SURF PI Delay Mods ..... ??

[Vortxrex]

My Surf Pi (Longboard) works best w/ the delay setting all the way down. (I am using a small gold nugget & a US nickle for targets).

There is a 15K resistor in series w/ the delay potentiometer. If I reduce this value to say 10K, won't that reduce the delay further? or am I going to damage the MC14093B if I reduce this resistance too far?

The delay potentiometer can be connected w/ 2 wires, so why can't we have two delay potentiometers & use a DPDT switch to TOGGLE back & forth between 2 different delay settings?

IE: Search w/ the minimum delay, then when you detect a target, you can quickly toggle to the second (longer) delay & see if you are still getting a good target signal, and if so its probably not a ring or coin.

Dave Frank

[taliesin]

My Surf Pi (Longboard) works best w/ the delay setting all the way down. (I am using a small gold nugget & a US nickle for targets).

There is a 15K resistor in series w/ the delay potentiometer. If I reduce this value to say 10K, won't that reduce the delay further? or am I going to damage the MC14093B if I reduce this resistance too far?

The delay potentiometer can be connected w/ 2 wires, so why can't we have two delay potentiometers & use a DPDT switch to TOGGLE back & forth between 2 different delay settings?

IE: Search w/ the minimum delay, then when you detect a target, you can quickly toggle to the second (longer) delay & see if you are still getting a good target signal, and if so its probably not a ring or coin.

Dave Frank

have you tried a 4011 instead of the 4093 ?

[Tepco]

Minimum delay you can achieve is determined mostly by coil parameters, (considering it is dumped properly), and minimum delay is very important to get sensitivity on small targets, jewelry peaces etc.Be free to reduce value of 15K resistor, up to the point when detector stops to respond completely (you wont blew up 4093 or something). At this point, integrator (first half of 358 chip, next to analog switch) will be in saturation, so set resistor for minimal value when detector can still respond, and you will get minimum sampling delay.

(exception is if you have very "fast" coil, then integrator offset may have to be adjusted using aditional parts to get minimum delay)

And you don't need any switch for this, just go as fast as possible under given conditions, and increase delay whit potentiometer when needed (useless control anyway)

[Vortxrex]

Hi Taliesin, I would like to try the 4011, what is the full part number of the one you used? I can detect a nickel at 13" & my little nugget at 4" This is w/ a 9.5" coil.

Tepco, I am going to drop the resistor value to 5K and see what happens. I still like the idea of just pulling a switch & going to a second "preset" delay to see if the target still sounds.

Dave Frank

[homefire]

Root around in this thread and you will find a few other things you may try.

http://www.geotech1.com/forums/showt...ghlight=xl+500

I did the same things to a XL-500 Garret.



I'll keeping a eye on this one. I bought a kit from SilverDog and had the same Ideas.

[taliesin]

Hi Taliesin, I would like to try the 4011, what is the full part number of the one you used? I can detect a nickel at 13" & my little nugget at 4" This is w/ a 9.5" coil.

Tepco, I am going to drop the resistor value to 5K and see what happens. I still like the idea of just pulling a switch & going to a second "preset" delay to see if the target still sounds.

Dave Frank

i used hef4011bp.it's a good idea the "preset".

[Vortxrex]

I dropped the resister value to 5.5K & added the switch & extra delay potentiometer in my second Surf PI. Here is what I found: the detector is most sensitive w/ the delay turned 3/4 of the way down (instead of all the way down) no surprise there. If the two delay settings are very far apart, you have to reset the threshold when you switch over otherwise it worked fine (takes a second for the circuit to stabilize when you switch over).

Homefire, enjoyed all the info in the Garrett post. Taliesin, I did put the HEF4011bp in my second surf PI, not sure if it made an improvement.

Dave Frank

[homefire]

Do you have a Scope?

On the Garret I was able to get sampling down to 7 or 10 uS.

