modification ATX

[Davor]

now who posted this picture ages ago with the suggestion a coil could be
balanced using this configuration ????............................

We had a thorough discussion on this coil here: http://www.geotech1.com/forums/showt...New-PI-machine
Aziz even made calculations with dimensions that balanced. The inner eye size is critical for balance. However, for PI a perfect balance is not a requirement.

[daverave]

i have wire like in the photo and mine is litz...cant believe garrett would use ordinary wire.

[Mechanic]

Hi Guys,

Here is the atx front end I don't have values for the tx and rx coils that I can find.

Cheers Mick

[Mechanic]

Ooops, the preamp ic is missing from the last file.
Please see the updated pic..

Cheers Mick

[Qiaozhi]

Ooops, the preamp ic is missing from the last file.
Please see the updated pic.. [ATTACH]34699[/ATTACH]

Cheers Mick

Do you know the value of the capacitor in the feedback loop of the preamp?

[Taktyk]

Mechanik,

You can photo pcb parts preamp and tx?

Regards

[Mechanic]

Do you know the value of the capacitor in the feedback loop of the preamp?

No unfortunately I have no values for the smd caps as I didn't want to remove them.

Taktyk, Here you go!

Modify and tinker at your own risk. If you blow it up, don't come to me

Cheers Mick

[greylourie]

Mechanic,

Thanks for posting. Its interesting to see the circuit of the preamp. Low value R input used there! wam posted some photos of the infinium which show same U3G diodes and a large poly(?) capacitor allied with tx circuit as seen in the atx image. (http://www.geotech1.com/forums/showt...275#post119275)

[Mechanic]

Hi Grey,

Thought it might be interesting for others to know what is in the front end, especially when fiddling with coils. Looking at the input protection(or lack of) I would say that the RX coil needs to be carefully nulled, and from memory it was nulled quite well. It beats me how they adapt a mono coil to it. I tried a ML type mono and put in a 500r resistor on either side(500r to rx input, 500r to Agnd) with some back to back diodes, but it was as deaf as a door post. I tested out a Garret mono but could not figure out how they configured the rx connection as it is done within the coil. Also there is no gnd on the tx side, both tx lines are active.
After field testing it though, my interest in the machine was very quickly lost... Nothing like swinging a 56k modem!!! And I could get nowhere near the sensitivity or depth claimed by others.

Cheers Mick

[WM6]

It beats me how they adapt a mono coil to it.

Probably this way:

[greylourie]

Hi Grey,

... Nothing like swinging a 56k modem!!! And I could get nowhere near the sensitivity or depth claimed by others.

Cheers Mick

Trying to recall the sounds of dialing up AOL, your description brings a smile.

[daverave]

No unfortunately I have no values for the smd caps as I didn't want to remove them.

Taktyk, Here you go!
[ATTACH]34701[/ATTACH]
Modify and tinker at your own risk. If you blow it up, don't come to me

Cheers Mick

why do garrett use such large pcb's in their pulse machines and spaced out components...maybe less cross talk between tracks and components ????

[Beav]

I was wondering the same thing daverave. I have been doing PCB design for 19 years and I can tell you there really is TONS of room remaining on the ATX board design, compared to the tight designs I'm used to doing. I have to meet strict EMC and EMI compliance
but there are techniques you learn over the years that make you better each design you do.
Beav

[Mechanic]

Probably this way:

Hi Wm6,

Could be something along those lines, however both sides of the tx coil need to be switched to the rx inputs while no tx is happening.

[Mechanic]

why do garrett use such large pcb's in their pulse machines and spaced out components...maybe less cross talk between tracks and components ????

Hi Dave and Beav, who knows why they are so large, perhaps they use auto-placement and autorouter!

Longer traces will have more potential for cross-talk and interference. The ATX board could be routed to 25% of its current size IMO.
One board I am currently working on is 60mmx75mm and has almost as many components and some room to spare!

Here is the timing pulse train...

Cheers Mick