Getting ready to build the MPP

[Qiaozhi]

I got a similar result - the minimum delay the coil can be sampled at is now about 12uS, depth in air tests about the same - I'll elaborate a bit in another post...

It depends on what type of target you were using for the depth tests. Sampling earlier allows you to detect targets with very short time constants, such as small gold nuggets or fine jewellry; and you will find a difference in depth for these objects. Unfortunately, sampling earlier also increases ground response.

[KRinAZ]

I tried a coil with a higher inductance and it would not allow me to use a short delay and it didnt make it any better at longer delays (> 30us) although there was no appreciable decrease in performance.

I got a similar result - the minimum delay the coil can be sampled at is now about 12uS, depth in air tests about the same - I'll elaborate a bit in another post...

[CAS]

Hmmm, thx Dan, I haven't put a shell around the coil yet - will do that & rig up a pole when I've mostly finished the electronics mods & tuning. I'm curious about how this coil would measure & perform with another basketweave turn, & I am experimenting with a mod to the coil template anyway, so I think I'll make an 8 turn version of the 8x10 & post it's test results...I'll also find the minimum stable & test with that delay & include in the results...may be a couple days...

I tried a coil with a higher inductance and it would not allow me to use a short delay and it didnt make it any better at longer delays (> 30us) although there was no appreciable decrease in performance.

[baum7154]

Hmmm, thx Dan, I haven't put a shell around the coil yet - will do that & rig up a pole when I've mostly finished the electronics mods & tuning. I'm curious about how this coil would measure & perform with another basketweave turn, & I am experimenting with a mod to the coil template anyway, so I think I'll make an 8 turn version of the 8x10 & post it's test results...I'll also find the minimum stable & test with that delay & include in the results...may be a couple days...

----------------------------------------------

By an '8 turn version' do you mean that you are going to have 8 wires in each of the slots or about 40 turns on the coil? That should put the coil well in excess of 450uh and I don't think it would decay in under 10us, just guessing. If you take a look at my post #31 in the CHANCE PI COIL thread the attached pictures show an 8" round coil with 39.4 turns of wire yielding a coil with 458uh of inductance. This coil will not allow my CHANCE PI to operate down at even 10us so a larger 8X10 elliptical coil would have even more inductance with the same number of turns.

By the way the increase in inductance after putting a shell or armor on the coil I believe comes from slightly crushing the open coil geometry with the wraps and is almost unavoidable. I have a couple of coils that were used in development of my 3:1 oval or RACETRACK style and after much handling un-armoured, they exhibited an increase of inductance I attribute to crushing. The change was from about 310uh to 325uh or about 5%. I am in the process of armoring with fiberglass a 4" X 12.5" 3:1 'Racetrack coil' and a picture of it ready for fiberglass is in post #232 of the CHANCE PI COIL thread. It now measures 328uh.

Best of luck,

Dan

[KRinAZ]

I would not add windings but instead push the sampling delay toward 6 or 7us instead. There is little to be gained I believe in slowing down your coil further. You have plenty of inductance at 324uh. One question is, have you armored your coil yet? If not, it will probably have the effect of slowing down your coil anyway.Regards,Dan

Hmmm, thx Dan, I haven't put a shell around the coil yet - will do that & rig up a pole when I've mostly finished the electronics mods & tuning. I'm curious about how this coil would measure & perform with another basketweave turn, & I am experimenting with a mod to the coil template anyway, so I think I'll make an 8 turn version of the 8x10 & post it's test results...I'll also find the minimum stable & test with that delay & include in the results...may be a couple days...

[KRinAZ]

Something for thought. A chart I posted in another thread awhile back. I kept the peak coil current constant for the test. Adding turns and increasing inductance might lower peak current and reduce the gain increase by the current ratio. With the same peak current the signal increased with the turns ratio squared. Signal strength should equal ampere turns. With a mono coil turns helps transmit and receive(squared). Gain in distance vs signal strength varies with coil size and shape. The chart shows what I got with my 8 inch round mono coil.

