Barracuda.....update(2)

[Geo]

Again here. I tried the file and now the simulation works ok.
But very bad pulses (for me).
Very small pulse width (4.5us) and the second sample pulse is very near to the first pulse (about 20us).
I don't know if there is any reason to be so near the pulses at the underwater detectors.
Does anyone know it, so to tell me???
Regards

[Geo]

Hi porkluvr, thanks for the file.
I will try it tomorrow. I have the pcb from barracuda but i used a LM-555 at 780 Hz for Tx pulses....(110us), a bc-558 transistor to reverse the pulses and after it to a 2x 4538. The first has a delay of 30us and after it the second has a delay about 200us. I suply all this with 0...+12V so after the 4538 i use 2 transistors to convert the pulses to -5V....+5V. Until now all are OK. But when i connect the sample pulses to the "Fet switch" the pulse goes to -1....+2V so can't turn the Fets on and Off. I can't understand why????
I use 2N3819. I have similar problem if i connect the pin10 (out) of the 4538 direct to the Fets......
Any idea........
Regards

[porkluvr]

my mistake

When i load the Barracuda!+TGT.asc and run simulation i take a message that says "could not open library file "FMMTA14.sub"
What i can do????
Best Regards

Whoops - that's because I neglected to include the FMMTA14.sub file with that little package. Here it is. Copy this text into a text editor and save it as:
C:\Program Files\LTC\SwCADIII\lib\sub\FMMTA14.sub (making any appropriate changes to the path name). That should do it.
************************************************** *******************
*Zetex FMMTA14 Spice Model v1.0 Last Revised 12/6/2002
*
.SUBCKT FMMTA14 1 2 3
* C B E
Q1 1 2 4 SUB38C
Q2 1 4 3 SUB38C 12.75
*
.MODEL SUB38C NPN IS=1.1E-14 ISE=7.1E-15 NF=1.012 NE=1.4758
+BF=220 IKF=.12 BR=15 IKR=.05 RE=1.3 RC=.5 RB=.3 VAF=150
+CJE=14.5E-12 CJC=4.14E-12 VJC=.515 MJC=.26 TF=1.15E-9 TR=75E-9
.ENDS FMMTA14
*
*$
*
* (c) 2005 Zetex Semiconductors plc
*
* The copyright in these models and the designs embodied belong
* to Zetex Semiconductors plc (" Zetex "). They are supplied
* free of charge by Zetex for the purpose of research and design
* and may be used or copied intact (including this notice) for
* that purpose only. All other rights are reserved. The models
* are believed accurate but no condition or warranty as to their
* merchantability or fitness for purpose is given and no liability
* in respect of any use is accepted by Zetex PLC, its distributors
* or agents.
*
* Zetex Semiconductors plc, Zetex Technology Park, Chadderton,
* Oldham, United Kingdom, OL9 9LL
************************************************** **********************

It produce very good pulses but now i have problem with the match of the pulses. You Know.... difference voltage (not symetrical at the 453, difference phase (180 degree reverse),etc. But never mind..... i will match it.

About non-symmetrical pulses: When you use the diode equalization trick with CD4538 or HC221, if you do not allow the "tpw" capacitor to recharge between pulses you will get short and long pulses instead of having all the same.
FOR INSTANCE: In Carl's Hammerhead, I increase the 330Ω resistor (R46A, in the July 2006 schematic) to 360Ω, because otherwise - when the secondary delay pot is set for shortest delay N3 and N4 may not be equal (in my simulation).

Alternatively, you can use seperate RC timing components for each channel if you want the shortest delay between each channel, but then your individual channels will probably have worse matching between their sample pulses.
Please show me your specific circuit if you are still having a problem

[Geo]

Hello GEO. I am sorry you could not get Barracuda to work. I am sure Barracuda is excellent underwater detector, but I think better options exist for land based searches. According to the Anchor website, the Barracuda setup has the attention of its designer, each and every time it is reproduced. And, that is with an accurate PCB and components. That does not shout "clone me from a buggared schematic".

Here, I have posted basic timing startup simulation for Goldscan 4 and also Hammerhead timing. As long as we have -reasonably accurate- models I think that simulation is an extremely useful tool.

