TD3X ( pi )

[DOOLEY]

thank you Thomas,

I had never seen that before.

audio websites have any previously discussed adding lots of capacitance
to whichever side is your output side,

not sure how to implement the solution you have shown in a dc coupled stage,
but worthy of consideration.

[6666]

earth field cancelling is something I saw on a post way back from sacho ( Bulgaria )
elegant solution I thought.

Hi Dooley do you still have the text of that post ? as I would be interested to read what is said in it , thanks.

[Thomas]

thank you Thomas,

I had never seen that before.

audio websites have any previously discussed adding lots of capacitance
to whichever side is your output side,

not sure how to implement the solution you have shown in a dc coupled stage,
but worthy of consideration.

Maybe some thing like this? when you find the values, hard wire it..

[greylourie]

Maxim has a page describing attributes and behaviour of switches (like 4066). Many interesting points, including "charge injection" and "make before break".

http://www.maximintegrated.com/en/ap...ex.mvp/id/5299

[DOOLEY]

hello 6666,

didn't have any text of explanation regarding the earth field cancel arrangement
that appeared on a diagram that was posted on here with "Sascho Bulgaria" written on it,
it was just self apparent what it was there for,
and what it did,
so used it.

section below........



hello Thomas,
been pondering how to use your suggestions.
done a rough mock up below......



only poss way I could think of to suppress the 4066 whilst still leaving
the end on end storage and hold capacitors alone when the switch is open.
although, this may not work.........

not sure theres an easy solution to modify my original arrangement to stop it
happening and still allow it to operate as intended.

[DOOLEY]

cheers greylourie,

checked out that link,
and had a quick browse through it.

will read in closer detail later.

hum,
so rather than mess about trying to cure an inherent "fault" with a 30 y/o
chip,
poss worth considering a replacement,
and by the sound of it pin for pin compatible replacements are available.

[DOOLEY]

had a request for an explanation as to how the earth field cancel works
via private message,
but cant post pictures on private messages,
so replying here......

hello 6666,

its quite simple really,

both samples get amplified x2 ,
then the earth field signal is taken off the 1st sample,
resulting in the difference x2.

quick example below........



so,
the 100mv 1st sample gets amplified x2,
so would be 200mv, if the lower 47k was connected to earth.

the 2nd sample is 80mv,
gets amplified by x2 resulting in 160mv.

then, this 160mv is taken from the 200mv,
resulting in 40mv difference ,
which removes the earth field and amplifies the target
signal by x2.

I have shown the op-amps in this example running on + and - supplies,
so my 1 volt reference voltage is not needed,
it was only needed in my design due to running the op-amps
on +8v and 0v.

the TL062's are only capable of sensing to down to about 0.62v without inversion.
(output drive also about the same, 0.65v from each rail.)
hence, on my design with 0v and +8v supplies I had to use a 1 volt reference voltage
on the 2nd sample op-amp to keep it above this region.

if + and - opamp supplies are used then this reference is not required and the 2nd
sample op-amp can be referenced to 0v.

point to mention,
the TL072 can only sense down to about 1.66v before inversion,
(drive is about 1.69v from each rail)
just thought id mention this before someone tries to use a TL072 here..........
(and some people thought all the TL 062,072,082 were all about the beast..........)

next point to mention with this system,
the input voltages need to be less than 1/2 of the supply voltage,
or the 2nd sample opamp will just be railed out all the time.

hence why I tweaked my preamp offset down to settle at about 2-3v.

if this system is used on a Pi with + and - supplies to its opamps,
and the signal voltage falls to 0v,
then this system works fine.

however,
this system only provides x2 gain,
where as a sampling integrator like used on the surf Pi variations
can provide x220 to x470 gain.

hopes this explanation helps.

[6666]

Hi Dooley thanks,
the quick example attachment is invalid.

[DOOLEY]

ah,
oh dear,
yup,
your not wrong, it's also changed name.............

try this,

[eclipse]

DOOLEY, the design and build (~ Whites 6000) are amazing and the audio is the most pleasant I've heard.

[bernte_one]

dooley would youn please update the complete shematic
maybe i can work on a layout during christmas holidays

[eclipse]

Isn't it better to have D2, D3, R4, and the coil connected directly to 12V?
Otherwise you will need to pass high current through the voltage regulating elements there.

R7 can be lower, 10R - 20R is typical. BTW I intend to use Mosfet Driver IC there, I think this is overlooked in many designs.
What is your current design using - Dual +/- voltage for the Op Amps and no Regulated 1V reference?

Thanks for sharing the design!

[6666]

dooley would youn please update the complete shematic
maybe i can work on a layout during christmas holidays

Hi Berte_one a layout would be very helpful, thanks.

[DOOLEY]

hello bernte one,

haven't given this detector any thought at all during the last few months,
been working on another project.

so complete detector schematic still spread out on 5 pages,
all connections between the different sections clearly labelled so easy to follow.
all the schematics are on post 1#

tried to fit them all on one page,
but ended up to big to export,
sorry.
( cheap / free program fault. lol. )

only one mistake to point out bernte,
on the audio section,
on IC1, the pin 4 & 8 connection is missing,
program error I think,
I updated and posted corrected section on post 23#

not altered the design at all since I built it,
what you see on the 1st post is exactly what I put in my box,
100% working.

used for a week in sussex not used much since,
I do more inland detecting in fields.

please feel free to give the design a go.

hello eclipse,
yes it is quite a heavy draw on the voltage reg,
but it coped quite well,
only resulting in about 16mv ripple on the 4v rail.

sounds bad,
but since the pre amp was referenced to the 4v rail worked fine.

reason I did it was during the breadboard stage I quickly realized that
regardless of what voltage I pushed in to the coil I got the same detection distance,
but the higher the voltage the more power it consumed.
so,
as I was going to regulate the rest of the circuit to 8v it made sense to have the coil
at 1/2 that,
saving the need for a charge pump.

sort of a split supply if you will.

worked well,
stable,
and good on power consumption.

ah yes,
R7 220ohm from 555 to mosfet..........
using the IRF840 I noticed that the high voltage flyback wasn't smooth
at its peak,
looked harsh,
quite jagged,
played around with R7 and found 220 ohm gave a smooth flyback,
it was a compromise between getting a nice flyback,
without to much delay,
and the highest flyback without harshness.

I also though about similar,
just using NPN PNP driver,
but with the 220 ohm in place,
and the TS555 being up to the job,
I didn't .

opamps running on the +8v rail and the earth/0v rail.
as the coil on 4v opamps quite happy.

oh, that 1v reference voltage,
as yes,
that's to stop the lower earth field cancel opamp going in to
inversion,
the TL062's quite happy to go down to about 0.65v,
but not any lower,
so used the 1v ref voltage to keep it above the inversion voltage.

hopes all this helps.

[DOOLEY]

hello all,
again.

regarding the schematic being all on one page..........

as I was using a limited program I suspect that's why it wouldn't let me export
the complete schematic on to JPG.

I ended up doing I the good old fashioned way.......
yup,
separate A4 sheets and a roll of selotape , ( other forms of clear sticky tape are available )
suck it together on the kitchen floor,
hey presto,
schematic all on one big sheet.