SMT Bandito Info/Build

[Silver Dollar]

Well it seems the coil I built is the problem. I went back to the original and it works fine.
I didn't shield it and used a centertap. I'll have to try it outside to see if it is usable.

I did find one suggestion for the oscillation on the IDX thread. It was suggested a 22k to
ground on the output of the first filter stage will reduce tendency to oscillate. I might try
that....

[anonymouse]

Hi Dear Silver Dollar
Did you have any PCB for SMT Bandido?(no SMD version,Deep version)
I'll build it
Regards

[Silver Dollar]

I have no boards for sale just one I got from Silverdog.

I need a good coil. I have one that I superglued that had ty wraps on it. After the glue dried
I couldn't get the ty wraps off. I'll have to try some solvents to see if I can reclaim that one.

I also have the Tracker coil. It's 2.7 mh and 2.4 mh but has a cap in the coil so fixed at 6.6 khz.
If I can get the half's separated I can remove the cap and experiment a bit. It's tough going as
it appears to be PVC glue so you almost have to cut the plastic apart.

I found some plans for an Omega coil for TGSL that might be fun to build. I'm not sure I have enough
wire left though...

[Silver Dollar]

Well I managed to open the Tracker coil and cut the cap out. So now I have a nicely
nulled and shielded coil to play with!

I tested it on the Minelab oscillator and got 5.9 khz at 23V P/P!

On the Bandito with no mods (220n and 20n) it gives 15.5 khz and 9.1v P/P
(with the TGSL caps it was 21.5 khz and 19v P/P)

Hmm the Bandito run seems a bit low. I might have to play with biasing on that
transistor. We might actually want a small TX the originals are usually around 9V.

I'll try to find some values to make the Bandito run at 10 khz and see how it works.
Right now at 15.5 khz it is just a little short at 8" for some coins and 9" on Silver
probably 8" for gold... I was able to null out the ferrite so I could leave it here if
the disc works ok...

[Silver Dollar]

Tried setting it to 12.5 khz but amplitude was only 5V! I tweaked the 1.2K back to 1.2
(as 2.49K was better at 10khz) now 9.2V but very low depth! It seems a strong TX chokes
the signal somehow. It seemed to run fine at 15Khz so maybe I'll go back to that...

I ended with 15.2 khz at 9.2V P/P and got 9" on silver and gold and 7" - 9" on coins.

I still haven't optimized the RX cap but it looks OK!

[Silver Dollar]

The null is 4V P/P with 15.2 khz but it still works good! Not sure if I should try to
lower it or change frequency closer to the designed one?

[Silver Dollar]

I tried many different frequencies by changing the TX cap and RX cap.
The null voltage got too big at some settings but then I lost the ability
to ground balance against ferrite. The ground and disc pulses look OK.

I went back to stock caps and the original Bandito Coil.

I tried messing with the components around the GEB circuit but nothing
changed. It seems ferrite is on the disc now for some reason. I'm not
really sure which components control the start of the disc range (or both)?...

[Silver Dollar]

Oops I found that I had changed the 39pf on the input opamp for some reason.
Now it's ground balancing fine. I guess the hidden target phase was changed by
making that 15pf instead of 39pf. I thought that was only for high frequency roll off...

I have the Tracker coil running at 13 khz and 7.7V p/p but depth is a bit low a little
more tweaking is in order!

[Silver Dollar]

The Bandito is very resistant to any tweaks! I found something funny, if you
remove the 470n's and short the connections it hardly works! I get 2" like that.
I'd imagine the other filters and opamps would do a bit better... Weird.

I did get a setup where it was very linear in volume depending on distance by
changing the 100K on the threshold opamp for 40K but I didn't like it as the
volume gets really low.

The threshold doesn't seem to do much, I'll probably leave it as a fixed resistor

It seems there is a bit more in there but hard to coax it out. I could up the TX
voltage but not sure it will help. I may try lowering the gain on the first stage.
It's counter intuitive but some depth might be gained. You just have to change
the caps too to keep the same overall filter response.

The chatter filter is the big limiter here. I can hear ticks that should sound as a
good signal but that filter stops it...

