dei 507m problem

Hi,

I Just bought and carefully installed a new DEI 507M in my 1990 GSX Eclipse.  The detector is located label-up and parallel to the floor pan.  The general problem is that it doesn't trigger when I jack a front tire clear of the floor. Both wire loops (white and black) are in the default position.

Test conditions: Red wire with 1A fuse connected to always-on +12V. Orange wire connected to "Ground When Armed" (measures 11.3V disarmed, ~0.08V armed). Blue wire connected via a 2.2K resistor to always-on +12V for purpose of testing.

Test: With the car fully down, I measure the voltage across the blue wire resistor as 0.0V -- no trigger current present.  I initiate the arm sequence and when it completes, the Orange wire goes to ~0.08V and the Blue wire resistor voltage goes to 6V (2.7mA draw). I jackup the passenger side front until the tire clears the floor.  The Orange wire is still at ~0.08V (as it should be), and the voltage across the Blue wire resistor remains 6V.

I see two specific problems:  One is the Blue wire doesn't wait for a trigger condition before it starts pulling on the alarm sense wire.  The other is there is no apparent change in detector state when a wheel is raised off the floor.  FWIW -- the quiescent current draw of the detector (Red wire to armed Orange wire,  Blue wire disconnected) is 1mA.

Am I doing something wrong? Or have I just been lucky enough to purchase a brand new defective 507M?

Thanks -- Trevor

PS.  I just made some measurements of the Blue wire voltage referenced to chassis ground.  When armed with a 2.2K resistor to +12V (as above), the Blue wire measures 6.18V.  When disconnected and disarmed (Orange wire not grounded), the Blue wire measures 10.73V. When disconnected and armed, the Blue wire measures ~0.08V. 

PPS. Guess I should also mention the target trigger sense is the hood-open input on the OE ETAC security controller.  Open hood is detected by pin switch grounding of the sense wire -- Blue/Black at connector C-16-2 pin22 of the ETAC.  My purpose is to mate the 507M blue wire to that sense wire and was expecting a simple open collector direct connection would do the job, but it didn't work, and that's how I came to be making these measurements and asking for input.

your ground when armed should bea constant ground and your meter will show 12v when activated.

if you have an existing alarm, connect the wires to the shock sensor, but diode isolate all wires.

Hi!

The "Ground When Armed" is a constant ground as long as the alarm is armed, otherwise it follows whatever is tied to it.  If you mean a voltage measurement from the red to the orange wire while armed should be the battery voltage, yes, it is.

According to the observed current/voltage behavior of the blue wire, it's open collector and is pulling on the sense wire at all times the system is armed.  Given open collector, inserting a diode isn't necessary, though I realize that's the standard practice.  The red and orange wires should never require diodes.

howie ll wrote:

---

This is interesting because I usually refuse to install them on the basis that they constantly false!

---

Great :(!  I bought it specifically because I was told it actually works well.  Maybe it's another of those sample specific situations.  The Eclipse ETAC ignores it anyway with direct connection to the hood-open sense wire (normally rides at about 0V when armed).  It responds to hard short to ground, but the 507M's Blue wire shows no sign of cutting it.  If I can get the darned thing to display a coherent before/after trigger behavior, I can throw together a relay interface for a genuine ground -- but first things first.

Too bad Directed doesn't provide electrical specs, not even for the trigger wire.  I'd have thought that would be automatically included with electronic apparatus -- couldn't find it online either.

Still waiting for return instructions from the Ebay seller.

I ordered a DEI 506T Glass-Break sensor at the same time as the 507M Tilt sensor (different vendor) and it just arrived today (slow as molasses shipping :().

I reused the "Ground With Arm" on the Black ground wire (no point in having it active all the time), connected the Red wire to always-on +12V, and my trusty 2.2K resistor between the Blue Wire and always-on +12V (just as with the Tilt sensor).  I then set it for max microphone sensitivity and gave it a whirl.  At arm, the resistor voltage briefly jumps to full BAT voltage (understandable power-up transient), and then returns to zero.  I clunked a set of keys against the driver door window (microphone sitting on the dash) and the Blue wire resistor voltage again jumped to BAT, and then back to 0V.

I then tried connecting the Blue wire directly to the ETAC Hood-Open sense wire (no resistor or diode).  I repeated the test and as soon as I clunked the keys against the driver window, the alarm fired :)!  Superb -- too bad the 507M Tilt didn't behave as nicely.  Since I doubt the Tilt trigger output is intended to be significantly different from the 506T, I would expect a "good" Tilt sensor should drop in just as cleanly and without any additional interface "glue".

I'll go ahead and complete the 506T install and cross my fingers for the replacement 507M.  The defective original will be off to the vendor in Thursday's mail.

howie ll wrote:

---

I missed something here why the resistor being used as a strap-up?

---

The combination of a test resistor for current measurement plus voltmeter allow me to see what the sensor trigger output is doing.  Since DEI doesn't specify the behavior and I've already had trouble with one, I prefer to test first instead of simply trusting it to work with my OE security controllers.  If I can get the 507M working properly with my Eclipse ETAC, the next step will be installing one with my 2008 xB controller.  I find it helpful to understand the interface behavior.

Received the replacement 507M today, tested its behavior, and now it's installed.  I raised the front corner far enough to trigger it and the tire was still in contact with the ground (default 1 degree sensitivity) -- looks good.  Guess I'll just have to see how it behaves over time -- hopefully, no false trigger problem.

Thanks for the help folks :)!

howie ll wrote:

---

This is interesting because I usually refuse to install them on the basis that they constantly false!

