Tweet
dfbowers, how to order your coil enclosure to EU ?
-
-
That is a "sweet" machine, it does work beautifully. Very encouraging.Originally posted by dfbowers View Post
Yes, I see Ivconic's complete details diagram of coil cable wires, it is as you made it. I have been following a different diagram (three wire + shield), so that may be an important factor that can help me.
To confirm: is your TX "ground" wire connected to the shield at both ends of the cable, and your RX ground wire connected to the shield only at the coil head? (In some ways, it seems the TX "ground" wire is just an unnecessary hitchhiker alongside the shield wire if it is connected to shield at both ends.)
I agree measuring inductance through the cable is a wildcard, cable adds 100 to 200 pF capacitance plus some inductance. Also meters probably are between 5% - 10% in agreement typically. Also how I placed the coils can affect it, but I tried moving them some to check if significant. Tried to keep them from nearby metal, but never sure...
In a related subject, I measured my 6 foot Belden cable at 200 pF capacitance, and a 3 foot RCA cable at about 95 pF capacitance. I've noticed I get very different behavior with these two cables, so I'm now checking how critical cables and their connections are. I think this is one area where I may not have things quite right yet.
I hope others can make a video of their air tests like dfbowers did -- really helps to see what is possible and what to expect.
-SBComment
-
Originally posted by simonbaker View Post
Simon,
Yes, that would be correct, the Tx "ground" wire is effectively connected to shield in the coil AND to ground on the PCB.. It does look like a hitch hiker.
The Rx ground is connected ONLY in the coil.. At least that's the way I read Ivconics document and the way that I did it. Ultimately, it would be the connection that goes to Rx J2-2 that makes it's way to ground in the coil. I did however make a boo boo in my first coil that I cast in resin and got RX J2-1 hooked to ground. It SEEMS to work just fine either way, but the designer must have had a reason to pick J2-2. At least it doesn't seem to make much difference in detection range. But..wouldn't it make sense not to have it connected to ground at all so as to produce a true differential signal?
To mix things up even more, the shields add a few hundred pF more to the whole equation. Depending if you put the shields on the bare wire or not makes a difference too. But, not much difference with the final product.
I did notice that others preferred final testing and nulling with cable and shields installed since that will be the final configuration.
Another observation.. Listening to the audio of others testing gives us a good clue of coil inductance since it's a divided signal from the Colpitts oscillator. Mine seems a little lower than the videos that Ivconic has posted, which would support your measurements of the coils that I sent you..
I think that 25cm on a U.S quarter is a very realistic goal for this design. I double checked mine today and I can get a good solid signal at 25cm.
DonComment
-
Originally posted by dfbowers View PostI've been pondering that for some time with no conclusion. When I look at the RX signal it "looks" cleaner grounded (at least the grounded sideBut..wouldn't it make sense not to have it connected to ground at all so as to produce a true differential signal?
) -- but generally the output of the LF353 looks same both ways; I assume there is method to the madness, but don't understand it exactly.
I checked my Belden cable hookup and I guess I do in fact follow the same connections; definitely something to keep clear on though.
It is nice the way we can infer the TX frequency from the audio, another reason the videos are useful. Your MD is proof that the TX frequency has some latitude as long as you have a compatible RX coil and resonant frequency.
Thanks for test with US quarter -- I think that should be another universal benchmark. The thing about your MD is the nice clean signal you get at the max range. My clean signal with quarter is probably well under 20 cm (on the rare moment the stars align), so a ways to go. However, I really will need to do a series of test in a low-noise environment which will take some work to pull off.
I usually null looking at the output of LF353, so obviously cable is involved for me. But your technique clearly works well.
-SBComment
-
i get allmost same induction of coils like your's(from bowers) still about 22cm but on weekend i'll try to work wit trimers to change a value of freaq.
ther's a lot of work but step by step and we'll have great stuff to play on
best regards
Comment
-
In a commercial Tesoro coil there is no "hitch hiker" wire. The TX uses the shield as one of the connections.Comment
-
-
Originally posted by climb View Post
Hmm.. Haven't thought about selling any yet. Maybe if enough people want one I might make a few more.. I would have to know what the most popular inner dimensions of the vertical tabs are for mounting a pole and how deep to make the shells for the coils. I customized mine for one of the Tesoro rods and to fit the dimensions of my own coils. You would also have to find a suitable strain relief.