I have found that, although it may be a simple matter for me to get the simulations to work, it is not so easy for someone else to set-up what I try to share. I try to supply all you will need, but setup will need your careful attention. Look at the .jpg for hints about finding missing components and filling in necessary parameters.

If there is something that you can't work out then please say so and I will try to help. I try to include all the "overhead" necessary to make the simulations work, but I can't think of everything.

Working with either the Goldscan 4 or the Hammerhead; each uses diodes in their timing networks so that one set of "RC" components can serve for multiple one-shots. This provides for symmetrical timing among multiple channnels (with the stipulation that adequate setup time exists between individual channels). In other words, several channels can have the same sample-width, but recharge time must be allowed from channel-to-channel.

You may notice that I call some model files ".txt", ".lib", or ".sub". It makes no difference what they are called. you may rename them as you like.

Hello porkluvr. Thanks for your files and interesting. There is a big problem with the timing so i constructed a small pcb with one 555 and two 4538. It produce very good pulses but now i have problem with the match of the pulses. You Know.... difference voltage (not symetrical at the 453, difference phase (180 degree reverse),etc. But never mind..... i will match it. The problem is that at the final... it will be a difference detector (litle from Barracuda....a litle from Whites.... a little from me ...etc)
Ohhh something else. When i load the Barracuda!+TGT.asc and run simulation i take a message that says "could not open library file "FMMTA14.sub"
What i can do????
Best Regards

[Geo]

hello Geo,
See my schematic in post 45, it work ok with Barracuda,
have nice work
Alexis.

Hi Alexis. I tried your schematic at LTSpice before to construct it. As porkluvr said the timing signal is not good (assymetrical), and has very small width (if i remember good the width of the sample1 pulse is only 4.5 us), I need a width pulse of 30us or, if the delay time is short, at least the same with the delay time.
Regards

[porkluvr]

other PI

Hello GEO. I am sorry you could not get Barracuda to work. I am sure Barracuda is excellent underwater detector, but I think better options exist for land based searches. According to the Anchor website, the Barracuda setup has the attention of its designer, each and every time it is reproduced. And, that is with an accurate PCB and components. That does not shout "clone me from a buggared schematic".

Here, I have posted basic timing startup simulation for Goldscan 4 and also Hammerhead timing. As long as we have -reasonably accurate- models I think that simulation is an extremely useful tool.

I have found that, although it may be a simple matter for me to get the simulations to work, it is not so easy for someone else to set-up what I try to share. I try to supply all you will need, but setup will need your careful attention. Look at the .jpg for hints about finding missing components and filling in necessary parameters.

If there is something that you can't work out then please say so and I will try to help. I try to include all the "overhead" necessary to make the simulations work, but I can't think of everything.

Working with either the Goldscan 4 or the Hammerhead; each uses diodes in their timing networks so that one set of "RC" components can serve for multiple one-shots. This provides for symmetrical timing among multiple channnels (with the stipulation that adequate setup time exists between individual channels). In other words, several channels can have the same sample-width, but recharge time must be allowed from channel-to-channel.

You may notice that I call some model files ".txt", ".lib", or ".sub". It makes no difference what they are called. you may rename them as you like.

[porkluvr]

Hello Alexis

hello Geo,
See my schematic in post 45, it work ok with Barracuda,
have nice work
Alexis.

Hello Alexis. You did nice work with the Goldscan 4, but I tried to simulate what you posted here about Barracuda and I saw only nonsymmetrical timing that I could not repair. Maybe I was in error, but are you sure about the wiring? Please double check. I will too.

If you can use LTSpice, I could send you my reconstruction so that you can understand my reasoning?

hello Geo,
See my schematic in post 45, it work ok with Barracuda,
have nice work
Alexis.

[Geo]

I made an other pvb with 2 IC's 4538 for the sample signal.
Until now i have not good results.
Today i have visitors at home, so at night or tommorow i will make the new experiment.
When it will work i am afraid that it would be a new detector

[Geo]

Hi All.
I just finished with the painting of my Lab and now i am ready to try some detectors.
For the first i tried the Barracuda with timing as porkluvr wrote at #49.
Pulses are good but...... there is a big mistake. Pulse of delay1 (first sample) don't begin after the finish of the Tx pulse but when Tx pulse begin.
It is not possible to work the detector with this timing.
I played with the LTSpice simulator without results.
So i will try with a difference circuit for sample pulses.

