Driveway sensor, model 917-1, mfg. around 1993 by "Detector Systems"

A few thoughts on these sensors. Generally, the loop is non critical and the unit automatically compensates for variation. The loop is somewhat standardized since one does not dig up the road to change the loop if the detector unit needs replacing. I seem to recall that the loop is typically about 4' x 8' and about 2 or three turns. One could easily tape a test coil to the driveway and try it out. Usually, the coil is buried about 3-4" into the asphalt. Try looking up the oldest sensor info you can find [Eagle signals?] and see what info it has on the coil. Usually, the connections on the quality sensors is by a single 'MS' type circular connector with 8 connections. You should be able to open the case and trace the pinout. Connections should be: AC input [line, neutral, ground], Output [NC, COM, NO] and LOOP. You could also check with your local roads department's electrical shop. Good luck!

Neil S.

Hi Neil I made a coil in the configuration you described out of 14AWG. wire. The coil is 4 turns of wire with a 4 X 8 rectangular shape. The perimeter is 96" approximately. I just realized that this is a larger perimeter than you suggested so I wonder if that might cause a stability problem such as I'm having? The funny thing is with this coil the control unit will sometimes detect the vehicle every time on the approach, (within 2 feet of the car even before the car passes over it. At other times there is no detection at all. I tried the different modes, pulse, 2 second delay, presence, and messed with the sensitivity. So I'm wondering if I need more or less turns or maybe a different loop size. I could be wrong because this was a long time ago but the driveway may have been about 8 feet wide. The cuts we made for the coil may have been much narrower than this coil. Those cuts could have been 8 feet by say 4 or 5 inches or so. Do you think that could have been possible.? Would configuration and size cause a reliability problem such as I'm having? This is so frustrating. I know that the system worked well at the time. I just can't remember exactly what we did. Thanks for any further input. Lenny

a y e r e

Lenny, I am a bit puzzled by your coil description. You say it is 4 x

8 and 4 turns, but has a perimiter of 96". Since 96" is only 8', I would think that the coil you made would have a perimiter of 288". Try and reduce the coil to 3 turns, but don't cut the wire off. The leads from the unit to the coil should be twisted together or you can get some funny results due to stray pick-up. Normally, the unit should not pick up the car until it is on top of the coil--it sounds like you have too much sensitivity--which is why I suggest reducing the coil turns. I have also never seen more than 2 or 3 turn coils, but that does not mean much. I cannot say I have seen any really narrow coils like you describe, but some on driveways are like 4' x 3' and the newer ones on the highway are about a 4' circle [neat machine that cuts the circular slot in the road surface]. Size and configuration are THE important thing in these coils, so it needs to be right for that unit. It is really not much different that getting a radio back loop antenna to track the tuning dial--a fair bit of cut-and- try.

Neil

d I e

o a

I

any

we

t

ge

sor

are

w e n e

Hi Neil You were correct. The coil is comprised of two 96" sections and two

48" sections for an approximate total of 288". You mentioned that the coil I constructed might make the system too sensitive. If that were the case then why does it trip the unit sometimes and just as the car approaches and then not at all at other times? I would think that too sensitive would cause false tripping to occur at random times as well. but I'm not really seeing that. but for certain, I'm no expert with this type of equipment. I'm going to take one turn off the existing coil however I didn't quite understand what you meant by "not cutting the wires". If I reduce the number of turns to 3 and not cut the wire I will have one lead longer than the other. That in itself I would think would cause an imbalance. Should I then just add the difference onto the shorter lead, twist them together and then connect to the short pair of twisted wires (loop), coming out of the unit?. Also one other thing concerns me. right now we're just testing this system with a coil laying on the ground and the controller connected to it right there. If I do get this thing working reliably, ultimately the coil will be buried about 150 -200 feet from the house. I planned to run two conductor sheilded cable, burried and grounded on the unit end out to the coil. Alot of work. What is the chance that the addition of this cable may alter the impedance of the system to the point of putting me right back to where I am now? Lenny

