Squirrel's Thoughts on the G2

SwankPeRFection

Corgi Lovers
Beginner User
Joined
Mar 31, 2020
Messages
1,398
Reaction score
1,409
Just me. The meter bar was low for a few seconds and then went high.

I think this was tested and they found it to definitely be an issue, but I can’t remember the details. With the new scale being only 6 LEDs, perhaps 2-3 on I/O us just outside the kill range and then anything above that is within it and that’s just how it’s going to detect. Without it being C/O, you just don’t have that longer time at 1-3 where you’re picking it up and getting closer and then eventually it pegs as you come into that range. Were you speeding and slowed down and all was fine or no and there’s no way of know if you’d have gotten tagged and pulled over or not?
 

Deadhead1971

Rock, rock, roll, Plymouth Rock, rollover
Advanced User
Joined
Dec 7, 2014
Messages
4,370
Reaction score
8,701
Location
Central NC
I think this was tested and they found it to definitely be an issue, but I can’t remember the details. With the new scale being only 6 LEDs, perhaps 2-3 on I/O us just outside the kill range and then anything above that is within it and that’s just how it’s going to detect. Without it being C/O, you just don’t have that longer time at 1-3 where you’re picking it up and getting closer and then eventually it pegs as you come into that range. Were you speeding and slowed down and all was fine or no and there’s no way of know if you’d have gotten tagged and pulled over or not?
I was not speeding really- about 5 mph over the PSL.
 

SwankPeRFection

Corgi Lovers
Beginner User
Joined
Mar 31, 2020
Messages
1,398
Reaction score
1,409
Yeah, that’s not enough to know if you would have gotten pulled or not. Even if the ramp up issue wasn’t there (if there really is one), I/O up close when you don’t see them and you’re the only one might as well be like a Lidar hit.
 

cihkal

Stepping out of the Matrix...
Premium Plus
Lifetime Premium
Corgi Lovers
Advanced User
Joined
Apr 21, 2014
Messages
4,199
Reaction score
8,091
Because you mentioned FPGA in the OP, I wanted to share some information related to it that I thought was really cool:


Makes me feel even better about the update process. I have some suspicions on how they get around a lower LE count compared to other RDs, but I'll save that for another discussion unless there's interest here.

I personally believe the memory next to the FPGA, the high-speed access memory, holds all the signal information.
 
Last edited:

Bossdad71

Premium Plus
Lifetime Premium
Advanced User
Joined
Jul 11, 2016
Messages
8,070
Reaction score
11,889
Location
Kansas
Did you remember writing this when you sobered up? :)
 

SquirrelMaster

Likes Unicorns 🦄
Administrator
ModSec
Premium Plus
Lifetime Premium
Corgi Lovers
Advanced User
Joined
Dec 3, 2015
Messages
4,651
Reaction score
12,492
Location
Liberal California
but I'll save that for another discussion unless there's interest here.

Theres always interest for tech discussions.



Did you remember writing this when you sobered up? :)
You caught me qt the wrong time... 3 vodkas, 2 rums, and 3 glasses of wine in.... Sober aint the right word
 
Last edited:

Gunbjorn

Incognito
Beginner User
Joined
Oct 10, 2015
Messages
121
Reaction score
261
Location
Far Above Sea Level
Theres always itnerheres for tech discussions.
Ah, yes. Let me know when the Squirrel dictionary is published and i can be the forum translator. :)
Post automatically merged:

Ah, yes. Let me know when the Squirrel dictionary is published and i can be the forum translator. :)
Great writeup btw!
 

SquirrelMaster

Likes Unicorns 🦄
Administrator
ModSec
Premium Plus
Lifetime Premium
Corgi Lovers
Advanced User
Joined
Dec 3, 2015
Messages
4,651
Reaction score
12,492
Location
Liberal California
Ah, yes. Let me know when the Squirrel dictionary is published and i can be the forum translator. :)
Post automatically merged:


Great writeup btw!
Lmao no dictionary needed yet. Im not THAT far gone...




Makes me feel even better about the update process. I have some suspicions on how they get around a lower LE count compared to other RDs, but I'll save that for another discussion unless there's interest here.

yeah that's basically how most things update nowadays. Sort of an A/B partition system. Everything can be done seamlessly without you even knowing an update is being done.
In fact, i almost garuntee not only the fpga, but also the main mcu is doing the same thing with how its dividing its flash.

Zero downtime should a bad image be flashed or a flash fail halfway.


I personally believe the memory next to the FPGA, the high-speed access memory, holds all the signal information
The non-volatile memory next to the max10 is for sure used for signal information. It can't be used to store the bitstream anyways as it would get lost on a reboot.
 

