Another P.I. link

hello xavier ,

ok , here goes online diagnosis , mechanic by trade and regularly get phone calls to diagnose car on phone ,

detector will bleep whenever the sample hold voltage in the capacitor drops below preset threshold , would check there is some voltage in the hold capacitor if there isn't fault is forward of this point.

and you say removing the last 741 stops the bleep , so we are ok to that point , well asuming last 741 output not shorted to +,

ok so 555 pulse ok ? pulse from 4538 ok ?

could be just delay set to soon , initialy set pots for delay and width to their maximum resistance , lower threshold ,bleep should stop , if it doesn't next steps,

using multimeter :
1st check threshold output varies , expected voltage based on your 12V supply is aprox 3V to 8v when knobs turned , if it's close to these values then this is ok,

2nd check output pin 6 U5A , if output is 0V or 12V then U5A shorted ,expected 2V to 10V depending where you have delay set ,wave metal infront of coil , and check voltage moves , expected variation is 0.01V or poss 0.1V ,

3rd check + input voltage of U5A also should be between 2V to 10V or there abouts , check for same variation ,

this is tough without a scope , i think i remember you didn't have a scope ??

i would check output of the TL071 for frequency same as the 555 , and set meter for dc voltage and check output is not either 0V or 12v indicating part faulty,

from looking at you pcb it all looks right , so i'm thinking faulty part , poss substitution of chips to diagnose , start with the last 741 just in case , then work forward ,

had a thought , take a good pic of both sides of the board as it is now and post , may help pin it down.



Linkback:

You are not allowed to view links.
Please Register or Login

http://www.thunting.com/smf/index.php/topic,3797.msg194128.html#msg194128

hello xavier ,

got it , youre threshold min voltage is higher than your U5A output ,

on the threshold circuit , short out or remove and link the resistor to earth , think it was a 10k , this will lower your threshold voltage to below signal voltage , and you should be up and running ,

ready to then play with your delay and with settings to get stability and distance perfected.

good luck.

Posted on: September 08, 2011, 05:42:43 PM

---

dont think coil is critical ,

the lower the uH the faster the coil , and the less delay you can use , and better responce to high conductive metals , but at the expense of signal deviation with target / sensitivity , also ability to drive more power into ground.

high uH coils are slower , may miss high conductivity targets , longer delay required before sample , but larger deviation with same target  , so more sensitive , but lower power driven into ground ,

it's all a compromise , common belief is 400uH or there abouts , but some designs are using 270uH , my coil is over 500uH

your coil is in the zone i was aiming for but missed , well done there.

looked at pics , only thing to say is they are some nice tidy tracks , if i venture to make my own i would like them like that my dear fellow.until then  verio board is my friend.

Posted on: September 08, 2011, 05:54:37 PM

---

forgot to mention , threshold voltage on this design has to have the ability to be lower than signal voltage for silence

Linkback:

You are not allowed to view links.
Please Register or Login

http://www.thunting.com/smf/index.php/topic,3797.msg194175.html#msg194175

Well now you have me confused 

Linkback:

You are not allowed to view links.
Please Register or Login

http://www.thunting.com/smf/index.php/topic,3797.msg194222.html#msg194222

Xavier

555 transistor is an npn , don't know how it ended up being a pnp ,

your 1st posted pic is 100% spot on.
your second posted pic is incorrect.


Hi Dooley
great thanks for clearing that up, all sorted
I will add the 100k trim pot in thanks

NOW
on to the important things
how did you go with testing your new ideas ?
hope you enjoyed your day off, it rained and hailed
here today .
and Frankensquirrel 2 with pin mods in orange.

Linkback:

You are not allowed to view links.
Please Register or Login

http://www.thunting.com/smf/index.php/topic,3797.msg194233.html#msg194233

hello 6666 ,

results time , after a day playing , aux add on board made and fitted , as per my description and just as you posted ,

using the offset pot on the TL071 set curve so it's flat on top , then sample 1 at the begining just after the bend and on the flat bit ,

sample 2 , you will have to play with depending on your coil , best way to describe the best place for sample 2 is
aproch coil with a piece of alu and it's where the lowest point of the dip is .

sample widths i'm using is 20uS

right then , tweek the offset pot to get the led's right on the change , then nudge it a little so green (high conductivity)
led is on and the red (low conductivity) is off ,job done , thats it.

so red led = alu , iron , steel
green led = gold , silver , and anything with nickle in it ( brass , silver coins , etc...)

sorry about copper , its the same responce as alu , so its a red led , so all old coins made off copper will show red.

sensitivity of the led's is not as good as what the audio responce is , but it isnt too bad ,

using a 20mm silver coin (nickle copper) audio goes from slow click to fast click / low note at about 40cm
but reaction of led's is at about 30cm
so if a target is a bleep she will show a definate ID of type of metal ,
but if its a clicking target , led's remain unchanged.

thinking about it i'm relying on the 741 to detect changes of only a few mV
if i were to do it again , i think i would add an extra op-amp on each sample , with a little gain , say 2X or 4X ,
so that the change the led op-amp sees is more definate,

all in all good result , 100% working , and it does what it was ment to , so quite happy , for the time being,
dark now , i'll get her out in the field on sunday , and only dig the green responces. 

xavier , did you get your threshold voltage to go below sample and hold voltage ??
you said you can get her to go quiet , so you must have , but now no responce ,
you have frequency at the output of the TL071 , so thats prob working ,
frequency from the 4538 to the 4066 so thats ok
and you have a dc voltage at pin 6 on the first 741 in the correct range so that sounds right ,

here is a thought , output from the first 741 , does the voltage go up and down as you alter the delay , it should , as you will be sampling higher or lower on the decay slope ??

and sample width , poss to small depending which end you have set the width pot to , try the width pot in the middle,
and the delay in the middle,
set threshold for hopefully slow click and check for responce ,
then hopefully , it will just be a case of tweeking the pots for the best responce ,

sorry to hear you are having problems .

Posted on: September 09, 2011, 02:19:48 PM

---

UPDATE !! UPDATE !!

just had her out in the garden in the dark , the audio sample point is now to soon , she is unstable and reacts to the ground , doesnt maintain slow click , but she was spot on when audio sample was later , in the "alu dip" as i mentioned before 6666 ,

so i know what i have to do ,

swop round the samples , the second sample does the audio and samples in the dip during the alu test ,
and the first sample only goes to the op-amp for the led's

so , simply going to swop the pulse wires round on the 4066 ,
then swop round the + input and - input wires to the led op-amp ,
re-tweek the delays / offset and poss width , and i will have a stable , easy to use machine again with working
red green discrim.

i'm happy now , i see the light , sunday morning tinkering , then off out to find treasure , whooooh......

Linkback:

You are not allowed to view links.
Please Register or Login

http://www.thunting.com/smf/index.php/topic,3797.msg194270.html#msg194270