ood

d I

s

ed

he

the

=A0to a

. I

many

at we

but

ange

ensor

t are

how

l

the

e

ld

e

pen

e,

the

y -

Neil Just an update. I removed one turn from the coil and added the difference onto the shorter wire. I twisted the wires together. so the lead in so to speak is now about 290 or so inches of twisted pair 14 AWG. There are four dip switches in the control box. One is marked "presence" and "pulse". Presence requires that the vehicle occupies the space for two seconds in order to trip the relay. For fast food restaurants and such I seem to remember that is the correct setting but for simple detection like I need I think we want the "pulse" setting. There is another switch marked "two second delay" and "off". I think that switch is related to presence and pulse and so I set that one "off". There are two other switches which I think are related to sensitivity however their markings are very ambiguous. There seems to be a high, med, and low combination of these printed on the circuit board but it makes no sense at all. I'm not sure what those settings are I did note though that when both these switches are in the off position. the vehicle is detected when the front end is about two feet from the coil. Sliding one of these switches to the right makes it so that the vehicle has to have one foot of its front end over the coil.. So it would appear that this combination makes the system less sensitive. Another interesting observation was that moving the coil around a bit with my foot would sometimes set off the relay. Maybe thats not a valid test though. Lenny.

Try 96'. (4'+8'+4'+8') = 24' per turn, four 24' turns = 96'.

--
Politicians should only get paid if the budget is balanced, and there is
enough left over to pay them.

food

and I

was

ated

the

m.

t the

ld

=A0to a

ay. I

he

ow many

that we

d but

trange

sensor

hat are

e how

n

ny

cal

e the

the

ould

the

?]

S'

open

ine,

l .

er

on

h the

a

may

t y t

xt -

So, Lenny, is the unit working more reliably with the fewer turn coil? or with the sensitivity set lower? I have a feeler out to the signals engineer for the City of Richmond to see if he has any data on coils and especially for your make/model. Give that a couple of days and I might have something for you. I don't know how long a pair you can run to the coil, but since the detectors are installed in the control cubical it is reasonable that the run to the furthest pick up coil at a large intersection could be as much as 150-200 feet. Disturbing the coil mechanically [kicking it with your foot] could quite concievably affect it enough to trigger the detector.

Neil

food

and I

was

ated

the

m.

t the

ld

=A0to a

ay. I

he

ow many

that we

d but

trange

sensor

hat are

e how

n

ny

cal

e the

the

ould

the

?]

S'

open

ine,

l .

er

on

h the

a

may

t y t

xt -

Lenny, here is the instruction manual for that detector. It lists the coil configuration data as well as sensitivity settings and so on. Let me know how it works!

Neil

st food

d and I

d was

olated

in the

tem.

e

cut the

ould

d =A0to a

elay. I

the

how many

k that we

ped but

strange

ay sensor

that are

ure how

can

any

tical

is

nge the

t the

could

, the

ls?]

'MS'

to open

[line,

cal

re.

rger

s

e on

s

ith the

be a

y may

ve

eet

em

m

d.

x

Try

ds

not

r d y

text -

ssories/Loop%20In...

How did you locate the manual?

Third item shown when I Googled "Detector Systems model 917-1".

Neil S.

Neil Last night I connected about 100 feet of 2 conductor 16AWG. sheilded cable to the test loop and ran it into the house. The sheild is grounded on the controller end. The modified coil with three turns and a 4 ft by 8 ft. perimeter now and a total length of approximately 72 feet is laying in the driveway.With the dip switch sensitivity combination I found the controller seems to detect a vehicle on the approach every time now. We're in New Hampshire so the real test will probably come when the coil is buried and the ground is frozen with ice and snow on top. Thank you very much for finding and sending me that manual. The interesting thing is that although it is indeed for a "917-1", it is marked 3M. I guess that 3M must have bought out detector Systems and simply put their own name on the unit. It does appear to be the same unit as mine though. In comparing the manual 's coil specifications to mine the manual calls for a coil consisting of an 18" x 54" loop with 5 or 6 turns of

16AWG. cable. Therefore the total amount of wire in that coil will be [(18+54) X 2] x 6 [(18+54) x 2] =3D 144 Then 144 X 6 =3D 864 854/ 12 =3D 72ft. So it seems that although their coil uses more turns of wire, the total wire length of theirs and my present test coil is the same. In your opinion, what is the significance of more turns as opposed to larger perimeter? I'm concerned of course with system reliability and naturally repeatability. If this thing fails to detect even one time it will be useless to us. And I certainly won't be digging it up in January. Would I be better off with building and burying their smaller coil with more turns or the one I'm trying out now? Thanks, Lenny.

