5 inch detection max ability. What went wrong?

[billr]

Guys I'm really having a hard time with this build. Could do with some assistance if for no reason other than a process of elimination.
I got to step 7 doing everything in stages and nothing went as it was described in the manual at that point. Turns out 4 components were faulty. Andy from diydetector has been awesome help resolving the issue, sent new parts , fitted and my readings on U7 matched his.

After getting his all clear to move on I completed steps 8 and 9 without fuss. Volts did change when passed over coil in step 8.

The best I can get is 5" with a pair of pliers and a soda can is detected up to 12" a coin or ring is limited to 2". Have adjusted trimmed and swore at PL2 pot but nothing makes it better than what I have detected.

One thing that does not match what is described in the manual is step 7 Set the DC voltage level to around -2.1V off pin 1. I could not get a minus value adjusting R14. it went from 1.00v to 4.99v IIRC. Finally settled for 2.3v which was what Andy said his showed.

Tried a variable dampening resistor to get the best match for the coil this had no impact on detection range.

Coil is 20c dia x 25 turns of 0.5mm cu coated motor wire. A very good coil on a Surf 1.2 I built and a Russian Pirate kit.

Last is the detector hums when away from metal and goes silent when metal passed over the coil. That seems a little odd to me almost in reverse of what is expected but if it normal disregard.

I want to ask if there anything that can be checked using a multi meter and a single channel scope, its all I have. If someone could take pin readings off each IC perhaps a fault may be shown or at least narrowed down. I dont really want to keep emailing Andy for help as he has gone above and beyond to get me this far after 30+ emails back n forth. I know the defective components were not my fault but seriously how much time should a guy selling a few bucks worth of kit devote to solving a problem.

If you can help I sure would appreciate it.

[waltr]

It is tough with only a single channel scope since many things depend on the timing from one event to the next.
Main one is time from TX pulse off to first Sample (PL2 adjustment).

Can you show us some scope pictures? These may help.

1- Output of pre-amp U2, pin 6
2- Gates of Q3 and Q2
If would be best to see U2 out and each gate on the scope.

Does your scope have an External Trigger?
If so then this can give you timing differences. Connect Ext to preamp out and ch1 to FET gates.
Now you can see the timing of the FET gate change when turning PL2.

[billr]

Hi waltr
Thanks for the reply.
You are always there to help me out. I'm sure I keep you busy.
One of the forum members has kindly offered to take a look at it. Going to mail it next week. He obviously knows what to look for and has the equipment to test. Where as I just fumble around hoping to get lucky.

[surfdetector]

Hi Bill,

As waltr says :
Can you show us some scope pictures? These may help.

1- Output of pre-amp U2, pin 6
2- Gates of Q3 and Q2
If would be best to see U2 out and each gate on the scope.

It would be great for you to post these picture as it might be helpful to someone else down the road with the same issues that you are having now. Not everyone is going to be lucky enough to be able to send their PCB off to another to have it fixed.
The pictures are a learning tool for the rest of the forum.
Cheers

[billr]

Hey Surf.
You're right. Help is a two way street.
Pics attached.

Does your scope have an External Trigger?
If so then this can give you timing differences. Connect Ext to preamp out and ch1 to FET gates.
Now you can see the timing of the FET gate change when turning PL2.

I don't think it has a external trigger. Least not what I can see on the 1 page manual.

[Qiaozhi]

It would be great for you to post these picture as it might be helpful to someone else down the road with the same issues that you are having now. Not everyone is going to be lucky enough to be able to send their PCB off to another to have it fixed.
The pictures are a learning tool for the rest of the forum.
Cheers

Bill is sending the PCB to me. So don't worry as I will post all the gruesome details here when I figure out what's wrong.

[waltr]

Bill is sending the PCB to me. So don't worry as I will post all the gruesome details here when I figure out what's wrong.

Great, Thanks Q.

[kt315]

Great, Thanks Q.

yuppy

[Qiaozhi]

Bill's PCB is now working as advertised.
When I first tested the detector I couldn't find anything basically wrong or faulty. The only issue was that the minimum main sample delay was set way too high, and that would greatly affect sensitivity.

