Baracuda Build Blog

[Koala]

Been out and got some new batteries.

1 new PP3 works fine but less stable. 8 new AA rechargeables don't work at all.

[morganton]

Been out and got some new batteries.

1 new PP3 works fine but less stable. 8 new AA rechargeables don't work at all.

Hi, what voltage do you have on pin 5 of the 7660 when you have your 8 X 1.2V rechargeables as the power source? It should be about -9.3 V or so. If it's below the threshold for the -5V 7905 regulator, the detector will not work. You could try another 7660, perhaps one made by Intersil. The 7660A can handle up to 12V if you are worried about the voltage.

[Koala]

Hi, what voltage do you have on pin 5 of the 7660 when you have your 8 X 1.2V rechargeables as the power source? It should be about -9.3 V or so. If it's below the threshold for the -5V 7905 regulator, the detector will not work. You could try another 7660, perhaps one made by Intersil. The 7660A can handle up to 12V if you are worried about the voltage.

Tried different batteries and all give roughly just under the battery voltage but minus on pin 5 which is good I take. (already using a 7660A)

Been doing some experiments on the detector only works when the batteries are nearly flat. Looking at the pulse to the coil when the battery is nearly flat I have a nice curve when running a fully charged battery there is a lot of random noise not ringing along the curve. Changed the damping resistor for a potentiometer does not matter where its set their is noise all over the place. Put an old battery back in and the trace is smooth no noise at all.

When its working well I get good detection on a gold ring of 9" with a 9" coil but just change the batteries for full charged and it will not detect a large spanner.

[morganton]

Tried different batteries and all give roughly just under the battery voltage but minus on pin 5 which is good I take. (already using a 7660A)

Been doing some experiments on the detector only works when the batteries are nearly flat. Looking at the pulse to the coil when the battery is nearly flat I have a nice curve when running a fully charged battery there is a lot of random noise not ringing along the curve. Changed the damping resistor for a potentiometer does not matter where its set their is noise all over the place. Put an old battery back in and the trace is smooth no noise at all.

When its working well I get good detection on a gold ring of 9" with a 9" coil but just change the batteries for full charged and it will not detect a large spanner.

That's good--it means your 7660 is working ok. I'm using 8 X 1.2V rechargeables with silverdog's kit and it works fine. You might check the voltages coming out of both the 79L05 (-5V) and 78L05 (+5V) regulators with fully charged batteries to see if they are correct. Pin 1 on the 78L05 is output. Pin 3 on the 79L05 is output.

[Koala]

That's good--it means your 7660 is working ok. I'm using 8 X 1.2V rechargeables with silverdog's kit and it works fine. You might check the voltages coming out of both the 79L05 (-5V) and 78L05 (+5V) regulators with fully charged batteries to see if they are correct. Pin 1 on the 78L05 is output. Pin 3 on the 79L05 is output.

Output from both look good. 5.1 and -5.1 give or take a few mV. The plus 5 has a .1 volt ripple on it

[porkluvr]

Output from both look good. 5.1 and -5.1 give or take a few mV. The plus 5 has a .1 volt ripple on it

I just took a look at a couple of different Barracuda schematics that I have on file. One version shows that -5V has a 220uF capacitor connected to Battery+, with no capacitor being connected to +5V. That does not strike me as a very smart thing to replicate...

An older version shows 220uF connected between +5 and -5V. That seems reasonable, although separate capacitors from +5V and -5V to ground might be a safer option for the average hobbyist.

This should raise the question of whether or not the +5 and -5V supplies are adequately bypassed on whichever PCB you have. What about it?

[morganton]

I just took a look at a couple of different Barracuda schematics that I have on file. One version shows that -5V has a 220uF capacitor connected to Battery+, with no capacitor being connected to +5V. That does not strike me as a very smart thing to replicate...

An older version shows 220uF connected between +5 and -5V. That seems reasonable, although separate capacitors from +5V and -5V to ground might be a safer option for the average hobbyist.

This should raise the question of whether or not the +5 and -5V supplies are adequately bypassed on whichever PCB you have. What about it?

