BARACUDA-Legend-1

[6666]

All
I have transferred some of the mods by Alex / APberg and Silver dog PCB to the old Barra schematic
it is not fully complete, I will complete it as I build it and get it working
I think I have done the drawing right so far, mods in orange colour dated 30 june 2011
I did it late last night and I am not well with flu

[ApBerg]

Hello 6666

Thanks.... and get better soon !

Regards.

Ap

[Janou]

Thanks 6666, i follow all mods

I receved the kit from Sliverdog, really nice PCB ( just perfect ) and very good kit !

I just try to read more information to find the "perfect" coil and after... testing
At the moment, the specification of coil from bernte_one seems OK

Value of C1 is not critical ? I find 3300uF in the kit

Regards and thanks for all theses informations !

[TH Jim]

Hello Ap

Yes I agree the offset adjustment on 5534 is critical to good performance. I noticed that it has a different scheme for this on the barracuda compared to the SurfPI or the one shown in the NE5534 datasheet, so today i changed barracuda type offset to a surfPI one. The change gave a much greater range of adjustment. Previously it was only adjustable within the lower half of the output of the 5534, now it covers almost all of the range of the signal. However this made hardly any difference to sensitivity. I have also tried using a smaller coil from a surfPI and this does improve sensitivity slightly but it is still nowhere as good as a surfPI.

One strange thing i have noticed is that the shape of the pulse on pin 2 of the 40106 seems to mirror the shape of the portion of the decay curve being sampled. The pulse on pin 4 is always a nice clean square. When the delay is adjusted to the longest setting then the pulse on pin 2 is nice and square (where the Rx signal is zero) and measures 30uS. But when it is set to a short delay (say 15uS after Tx switch off) it becomes distorted and shortened to as little as 15uS

Has anyone else observed this happening on pin 2 ?

The info given by Alexismex said the sample window was 45uS wide, mine is 30uS at best. Does anyone else have a 45uS sample window?

If i dont crack this problem tonight i will post some scope pics tomorrow.

Regards, Jim.

[ApBerg]

Hello Jim,

One strange thing i have noticed is that the shape of the pulse on pin 2 of the 40106 seems to mirror the shape of the portion of the decay curve being sampled. The pulse on pin 4 is always a nice clean square. When the delay is adjusted to the longest setting then the pulse on pin 2 is nice and square (where the Rx signal is zero) and measures 30uS. But when it is set to a short delay (say 15uS after Tx switch off) it becomes distorted and shortened to as little as 15uS

Has anyone else observed this happening on pin 2 ?

Yes with me the same, ..I first make the square at pin 40106-2 at its full wide ( the same fig. as on pin 4) , look at the scope photo a page ago... then with the offset I look for the most sensetive setting..

Yes... strance but it works ,also for you I hope !
Made now 6 proto bara's al with sens. over 30 cm for a gold ring with coils from 20 and 22 cm dia. So a little better than the surf... but the surf only takes some 50 mA and the bara some 100 mA..

Best regards

Ap

[TH Jim]

Hi ApBerg and other friends,

Have attached some photos of my delay pulses. The top trace is the output from NE5534, the bottom trace is the delay pulse from 40106/2. I was incorrect yesterday when I said that the shape of the delay pulse was a mirror of the decay curve - it is in fact as you can see below a copy instead. Anyone have any ideas why the delay pulse is distorted like this? Would expect to see some feedback on the base of a BJT, but not on the high input impedance of a FET? Needless to say the signal coming out of gate A is a similar shape too.

ApBerg, I'm glad to hear you have the same result and still have good performance, I won't worry about this being the cause of my poor results.

As for the SW1, i think Smitty is correct in post 56, i followed his instrucrions and then found that the delay is set to a nice point over the decay curve (photo 2) and that the switch then puts the machine in motion / non motion modes. Very handy

My threshold still needs to be turned completely to one side, that doesn't seem correct.

My barracuda is now almost as sensetive as a surfPI, but if its gonna draw twice as much current as the surfPI - THEN I WANT IT TO PERFOM BETTER!!!

Regards to all, Jim.

[plumabob]

For Apberg,

30cm on a gold ring? Can we have your coil specs, type of wire used?

Thanks

Bob

[DIRCEU]

Hi! ApBerg

30 cm for a gold ring you said.
Please tell me if you broke this gold ring at what distance can you detect it?
with the Barracuda and Surfmaster.


Many thanks.
Dirceu

[porkluvr]

Hello Ap

One strange thing i have noticed is that the shape of the pulse on pin 2 of the 40106 seems to mirror the shape of the portion of the decay curve being sampled. The pulse on pin 4 is always a nice clean square. When the delay is adjusted to the longest setting then the pulse on pin 2 is nice and square (where the Rx signal is zero) and measures 30uS. But when it is set to a short delay (say 15uS after Tx switch off) it becomes distorted and shortened to as little as 15uS

Has anyone else observed this happening on pin 2 ?


Regards, Jim.

The reason you're seeing that difference between the base signals on the two sample FETs is because the non-inverting opamp input (primary sample) has a very high impedance but the inverting input (secondary sample) has an impedance approximately equal to the input resistor; 1k in this case.

That underscores a difference between the Baracuda and most other PI's where other PIs usually feed the primary sample to a non-inverting input with the secondary sample being fed to an inverting input.

Loading on the FETs would be less apparent if the samples were reversed because the primary sample would be seeing a high impedance. Although there would still be some curvature to the gate drive signal due to the shape of the decay curve, it would be much less obvious.

Ideally, a differential amplifier would have equal impedances on the two inputs but because of the nature of pin-2 being a 'virtual ground', that is not true.

I hope I clarified that.

[yhl]

help me

LF412cn, / NE5532 LM358 and TL082 can be used instead?

[ApBerg]

Hi! ApBerg

30 cm for a gold ring you said.
Please tell me if you broke this gold ring at what distance can you detect it?
with the Barracuda and Surfmaster.


Many thanks.
Dirceu

Hello Dirceu.

You mean if you make a cut in the ring so that it is not a complete circle anymore ?
I have not tested that yet with this Baracuda.... however some 25 year I have discovered that a not complete ring is very hard to detect.(it was broken so it was not a closed circle any more)
On a detector-rally day had some fun... new detectors that were very bad detecting a that gold ring ..ha..ha.. my selfmade detectors were much better...(closing the the ring so it became a full circle..)
Off course later I have told the trick !

Best regards
Ap

[ApBerg]

LF412cn, / NE5532 LM358 and TL082 can be used instead?

I use a lm358 in place the lf412..

[6666]

Have a look at this video
he cuts the can ring

http://www.youtube.com/watch?v=PBKBkHtzVDY

[6666]

When building the bara
make sure you get the 2222a in the right way

[DIRCEU]

Hi! ApBerg.


My surfmaster can detect a gold ring at a distance of 27cm.
I'm trying to detect a broken gold ring with various types of coils but without success.
Today I will start building the Barracuda. I use the LM 358 because I'm not finding the LF 412.
Anyway thank you for answer my question

Best regards
Dirceu