Hi Alec, in post 128 there is a schematic of the PCB. I am interested in pins 4 and 5 that go to the mic. I am trying to determine which lead is positive and which negative. I have tested (continuity) between both pins of the mic to pin 4, pin 5, gnd, c4 neg side with no indication of which pole is going to ground. the speaker is easy, pin 9 to speaker + you get a tone. why not with the mike???

 

sometimes it is frustrating

 

ok, I screwed up, I pulled the 56k and accidently replaced it with a 5.6 ohm resistor, which led me to 8.75 v out. NOW I have replaced the 5.6ohm with a 5.6K and still have the same 8.75v out. ARRRGHHHHH! anyway to test the transistors on the board?

 

I am interested in pins 4 and 5 that go to the mic

They connect to the mic via capacitors, not directly. The '+' side of the mic should connect to R2 and the 'gnd' side of the mic should connect to R3.

ARRRGHHHHH!

These things happen! Simulation shows the transistors are probably OK, but the TL431 is probably fried .

 

WELL WELL WELL....Hi Alec. I started from scratch, etched some new boards, meticulously checked all components before installing (almost). the regulator in bypass mode puts out a wonderful 4.95v. the pir circuit works triggers everything as it should. the pcb works. BUT. the ldr circuit doesn't. so here's what happened. I asked for 4 LM339 chips at the electronics store, and I got 4 chips. after I built my board I noticed 3 have markings, Motorola lm 339, one (the one I grabbed and didn't check) has no markings. AND, the case has one slight (very slight) difference. (noticeable under a magnifying glass). So besides another lesson learned. that would be double check before you solder. I am wondering, since this chip has no markings and is slight different. is there ANY way to determine if it actually is a comparator chip? any kind of tests I can perform while it is soldered to the board. de-soldering is not actually the most fun thing I have ever done : )

 

Oh dear. You seem to be having more than your fair share of gremlins .

is there ANY way to determine if it actually is a comparator chip?

I'd start by checking LM339 datasheets from several manufacturers (if available) to make sure there are no chips sold with that code but different pinouts.
Assuming the pinout is correct and as per the schematic, and that pin 3 is getting 9V, I'd adjust Trim1 over its whole range and check for high/low switching of pin 2, both with and without light reaching the LDR.
I'd also measure the voltages at pins 2,3,4.
As for desoldering, do you have a solder sucker or desoldering braid/wick?
You might want to consider using a socket for the comparator IC.

 

don't waste your time with this one. the PROBLEM stemmed from a bad solder joint on the variable trim pot. it LOOKED good. however it wasn't sufficient to move current. and another lesson learned! looks can be deceiving! it now works as designed. as simple as this project may seem, I have learned a great deal with your guidance, and it is greatly appreciated. thank you!


yes I have a desolderer and this thing worked the first time till I tried to raise the voltage. I went ahead and desoldered and replaced a known lm339. look at my board. I HAVE MEASURED EVERYTHING MEASURABLE. however, I think I may have screwed up.

please note: r3 is 10k, r8 is 5k6, and r2 is replaced with a trim

 

If your pic is a view of the copper side of the board then it is the mirror image of what it should be.

 

yes, but... it is a direct transfer. it prints right reading for the circuit and transfers to the copper side as if you can see through the board. in other words, you set the copper side to the printed copy. the component side is printed mirror image so when it is transferred to the board it is now right reading. the process uses a laser printer toner transfer. it works well, though there is some gain/spread in the circuit terminals. I will make an adjustment to those as the white dots close and the connection terminals get a bit close. it took a couple dozen (32 to be exact) tries to get the temp, the pressure and the time down but I can do a clean transfer with an iron in a minute. those little white circles are handy when drilling! my first lesson in electronics has been a blast, learning to read a schematic, the functions of the components, learning to design the circuit board, etching, drilling, learning to solder, AND to UNDO solder, trouble shoot and my favorite, plugging it in and getting the correct results! every step has been frustrating, fascinating and yes FUN.

the other day at the electronics store I noticed copper tape (like duct tape) I assume you simply tape it to, maybe heat laminate it to a board??? don't have all the details yet. have you ever experimented with it?