---

Having just installed the 507M, I'm now trying to verify correct operation -- relative, that is, to your mentioned  false triggers.  Except for when the vehicle is being driven, I'm keeping the alarms armed (including overnight in my garage).  Nothing unexpected has happened yet and so I'm wondering if there's any pointers or insights you could share with me.  If there's a problem, I'd prefer to identify it sonner than later.

Thanks -- Trevor

Wow, install it and it fires just sitting there -- ouch!  Thanks for the run down, it's interesting, all the things that can go wrong :(.  Just a thought, there is one thing probably a little unusual about my install (longitudinal) -- the 1990 Eclipse ETAC Security Controller doesn't provide a "Ground With Arm" (GWA), so I created one using an open collector transistor to yank on the Orange wire when the system is armed.

That's something I'm wondering about.  The install instructions say to use a "True" ignition line if no GWA is available, however, of course, it needs to go to ground when the Ignition is OFF.  I just started modding my 2008 Scion xB OE Security and like the Eclipse, I see no sign of a deliberate ground on any of the ignition wires -- they appear to be either HOT or not.  It strikes me as a question mark whether an OFF Ignition Switch output satisfactorily grounds the DEI 507M Orange wire.

For the Eclipse, I'd previously synthesized an Armed Status signal (not provided by the ETAC), and so it was fairly straightforward to generate a GWA.  With the xB, I'm back to square one (no identified Armed Status)., though I could detect the OE Security flasher (Eclipse didn't have one -- does now :)) and synthesize an approximate Arm Status and hence OC GWA.

How do people in the business handle the Orange wire connection?  I'm beginning to wonder if it might be a crap shoot.  The quality of that ground during Arm might be a factor in 507M performance (just a thought of course -- I'm new to alarm systems).

Thanks Howie :)!

With the Eclipse, I tried to use whatever the OE Security Controller (ETAC) made available to get a clean "factory-like" result.  It took a lot of head scratching and a couple false starts to develop a clean interface between it and my aftermarket Keyless Entry Controller.  After which, further adjustment was needed to support aftermarket sensors.  From what I've read, people generally just replace the ETAC with an aftermarket model :(!

With the Scion xB2, I'm again trying to leverage off the OE Security Controller.  Keyless Entry is already provided, so it's a much simpler problem than the Eclipse, but I still want to achieve as clean as practical an interface with the aftermarket sensors.  However, as you pointed out, one can't afford to do R&D with vehicles when your livelihood depends on turning the jobs around.  Fortunately for me, these are my personal vehicles and I can pretty much take whatever time I need to get the result I'm looking for.  Still, once again, from reading, people generally install an aftermarket Security Controller in the xB2 :(!

Guess I just hate the idea of replacing a perfectly good OE Security Controller, even if it is somewhat limited :)!

jcs091570 wrote:

---

Hi I have the 507t and never a problem. I believe after setting your alarm system, the 507t takes 2 minutes before becoming active. Do in part to it needing to adjust to any inclination. Cutting the black and white wires make it more sensitive obviously, but could result in false alarms. I would scrap the 506t and go with a prox sensor. Just my thoughts.

---

FWIW -- The 506T is working great!  I've also installed a DEI 504D Shock Sensor which also seems good, just needs a spot of "glue" for the horn warning (working on it :)) .  I considered the Proximity Sensor, but was recommended against it due to apparent issues.  Since parking lots invite people to get especially near adjacent cars, I can quite believe they could be a hassle.  Between the Shock and Glass-Break, I think the interior is covered.  The Tilt covers jacking and towing, so I'm thinking it's sufficient.

PS. The 507M defaults to quick arm and one degree sensitivity -- no falses yet.  I'll be watching these guys' for misbehavior, but initial impressions are good.

It sounds like you also have a specific security controller installed.  I'm using my OEM controllers and so the Aux etc functions aren't available -- it's just a hardwire.

That's a good question on possible issues associated with the 506T and the windows down.  I normally leave them in a top-open vent mode rather than fully down.  Still, it's something I need to keep an eye on.  This is my first experience with electronic security sensors, so I'm in learn mode.  That also goes for the "glue" I apply -- sometimes it doesn't work entirely as anticipated :).

My thinking regarding 504D attachment was to go for maximum impulse delivery (max suspended magnet movement at time of "shock").  My expectation is the softer the mounting, the less energy delivery, and so reduced likelihood of detection.  It's too bad DEI doesn't provide specific installation tips since ultimately, the nature of the sensing determines the optimal mount.  Eg -- is a taller primary impulse better or worse than a series of distinctly smaller vibration impulses?  All I know is I can fairly lightly rap my bumper and have the warning horn blow -- my guess is that's OK.

The Proximity Sensor seems perfect for a show situation (ie. "keep away"), but I'm not so sure for around town.  Though as you say, with the right controller functions it could be more flexible.

PS.  The 507M for the 2008 xB arrived today, so I've got to get the "glue" box built (GWA) so I can install and test it. Hopefully, it will be as well behaved as the one now installed in the Eclipse :)!

trevors wrote:

---

My thinking regarding 504D attachment was to go for maximum impulse delivery (max suspended magnet movement at time of "shock").  My expectation is the softer the mounting, the less energy delivery, and so reduced likelihood of detection.  It's too bad DEI doesn't provide specific installation tips since ultimately, the nature of the sensing determines the optimal mount.  Eg -- is a taller primary impulse better or worse than a series of distinctly smaller vibration impulses?  All I know is I can fairly lightly rap my bumper and have the warning horn blow -- my guess is that's OK.

---

Just a follow up on the 504D install:

Thanks "jcs091570" -- between your and others inputs, I changed my mount to velcro and am glad I did.  Not only is it much more sensitive, but more importantly to me, the response is more uniform around the vehicle.  Thanks also for your other comments -- something to think about :)!