with false alarms and LEO within a fraction of 0.001 but still will alert you to the threat. Question, what would happen if LEO engineers develop radar guns that would allow LEO's to "tweak" the gigahertz numbers so that the radar gun would match the Shell Gas Station false alarms as Radar Roy mentioned with that specific location?

OK first off, I don't ever see radar guns being made self-tunable just to defeat the 9500i. They would also have to add a method to the radar guns to display the frequency of the Shell Station so that they could match it. I just don't see it happening.

Second: I'm quite certain that the 9500i does not have the capability to resolve the frequency down to .001 GHz (1 MHz) even though the frequency display on the unit goes out to three decimal places. I have not had a 9500 to play around with yet, so I don't know if it is any different, but it supposably uses the M4 RF receiver like the X50 and RX-65.

First, I did a little bench test on the RX-65 in this thread:
http://www.radardetector.net/viewtopic.php?t=348

And I expanded on it at the bottom of this one:
http://www.radardetector.net/viewtop...r=asc&start=30

Here's the images from the 2nd test (RX-65 VS Freq Counter Images):
http://www.kc8unj.com/radar/freq/

So for example, throughout the entire K-Band range you'll only have these "segments" (or slightly different frequencies with the same difference between them, depending on how the individual detector has self-calibrated:

24.050
24.066
24.082
24.100
24.116
24.132
24.150
24.166
24.182
24.200
24.216
24.232

Notice that there are 12 segments for K-Band, with a frequency resolution of 14-18 MHz. Like I said in the other thread that's about .1% for a 15 GHz LO, which is pretty darn good.

Also: even if they were capable of resolving the frequency with more precision, they might not WANT to resolve the frequency any closer. Think about this, if they resolved the frequency to 1 MHz: one day there's a known false at a certain location, and you lock it out. But, even high-quality gunn oscillators drift in frequency maybe 1 MHz for every few degrees change in temperature, cheap oscillators in motion sensors likely drift even more. So it drifts a few MHz in frequency by the next time you drive by, and the detector alerts to it. It would defeat the purpose of the feature if you had to lock it out again. By using larger "segments" they are probably able to avoid this in most cases.

So, assuming the 9500 frequency resolution is similar to the RX-65, and it works the way I think it does:
If one were to lock out a known K-Band false frequency for a certain location, it locks out one of the 12 segments. If a cop happened to be operating K-Band at that location at a later time, then you would have a 1 in 12 chance of missing the alert. Possibly a bit more, since one would think that the frequencies would more tend be towards the center of the band than the edges, but I don't have any real data on that. I don't think that is too bad, considering ANY type of "city mode" is technically going to compromise your protection somewhat, but that's the price we pay for a quieter drive.

Jim

[happya$$]

I assume everyone knows how the false alarms are eliminated with the new 9500i. It marks the spot of location with a GPS system with the specific band down to at least 0.001 gigahertz (plus or minus 0.001). Then when you pass by this same location, 9500i stores this information and does not alert you the next time. I have noticed with the RX65 there is consistancy with false alarms and LEO within a fraction of 0.001 but still will alert you to the threat. Question, what would happen if LEO engineers develop radar guns that would allow LEO's to "tweak" the gigahertz numbers so that the radar gun would match the Shell Gas Station false alarms as Radar Roy mentioned with that specific location?

OK first off, I don't ever see radar guns being made self-tunable just to defeat the 9500i. They would also have to add a method to the radar guns to display the frequency of the Shell Station so that they could match it. I just don't see it happening.

Second: I'm quite certain that the 9500i does not have the capability to resolve the frequency down to .001 GHz (1 MHz) even though the frequency display on the unit goes out to three decimal places. I have not had a 9500 to play around with yet, so I don't know if it is any different, but it supposably uses the M4 RF receiver like the X50 and RX-65.

First, I did a little bench test on the RX-65 in this thread:
http://www.radardetector.net/viewtopic.php?t=348

And I expanded on it at the bottom of this one:
http://www.radardetector.net/viewtop...r=asc&start=30

Here's the images from the 2nd test (RX-65 VS Freq Counter Images):
http://www.kc8unj.com/radar/freq/

So for example, throughout the entire K-Band range you'll only have these "segments" (or slightly different frequencies with the same difference between them, depending on how the individual detector has self-calibrated:

24.050
24.066
24.082
24.100
24.116
24.132
24.150
24.166
24.182
24.200
24.216
24.232

Notice that there are 12 segments for K-Band, with a frequency resolution of 14-18 MHz. Like I said in the other thread that's about .1% for a 15 GHz LO, which is pretty darn good.

Also: even if they were capable of resolving the frequency with more precision, they might not WANT to resolve the frequency any closer. Think about this, if they resolved the frequency to 1 MHz: one day there's a known false at a certain location, and you lock it out. But, even high-quality gunn oscillators drift in frequency maybe 1 MHz for every few degrees change in temperature, cheap oscillators in motion sensors likely drift even more. So it drifts a few MHz in frequency by the next time you drive by, and the detector alerts to it. It would defeat the purpose of the feature if you had to lock it out again. By using larger "segments" they are probably able to avoid this in most cases.

So, assuming the 9500 frequency resolution is similar to the RX-65, and it works the way I think it does:
If one were to lock out a known K-Band false frequency for a certain location, it locks out one of the 12 segments. If a cop happened to be operating K-Band at that location at a later time, then you would have a 1 in 12 chance of missing the alert. Possibly a bit more, since one would think that the frequencies would more tend be towards the center of the band than the edges, but I don't have any real data on that. I don't think that is too bad, considering ANY type of "city mode" is technically going to compromise your protection somewhat, but that's the price we pay for a quieter drive.