When I developed the Geotech Baracuda I wanted to keep as closely as possible to the original design, otherwise it would no longer be a Baracuda. The main focus was on fixing the various issues that others had found, such as intermittent failures of the 40106 IC. This meant that I had specified the mosfet as an IRF9530, whereas I notice that Andy has been supplying the kits with an IRF9640, which is actually a better choice. The IRF9640 has a breakdown voltage of -200V as opposed to the IRF9530 with -100V. The other thing I had forgotten about this design is that the damping resistor needs to be a much lower value than with other designs such as the Surf-PI, Hammerhead, or the Minipulse Plus.
The test coil I used had an inductance of 377uH and a resistance of 1.7 ohms. It was wound using 1mm thickness (with insulation) hookup wire on a 22cm diameter. This is not a fast coil by any means, but good for coin shooting. Normally this coil requires a 680R damping resistor, but the Baracuda needs this to be 330R.

I did some comparative measurements on a Geotech Baracuda board fitted with an IRF9530, and obtained the following results:
Power supplies: +5.00V and -4.93V
TX pulse width = 87us
TX pulse rate = 646pps
Preamp comes out of saturation at 30us
Mosfet avalanches at -100V
Main sample pulse width = 49us
Minimum sample delay was set to 30us

Detection ranges (in air):
Victorian penny = 12"
50p coin = 9"
U.S. Nickel = 7"
Coke can = 21"
?2 coin = 12"
2p coin = 10"
1p coin = 9"
Pliers = 12"

After nulling the preamp, adjusting the minimum sample delay, and lowering the damping resistor to 330R, the measurements on Bill's board were:

Power supplies: +5.06V and -4.96V
TX pulse width = 83us
TX pulse rate = 662pps
Preamp comes out of saturation at 17us
Mosfet avalanches at -230V
Main sample pulse width = 47us
Minimum main sample delay was set to 20us

Detection ranges (in air):
Victorian penny = 14"
50p coin = 12"
U.S. Nickel = 10"
Coke can = 24"
?2 coin = 13"
2p coin =11"
1p coin = 10"
Pliers = 14"

When I returned to Bill's PCB, after testing my own Geotech Baracuda board, I discovered that it had mysteriously stopped working. I traced the fault to U1, which was replaced with a 2N3904, as I did not have a 2N2222 available.

I have attached 3 scope images of the preamp output with various damping resistor values. Why Bill had the problem where the detector went quiet over a target I do not know, as it is working ok now. The only other minor issue I found was that U7 was slightly intermittent in its socket. If you pressed on it the detector stopped working, so I removed it and re-seated.

The confusion appears to have been caused by an error in the Build Document where it instructs you (if you don't have a scope) to adjust R14 so that the dc voltage at U7 pin1 is equal to -2.1V. The minus sign should not be there. The minimum sample delay is set by R14. On Bill's PCB, setting the voltage at +2.1V gave a delay of 25us. A voltage of +2.5V gave a delay of 20us. Using this dc voltage to set the minimum delay is only a rough guide anyway, as component tolerances will make this an unreliable measurement. The only real way of setting the delay is using a 2-channel scope. However, if you have a single-channel scope with an external trigger input, you could use that to trigger it from the TX oscillator output.

I have updated the Build Document which is available here -> https://www.geotech1.com/forums/show...roject-Details
If you have an older version of the Build Document, then delete it and download the latest one.

Bill received the fixed PCB today and has confirmed that the detector is working as expected.

[Carl-NC]

Bill received the fixed PCB today and has confirmed that the detector is working as expected.

Service with a smile Well done.

[billr]

Bill received the fixed PCB today and has confirmed that the detector is working as expected.

I did and I have.
And I want to say for the record in the thread I started. Thank you so much for helping me out with this. Very kind of you.
There is a lot of good info in your post above and slowly, gradually I am starting to understand the principles of what the build manuals are saying.
I still have the MPP to complete once I purchase a 2 ch scope. The Barracuda with all it's problems has been frustrating but at the same time a good learning experience. I think that may be an asset later on.

[surfdetector]

Good news regarding the PCB fix and some valuable measurements in the post above. My Barracuda PCB is still on the bench, I shall have to do some comparisons.

[surfdetector]

In reviewing the Barracuda schematic I have noticed that R14 is listed as 25K, whereas in the list of materials it shows as 20K and I am wondering which one was the true value?
Thanks.

[kt315]

25k is rare value type, so people correctly write 20k. no true value. just a choice.

[waltr]

In reviewing the Barracuda schematic I have noticed that R14 is listed as 25K, whereas in the list of materials it shows as 20K and I am wondering which one was the true value?
Thanks.

R14 is a pot with a 10k pot in series.So 20k or 25k doesn't make much difference since the adjustment of R14 should be somewhat near center.
As KT said, it is a choice not a hard value.