Many thx green, excellent test & results presentation, & food for thought

[baum7154]

I would not add windings but instead push the sampling delay toward 6 or 7us instead. There is little to be gained I believe in slowing down your coil further. You have plenty of inductance at 324uh. One question is, have you armored your coil yet? If not, it will probably have the effect of slowing down your coil anyway.

Regards,

Dan

[green]

So I'm thinking of another version of the basketweave coil - since it seems to be achieving what we're all wanting from a fast coil - sub 10uS recovery and sampling. I was surprised that it achieved 5uS - this is just past the knee in recovery where the pulse flattens out. I tweaked the MPP for minimum delay of 9uS and it did give response to small lead targets that get no response at original 20uS delay. But better sensitivity to small low conductive targets would be nice, so thinking of more windings for the next coil. Here's the state of my RevC MPP: did mod's to: add MUR460 diode to Tx circuit; swap R29 for 1M pot with 15K series resistor for SAT adjust; swap R33 for 5K pot with 180 ohm series resistor for sensitivity adjust; adjust R6 for 9uS sample delay minimum & replace 50k delay pot for 100k - delay range is now 9uS to 38uS; swap R11 original 22K resistor to 6K8 resistor for 15uS sample pulse width & better sensitivity; add 100nF decoupling caps to U6 & U7 as per the RevD MPP. So - since I am achieving sub 10uS sampling with these mods & this coil (the 8x10 basketweave I gave specs for in a previous post), & since the coil can be sampled at 5uS, & given that my MPP will adjust down to a 7uS delay (staying at 9uS may be more stable though), then it seems I have room to add windings and "slow" the coil to 7uS or 9uS this MPP is capable of, & try to get some added sensitivity to small gold. - But, this [building a new coil & template] takes a lot of time so I thought I'd put this out to you all first to see if there are flaws in this logic / approach, or if I should be doing some other mod first / instead. I look forward to any thoughts you all have!

Something for thought. A chart I posted in another thread awhile back. I kept the peak coil current constant for the test. Adding turns and increasing inductance might lower peak current and reduce the gain increase by the current ratio. With the same peak current the signal increased with the turns ratio squared. Signal strength should equal ampere turns. With a mono coil turns helps transmit and receive(squared). Gain in distance vs signal strength varies with coil size and shape. The chart shows what I got with my 8 inch round mono coil.

[KRinAZ]

So I'm thinking of another version of the basketweave coil - since it seems to be achieving what we're all wanting from a fast coil - sub 10uS recovery and sampling. I was surprised that it achieved 5uS - this is just past the knee in recovery where the pulse flattens out. I tweaked the MPP for minimum delay of 9uS and it did give response to small lead targets that get no response at original 20uS delay. But better sensitivity to small low conductive targets would be nice, so thinking of more windings for the next coil. Here's the state of my RevC MPP: did mod's to: add MUR460 diode to Tx circuit; swap R29 for 1M pot with 15K series resistor for SAT adjust; swap R33 for 5K pot with 180 ohm series resistor for sensitivity adjust; adjust R6 for 9uS sample delay minimum & replace 50k delay pot for 100k - delay range is now 9uS to 38uS; swap R11 original 22K resistor to 6K8 resistor for 15uS sample pulse width & better sensitivity; add 100nF decoupling caps to U6 & U7 as per the RevD MPP. So - since I am achieving sub 10uS sampling with these mods & this coil (the 8x10 basketweave I gave specs for in a previous post), & since the coil can be sampled at 5uS, & given that my MPP will adjust down to a 7uS delay (staying at 9uS may be more stable though), then it seems I have room to add windings and "slow" the coil to 7uS or 9uS this MPP is capable of, & try to get some added sensitivity to small gold. - But, this [building a new coil & template] takes a lot of time so I thought I'd put this out to you all first to see if there are flaws in this logic / approach, or if I should be doing some other mod first / instead. I look forward to any thoughts you all have!