Jim

Good stuff Professor! So at best there is an 8.33% chance that a real K band threat could be missed if True Lock is used which in my opinion is pretty damn good. I will take that 8.33% chance for a more silent non annoying drive! I can not imagine having a detector to when it alerted I knew 100% it was a LEO operating radar

[microbb]

Based on Jim's previous posts the following quote of an Azonehits caught my eye.

During the test, the 9500i lockout the auto soor sensor at 24.110 which was documented on video. From experience, this frequency changes and even got 24.116 and 24.122 with the 9500i at times. As far as the frequency lock is concern in plus or minus frequency, i just don't have a clue. The MPH radar frequency during the test was reading 24.172 with the 9500i and 24.166 with my X50.

Whereas on the previous Bel/Scorts Jim was seeing a 16-18MHz segmentation, the pattern in the quote appears to be 6Mhz, assuming the 9500i had not re-calibrated. I also asked Riley if a lockout and real threat he had tested against, displayed the same Freq. on his 9500, and he stated "yes" and he had double checked it to make sure. There is also the question as whether the internal frequency being processed has greater resolution that is what is being displayed.

Something I have not seen mentioned is the possibility of a GPS weighting factor coupled to the frequency. The accuracy of the GPS chipset should be at least 10M and possibly 5M. A possible confidence factor based on a series of concentric distance rings to a locked out location could be part of the equation. So a simplistic scenario i.e. that a 300M radius is the dividing line for a designated frequency may not be the case. This could be a means of controlling for the cheap door sensors that are prone to drifting, so that only ignoring a particular piece of spectrum may not be the only thing the 9500 is doing.

It may be constantly examining things like:

1. Distance to lock-out target.
2. The stored signal strength at the original lock-out postion, and a calculation of what the signal strength should be at current position.
3. Closing rate to that lockout position, and an estimate of what the signal level ramp up rate should be.
4. Direction of travel.
5. How close the frequency is to the original lock-out.
6. What other in band signals are present and how do the above factors correlate in friend/foe ID.


Unless the 8500 had been designed using a microprocessor and DSP with a significant amount of processing & memory headroom, then most likely there has been a significant processor/DSP upgrade to handle all the new tasks being asked for in the 9500. My guess would be the receiver portion & DSP remains the same, but the processor gets a major boost.

microbb

Relatives don't let relatives, drive with Cobras. :twisted:

[jimbonzzz]

Based on Jim's previous posts the following quote of an Azonehits caught my eye.

During the test, the 9500i lockout the auto soor sensor at 24.110 which was documented on video. From experience, this frequency changes and even got 24.116 and 24.122 with the 9500i at times. As far as the frequency lock is concern in plus or minus frequency, i just don't have a clue. The MPH radar frequency during the test was reading 24.172 with the 9500i and 24.166 with my X50.

Whereas on the previous Bel/Scorts Jim was seeing a 16-18MHz segmentation, the pattern in the quote appears to be 6Mhz, assuming the 9500i had not re-calibrated.

Hard to say. But if it is indeed 6 MHz, one would end up having to lock out this particular motion sensor at least two different times, since this same motion sensor displayed a 12 MHz difference on the 9500i at different times (24.110-24.122). This would result in two 6 MHz segments (a total of 12 MHz) being locked out.

[nascar]

I think you guys need to look at this in more "real world" situations.

1. The lock out area is very small.

2. The LEO would have to IO you while you are in the small locked out area. And even in this area I believe the signal strength of the LEO's radar would alert the 9500i even on the same band. Most door openers are facing downward to detect people, not cars on the road. So the door opener signal would be much weaker then the LEO's aimed at you.

I feel the real situation would be that you would get alerted to the LEO's radar well before you ever reached the small locked out area. Thus you would be alerted to his radar.

One experience I had was I previously locked out a K band door opener and before I reached the locked out area I was hit with LEO Ka. I know this is a different band than the door opener but it illustrates my point that the Ka from the LEO overrode the locked out the small K area as I proceeded through it. The 9500i indicated both the locked out K as well as the LEO's Ka.


I might be missing your point, but this is what I feel will happen. I wish I had a K band radar gun at my disposal to do my own testing.

[compu44]

Realistically, I mute so many K-Falses on my Rx65 I stand just as bad of a chance of missing an alert on it really.

Is there any conclusive evidence either way as to whether the signal strength is taken into account when determining locked out false from LEO source?

[microbb]

I don't think there's conclusive anything yet. There does appear to be a high confidence level from experienced users that the 9500 is doing what it needs to do, per the marketing claims made for it.

My post was aimed more at understanding how it does it, and the need to examine a wide variety of possibilities and internal capabilities, should it exhibit uncanny behavior that does not fit within a simplistic model.

As an example, assume the lockout radius is 300M on K-band, and a door opener at 24.166 has been locked out. At 310M the 9500 will alert to 24.166 but at 290M it will not(taking into account +/- 10M GPS accuracy). That is a very simple model of how the 9500 MAY work.

But assume in the above example if you were driving through an intersection E<>W when you locked out a false at 24.166. Returning to the area on the same bearing, the door opener has drifted to 24.160, but the 9500 considers this within a spectrum window of acceptance and does not alert. However if you approached the intersection from N<>S, 24.166 would not alert, but 24.160 would because of a difference in bearing to lockout target.

So far owners seem pretty happy with it no matter what's going on inside.

microbb