Plus, they don't come out perfect..
but good enough for coming out of the wifes oven!
Comment
-
Quick Test Coils
And I'm still in the same boat as lunamay.Originally posted by lunamay View Posti get allmost same induction of coils like your's(from bowers) still about 22cm but on weekend i'll try to work wit trimers to change a value of freaq.
ther's a lot of work but step by step and we'll have great stuff to play on
best regards
I did a quick test with dfbowers coils and my latest PCB. The coils null very well and no beeping. I do not magically get better depth; too bad. However I know my RX coil capacitor C6 is around .013 uF and could use some tuning.
So from quick test, it seems maybe that my TGSL circuit is not working optimally. This is a very helpful result thanks to having some proven coils to work with. But I won't jump to conclusions, as I need to really check out some things. Not even sure I have the coils arranged perfectly.
QUESTION
I noticed something about nulling that I'd like to get some more help on. With dfbowers coil and also my adjustable coil, it seems that the correct null phase is obtained on the "outside" of the null, in other words, when the coils are farther apart. I thought that Qiaohzi mentioned that usually the correct phase is obtained on the side where the coils have more overlap.
I believe I have the correct phasing of the coils because if I flip one over, I get the double-beep with the target.
I interpret the correct phase as being the one where the voltage on the capacitor C15 is positive, and the TX and RX signals look sort of 180 deg out of phase.
If you have a small enough null voltage, a slight negative voltage at C15 is probably OK, but you risk the JFet becoming forward biased if you go too negative. So you probably could pick the "in phase" relationship if you get better detection with that nulling -- but not what I would prefer.
What do other people find regarding this when nulling their coils? Which side of null do you choose and which has more overlap?
-SBComment
-
Well, like I said, divide and conquer.. You may have a good set of coils that you made and not know it yet. How does the phase relationship look between J1-1 and U101a pin 7? In the image attached I used the same coils, but misadjusted slightly to show phase relationship.Originally posted by simonbaker View Post
I remembered picking though all my 15nF caps until I found the one closest to 15nF. Maybe next thing to try is adding 2 nF or so to bring resonance a little closer?
Also, when I nulled my coils I tried to keep the phase at 20 degrees, not 220, but I think the circuit would work OK either way.
Maybe we can get an idea of phase relationships by how your GB and disc pots need to be set. With mine, it will GB with the GB pot set to almost exactly midrange (minimal response to a ferrite rod on pin 5 U106b). Foil will disc out with the disc pot set at midrange. ( I need to find out if it's a linear pot or not).
Simon.. Do you have a list of your component selections? Maybe we can compare.
DonComment
-
19mm is a common inner dimension for the stem fixing.Originally posted by dfbowers View Post
If the coil inductances are different to the expected values then the correct null could possibly be on the other side.Originally posted by simonbaker View Post
20 degrees is the correct phase relationship for a DD coil, and 200 degrees for a concentric.Originally posted by dfbowers View PostComment
-
simonbaker could you send me on pw some parameterson photo?Comment
-
Originally posted by lunamay View Post
Hi.. That coil pattern was posted by Pip3c
http://www.geotech1.com/forums/showt...12692&page=168Comment
-
Ok, this is good stuff.Originally posted by dfbowers View Post
---------
NULLING
I definitely have always chosen to null to the 220 deg type of relationship using J1 and U101 pin7 (different to your setup), mainly because I thought that was what we were all told to do, going way back. Maybe I got fooled by some photos for concentric coils
.
However there is some logic to the 220 deg choice. If instead you null to the 20 deg, you tend to bias the detector JFet source pins negative, which doesn't seem like a desirable effect, since it increases the chance of forward-biasing the JFet.
On the other hand, choosing the side with more overlap should improve the geometry of the magnetic field stuff.
