Junk filtering tests

How about doing some tests to see how well various detectors filter out false signals like Cobras? I was just driving earlier today and noticed how my older Bel 980 responded to a nearby Cobra, and wondered how others would handle the same situation. Basically power up a known source of interference and drive towards it, noting any signals. Perhaps also include a real gun nearby and make sure one is ignored while the other is alerted to.

And if we could duplicate an infinity speed system for example, similar tests on the laser detection side of things would also be interesting.

That might be interesting, I remember some of the magazines used to do something like this...

Quote:

---

Originally Posted by jimbonzzz

That might be interesting, I remember some of the magazines used to do something like this...

---

Car and Driver tests from the 80's and 90's.

J/

The Pro 78 is a top performer at filtering. With POP Off and Filter 2, wouldn't surprise me if it filtered just as well as a V1.

But I say bring on the tests!

Some Cobra models present especially nasty problems for any RD to filter out. For example, a couple of weeks ago I was headed to town and I decided to use my STi that particular day. As I approached the local highway and local corner store near where I live, I got a 35.5Ka band alert. I thought, "Cool! I am gonna follow that GSP and do some range tests." I drove closer to the corner store and noticed a pickup truck pulling out and heading up the highway. The 35.5 signal disappeared, yet I did notice the familiar shape of a Cobra RD model stuck on the truck's windshield. I could also tell, due to the small size of the Cobra, that it likely was one of the cheap models sold at my local Walmart. I quickly turned left, instead of right to go to town, and followed the truck for about a mile or so. Here I was driving roughly 1500 feet behind the pickup truck and my STi was generating a continuous alert which varied between signal strength 1 and 3. There were no other vehicles on the road. I backed off from the pickup truck to around 3000 feet and the signal disappeared. So, there you have it! Not even Bel's top-of-the-line RD model is immune to getting falses from a cheap Cobra RD model! And yet the STi might have filtered it out if this Cobra RD has been in a vehicle which was approaching me since then the STi would have been able to see the Cobra's full set of LO emissions and realized that the 35.5 signal was a false.

So, after this experience, I too would like to see GOL perform some RD filtering tests. And after this experience I realize that such tests would have to include both approach tests as well as trailing tests as well. :)

Yes. You can see that VR are pushing the Ka sensitivity to the point where the Ka guard has to work well. This would be a very important test imo. The fact that got me off side with bell detectors was that the RX65 and the 990 int had very very very poor Ka filtering which made the Bell detector way to chatty on Ka and really turn me off bell. When i first started running my V1 .813 the Ka guard was stellar which in turn made New Zealand's very first Zombie.
This is IMO a Big sales point and with more education to the masses you never know the whole of the USA if not the world will become V1 Zombies. :zombie3::big_boss:

Yeah, the V1 may be "chatty" but it certainly is not false-prone.... In fact I've only had once or twice of all my time with my V1 where it alerted to a non-radar source. My Whistler 1793SE and Cobra XRS-9400 talked a LOT less on Ka band, but falsed a lot more. When you see that Ka light come on with the V1, you KNOW you're in for trouble.


The V1 is just simply waaaaay too sensitive on X and K bands, particularly off axis. Using 4-dot big-L is not actually all that absurd, as I've found that the V1 at 4 dots of X or K is just about the strength at which other detectors START alerting anyway!

Interesting... Wonder how we could do this and keep it consistent?

Quote:

---

Originally Posted by Ovencleaner

The Pro 78 is a top performer at filtering. With POP Off and Filter 2, wouldn't surprise me if it filtered just as well as a V1.

But I say bring on the tests!

---

You rarely hear the V1 being a good filterd RD, but I agree.

Quote:

---

Originally Posted by Rottman

You rarely hear the V1 being a good filterd RD, but I agree.

---

Its Ka guard with POP Off almost never falses is what I hear....

Quote:

---

Originally Posted by thestaton

Interesting... Wonder how we could do this and keep it consistent?

---

7000 runs up and down the "test strip". :rolleyes:

Each detector for a base run or 2 to get benchmark, then mount 1 detector center windshield and test its performance each time with a new detector mounted beside it. Repeat with another detector mounted center windshield etc..

Think of the size of the spreadsheet!!
J/

I would of thought that a real world test would be better. Like grab a well known cobra leaker stick in a car. Grab another car put the test detectors in it and take it from that.
What I'm pretty sure of is that if you start using a real police radar with the cobra and a test detector eg the V1. The V1's gonna show 2 bogeys as the filtering seems to go out the window with real cop radar going (interesting eh) . Better than a J thou.

Not as easy as that.... I cannot find a detector that induces a repeatable false on the V1.... I KNOW my Cobra at far distances will do it, but up close the V1 is silent to its presence.

Quote:

---

Originally Posted by jdong

Not as easy as that.... I cannot find a detector that induces a repeatable false on the V1.... I KNOW my Cobra at far distances will do it, but up close the V1 is silent to its presence.

---

That's a great point. I actually picked up a Cobra so I could see it's effect on my other detectors, but couldn't get the V1 to fall for it at all (with Ka-Guard ON). I didn't even think of trying it from farther away. Thanks!

Quote:

---

Originally Posted by ghz1

Like grab a well known cobra leaker stick in a car.

---

Do we know for sure which Cobras are the well known leakers?

Quote:

---

Originally Posted by Stealth Stalker

Do we know for sure which Cobras are the well known leakers?

---

The Cobra esd7000 is a bad leaker.

Quote:

---

Originally Posted by djrams80

The Cobra esd7000 is a bad leaker.