 

trim question. I have a cheap trim pot that adjusts time. it goes around in one turn probably 270 degrees. the problem is there isn't much latitude/control in timing. would something like a 3 turn trim give better results? I am assuming that with a longer sweep the increments in resistance would be smaller. EXAMPLE: 1 turn to go from 0 to 60. at the half way point you are at 30. with a 3 turn it would seem that you would have a smoother transition/more control over the speed? make sense??? it is the timer trim of the PIR. if you move it a hair or even half a hair it goes from 5 seconds to 45 seconds. how many ohm should be used?

 

I've not tried the copper tape. Other members here may have?
A 3-turn trimmer will certainly give better reolution. A 10-turn trimmer would be better yet (and probably little difference in price). You would have to measure the PIR's trimmer to know the resistance value.

 

500k, I got a 15 turn. BIG difference! and soldering is not as easy as it looks. it can "look" GREAT, but not have a good connection. but I did learn how to trace it back! so alls good!

 

it has been determined that the breadboard was grounding somewhere and goofing up the voltage readings. once I used a new board everything works fine. though there was no indication of continuity/ground on the old board. Alec, do you ever build stuff with micro processors? I have not considered them do to the cost. but I would like to make something else. I would like to make a light go on for 10 seconds every quarter hour. I'm pretty sure it can be done with say a 556 timer, but I am not sure how to determine how much power will be required since it would have to run all the time. UNLESS, well....? would the relay coil that I used originally have sucked up lots of power?

crystal oscillator ?????

 

ok, I thought it out.....what do you think? beside the PIR will go the interval timer. kinda between the PIR and comparator circuit. power will go to the 556 interval timer circuit, however, like the PIR it won't send current unless it is dark and it will also be 5v regulated. instead a light it will operate the PCB, once every hour for 10-12 seconds.

now my question is...there may be a time when the PIR is activated on the hour at the same time the 556 is sending current. now it is the same power supply. but it won't be receiving twice as much power???? correct?

in a previous version we made the original interval timer we made it close the relay longer from 3 sec to 6 sec. do you know the equation for determining the time off, then the pulse on?

it doesn't need to be exactly and hour but the 10-12 seconds is critical.

 

A 555 or 556 circuit is ok for time periods up to a few minutes, but is not a good choice for time periods ~1 hour. It would be better to use a clocked counter such as a CD4060, or else a microprocessor. In answer to your post #153 question, no, I don't do much with micros. Other forum members could help you with that. A micro would also cope better than a 555 with the 10-12 sec period if it is critical. As for cost, a suitable micro these days would cost no more than a few discrete components, but you would need a programmer for it (quite a few bucks).

 

well, the 10 -12 seconds is kinda critical, but....if I can get it + - 10% or 20% I think it would be fine. discrete components is my preference. do you have a CD4060 schematic I can review and add to my learning experience? Thanks!

what kind of micro are you talking about? aurdrino? (sp?)

 

ok, I read almost all posts on cd4060. they seem to be great for longer durations however they also seem to be 1 hour ON 1 hour OFF. similar to the 555. might there be a 556 version of the cd4060. where one half would be the hour and the other half the seconds. or some way to get from hour OFF ......12 seconds ON, then RESET?

 

they also seem to be 1 hour ON 1 hour OFF.

Not necessarily. One CD4060 plus a couple of AND gates can be configured to do 12 secs on, ~1hr (= 12 secs x 320) off.

 

let me go learn about AND gates.....

 

The 4060 has multiple outputs, so AND gates can be used to logically combine them and so define a chosen time delay (e.g. the Q14 and Q12 outputs combined give a delay of (256+64) x 12 secs = 64 minutes).

beside the PIR will go the interval timer. kinda between the PIR and comparator circuit

I'm unclear why you'd want to do that. If the timer is only powered up when the PIR is triggered it will presumably stop when the PIR output ceases?