Isst

Learning to Fly
Beginner User
Joined
Jun 8, 2013
Messages
98
Reaction score
85
Location
Southeast
Haven't typed up a detector review/thoughts in a while.... I am just the right amount of inebriated for it however, so lets go!

First of all, what a HUGE improvement over the G1 this is. I was very pleasantly surprised.

Lets start with the temporary bad stuff and get them out of the way:
Quality control
This is the big negative to the G2. The issues with it dying and not waking up (my take on that issue here) as well as the displays delaminating.
As much as both of these specific issues suck, they don't really speak to the design of the detector and really don't matter much as if it happens, you'll be taken care of by VR.


Packaging and Accessories
The G2 is packaged pretty well. It comes in a nice box, padded with foam, and chock-full of accessories. They include so many things in the box!
Not only do you get the detector and mount, you get two choices of power cables to go along with it (coiled or straight), an extra mount if you want to hide the detector up on the headliner, a hardwire kit, and spare suction cups. Super cool to see that much extra crap thrown in.

My one complain here is that they fit the product box perfectly inside a brown shipping box. This feels great but sucks when inevitably the carrier uses a baseball bat as the primary tool to load and unload the package. Wish it was a bigger box with maybe some packing peanuts or bubble wrap so the product box doesn't get damaged during shipping.
View attachment 153440


Mounting
Overall, the mount is the same as the G1. I still do not like the way the windshield mount locks to the detector. On the G1, some units had their paint/coating rubbed off on the top ribs from being taken out of and put into the windshield mount. Probably will see the same thing here.
The windshield mount is alright. It seems a bit more wobbly and less stable than the G1 was. It would bounce around a lot on my windshield. The visor mount also aims it down on my car so not really a good option. I think that if I run the G2 permanently, I would use a blendmount for it.

Overall Build and Look
This thing looks super sleek. That combined with the metal case make the thing look and feel amazing. It definitely shows quality on the outside. The display up front kind of takes away from that but is not too bad. I like that its now dark with no letters or visible text on it until you power it up but I wish it didn't have the faint purple see-through layer. Could have gone for a translucent black layer instead.

A big change is the lack of knob. While the knob was awesome, I get why it is gone. Its far cheaper to use buttons instead. It's no less different to me to use than before however.

Overall, It feels nice, sturdy, and premium.

View attachment 153441


Display
As I previously mentioned, I like the new display. It looks far better than the G1 did. That being said, it kind of sucks to see they didn't use this huge refresh to add a frequency readout. That would have made this detector far better.

Back to the display, it is easy to see during the day and doesn't bother you at night thanks to the automatic brightness option. Really good addition to the detector.
One of my biggest complaints about the G1 was not being able to distinguish what band the thing was alerting to by looking at the LEDs on the display. This solves that by making the LEDs light up the letter of the band. Super awesome change.

I guess I do have to mention this; there's now only 6 strength bars instead of 8.... cool..

A huge complaint I have is how unintuitive it is to program. The settings are not intuitive to get into and impossible to set without looking at a reference table. The table as well does not come with the detector. Its user manual (and what I could find on their website) say absolutely nothing about how to turn different options on and off using the detector itself.

Pic of the purple cutouts and the UGLY "Z" shape of the bogey counter:
View attachment 153447


Performance
Here is the fun part. I did some basic testing against my collection and some more in-depth testing against my main detector, the R7. The testing I did was on windy. hilly, mountain roads as that's what I'm mainly on due to where I live.
Long story short, the G2 curb-stomped my R7. On direct off-axis, the G2 gave a marginal but significantly better alert than the R7.
On testing the range over a road with valleys and hills as well as curves, the G2 made the R7 its bitch. Against 34.7, the G2 first picked up the radar signal a whole mile before the R7 picked it up.
Not only did it consistently beat the R7 around curves, but it also picked up further over valleys and hills. I'd say it is better suited at tricky terrain from my experience.

False alerts are far fewer than the G1 was. I live in an area full of cars with BSM systems. For the most part, the G2 does a decent job of filtering them out. I do not get annoyed with it because of how occasional the alerts are. It is about on par with my R7 in terms of annoyance. That is, until my car's BSM triggers it. It for some reason doesn't filter out my car so at slow speeds in city driving, it constantly alerts to my BSM.

There are some issues that need to be mentioned.
First, the ghosting is ****ing annoying. It is not productive to my chakra to get a Ka alert that then turns into 9 bogeys. This happens way too often.