---

Just our luck, I'm betting that is their best seller too! :rolleyes:

Quote:

---

Originally Posted by jdong

Not as easy as that.... I cannot find a detector that induces a repeatable false on the V1.... I KNOW my Cobra at far distances will do it, but up close the V1 is silent to its presence.

---

I have certainly passed some that are. One I know was the letter-number brand that is unmetionable here run by a friend. Just need to go shopping for some of the junkiest detectors you can find. :)

The dirt cheap ESD-7000 leakage is horrible since it uses an old sytle Gunn oscillator within its tiny antenna.

The ESD7000 may be a K band leaker too. It was one of the smaller Cobras with the black case that was causing full scale K band alerts on my 985.

And they're selling the ESD7000 at Fry's dirt cheap. :mad:

GTO_04

for some reason my bel v985 if in the same car as my v1, will constantly go off on ka band from the v1. does that mean my v1 is leaking?

Quote:

---

Originally Posted by 06blackgto

for some reason my bel v985 if in the same car as my v1, will constantly go off on ka band from the v1. does that mean my v1 is leaking?

---

Yes. All RDs leak emissions from their local oscillators. Your V985 also leaks, but your V1 is ignoring it even though this could seriously hamper the V1's detection range.

Yeah in my experience, the V1 is one of the most resistant detectors to Ka leakage. It's pretty clear Valentine put some extra effort into addressing the issue of Ka falsing from other oscillators in the vincinity, but indeed this almost certainly means the V1 is ignoring a certain spectrum of Ka band to filter out that detector. The million dollar question is, if a police radar gun is sitting in this frequency, would the V1 turn a blind eye? :)

Quote:

---

Originally Posted by jdong

Yeah in my experience, the V1 is one of the most resistant detectors to Ka leakage. It's pretty clear Valentine put some extra effort into addressing the issue of Ka falsing from other oscillators in the vincinity, but indeed this almost certainly means the V1 is ignoring a certain spectrum of Ka band to filter out that detector. The million dollar question is, if a police radar gun is sitting in this frequency, would the V1 turn a blind eye? :)

---

No it would not. It's years of using the same design detector which has a detector that's almost Ka false fear.

Quote:

---

Originally Posted by jdong

Yeah in my experience, the V1 is one of the most resistant detectors to Ka leakage. It's pretty clear Valentine put some extra effort into addressing the issue of Ka falsing from other oscillators in the vincinity, but indeed this almost certainly means the V1 is ignoring a certain spectrum of Ka band to filter out that detector. The million dollar question is, if a police radar gun is sitting in this frequency, would the V1 turn a blind eye? :)

---

Jimbonzzz could explain this better, but basically a RD looks for the primary LO frequency and/or the first harmonic whenever it detects Ka band radar. If the RD sees the LO or harmonic then it assumes that the Ka signal is a false higher harmonic. This is a simplistic way of describing what occurs during signal processing, but it should give you the basic idea. So, as long as there is no other radar detector around, then the RD will alert to the radar gun, assuming that the RD is a Belscort, Whistler or Valentine. This might not be true for cheap import models which may employ strong notch filters in order to ignore specific Ka frequencies. Belscort, Whistler and Valentine own all of the patents for false signal rejection. This doesn't leave any alternatives for the cheap import RD manufacturers.

Quote:

---

Originally Posted by jdong

Yeah in my experience, the V1 is one of the most resistant detectors to Ka leakage.

---

The others must be horrible then, because on my trip to Floridia and back, I encountered at least six Ka false alerts that were attributed to nearby RDs. I'm not good enough at visual RD identification to confirm that they were Cobras, so I'm not sure what they were. Are there other RDs besides Cobra that cause false Ka alerts?

What revision V1 do you have? I've seen videos of the 3.813/3.826 falsing to Ka on Cobras, but I have not been able to get any Cobras to set off my 3.863, even with POP enabled.

Quote:

---

Originally Posted by jdong

What revision V1 do you have?

---

Not sure. Bought it new from VR in July. I'd have to find the box to find out, I guess. It's in default settings, A mode.

Quote:

---

Originally Posted by MEM-TEK

Jimbonzzz could explain this better, but basically a RD looks for the primary LO frequency and/or the first harmonic whenever it detects Ka band radar.

---

Pretty much... except that the "first harmonic" is the same as the fundamental LO freq. Two times the fundamental is the "2nd harmonic".... but I've made this same mistake on occasion in the past, so it's all good ;) If someone wants a real explanation they can go here: Radar Detector Filtering - Guys of LIDAR

Of course, for best results a modern detector would have different algorithms and schemes to deal with different "platforms" of leaky detector models.

I once heard a story about a certain detector engineer at who (of course) drove with his detector in his vehicle for testing purposes. Out on the road, when he encountered a new false alert from a leaky detector, he would try to identify the vehicle and either follow them to their destination or try and flag them down, and then offer them $$$ for their detector. Then, he could take it back and analyze it in the lab to develop the best scheme to prevent the false alerts caused by that particular unit. :D

Could it be Mike B. and how FILTER,FILTER 1,FILTER 2 started. I use Filter 2 95% of the time and its super quiet on the interstates. I have seen as many as 2 Cobras in the same car side by side :mad:

Hah, that could be the next big thing in false alert rejection: "Mark this signal as a false", which the detector then takes a sample of time vs frequency (i.e. STFT) for a bit and uses the data for bayesian analysis....


Crap. I call patent on the idea! *Goes launch the Jdong One product line*

Quote:

---

Originally Posted by Jag42

Could it be Mike B. and how FILTER,FILTER 1,FILTER 2 started. I use Filter 2 95% of the time and its super quiet on the interstates. I have seen as many as 2 Cobras in the same car side by side :mad:

---

I too found Filter 2 mode on my Pro78 to be rather effective with marginal impact on overall performance.