[Silver Dollar]

I tried to run the TX on + and - 5V to get a larger TX and got 13v P/P but it was pulling the
supply down and sounding continuously. I couldn't even get an idea if it gained any depth as
it was too noisy...

13 khz seems to work the best with the Bandito board and Tracker coil. It's modded a bit and
gets 9" on gold, 10" on silver and 7 - 8" on coins. Time to package it up and take it outside!

Another curious thing, I cut the gain in the first stage in half and when I adjusted the sens for
the edge of chatter I got the exact same depth. I wonder if even lower gain will reduce the
noise so allow a higher sens and be able to see a little deeper?

I also removed the diodes in the second stage with virtually no effect.

The signal getting to the threshold opamp was about 55 ms wide. I'd like to know where that
55 ms is coming from, that's where some changes might have an effect....

[Silver Dollar]

I have a couple of ideas. The first is that the input config has a great effect on the signal
and noise. I changed the gain resistors from 5k and 100k to 1k and 20k to lower the noise
but that might be having a loading effect on the coil. I'm not sure the equivalent noise resistor
for the MC33178. Maybe try 10k and 200k just for giggles? On the TGSL eliminating the 2 resistors
on the RX positive side helped depth a lot but here it hurt it....

The other thing I'd like to try it to eliminate the diode and 1meg from the comparators to the final
opamp just to see what kind of signal I get. It looks like the noise is hiding under that diode so it might
become very noisy. On TGSL they add the 1meg for loud audio but I found reducing the gain of the
last opamp gives very proportional volume.

It seems the whole audio chain is tuned and so it's hard to make huge improvements but easy to hurt
performance. I'd like 1" more on all targets but might not be able to get it....

[Silver Dollar]

Well after lots of work I've arrived at the most optimal values for most every part on the board.
I'm using the Tracker Coil (TX 2.7 mh RX 2.4 mh) and have the max depth and stablest config
I could tune. The last mod allows for volume change with depth. You might think your missing
some targets due to low volume but they are not there. There's a wall you hit and no signals
get through it. Motorboating lives at extreme sensitivities and I couldn't find a way to get rid
of it to hear those faint signals. The volume mod stops the tendency to run on after a target
goes by. So I'll clean the board one more time and start working on the front panel and putting
it in a box!

Here's the final schematic. There's some funny values arrived at by lots of experimenting;

[anonymouse]

Hi

this is the final change for bandido?
now,we can design PCB for it?


Please put your test results here
which is the best?(TGSL,IGSL,IDX-Pro,SMT Bandido)


Thank

[Koala]

Hi

this is the final change for bandido?
now,we can design PCB for it?


Please put your test results here
which is the best?(TGSL,IGSL,IDX-Pro,SMT Bandido)


Thank

All very similar. Each build and coil is slightly different but on average all the same. IDX is probably slightly easier to get going good. and the IGSL slightly better discrimination.

Out of the Tesoro none digital range Monte preferred the Bandido II µmax for trashy areas (Low noise, quick recovery and true no motion) and as a metal detector dealer he had the choose of any detector

Technically there should be a small advantage in the lower noise of SMT not seen in the real world yet

For a first build go for a tried and tested one IDX or IGSL and get a board from silverdog as they are proven to work unless you detect on the beach then you would be better with the Surf PI, Baracuda and the new MiniPulse.

Wish you much success in what ever you build

[Silver Dollar]

I had a bit of a breakthrough today. I had just about changed every component on the board
trying for more depth. Well I was looking at the IDX as a possible board for my daughters little
MD. I got a BH coil that has a cap in it so wants to oscillate at 6.5 Khz. I built the IDX oscillator
tonight and found it works quite well on the Bounty Hunter coil.

So anyway I was studying the schematic to see the differences from Bandito/TGSL to IDX and
noticed the cap after the input op amp is 22n on IDX and 220N on the other MD's. I changed it
on the Bandito and Bang 1 - 2" more on all targets. Silver is solid to 13" and cracks to 15". Gold
goes to 12" and the the coins go 1 - 1.5" deeper too!

I'm not sure the function of that cap. It could be a low pass or high pass but changing it sure
makes a difference! The sound is a bit crunchy towards the max depth and haven't tried it in the
ground yet but it appears promising.