[porkluvr]

whoops-almost forgot...

Sorry folks, I left out some important information...
When you open Barracuda!+TGT.asc or Quick Barracuda.asc, you will probably get several error messages stating something like: "... can't find IRF9530 asy".

You need to use the PMOS symbol that I have included. To add:
click the "Add Part" icon (the one with an AND gate symbol) and navigate to the folder that you unzipped. Use the PMOS symbol. After placing the PMOS symbol,
right click on it to open its attribute box, and edit it as shown in my new picture.

Better still, is to start a new folder inside the LTspice installation folder: C:\Program Files\LTC\SwCADIII\lib\sym\3RDP. Inside the 3RDP folder you can create subfolders
to organize different types of parts. Place the *.asy parts that I supplied in a ...\sym\3RDP\WHATEVER folder. Put the complete "CD4000 asy" folder in \3RDP, also.
Do the same with xopamp.asy and LF442A.asy. Something like \3RDP\amps would be good. Something like \3RDP\POWER would be a good place for the LTC1044A.asy symbol
because it is a modified version of what was supplied with LTspice.

I suggest that you rename "CD4000 asy" folder to CD4000; I changed it and should not have. Change it back.

You will have to restart LTspice before the new symbols will be available.

You will also need to place LF442A and the CD40106 in the schematic. After placing, right click the CD40106 and edit it like in my picture. The LF442A should be OK.
It was created from the Xopamp.asy symbol and needs no further editing, I think. Once you place and edit a part once, you can copy and paste it many times.

If you leave all model files in the same folder as the simulation.asc file, then simply use the ".inc whatever.sub" statement (already present) without indicating "drive:\pathname\".

It's hard, I know - but I try to make this easy. If I left something ELSE out, please let me know. I'm sorry, but I lied when I said this was for everybody.
I forgot that you've gotta have more than a little geekness or you will give up too soon.

I'll try not to bother you any more with this unsolicited junk. Good luck to yous.

[porkluvr]

Barracuda simulation file

Here I have posted a ZIP file that includes LTspice simulation file of a (simplified) Barracuda circuit (as best as I can determine) and all models and overhead required for a reasonably intelligent person who can operate a computer to be able to run "straight out of the box". Of course, you will need to have LTspice installed on your PC. It is freeware available at: http://www.linear.com/designtools/software/.

One more stipulation is that you should also insure that spice models for 2N5484 and 2N5485 are available. I have included them, and stated where they should be put. See the standard.jft.txt file for easy instructions. They may already be included with the LTspice download.

It is good that somebody took the time to reverse engineer this detector, but I think there were several important areas where mistakes were made. I do not believe that a detector made from KT315's drawing or from the subsequent revisions, would work effectively. No offense intended to KT315, but whoever gave you that schematic made some serious errors. There were nonsensical or illegible parts values, a reversed amplifier (U2A), and, what else... ? Oh yes: no, that resistor (R19) cannot go to V- (jeez Louise)! I don't want to throw stones, because I might not have done any better.

Fortunately, once given a starting point (thanks to KT315) and enough time - it is sometimes possible to adjust a circuit through trial and error until a working circuit is found. I believe that is what I have done, but, I could not perfect the timing and synchronization the way I would want it. I do not know what the original design was, but I am pretty sure that It was not what I started with. So, what I have derived may be functional - but I could not tweak this circuit to where I would say that it is elegant; that is beyond my resources and enthusiasm.

I leave further tweaking to you, the die-hard builder. That's not me, because although I respect this circuit's simplicity, I would prefer a PI with more adjustment features. If you are going to carry a PI detector around all day, I believe it should be one with a better adjustability range than what I could make of this one. The Barracuda is durable, and good to extreme depths, but, with what I know - I was unable to hack this circuit to perform at a level where I would want to construct a clone. I don't like my secondary delay period adjustment and I don't want to play with it any more.

Have some of you have already built this, but it did not work? Maybe now, you can get it to work. Please, don't blame me if if doesn't perform to your expectation. And don't blame Anchor company. For somebody who has not already built a Barracuda clone I would say: this circuit is unproven. Why bother? There are other proven designs available for hacking. I think the best features of the Barracuda are its low cost and waterproof durability. It works great underwater and on beaches where you do not want or need a bunch of adjustments! The manufacturer knows for sure what the circuit is and how to tune it. I don't, and you probably don't either.