[Tepco]

I dropped the resister value to 5.5K & added the switch & extra delay potentiometer in my second Surf PI. Here is what I found: the detector is most sensitive w/ the delay turned 3/4 of the way down (instead of all the way down) no surprise there. If the two delay settings are very far apart, you have to reset the threshold when you switch over otherwise it worked fine (takes a second for the circuit to stabilize when you switch over).

Homefire, enjoyed all the info in the Garrett post. Taliesin, I did put the HEF4011bp in my second surf PI, not sure if it made an improvement.

Dave Frank

If you want to totally NUKE your Surf, this is one bombastic timing modification: Reduce pulse width to 20uS, inter-sample delay to about 30uS, about same as samples (OK, now sampling time is longer than a pulse) and increase repetition frequency to 5-7kHz (OK, now you think i'm crazy, but let me explain). In total, value of 3 resistors and one capacitor have to be changed.

For what this is good? If pulse width is reduced 3 times, this will generate 3 times less coil current, producing 3 times less field intensity, and consequently 3 times less target signal. Except that, all ext. interference and internal detector noise will decrease signal to noise ratio 3 times. Not looking good at all. Except for one less obvious fact:

Decreasing pulse width 3 times will decrease energy stored in coil 9 times, less energy, easier to get rid off it (small decrease in delay), but most important, you can now increase pulse repetition rate 9 times, and still have same current consumption. As a result, detector can get and integrate 3 times more information from target (YUP!), and all noise will be integrated 3 times longer, so, there will be about 3 fold increase in S/N ratio.(This is a bit simplified but accurate enough)

Results: well, still will not detect Earth's core, but some 2-3 " increase in range for typical objects (only thing you need if you are not alone on the beach

Try to adjust one detector according to this and compare results, you will see what i'm talking about. Also worth trying whit SH500.Note: this is for beach machines only, this way you will actually loose some sensitivity for large objects (beercan sized or larger).

At first opportunity (when i get 2 surf boards clear) i will post more on this in separate topics, whit comparison data etc (my only surf at this point is epoxyed, sorry)

[Vortxrex]

I actually have one more Surf PI board & all the parts to build it with. I think it would be interesting to do some more mods with it. I have done some experimenting w/coils & offset, but still only getting about 12-13" on a US Nickel.

Dave Frank

[plumabob]

Great idea Vortrex,

I was curious so I dropped in a 7K resistor and my delay dropped from 12us to 10-11us on my scope. Coil went dead at the lowest settings but an adjustment on the damping resistor brought it back to life. Depth increased slightly ( 13-14 on a nickel ) but I saw a bigger change in sensitivity to my small gold test targets. I'm a salt water beach hunter so this is about the lowest usable delay setting I will care to try. Anything below 10us is relatively useless.

Great job!

[homefire]

Great idea Vortrex,

I was curious so I dropped in a 7K resistor and my delay dropped from 12us to 10-11us on my scope. Coil went dead at the lowest settings but an adjustment on the damping resistor brought it back to life. Depth increased slightly ( 13-14 on a nickel ) but I saw a bigger change in sensitivity to my small gold test targets. I'm a salt water beach hunter so this is about the lowest usable delay setting I will care to try. Anything below 10us is relatively useless.

Great job!

The coil is limiting you. Is it shielded?

[plumabob]

nah the coil is fine. I had to bump the damping resistor from 572 to 668 ohms. Coil is shielded.

Bob

[pelanj]

If you want to totally NUKE your Surf, this is one bombastic timing modification: Reduce pulse width to 20uS, inter-sample delay to about 30uS, about same as samples (OK, now sampling time is longer than a pulse) and increase repetition frequency to 5-7kHz (OK, now you think i'm crazy, but let me explain). In total, value of 3 resistors and one capacitor have to be changed.

