Cheap Motion sensors , chinese I think

[(*steve*)]

Nope, still getting the same message.

I'm approximately half way between Beijing and the Ronne ice shelf.

I find that search engines differ greatly in their utility. In my experience Google is more consistently reliable in finding datasheets (and I look for them a LOT)

Now you have the English version of the datasheet which gives you a sample circuit and the formula for calculating the delays, have you identified the resistor(s) and capacitor involved? Can you calculate what the delay should be based on their values? Does it match what you're seeing?

If the capacitor is large in value, and especially if the resistor is also large, the delay may be significantly larger due to leakage and input currents. The times you are reporting are pretty long for a simple RC delay.

Regardless of all that, if you have correctly identified the chip then it should be a simple case of locating and replacing the correct resistor and/or capacitor.

 

[davenn]

I don't use Google .

well maybe you should ... it would solve a lot of your problems

 

[(*steve*)]

R14 and C15 look to be the ones.

Looks like you were on the right track. Can you measure either/both of the original values?

 

[AprilSteel]

Appologies steve. I was in a shitty mood yesterday. When I came back to delete my post I found I could not so please overlook it . You were a bit sharp .
Further progress on this .
Pin 6 of the IC controls time on it seems and they put it as Ti =approx 24xR9xC7
The device is marked differently from the datasheet example and for any coming after me it seems pin 9 is connected to R13 and C17. I had the wrong ones .

Also my device does not have a LDR contol built in but it seems I can add one

 

[AprilSteel]

well maybe you should ... it would solve a lot of your problems

Nah Google wants to know every time I go to the loo. Stuff google and microsoft . I'm a linux user and use start page . If that doesn't do it Ill ask someone smart , like yourself.

 

[AprilSteel]

This losing editability of posts is a pain.

Got it wrong again above and I would like to change that post but you won't let me . Makes for confusing reading later.

OK so its Tx I'm interested in of course . Which is time the output stays high.
Tx=approx 24576 x R10 x C6

R10 is connected to pin 3 and the cap on 4 . On my device thats R14 and C15 so I'll halve R14 and see what I get . R13 was marked 684 and measured about 660K
R14 is marked 303 and measures 29.8K
Can't measure the caps yet as my meter is not working on caps.

The variable Pot measures 220K
These readings are just for me as it saves recording them elsewhere . I can always come back here and get them.

 

[(*steve*)]

R13 was marked 684 and measured about 660K
R14 is marked 303 and measures 298K

The first digits of the marking are the first digits of the resistance. The last digit is the number of zero's following.

684 = 680 000. Your measurement of 660k is well within the bounds expected.

303 is 30 000. Given your measurement, are you sure it's not marked 304, of maybe 3003? Or perhaps the measurement was 29.8K?

 

[AprilSteel]

Yep corrected above 30K. Explains why I got nowhere with 150K across the resistor to halve its value . Tomorrow . So where in the world is the Ronne Ice shelf and I'm guessing you are in Melbourne or so but I don't understand "phear the antipodesi" are you greek?

 

[KTW]

The trim pot should be a 204, 200k.

 

[(*steve*)]

The Ronne ice shelf is at the head of the Weddell sea. I am neither in Melbourne or Greek, however the phrase in question is rather appropriate for large swathes of the USA. My location is specified in an unusual way as a bit of a puzzle. Several people have solved it, but you can't use maps.

 

[AprilSteel]

You have all the right pieces in all the right places, a component is failing or isn't making a connection.
You shouldn't have to substitute anything but more likely test and repair or replace something.

No I hav'nt actually . Turns out the resistor in question is a 30K . I misread it as 300K so I substituted a 150K and wondered why it didn't turn off at all. Today I will attack it again and put a 15K in place and see . If I can get the timer adjusted down I might then try playing with adding an LDR per the datasheet . Its nice to have a datasheet in English . Thanks *steve*

 

[AprilSteel]

OK now I experimented a few times with old ones without success so I opened a new one , tested that it turned off properly after whatever time . This one turned out to be 2min 30 secs min setting.

Then I bridged the resistor with a 15K and obtained a final value of about 10K .Tested that resistance back to pin 3 OK . Turned it on and it would not turn off. Took the bridge off and tried again and it would not turn off .

So there is something going on here to blow the chip or at least a more sophisticated approach is needed to understand its operation . I discounted static as I have not had any problems with static up to now on my present bench setup . However I have pretty much everything on the bench all at once so anything is possible . Ill put up a shot of it , itll probably give you a laugh anyway.

The caps on Ti and Tx are 10n on the example and I'm guessing they are the same on this board. Got me beat what the RT stands for and its purpose?

 

[AprilSteel]

Anybody still interested in this . I still hav'nt found an answer?

 

[davenn]

I have deleted a bunch of unhelpful/rubbish posts from KTW and done other tidying up as a result

 

[(*steve*)]

I've done my best based on the datasheet and the not-too-good photographs.

The possibilities that still exist:

1. The datasheet is wrong
2. The chip is not the one described by the datasheet.
3. I made a mistake reading the board.
4. I made a mistake reading the datasheet.
5. There is something odd that you're doing.
6. Other.

Edit: inserted new option 4

 

[KTW]

You appear to be out of options so try a 2.2k resistor parallel to R14 to change the timing.







[mod note ... bad comment deleted]

 

[AprilSteel]

Well thats what I have tried but using 15K across it and it did not work . That drops the resistance seen to about 10K as per the data sheet so 2.2K seems very low . I don't think I will try that.
30Kparalleled with 2.2K is what A few hundred ohms? Thanks Mod
Its easiest for me to remove the surface mount altogether and just fold a resistor to suit . So what value do you suggest I use ?

 

[duke37]

1/30k +1/2k2 = 1/2k05 so more than a few hundred.

 

[AprilSteel]

1/30k +1/2k2 = 1/2k05 so more than a few hundred.

My maths are pretty hopeless but I joined a 33K with a 2K and measured as 2.5K so it seems a bit low but I'll give it a run thanks

 

[KTW]

If you had taken my suggestion, this is what you would have.

This is a 50k trim pot adjusted to 2.2k and bridged to R14.
20k bridged to R14 will buy you 35 seconds.