SMT Bandito Info/Build

[Koala]

Firebox http://www.youtube.com/watch?v=1HiHej59W2Y do not look to shabby even though its maxed out. Homemade coil by the look of it and minelab mod visable on the board.

Even with any mods the Bandido ii µmax should out preform the original Bandido for depth slightly either something amiss with the board or we don't know the correct specs for the 5 pin µmax coils.

[mschmahl]

Here is a link for the german forum http://www.detektorforum.de/smf/eige...-t54227.0.html where firebox write some info´s.

[Silver Dollar]

Well I did some testing today. I had cut the original cable as it was 8 foot long and
I put a connector on it today and tested the Original Bandito and the SMT one.

The Original wouldn't ground balance on a piece of ferrite but might need an internal
adjustment for the new short cable. But it got;
9 - 10" Gold
12" Silver
All Coins 9"

The SMT Bandito nulled ferrite ok and got;
10" - 11" Gold
12 - 13" Silver
9" All Coins

With the new 9" DD coil I just made;
10" - 11" Gold
11" - 12" Silver
9" - 10" all Coins

The new coil is a 6.0 mh TX and 6.5 mh RX center tapped with no shielding. Some random beeps were
apparent. Here's a picture;



The null was around 200 mv at pin 7 of the first opamp but went way up while the hot melt was applied
then back down to 73 mv after it cooled. That might be too low. We are aiming for 4 mv at the coil which
should be 200 mv at the output of the first opamp. I saw 500 mv with the original coil...

[Koala]

Sounds good my IDX gets just under 16" with a 27cm DD but looses lots in the ground.

Ivconic made both normal and center tapped and found the normal had more depth but center tap had better discrimination but that detector originally designed for a center tap coil.

[Silver Dollar]

Well I saw 16" and loud audio with the TGSL once but I'm back to using the SMT Bandito
as it fits in the Tracker case. The plan is to get the coil working then tweak the board a
bit more and see what it can do. I tried the Minelab oscillator but couldn't get the AB+C
mode working as not enough -5V. The Original Bandito is still less noisy too, I'm not sure
how to fix that. Reading the TGSL thread again it can be the pot, the coil, the board or most
anything. It can go away for 45 minutes then come back etc....

Experimenting I had echoes in the taget signal, sustained tones after detection and rhaspy
tones. There's a lot going on there. It's like a fractal thing the closer you look the more complex
it gets! Well I think it will be usable as is. I'll try it on real ground soon and see what happens!

I mainly used the center tapped coil as making a sheild is tough! I have some shielded coils that
don't fit in the housing! You loose 1 or 2 DB with the centertap but gain the loss with shield so
is probably a wash. I just hope it is OK on wet grass...

I thought I understood how things worked on these devices but found the real world sometimes acts
differently than you expect. For instance I was playing with the 4.7uf caps on the TGSL. I matched a set
with the highest value I could find (To get the lowest frequency so as to let more signal through). The
matched ones were 4.5uf but the ones I took out were 4.9uf! I tried the 4.5's and they seemed to work better!

[Koala]

Well I saw 16" and loud audio with the TGSL once but I'm back to using the SMT Bandito
as it fits in the Tracker case. The plan is to get the coil working then tweak the board a
bit more and see what it can do. I tried the Minelab oscillator but couldn't get the AB+C
mode working as not enough -5V. The Original Bandito is still less noisy too, I'm not sure
how to fix that. Reading the TGSL thread again it can be the pot, the coil, the board or most
anything. It can go away for 45 minutes then come back etc....

Experimenting I had echoes in the taget signal, sustained tones after detection and rhaspy
tones. There's a lot going on there. It's like a fractal thing the closer you look the more complex
it gets! Well I think it will be usable as is. I'll try it on real ground soon and see what happens!

I mainly used the center tapped coil as making a sheild is tough! I have some shielded coils that
don't fit in the housing! You loose 1 or 2 DB with the centertap but gain the loss with shield so
is probably a wash. I just hope it is OK on wet grass...

I thought I understood how things worked on these devices but found the real world sometimes acts
differently than you expect. For instance I was playing with the 4.7uf caps on the TGSL. I matched a set
with the highest value I could find (To get the lowest frequency so as to let more signal through). The
matched ones were 4.5uf but the ones I took out were 4.9uf! I tried the 4.5's and they seemed to work better!