Actually, with coils I test, I don't notice a significant difference in detection depth on either side of the null. Of course, maybe the difference would be there if everything else about my MD were perfect. But my theory is it shouldn't make a huge difference, although I'm certainly open to seeing the proof in the pudding.
With some coils, I actually once ran into forward-biasing the JFet when I turned the Disc pot, so the null point can matter sometimes. With dfbowers coils and my adjustable coil, regardless of which null side I choose, the voltage on the JFet source pins never goes much negative.
So maybe I'll test more with nulling with the 20 deg relationship and see what I find.
---------
DISC
I usually test discrimination with a US Nickel, and my latest rejects the nickel a little over half-way along the linear pot range. That part seems to work well. My disc phase range is about 80 deg, that seems acceptable. Theoretical range I believe is 90 deg.
--------
GB
My ground balance circuit is maybe a problem. First of all, its range is maybe 30 deg (theoretically it should be 90 deg I think, I'd like to see 50 or above).
I get quite shallow detection of my ferrite ring target, but the GB pot seems to reject it at about half-way I would say. It is a little uncertain where it happens, there is some randomness, maybe noise.
For now I test with the GB wide open (no rejection) just to figure out this depth problem.
-------
PARTS
My components are mostly bought online from discount sellers like Electronic Goldmine, All Electronics, HSC Electronic Supply, Circuit Specialists Inc or Fry's Electronics and Radio Shack locally. I'm cheap.
My resistors are fairly good quality. Capacitors may be questionable. ICs may be very questionable. Pots - not the very best quality, and I think for a working field detector I would upgrade to premium pots for sure.
My RX cap C6 soldered in is low, about 13 pF, but I soldered in an IC socket slice for adding caps. I don't know how much noise that might add, but I thought the coil noise overwhelms circuit board noise -- an assumption I may revisit.
My oscillator is running just under 14 kHz, maybe 13.8 to 13.9 kHz, even with dfbowers coils. The capacitance measured on board around C2 is about .023 uF. C1 is .22 uF. C1 and C2 in series calculates to about .0208 uF, which with dfbowers TX coil of 6.38 mH calculate a resonant frequency of 13.8 kHz, so it makes sense (I guess).
I do not use back-to-back 10 uF caps for C14 and C17, rather I use single 4.7 uF nonpolar caps.
Instead of two BC547 transistors in the audio circuit, I use a single Darlington MPSA13 transistor -- given how critical Ivconic found the audio circuit (puzzling why), maybe I should fool with using separate transistors. I don't have exact BC547s, just 2N2222.
Other than that I stick with the original parts.
Now I did add the custom voltage divider circuit between pin J1 and the GB and Disc circuits as mentioned previously. It creates a small phase shift, like 3 to 5 degrees, but it seems negligible to me, and I don't think it causes a loss of sensitivity somehow -- but I'll probably try bypassing it as a test also.
--------
COIL
One thing I've notice is Ivconic and dfbowers and others have coil start and end in the middle of the straight section, whereas I always put it at the corner. However I don't see yet that this should matter much.
---------
I will continue some other observations in another message.
-SBComment
-
For now I test with the GB wide open (no rejection) just to figure out this depth problem.Originally posted by simonbaker View PostOk, this is good stuff.
---------
NULLING
I definitely have always chosen to null to the 220 deg type of relationship using J1 and U101 pin7 (different to your setup), mainly because I thought that was what we were all told to do, going way back. Maybe I got fooled by some photos for concentric coils.
However there is some logic to the 220 deg choice. If instead you null to the 20 deg, you tend to bias the detector JFet source pins negative, which doesn't seem like a desirable effect, since it increases the chance of forward-biasing the JFet.
On the other hand, choosing the side with more overlap should improve the geometry of the magnetic field stuff.
Actually, with coils I test, I don't notice a significant difference in detection depth on either side of the null. Of course, maybe the difference would be there if everything else about my MD were perfect. But my theory is it shouldn't make a huge difference, although I'm certainly open to seeing the proof in the pudding.
