Printable View
I am using a Pic18 I picked up few some years back. neat thing it had was a full USB supported interface for programming. I think You can still buy them for about 3-5 bucks now but there is allot better out there. the one I am using has a meg of flash on itQuote:
Originally Posted by kpatz
Solion
I just thought of something. Some part of me wants to call it the "Spyder system"
Lets take 1 Blinder.(check)
Take a handful of IR photo transistors or photo diodes (whichever they use) and pigtail out sensors for every light housing and possible "hit" point on a car.
Mmmmm me thinks this will be an interesting year
Solion
:)
Solion just touched on a thought I had while reading this thread.
First off, some background. I just finished a (partly successful) MIRT build involving 144 high power IR LED's, a constant current source, and a PICF184550 (plus a bunch of nice goodies like an LCD, etc).
My thought was why not use a photodiode as a receiver and use it to trigger your LED array. The key here is to appropriately filter the output of the photodiode to remove DC bias (from the environment) and external light sources. Since we know that the pulses we are expecting are square wave, a highpass filter with a high corner frequency should do. Stick the output of the filter to an interrupt pin on the PIC and have it pulse the LED for a short duration (couple mS maybe). Since there will be a brief delay to handle the interrupt (usually 5-8 instructions which = 3.5 uS) the LIDAR gun will receive two pulses for every one that it sends. The second pulse will be about 3.5 uS delayed.
It was my understanding that LIDAR guns take several samples for each distance reading. Since the gun now receives two valid pulses (the 3.5 uS delay equates to an added distance of about 500 m, which is within acceptable limits) it has to throw out that sample.
If you upped the clock speed of the PIC to 24 mhz (easily handled by the PIC I am using) you could return the pulse as quickly as .875 uS later for an equivalent added range of 125 m.
This is how I understood the ZR3 to work...
Actually for the test, I have hacked the laser trig from my old cheap laser/radar detector for for my jamming array. It gives me an isolated 1.43v to flip a transistor (2n4401) seemed to be an easy way to do it without having to design a detection circuit.
As for what you are saying about the delay. Yeah, that was one of my concerns as well, I think however I will probly drive the clock higher like you suggest. 3.5us may not seem like allot, But int he world of computing it's forever.
What I think I may do ( and still keep it under budget ) is setup a secondary array of IR photo-diodes that when hit will flip the jammer heads to an ultra simple 1kz pulser (always on) then roll back to the PIC when it executes.
However, before I do that I will be testing to see if I gain any success at all at jamming. short of the brute force 2khz omygodIamlookingintothesun I may be outta luck.
Really cannot wait to see if I can have any effect on Lidar guns. this really is the 15.00 and a soldering iron sort of project.
Thanks for the input
Solion
I've built a detector front end using the LT1328 IrDA receiver chip. So long as the photodiode has decent optical filtering for 900nm-ish peak efficiency there is very little noise. It is possible to make GHz bandwidth receivers with a few high grade, low bias current op-amps for even better detection. The 1328 gives a full scale output so it would be pretty easy to scatter a bunch of these detectors all over the car and trigger the jammer heads already installed on the car.
I wouldn't be too concerned about the interrupt handling delay. You want a certain amount of delay - its the presence of a second delayed signal that confuses the LIDAR gun. You just need to ensure that the delay is not so much that the gun can filter it out as a "jammer". My rough calculations based on 8 mhz would cause a secondary pulse to arrive 3.5 uS delayed which, based on the speed of light, works out to be the same as if the LIDAR was ranging a car that was an added 500m away.
I'm sure this is close enough that it will not be filtered out as Jammer noise.
Another concern is if the LIDAR detects the LENGTH of the pulse as well. If you are sending a 2 khz Brute-Force signal, what duty-cycle will you use? The advantage of using an IR photodiode and the interrupt pin is that you could trigger on the rising and falling edge - reproducing the source signal precisely.
UPDATE: Scratch that, research tells me that the LIDAR pulses are extremely short. On the order of a few nS! I think it is unlikely that the guns would do any sort of measurement on the duration of these pulses. Based on this, the interrupt handler for the pulse should be nothing more than turning the LED on and then immediately turning it off. This is good news; LED's can be overdriven substantially when run at low duty cycles. The duty cycle for a single instruction LED pulse (based on an 8 mhz clock) on a 1khz freq would be .05%!!!
An LED designed for a current of 100mA could easily handle 1A - substantially increasing it's output.
Right on brother. Im an engineer as well and this project intruigues me! I built an OBD-I reader with the help of other guys and love it!! I would be very interested in learning more and helping if I can.Quote:
Originally Posted by Solion
Im with an optical networking company that produces MEMS devices, CW light sources using laser diodes. We throw the things away by the case full due to hermeticity issues etc. Im going to dig around and find out the specs on these things.
What I dont understand is why you are trying to find a pulse rate? Why not swamp out the receiver circuit of the gun? Is it to avoid jamming notification at the gun? Your LEDs laser or not will always be brighter than the reflected signal.
I feel stupid potentially having throw away diodes that can build jammers. Im hoping they are closer to the 902nm that is needed.
Please keep posting on the DIY project! This is exciting.
The biggest problem with swamping out the receiver rather than pulses to trick it is the amount of light output required. You're broadcasting light out in a general sweep of an area to overcome the reflection of something which by comparison is more focused. I saw a web page, a little dated by still relevant which contained all the light output calculations and requirements for such a device and it can and has been done but its much more expensive and more prone to failures.
I'll see if I can dig it up, it was great article.
If you use delayed return pulses they'll have to be extremely close to the gun's returned pulse, as the gun itself only responds to the first returned pulse. So if the gun's own pulse returns before your jammer's pulse, guess what, the gun ignores your jammer's pulse!
You need to use an algorithm that anticipates the arrival of the next pulse (by determining the pps rate) and then sending pulses out AHEAD of the gun's next pulse. That way the gun measures based on your jammer's pulses instead of the gun's pulses and you can achieve a successful jam by varying the timing of the return pulses.