Neil Last night I connected about 100 feet of 2 conductor 16AWG. sheilded cable to the test loop and ran it into the house. The sheild is grounded on the controller end. The modified coil with three turns and a 4 ft by 8 ft. perimeter now and a total length of approximately 72 feet is laying in the driveway.With the dip switch sensitivity combination I found the controller seems to detect a vehicle on the approach every time now. We're in New Hampshire so the real test will probably come when the coil is buried and the ground is frozen with ice and snow on top. Thank you very much for finding and sending me that manual. The interesting thing is that although it is indeed for a "917-1", it is marked 3M. I guess that 3M must have bought out detector Systems and simply put their own name on the unit. It does appear to be the same unit as mine though. In comparing the manual 's coil specifications to mine the manual calls for a coil consisting of an 18" x 54" loop with 5 or 6 turns of

16AWG. cable. Therefore the total amount of wire in that coil will be [(18+54) X 2] x 6 [(18+54) x 2] =3D 144 Then 144 X 6 =3D 864 854/ 12 =3D 72ft. So it seems that although their coil uses more turns of wire, the total wire length of theirs and my present test coil is the same. In your opinion, what is the significance of more turns as opposed to larger perimeter? I'm concerned of course with system reliability and naturally repeatability. If this thing fails to detect even one time it will be useless to us. And I certainly won't be digging it up in January. Would I be better off with building and burying their smaller coil with more turns or the one I'm trying out now? Thanks, Lenny.

Having trouble posting. I don't know if this is going through. Neil Last night I connected about 100 feet of 2 conductor 16AWG. sheilded cable to the test loop and ran it into the house. The sheild is grounded on the controller end. The modified coil with three turns and a 4 ft by 8 ft. perimeter now and a total length of approximately 72 feet is laying in the driveway.With the dip switch sensitivity combination I found the controller seems to detect a vehicle on the approach every time now. We're in New Hampshire so the real test will probably come when the coil is buried and the ground is frozen with ice and snow on top. Thank you very much for finding and sending me that manual. The interesting thing is that although it is indeed for a "917-1", it is marked 3M. I guess that 3M must have bought out detector Systems and simply put their own name on the unit. It does appear to be the same unit as mine though. In comparing the manual 's coil specifications to mine the manual calls for a coil consisting of an 18" x 54" loop with 5 or 6 turns of

16AWG. cable. Therefore the total amount of wire in that coil will be [(18+54) X 2] x 6 [(18+54) x 2] =3D 144 Then 144 X 6 =3D 864 854/ 12 =3D 72ft. So it seems that although their coil uses more turns of wire, the total wire length of theirs and my present test coil is the same. In your opinion, what is the significance of more turns as opposed to larger perimeter? I'm concerned of course with system reliability and naturally repeatability. If this thing fails to detect even one time it will be useless to us. And I certainly won't be digging it up in January. Would I be better off with building and burying their smaller coil with more turns or the one I'm trying out now? Thanks, Lenny.

Yes, Lenny, all three posting came through okay. I would certainly reconfigure your coil to match the one specified in the manual. Since the wire length is similar, it would not be a big issue to to change it now and try that out. The manual does specify a nominal loop inductance of 30 to 1000uH, so there is a lot of margin designed in. If you have an inductance bridge, you could measure the inductance of your existing coil and see where it falls in that range. As far as operating in the cold/wet/ice, one will have to see, but since it is self tuning it should be fine. You say it must operate EVERY time, but that is expecting too much of it. I would expect better than 99.9% sensing of 'normal' cars but motorcycles and the like can be 'IFFY'. You will also have to expect the occasional false trip due to abrupt weather changes [sudden rain storm for example].

Neil

nesesu wrote:

Heh. Google Groups' quasi-annual takedown for maintenance. Roughly 10AM - 3PM today.