[KRinAZ]

Well I tried 2 different diodes, they make no difference in the recovery of the pulse at TP3 - I tried the RM11B, then a RH1S (600v 1a trr=200ns) & even verified with my diode checker they are good, but the waveform is no different when either of these is added in series between Q1 & R3. I was expecting faster recovery and/or higher ohms for the damping resistor - the diode addition I did matches the RevD schematic. Any thoughts? Am I correct in what I'm expecting from the diode addition? Is the MUR460 going to make a difference over these diodes I tried? Thx for any ideas!...

OK, I put the MUR460 in the circuit today and the short answer is yes, it makes a big difference - now when adjusting for critical damping I had just a bit of overshoot at 2K2 and when I went to 2K1 there was no hint of ring, no overshoot, so I left it there. Signal stable & can be sampled starting at 5uS, very nice.

[KRinAZ]

Here are some pictures. Control box is not attached yet. Work in progress. The PCB is the MPP front end. I will built this in the coil housing. The plastic handle and connecting parts are developed with Autocad and 3d printed. Never used ACad before but the advantage of a project like this you obtain a few skills along the track. The PCBs are made in China and cost little more than an apple and an egg. Next challenge is to learn programming the MBED. Cheers, JCobb

[ATTACH]33170[/ATTACH][ATTACH]33171[/ATTACH][ATTACH]33172[/ATTACH][ATTACH]33173[/ATTACH][ATTACH]33174[/ATTACH][ATTACH]33175[/ATTACH][ATTACH]33176[/ATTACH]

Hmmm, nice job & thx for the pics, think I'll watch for a good deal on a monopod...

[Achillion]

Interesting, thx for the idea - how do you hold it to swing it - any pics?

Here are some pictures. Control box is not attached yet. Work in progress. The PCB is the MPP front end. I will built this in the coil housing. The plastic handle and connecting parts are developed with Autocad and 3d printed. Never used ACad before but the advantage of a project like this you obtain a few skills along the track. The PCBs are made in China and cost little more than an apple and an egg. Next challenge is to learn programming the MBED. Cheers, JCobb

[KRinAZ]

Something I posted awhile back. The fast diodes isolate the fet capacitance and increase the circuit resonance allowing a higher Rd. The scope pictures are across the coil with no Rd or input resistor. I find Rd=circuit resonance x inductance x pi is close. Don't know if the two examples used the same coil but a 4 times increase in resonance should allow a 4 times in Rd. The diode helps to sample sooner but with all else the same you loose a little signal due to voltage drop across the diode when charging the coil. Use caution, the high voltage can damage the scope when measuring across the coil.

thx green, I think I read your post on this last year. What you say about the effect of the function of the diode makes sense & I'm perplexed on why adding the diodes I tried had no effect in my circuit. I did just get my hands on a MUR460 though so I'll try it soon, maybe the diodes I tried were no good for this circuit...

[KRinAZ]

I'm using a carbon monopod (used for photo camera) which works well. It's strong and the lower shaft is carbon which is ok for metal detecting. Brand name is Benro.

Interesting, thx for the idea - how do you hold it to swing it - any pics?

[green]

& by the way, with no diode & on the bench anyway, it looks to me like the signal recovery is stable & could be sampled at about 8.5uS from when the coil signal just starts to collapse - again observed at TP3 - & right about 11.5uS from when the TX pulse at TP1 just begins to fall. Seems to me the important part is how long after the [TX] coil begins to collapse - can I start to sample the signal for targets & from when the coil signal just begins to fall/collapse at TP3 to when stable - I'm under 10uS now.

Something I posted awhile back. The fast diodes isolate the fet capacitance and increase the circuit resonance allowing a higher Rd. The scope pictures are across the coil with no Rd or input resistor. I find Rd=circuit resonance x inductance x pi is close. Don't know if the two examples used the same coil but a 4 times increase in resonance should allow a 4 times in Rd. The diode helps to sample sooner but with all else the same you loose a little signal due to voltage drop across the diode when charging the coil. Use caution, the high voltage can damage the scope when measuring across the coil.