Yeah, if you are referring to how the 220uF cap is routed on the PCB, see post #8 of this thread. http://www.geotech1.com/forums/showt...run-on-9-volts

I rerouted the cap the way it should be on my barracuda PCB. Although I didn't see any difference in operation, I also wonder if the way the cap is routed on the PCB may be causing some kind of long term problems with the 7660. In any event, I kept the change I made in accordance with how the 7660 data sheet says the 220 uF electrolytic should be connected. The data sheet says the cap "must" be connected with the positive lead to ground and negative lead to pin 5.

The 78L05 (+5V) data sheet shows the input with a .33uF cap to ground and the output with a .01uF cap to ground. http://www.jameco.com/Jameco/Products/ProdDS/51182.PDF Note 4 to the "Electrical Characteristics" says the .01uF cap is minimum load capacitance to "limit high frequency noise".

[morganton]

Koala, what I would probably do next is check the current draw of the circuit using your flat battery and then check it when using the rechargeables (if your meter can measure amperage) to see if there is any kind of big difference. I did this for mine and I think it was about 150mA.

[Koala]

Not to far of 177mA been doing some experimenting made a smaller coil a now at least I have some detection on full battery iron and copper but poor range gold not at all. Coil trace on the scope is very noisy. Random jittering all over the trace but nice and smooth with a nearly flat battery.

[sinclairuser]

Not to far of 177mA been doing some experimenting made a smaller coil a now at least I have some detection on full battery iron and copper but poor range gold not at all. Coil trace on the scope is very noisy. Random jittering all over the trace but nice and smooth with a nearly flat battery.

just an idea koala, try changing the 2n2222, i had a metal can one(as per kit) in my 2nd barra to start with, but it was unstable with higher currents, after changing 7660, 7805,7905 and the 40106, for replacements, a cheap plastic 2n2222 did the trick.
was going to put another metal can original in, but left it as it worked well, though you will obviously have to turn round the plastic 2n2222 compared to the can type.

[Koala]

Cheers sinclairuser 2n2222 is about the only part I have spares of so easy to change. Will have to be Monday now as will be out with the IDX tomorrow.

Was going to try some different wire for the coil. Strange as I can see the fault. The trace on the coil is really bad with fully charged batteries but really nice with about 6 volts. Kinda like little blips dancing all over the place.

When I hit the Goldilocks zone with the battery it works very well.

[sinclairuser]

yes it works but not right, and the scope looks noisy with better batteries, does yours have an instibility in regard to your hand and the board?.
mine did all you describe and what i put above, a new plastic 2n2222 cleared it all up mate, worth a try.

[Koala]

Made a new coil out of standard copper wire just to eliminate the unknowns. Don't know the quality if the wire I used.

Already using a plastic 2n2222 but changed as it's the only part I have spares of.

Now works on a full PP3 battery (moving forward thanks) and detects gold but kinda backwards. If you hold the ring near the coil and move it quickly away then it will sound. Depth on gold is only a couple of inches was 9" no amount of fiddling with the setting helps. Where as iron is the same depth as with a flat battery

It's always done this but the threshold sometimes fades in and out like a siren swithing on and of or tweaking one of the pots temporarily stops it.

Still will not run on rechargeables even if I remove some of the cells to fetch the volts down to below 9V



Have to look at the trace again but that will have to wait until tomorrow

[Koala]

Looked on the scope and there is noisy all over the coil trace.

Volts at the coil look about the same with a 9V battery and rechargables must be more energy in the coil though as with rechargables the coil has a slight buzz.

Noticed that I turn the detector off as the capacitor discharges the volts go up at the coil by 35 volts ?

Really impressed with the detector when it's working it's works well.

[sinclairuser]

yes it seems the barra can be quite awkward when it wants, the first one i built is spot on(all components as supplied), the second one was quite simply not as easy to build, all components on this one are higher rated or millitary spec(metal can trans(except 2n2222), ceramic chips, even the 7660 and tip darlington are metal can(expensive)).
but even then it has a skittishness when batteries are low, i have reworked this one 4 times now, while testing andy's coil i have used the first one built because of this problem(so as not to confuse results), the reason i keep pratting with the second one is my "gut" tells me its something simple.