Replies to Minipulse Plus

[multieagle]

Am looking forward to this as my next project. I have a question regarding coil connections. Perhaps it is mentioned elsewhere in this thread, but have not stumbled across it whilst speed reading through the material. It relates to the three PCB connections. First, am I correct that "Screen" refers to the coil and cable shielding? My coils are constructed with a copper shield. The shield has a wire soldered to the coil ground wire, which is in turn connected to the cable shield. Thus at the PCB there are only two connections: shield to negative and central conductor to positive. Can this type connection be used with the Minipulse? Or must I used a two conductor cable from the coil, with the coil shield connected to the cable shield, then connected to the third (Screen) PCB connection?

Or, have I completely misinterpreted it?

[joop]

Just some info after a little test... with delay set to a minimum, which is 6 uS in my case, I need a 250k pot to reject pull tabs.

[Davor]

Davor have you built your time base ? how does the 4017 cycle through with the clock connected to the reset ?

Maybe I dont understand how a 4017 works,
the counters need to be reset at some point in the sequence,but if the 555 pulse is connected to the reset pin (15), (not the clock pin 14), the 4017 will reset the counters back to zero with each 555 pulse, how do the counters cycle if the 555 is continuously resetting the 4017 with each pulse.

Hmmm is U4 somehow running as a independent clock
how does this thing operate Davor ?

6666, I'm not ignoring your posts, it is only that your repeated questions kinda confuse me. I answered them already, but well...
For clarity, yes, there are actually 3 time bases. The main one is set by the 555 and it controls the rest of shebang. The Schmitt triggers are responsible for a second time base. I need a time base for samples, and as the samples are mere multiples, the timebase is constant for sample, twice the GB, and EF pulse. The third time base is a bit hidden, as it is the first pulse of a time base which is variable to account for different GB exponents. I designed it to operate as a ratio against the sample duration, and thus the GB setting remains valid for some range of sample durations.

We had a lot of bad weather here, and as my "lab" is a table under the stars, I can make something only if the weather is nice. Trust me, I'm already twitching in expectation to see this thing working.

[Qiaozhi]

Just a quick note on Rev C build document.
At Step 3 Q2 (2n3904) is not mentioned to be installed, but Rev B it is...so go with "B"...

I have updated the errata list for REV-C to include this information.

Although U2 was in the list of components in Step 3 of the REV-B document, it was removed for REV-C because of the sync problem (which has now been resolved).

[johnandles]

Just a quick note on Rev C build document.
At Step 3 Q2 (2n3904) is not mentioned to be installed, but Rev B it is...so go with "B"...

[joop]

Ran the time constant for an aluminum pull tab. 10.7usec.

Thanks, will keep that in mind.

[Qiaozhi]

George, could you please post a sort of specification on timings in form of spans; delay to first sample (min, max), first sample duration (min, max), delay to EF (min, max)
Thanks.

The main sample delay is set by (Reject pot [50k] + R6 preset [100k]) * C8 [220pF]. If the Reject pot is turned to zero, and you adjust R6 to lower its resistance, it should be possible to get the minimum main sample delay down to 6-7us, although the datasheet for the MC14538BCP specifies 10us as the lowest value. The maximum sample delay will depend on where you set the minimum delay. So, if R6 is adjusted to 45.45k to give a 10us delay, then (45.45k + 50k) * 220pF = 21us (max). If you want to have a bigger maximum delay (which would probably be a good idea), then increase the value of the Reject pot and/or C8.

The secondary sample delay is set by (R10 [10k] * C10 [10nF]), which equals 10k * 10n = 100us. This value is added onto the main sample delay to give a range from 110us to 121us (if the minimum main sample delay is set to 10us). Again, if you want the secondary sample delay to be at 200us (for example) then increase R10 to 20k.

Both the main and secondary pulse widths are set by (R11 [22k] * C11 [2n2]), which equals 22k * 2n2 = 48.4us. However, due to the diodes D4 and D7, the actual pulse width is 58us.

As you can see, the range of values depends on how you adjust the minimum main sample delay.

[green]

Pulltabs are still very loud. So I am thinking about making the reject pot much larger, maybe 250k?

Ran the time constant for an aluminum pull tab. 10.7usec.

[green]

A 0.47 gram nugget (7.25 grains) can be emulated with a 9mm x 9mm aluminium square from the side wall of a Cola can. Your 10mm x 10mm square of aluminium foil is much thinner, so I doubt you'll be able to detect it.

Thanks for the information. It's the first time I remember seeing a target size referenced to a nugget size. I recorded some data with the two targets. Did it with two different coils to see if the time constant stayed the same. Got .6usec time constant for foil and 2usec time constant for the can.

[joop]

A 0.47 gram nugget (7.25 grains) can be emulated with a 9mm x 9mm aluminium square from the side wall of a Cola can. Your 10mm x 10mm square of aluminium foil is much thinner, so I doubt you'll be able to detect it.

Thanks! I cut a 10x10 piece out of a alu can. I can detect it with a delay of 6-7 microseconds. My coil has an average diameter of 21 cm, so it might be a bit big for small stuff. I am pleased I can still detect the aluminium piece.

I think this is good test! The coil is 360 uH with a 280 Ohm resistor, which I optimised by watching the receiver opamp output. I will try and see if this test can also be used to optimise the coil resistor....

I also took a 15x15 piece of alu from the can, to see if I could reject it. With the 100k reject pot I can barely reject it at full scale, which is about 40 uS. Pulltabs are still very loud. So I am thinking about making the reject pot much larger, maybe 250k?

[Davor]

George, could you please post a sort of specification on timings in form of spans; delay to first sample (min, max), first sample duration (min, max), delay to EF (min, max)
Thanks.

[Qiaozhi]

Just wondering.... Should I be able to detect a small piece (1x1 cm) of ALU foil with the minipuse plus? It does not give a sound, even with the delay set below 10 microseconds. I believe I read somewhere that aluminium foil is a good indication for gold sensivity....

A 0.47 gram nugget (7.25 grains) can be emulated with a 9mm x 9mm aluminium square from the side wall of a Cola can. Your 10mm x 10mm square of aluminium foil is much thinner, so I doubt you'll be able to detect it.

[ivconic]

Just wondering.... Should I be able to detect a small piece (1x1 cm) of ALU foil with the minipuse plus? It does not give a sound, even with the delay set below 10 microseconds. I believe I read somewhere that aluminium foil is a good indication for gold sensivity....

I know i missed something to do, now i know what; to test it on Al foil, damn!
I will do that next time.

[joop]

Just wondering.... Should I be able to detect a small piece (1x1 cm) of ALU foil with the minipuse plus? It does not give a sound, even with the delay set below 10 microseconds. I believe I read somewhere that aluminium foil is a good indication for gold sensivity....

[ivconic]

Ooops! I can't post 2 videos in one reply!?

"...You have included a total of 2 videos in your message. The maximum number that you may include is 1...."

Ok, here is the second one (Rejection):