Second is the laser detection. As it is on most detectors, laser detection is useless here as well. This, however, is not for the same reason as other detectors. First of all, once your detector alerts you to laser, if it is not a false, it means it's too late. Now that that is out of the way, the main reason that the G2's laser detection is useless is because of how sensitive it is and how little filtering it has. Every little thing sets this off - weird HID headlights, phone's IR sensor and autofocus, etc. The worst thing about it's lack of filtering is that it falses to my car's laser collision avoidance system. It is very much a boy who cried wolf system.
According to the patents, however, the lidar sensor is connected to the FPGA so it should be EXTREMELY simple to create a basic counter to filter out non-police lidar pulses. And this can all be done with a "software" update. I have hopes, but in its current state, it is unusable.

Finally, the most important issue is the lazy Ka alert. The G2 detects Ka almost instantly (usually). It, however, appears to average the signal over a period of time. This means that if hit with IO on Ka, instead of getting a full-tilt alert, you get a small blip for a second or two and then an immediate ramp up to full blast. I have observed this in my lab testing and others have seen it in real world situations. This is a HUGE issue.

Technical Things
I don't mean this section as a technical discussion, I just wanted to post some thoughts I had on the technical side of things that didn't fit anywhere else.
My hardware version is REV C on the main board and REV E on the RF board.
We have seen some really dodgy things from the G2 internally. There have been bodges, dirty solder joints, flux everywhere, etc. My unit had almost none of that crap. Overall it was very clean inside and there was very little after the fact hand soldering as we had seen before. Another thing was that the horn itself seemed cleaner. The first G2 I had was full of metal shavings embedded into the radar absorbing rubber. This one was clean.
View attachment 153445


The one bodge that it did have was pretty bad though.
There was a capacitor soldered in manually with a terminal less than a millimeter away from a power rail..... I ended up desoldering it, cleaning the area up, and resoldering it so it was a bit further away from the resistor. Here you can see how poorly it was soldered in originally.
View attachment 153446


One thing I want to note is that in my attempts to figure out the power issue, I noticed that the unit draws virtually no power when it is shut off. It looks like they are not just shutting off the FPGA and microcontroller when you power it off but are also shutting off the internal power rail. This effectively means you can leave the thing plugged into your car (but powered off) when you're not driving for a while and it will have virtually 0 effect on draining your car's battery. It's the little things..... :)

It was nice to see the bluetooth inside of the unit now but I wish the antenna for it was routed a bit smarter. Seems to work fine though.
Also super cool to see a FPGA inside the detector. Hopefully they can use it to its full advantage.. A FPGA is far from the cheapest component out there. Them using one shows they actually care enough to make this thing a beast and not the keep their margins high. Kudos here.

Unfortunately, there is still no GPS on the unit for at least low speed muting but I get that it's hard to do that with a metal enclosure. GPS would take forever (if at all) to get a satellite lock.

Final Thoughts
I started off by saying the G2 is a huge improvement over the G1. It definitely is. It's a whole other beast... Aside from it being completely better than the OG model, it also seems to be better than most other units out right now.
It has great range, good filtering on radar signals, good build quality overall (besides little temporary QC shit), and was obviously designed with attention to detail.

The detector (once set up) can definitely be used standalone without any issues (or with an app if that's your kink). Yes, the apps may make it far better, but it by itself is rather good. It still lacks a frequency display which sucks for running it by itself, but for people that aren't huge enthusiasts or like to run apps as their primary unit, that's not a big deal.


The G2 definitely gets my humble recommendation with the caveat that Ka IO is lazy currently. Once the Ka IO issue is fixed, the G2 will replace my R7 as my main detector until something better comes along. If you're thinking about getting a G2, by all means, go for it. It's a good little unit.
 

cihkal

Stepping out of the Matrix...
Premium Plus
Lifetime Premium
Corgi Lovers
Advanced User
Joined
Apr 21, 2014
Messages
4,199
Reaction score
8,091
Have to say now, after using a V1G2 since "pre-release" that the laser filtering has been kicked up big time!

The Gen 1 was a falser. The Gen 2 started that way until the update, and boom now it works very well in my experience. I would say they're using the FPGA for filtering, based on this (below docs 😘*) and real world experience. Very cool!
 

Attachments

  • IMG_20200728_184417.jpg
    IMG_20200728_184417.jpg
    574.9 KB · Views: 28
  • Screenshot_20200328-153237~3.png
    Screenshot_20200328-153237~3.png
    501.2 KB · Views: 28
Last edited:

Deacon

TXCTG
Premium Plus
Lifetime Premium
Advanced User
Joined
Nov 13, 2016
Messages
12,828
Reaction score
17,822
Location
Hill Country, TX
I would say they're using the FPGA for filtering, based on this and real world experience.
You’re basing the specific claim that they’re using their FPGA for filtering based on the idea that filtering exists? Maybe they are, but I don’t see how you can say that so matter-of-factly.
 