Don't forget the original Bandido II µmax had a red zone at the end of the sensitivity knob and if you use the machine in the red zone you get an occasional beep.

I agree that getting the shield right is hard. Depth isn't everything out in the field stability and target separation are equally as important.

[Silver Dollar]

Hmm I'm looking to tune the coil a bit but am confused. The TX is around 10 khz but the RX calculates to 10.867 khz.
On TGSL the TX vs RX frequency is about 1.5khz difference...

[Koala]

Hmm I'm looking to tune the coil a bit but am confused. The TX is around 10 khz but the RX calculates to 10.867 khz.
On TGSL the TX vs RX frequency is about 1.5khz difference...

You and me both I read through most forum post I can not find any measurments of the newer Epsilon coils designed for µmax machines. Seem most of the work was done on the Tesoro years ago and interest has been lost in all except for the TGS

[Silver Dollar]

Well you can get the Bandito to 14 khz by using a .022 instead of .047 in the TX. It showed the best
range before adding a RX cap on gold and coins but not so good using the .015 of the TGSL on RX. Could
need some Disc and GEB work to perform better.

The SMT Bandito draws 15.3 ma when not sounding. Changing the TX from 10 khz to 14 khz didn't change
the draw any but the Peak to Peak value for TX went from 8.5v to 10V...

[Silver Dollar]

Well I'm doing some playing around today with different frequencies and found that the higher null
signal gives greater depth?? The detector sounds sick today though a bit quiet and "scrunchy" not
sure how to get a clear tone? I'm looking at the ground signal right on the 100n cap and see big jumps
with the random beeps. Power to the chip bad?

[Silver Dollar]

The original was acting a bit funny too. Seems there was a bit of noise in the air.
I did find pressing on the first filter chip caused a low frequency oscillation and the
sound got stronger! Measuring the gain resistors and one reads right the other reads
low so it seems there is a mechanical problem in that chip. I didn't buy spares either...

I can try a 27L7 or LM358 as I have those in SMT... I wanted to try the original filter
setup anyway. That uses a servo like filter with a 358 and 27L2...

[Silver Dollar]

Well I tried the servo filter but the board didn't work well. So I tried the LM358
in the original circuit but less depth. So I put the MC33178 back in. Then it
worked as before but with weak sound. I get a bit of chatters too forcing me
to use low sensitivity. The average noise looks OK (40 mv peak to peak at last
LM393) but occasional big hits come through on a random basis.

I thought things seem worst after removing the TX while trying the Minelab
oscillator so maybe the new parts are a bit different? I had noticed a larger TX
at 14.7 khz (TGSL frequency) so changed the caps in the TX to get 14.5 khz (10.5V P/P).
The beeps got a bit high pitched so I had to move the diode on pin 11 of the 4024
to pin 5. But now I get a kind of echo after a target goes by. I did seem to get silver
at 15" though so will persue this setup to see if some depth can be gained.

The gain seemed to come from the input opamp. I saw Davor added a different opamp
to IGSL and balanced the input by putting a cap across the gain resistor to ground so
I did the same. I changed the 39 pf to 15 pf and added one to ground then saw 15"
on silver...

There could be a problem with the ground balance but it looked OK. The Disc is defiantly
shifted as I was able to disc out the gold. A few caps might fix that.

I might try using a 7660 for the negative voltage (as the original Bandito does) mounted
on the back side of the board thus using the ground plane as a shield to try and reduce
noise.

[Silver Dollar]

This thing keeps changing. Today I added the Threshold pot and now I have 100mv noise at the
last LM393 and I only get 5" on gold...

[Silver Dollar]

Today 8" and noisy. I see 10 hz oscillation on the filter opamps. I tried changing the front end gain
and not much difference. It does loos range by lowering but doesn't gain by raising the gain. I saw
double the gain on the TGSL and expected it might be useful here...

The pahse from TX to RX seems to change too. Today it is 236 degrees. I saw 40 degrees the other day...

Not sure how to calm this thing down. I added 10 ohm resistors to each power pin on the opamps
then a cap from the pin to ground hoping to isolate and quiet down the power to them. The original
doesn't use any but it is on a substrate with all resistors and etches in it probably to fight chatters...

[Koala]

Have you tried turning the threshold all the way up when in disc mode