With some coils, I actually once ran into forward-biasing the JFet when I turned the Disc pot, so the null point can matter sometimes. With dfbowers coils and my adjustable coil, regardless of which null side I choose, the voltage on the JFet source pins never goes much negative.
So maybe I'll test more with nulling with the 20 deg relationship and see what I find.
---------
DISC
I usually test discrimination with a US Nickel, and my latest rejects the nickel a little over half-way along the linear pot range. That part seems to work well. My disc phase range is about 80 deg, that seems acceptable. Theoretical range I believe is 90 deg.
--------
GB
My ground balance circuit is maybe a problem. First of all, its range is maybe 30 deg (theoretically it should be 90 deg I think, I'd like to see 50 or above).
I get quite shallow detection of my ferrite ring target, but the GB pot seems to reject it at about half-way I would say. It is a little uncertain where it happens, there is some randomness, maybe noise.
For now I test with the GB wide open (no rejection) just to figure out this depth problem.
-------
PARTS
My components are mostly bought online from discount sellers like Electronic Goldmine, All Electronics, HSC Electronic Supply, Circuit Specialists Inc or Fry's Electronics and Radio Shack locally. I'm cheap.
My resistors are fairly good quality. Capacitors may be questionable. ICs may be very questionable. Pots - not the very best quality, and I think for a working field detector I would upgrade to premium pots for sure.
My RX cap C6 soldered in is low, about 13 pF, but I soldered in an IC socket slice for adding caps. I don't know how much noise that might add, but I thought the coil noise overwhelms circuit board noise -- an assumption I may revisit.
My oscillator is running just under 14 kHz, maybe 13.8 to 13.9 kHz, even with dfbowers coils. The capacitance measured on board around C2 is about .023 uF. C1 is .22 uF. C1 and C2 in series calculates to about .0208 uF, which with dfbowers TX coil of 6.38 mH calculate a resonant frequency of 13.8 kHz, so it makes sense (I guess).
I do not use back-to-back 10 uF caps for C14 and C17, rather I use single 4.7 uF nonpolar caps.
Instead of two BC547 transistors in the audio circuit, I use a single Darlington MPSA13 transistor -- given how critical Ivconic found the audio circuit (puzzling why), maybe I should fool with using separate transistors. I don't have exact BC547s, just 2N2222.
Other than that I stick with the original parts.
Now I did add the custom voltage divider circuit between pin J1 and the GB and Disc circuits as mentioned previously. It creates a small phase shift, like 3 to 5 degrees, but it seems negligible to me, and I don't think it causes a loss of sensitivity somehow -- but I'll probably try bypassing it as a test also.
--------
COIL
One thing I've notice is Ivconic and dfbowers and others have coil start and end in the middle of the straight section, whereas I always put it at the corner. However I don't see yet that this should matter much.
---------
I will continue some other observations in another message.
-SB
Probably a safe thing to do.. but may want to do some test with GB set like its used so as to get a like comparison. How does your circuit respond in the AM mode?
My RX cap C6 soldered in is low, about 13 pF
You meant 13 nF for C6, right?
I do not use back-to-back 10 uF caps for C14 and C17, rather I use single 4.7 uF nonpolar caps.
I built 2 TGSLs, one with back to back 10uF caps, one with single 4.7uF cap for each channel.. I can't tell any difference in audio or detection range. Anyone else know if there is a difference?
Now I did add the custom voltage divider circuit between pin J1 and the GB and Disc circuits as mentioned previously. It creates a small phase shift, like 3 to 5 degrees, but it seems negligible to me, and I don't think it causes a loss of sensitivity somehow -- but I'll probably try bypassing it as a test also.
I didn't see the thread on using a voltage divider. Does it provide some benefit?
One thing I've notice is Ivconic and dfbowers and others have coil start and end in the middle of the straight section, whereas I always put it at the corner. However I don't see yet that this should matter much.
I pondered the differences too. Most commercially made coils have the cable exit somewhere away from the center. There must be a reason. Also, I don't know if it makes any difference on where the gaps in the shields go.. it just seemed to be a convenient place for the wires to exit.
Anyone have any ideas?
DonComment
Comment