Rags

Tumbleweed
Advanced User
Joined
Oct 26, 2011
Messages
2,080
Reaction score
1,750
I don't know what they are using to filter lidar, but whatever they are using, it's working extremely well. I have had no false alerts on Lidar at all with mine. My V1G1 is most definitely noisier with false alerts, so they are doing something. Whatever it is, it works. It works so well that I am confused by the assertion that the V1G2 has issues with false alerts in other posts. I fully expected to have those issues based on what others have asserted, but so far so good.
 

jdong

Premium Plus
Lifetime Premium
Advanced User
Joined
Jun 5, 2013
Messages
7,406
Reaction score
11,467
You’re basing the specific claim that they’re using their FPGA for filtering based on the idea that filtering exists? Maybe they are, but I don’t see how you can say that so matter-of-factly.
I think the laser diode is connected to the FPGA so more likely than not, the FPGA has to be involved with filtering laser, right?

With that said, it doesn’t necessarily mean there’s anything fancy going on. It could be just counting a minimum number of pulses and making sure the pulse rate is within a certain range, which is how a lot of other low-false-alert laser detectors like the Redline have worked for ages.
 

Deacon

TXCTG
Premium Plus
Lifetime Premium
Advanced User
Joined
Nov 13, 2016
Messages
12,828
Reaction score
17,822
Location
Hill Country, TX
I think the laser diode is connected to the FPGA so more likely than not, the FPGA has to be involved with filtering laser, right?
It certainly could be. But with all this talk of FPGAs with Theia, it seems like even a basic FPGA is assumed to possess godlike powers that I’m not convinced are necessarily accurate.

With that said, it doesn’t necessarily mean there’s anything fancy going on. It could be just counting a minimum number of pulses and making sure the pulse rate is within a certain range, which is how a lot of other low-false-alert laser detectors like the Redline have worked for ages.
Could be, with the right programming.
 

cihkal

Stepping out of the Matrix...
Premium Plus
Lifetime Premium
Corgi Lovers
Advanced User
Joined
Apr 21, 2014
Messages
4,199
Reaction score
8,091
It certainly could be. But with all this talk of FPGAs with Theia, it seems like even a basic FPGA is assumed to possess godlike powers that I’m not convinced are necessarily accurate.


Could be, with the right programming.
I didn't assume that, thanks Deacon lol.

Jdong explained it for ya, doesn't have to be fancy filtering.

FPGAs aren't gods, just in this case they can count quickly.
 

Attachments

  • Screenshot_20200328-153237~3.png
    Screenshot_20200328-153237~3.png
    501.2 KB · Views: 12
Last edited:

jdong

Premium Plus
Lifetime Premium
Advanced User
Joined
Jun 5, 2013
Messages
7,406
Reaction score
11,467
It certainly could be. But with all this talk of FPGAs with Theia, it seems like even a basic FPGA is assumed to possess godlike powers that I’m not convinced are necessarily accurate
Haha so true. FPGAs can do boring things too, and in this case that’s very likely what’s going on. Perhaps in the fanciest form, they are counting pulse rate and count on the FPGA and sending that to the MCU to apply a few lines of C code to filter. Certainly no Theia level magic.
 

biolink

Learning to Fly
Beginner User
Joined
Dec 5, 2012
Messages
383
Reaction score
837
A huge complaint I have is how unintuitive it is to program. The settings are not intuitive to get into and impossible to set without looking at a reference table. The table as well does not come with the detector. Its user manual (and what I could find on their website) say absolutely nothing about how to turn different options on and off using the detector itself.
Do you think this was intended to keep the casual user out of programming so they don't deactivate a feature? This would certainly reduce calls to VR from angry users claiming they didn't pick up an LEO. Just a thought.

I do like the fact that the app makes programming a breeze. I do agree, though, the chart and manual programming is a nuisance.
 

SquirrelMaster

Likes Unicorns 🦄
Administrator
ModSec
Premium Plus
Lifetime Premium
Corgi Lovers
Advanced User
Joined
Dec 3, 2015
Messages
4,651
Reaction score
12,492
Location
Liberal California
Welp looks like VR has finally fixed the complaints holding this back from replacing my r7. (https://www.rdforum.org/threads/105782/post-1527164).
Only complaint now is the laser filtering.


At this point, in my opinion, the only advantage the r7 has over the g2 is a frequency display! With that said, G2 has replaced the R7 on my daily's windshield.
 

jcaptchaos2

Learning to Fly
Beginner User
Joined
Dec 27, 2015
Messages
171
Reaction score
223
They both are incredible
 

Discord Server

Latest threads

Forum statistics

Threads
90,936
Messages
1,384,070
Members
22,877
Latest member
MilSpec
Top