Impatient posters using a Web interface to Usenet should poll ANOTHER one of those *before* posting *again*. (Google is NOT the only game in town.) Here are some with reasonable latency:

Neil When I said "every time" I wasn't thinking of motorcycles. My wife's business which is run out of part of our home is a gift shop catering primarily to women. The shop is off the main road. So considering the clientele we're really not expecting too many bikes in here, and if the occaisional one does come through and does not get detected then its not the end of the world. No, I was referring to the mass of a typical automobilie. The issue is that there are sometimes long periods of time when there is no business. she doesn't want to have to stay down there waiting for a customer to show up. This sensor will allow her to turn out the lights and be in some other part of the house until the alarm sounds. She can then run downstairs, turn on the lights and be "open for business" And so far with the loop laying on the ground its worked every time. You were right on the money with your idea of three turns and a 4 X 8 footprint which is the 72 feet the book talks about. I don't have an impedance bridge.so I can't directly measure inductance, however the instructions did mention an operating frequency of between 30 and 60 KHZ. I suppose that I could look at this with the scope, determine exactly what it is and then figure out a way to determine inductance if I really want to. But for the time being I'm going to rewind the coil for the 6 turns the instructions call for. So now I need to ask for your advice again. Last year I rewired my well. I pulled up 175 feet of the old wire and replaced it. The conductors in that wire were continuous however the insulation was scuffed and scraped and in some cases down to bare copper and no longer servicable for well application. However I kept it around to use for projects, such as this one..Well wire if you've ever worked with it is typically three twisted conductors made from 14AWG. This wire is apparently impervious to moisture because it runs right down the well casing throught the water to the submersible pump. My present test coil is made from some of that old wire.unwound to single srand. The scuffed parts were taped up in places for purposes of this test but it is not in good enough condition to put directly into the ground. I have a gravel driveway and ultimately I will need to bury the final coil amidst this abrasive material.. I could get some more of this well wire or the instructions say that you can directly bury THHN or a similar type of wire. My concern though is the dirt, stones, sand, etc. which would be in constant cantact and abrading this wire. I think that I would like to encase it somehow. ideally in PVC. Then have some kind of an underground junction box to connect my sheilded cable to. Do you have any ideas as to how to do this? Lenny

Lenny, I think that the 'well pump wire' is stranded type TWU . In any case it generally has a much thicker insulation intended exactly to withstand high abrasion situations such as sliding in and out of well casings or being directly buried in the ground. It would not survive being shallow buried in a gravel drive for very long. If you can come up with the needed length of normal TW or XLPE or the newer Nylon coated wire, then I would suggest you buy some 1/2" PVC conduit, couplings, glue, 4 90 degree sweep bends and an outlet box with two

1/2" hubs on one end and one hub on the other end as well as a blank cover for the box. You then form the PVC pipe into the rectangular coil form and pull in the wire around the loop the required 5 or 6 times and splice it to the shielded lead in in the box. Please email me directly with your email address and I will send you a sketch of how to form the pipe. Naturally the pipe should be surrounded with coarse sand to pad it from the stones, but since you should be staying within 3" - 4" of the surface, you don't have a lot of margin for that.

Neil

Neil The well wire is solid 14 AWG. I looked at it again and it is marked

600V type "THW". and I definitely agree that it probably would not survive the gravel driveway for very long. I was thinking about a similar idea such as yours but it seems like it would be very hard to pull, or for want of a better term, "sew" 6 wires into a loop one at a time through all that pipe and those sweeps. Its virtually impossible to rtry to add wire to a piece of pipe that already has wire in it.The only way I would think that you could do it would be to pull 6 individual labeled numbered wires through the pipe, leave all the fittings loose until all the wires are pulled through. At this point you would glue the pipe together. Then you would have to splice the wires together, 1 - 2, 2 - 3, etc. in the junction box to form a coil with 6 turns. Then I'll have a continuous loop.. The only problem I could envision is that it might be somewhat difficult to get the fittings together with the wire in there. I don't know I guess its worth a shot. I've seen PVC boxes with seals for this type of underground work at some of the electrical supply houses. I'm going to a hamfest tomorrow. Maybe I'll run into something there. You never know. I really appreciate all the time you're taking on this with me. BTW my email is: snipped-for-privacy@yahoo.com Best regards, Lenny