For what this is good? If pulse width is reduced 3 times, this will generate 3 times less coil current, producing 3 times less field intensity, and consequently 3 times less target signal. Except that, all ext. interference and internal detector noise will decrease signal to noise ratio 3 times. Not looking good at all. Except for one less obvious fact:

Decreasing pulse width 3 times will decrease energy stored in coil 9 times, less energy, easier to get rid off it (small decrease in delay), but most important, you can now increase pulse repetition rate 9 times, and still have same current consumption. As a result, detector can get and integrate 3 times more information from target (YUP!), and all noise will be integrated 3 times longer, so, there will be about 3 fold increase in S/N ratio.(This is a bit simplified but accurate enough)

Results: well, still will not detect Earth's core, but some 2-3 " increase in range for typical objects (only thing you need if you are not alone on the beach

Try to adjust one detector according to this and compare results, you will see what i'm talking about. Also worth trying whit SH500.Note: this is for beach machines only, this way you will actually loose some sensitivity for large objects (beercan sized or larger).

At first opportunity (when i get 2 surf boards clear) i will post more on this in separate topics, whit comparison data etc (my only surf at this point is epoxyed, sorry)

Which resistors and capacitor would one have to change and what values would you suggest? I have one kit and one extra board waiting to be finished. Thanks!

[Tepco]

My detector is not exactly Surf, but similar, just works whit different timing. I believe this component values will be good for Surf and mod's will work on different boards and SM incarnations.

First, you have to change TX timing, whit these values you will have about 5.2kHz repetition and 18uS pulse. (components referred to "original" White's schematic, check what you have on your PCB version)

Change R2 (3K3), between 555 pin 6&7 to 1.2K (or 1.5K) this will shorten TX to 18 (22us). Change R1 (100K) between 555 pin 8&7 to 10K, to increase frequency.

Now play whit sample timing. In original, samples are very long (like 55-60uS), i think unnecessary long. (response from any meaningful target will reside in about half that time, this just picks up more interference keeping integrator on unnecessarily) And, time between pulses is now only 200uS, everything must be "packed" within, so reduce R28 (220K) from pin 13 4093 to+V to 68K.This will shorten inter-sample time to only about 25uS, and you will get sequence like this: TX 18-22uS, delay, coil dependent, as short as possible, 10-15uS (better below 10), 60uS first sample, about 25uS delay, 60uS second sample, and some headroom left before next TX to accommodate delay control. So just "critical", to avoid sample overlapping, or overlapping second sample and TX.

Now this will be almost inaudible at 5kHz (i have VCO in my detector), and best way to fix this problem is to removeR27 (7660 sync, if you have it), add something like 100pF between pin 7&8 or 7&3 7660 (may vary whit different chip manufacturers, adjust pitch you like, some board versions already have it whit unsynced 7660). Then disconnect pin 12 4066 from pin 11 4093, and connect it thru series connected 100nF cap and 10K resistor (values not critical) to pin 2 7660, and add 100K resistor between 4066 pin 12 and GND to get audio tone. Will work on both versions, when 7660 is powered from 5V like original, or from 12V and postregulated (better solution). Anyway, now 7660 will generate audio signal. I hope this will not introduce interference on some particular board designs.

You can go even faster, replacing R26 and R29 whit 27K (this will provide ~25uS sample time) and increase frequency even more (varying R1), and if current consumption is not an issue, increase pulse width (R2). This mod may look unusual, but my detector is set like this: 20uS TX, some 6uS (coil), 20uS first sample and 20 uS second one, no intersample delay at all, operating at 10kHz.And few remarks:

-It is handy to have an oscilloscope to check adjustments.
-Optimum dumping resistor value may vary somewhat due to shorter pulse.
-Compare detectors whit same coil and delay setting (always at minimum)
-If necessary, make integrator offset adjustment to prevent it from saturation at minimal delay.
-I don't guarantee performance of this modification in bad ground, but in lake or salt water works perfectly.
-You will notice actual decrease in "end range" for large objects, car body or sheet metal etc, but for jewelry-coin sized objects (considering you have fast enough coil) this mod will blow away about any commercial PI detector.
-If you like "softer" threshold, just replace MPSA13 whit small NMOS like BS170 or 2N7000, you can tune it to constant background tone, and interference will be less audible.

Good luck whit mods, and please report comparison results.
(I just got one old leaked SH for free, will try similar mod on it)