If you will scuba, then buy a Baracuda, you will be very happy. But, if you hunt on land; until we can get a better knowledge of the Barracuda circuit (and that means somebody taking one apart and telling us what is REALLY inside*), then, build a Hammerhead (or one of the others)! (Just my two cents worth.)

I think my package should work, but if somebody has trouble simulating this I will help if you ask. (Oh, but please consult the help files, first.)

*Please supply timing diagams and photos if the circuit is really strange

[nelson]

HI GEO, CAN YOU TELL US A LITTLE MORE ABOUT YOUR EXPERIENCIES WITH BARRACUDA.

BEFORE THIS POST I HAD PLACE ANOTHER WITH SOME QUESTIONS THAT I HOPE YOU OR SOMEONE ELSE CAN ANSWER ME.
I LL BE WAITING FOR THAT, BEFORE I START THIS PROJECT.

BEST REGARDS

NELSON

Thanks KT...
It is time to check the pcb. Now my Barracuda is ready, when i will take the ICL7660 (i have order it) i will try it. It is a classic schematic so i am sure that i will find the errors and to make it to work perfect ( because as i wrote Barracuda was one of the best PI detectors).
Regards

......porkluvr can you simulate again the circuit with editing the errors who KT315 sent here???

Regards

[nelson]

HI, CAN ANYONE CONFIRM PERFORMANCE OF THIS DETECTOR.

I HAD SAW THE REVIEW POST BELOW AND NOW I M CONFUSE, CAUSE SOME PEOPLE SAID THAT THIS DETECTOR DOESN´T WORKS HAS THE REVIEW TALKS ABOUT.
ALSO THE REVIEW SAID THAT THIS MACHINE IN SOME WAY CAN DISCRIMINATE THRU THE USE OF DIFERENT SOUNDS AND THAT CAN DETECT A DIME AT 7 INCHES.

PLEASE ANY COMMENTS OF PEOPLE THAT HAS BUILD THIS DETECTOR WILL BE VERY HELPFULL BEFORE I STARTED THIS NEW PROJECT,
REGARDS

NELSON

Im sorry, but has i understood this detector is the same show on this link review?

http://www.losttreasure.com/fieldtests/ArchiveView.cfm?ID=lt19950844

Please be more specific.

Thanks adn regards

Nelson

[porkluvr]

It's alive, it's alive.

Thanks Alex, I had -almost- figured out how to fix this thing and your timing values in the oscillator were just what I was looking for. Taking Alex's changes to the oscillator, I tried simulating his entire timer and PW circuit but ran into some problems. So, instead I try and fix the original circuit.

This is tricky because most of the adjustments that can be made are interactive, and somewhat unpredictable. I do not intend to build the Barracuda because I have unfinished Goldscan 4 and Tesoro "Golden Ticket" (ahem) that need attention. But I offer this for the benefit of anybody who has already built a Board using the existing schematic.

As I said before, the adjustments are interactive. It is easy to really screw up the timing, even to the point of making the primary sample pulse occur after the secondary. You don't want to do that, and it is HIGHLY RECOMMENDED to use an oscilloscope. Also, the PW inverters can fail to trigger for certain RC timing values. In particular, the secondary delay range of adjustment is limited.

Also please notice, and this is very important: the receiver circuit works on a negative input. The primary sample pulse must clock the inverting input of the second opamp. Otherwise a good target will cause a negative signal to the output transistor!! My previous post showed the two sample pulses in the opposite order. That can happen if you are not careful.

I will get this circuit simulation zipped and posted so it can be used without having to find third party models, but it will take me a couple of days. You, the reader, may want to see for yourself just how tricky this Barracuda can be to adjust.

I do not guarantee that my circuit is accurate, or even suggest that it is a good one - so just use it for reference. There is an improvised target return to exercise the receiver. It's not perfect; the operator/technician must make sure the "target return" (-) peak does not coincide with the secondary sample pulse.

I think Carl Moreland's Hammerhead is more versatile and easier to tune - but the Barracuda comes with a neat waterproof box if you buy one!