Impulse 2

[dbanner]

I have a question. Can improvements be made to the tx coil driver circuit without having to alter the code?

[Qiaozhi]

thanks.
However, I could not find where label "D", the pre-amp output, goes.
Guessing into a PIC ADC, maybe pin 12?

Thanks for spotting that. Any other issues, please let me know.
Version 1.1 attached.

[waltr]

Here's the schematic of the Impulse2 PI (see PDF attachment).

thanks.
However, I could not find where label "D", the pre-amp output, goes.
Guessing into a PIC ADC, maybe pin 12?

[6666]

Thanks Q .

[Qiaozhi]

Schematic of Impulse2 PI

Here's the schematic of the Impulse2 PI (see PDF attachment).

I originally tried to match the reference designators to the PCB, but ended up with some conflicts. Also, there were some components with no associated silk screen, so I gave up and auto-numbered the symbols.

[bernte_one]

mine works good enough, what did you expect from such a small circuit?

by playing with pulserate und delay Timing you can get small Separation of iron based objects and non-fe Targets
tried with a 35cm diy monoloop coil

i gues i can achieve better results with basket or planar coil

but it is not on me to make promotion for this detector, the designer should do this

[qute1]

nice to know hope you had a good holibob.

[Qiaozhi]

Any new on mods or improvements for the impulse?

Not yet. Just got back from holiday, and trying to find time to get Kicad up and running.

[qute1]

Any new on mods or improvements for the impulse?

[dbanner]

The next obvious step is to see whether a new front end design can be developed that plugs right into the same micro, if that is at all possible. Maybe not.

[Tim]

Thanks all for your interest in this, especially Qiaozhi for his effort.

[dbanner]

Yes, I've back-engineered the schematic. Since seeing gtdavid's efforts, I wanted to create the schematic using KiCad. The Software Manager for Linux Mint had version 4, but I noticed that gtdavid was using version 5. Consequently I now have version 5.0.0.-rc1-44a33f2-62-ubuntu16.04.1,release build loaded on my computer. However, I won't be able to do anything with this for at least 2 weeks, but after that I will post the details here once I can get my head around how KiCad works. Unfortunately the documentation hasn't kept pace with the software, so some things do not work as expected. The libraries are a case on point.

As far as the design is concerned there's nothing surprising here.
The circuit is protected from someone connecting the battery back-to-front by a series diode. The 5V supply comes from a linear regulator, and there are no switching regulators. The Microchip PIC (16LF1825) drives the MOSFET (IRF740S) via an npn transistor (BC847). The preamp is based on a TL072 with 2-stages of ~1000x and ~25x. So there's a huge gain at the input to the 10-bit ADC to make up for the lack of resolution. Audio output is generated by the PIC. Three tactile switches are read by the PIC to control the menu system, and the display is a Nokia 5110 LCD (84 x 48, 4032 pixels). Transmit pulse rate is 175pps.

As you can see, this detector is a demonstration of what you can do with a just a few components, although [admittedly] some of these have hidden complexity. It is not a deep detector, nor is it a detector suitable for searching for gold nuggets. I guess it's really a small inexpensive, not power hungry, beach (holiday) detector. It's also non-motion, and there's no Earth field elimination.

Anyway, it's an interesting device for a cheap price. It also works much better than I was expecting, so kudos to the designer.

Very interesting. Not bad for the price, and as you said, good for holidays at the beach for not so deep targets. Don't want to be digging too deep while sipping cold beverages. Cheers and congrats to Prince Harry, me being from the Commonwealth.

[6666]

BTW even if its hand drawn on a napkin is fine

[6666]

Anyway, it's an interesting device for a cheap price. It also works much better than I was expecting, so kudos to the design

Thanks for the summary.

[Qiaozhi]

Thats not bad, anybody nutted out the schematic yet ?

Yes, I've back-engineered the schematic. Since seeing gtdavid's efforts, I wanted to create the schematic using KiCad. The Software Manager for Linux Mint had version 4, but I noticed that gtdavid was using version 5. Consequently I now have version 5.0.0.-rc1-44a33f2-62-ubuntu16.04.1,release build loaded on my computer. However, I won't be able to do anything with this for at least 2 weeks, but after that I will post the details here once I can get my head around how KiCad works. Unfortunately the documentation hasn't kept pace with the software, so some things do not work as expected. The libraries are a case on point.

As far as the design is concerned there's nothing surprising here.
The circuit is protected from someone connecting the battery back-to-front by a series diode. The 5V supply comes from a linear regulator, and there are no switching regulators. The Microchip PIC (16LF1825) drives the MOSFET (IRF740S) via an npn transistor (BC847). The preamp is based on a TL072 with 2-stages of ~1000x and ~25x. So there's a huge gain at the input to the 10-bit ADC to make up for the lack of resolution. Audio output is generated by the PIC. Three tactile switches are read by the PIC to control the menu system, and the display is a Nokia 5110 LCD (84 x 48, 4032 pixels). Transmit pulse rate is 175pps.

As you can see, this detector is a demonstration of what you can do with a just a few components, although [admittedly] some of these have hidden complexity. It is not a deep detector, nor is it a detector suitable for searching for gold nuggets. I guess it's really a small inexpensive, not power hungry, beach (holiday) detector. It's also non-motion, and there's no Earth field elimination.

Anyway, it's an interesting device for a cheap price. It also works much better than I was expecting, so kudos to the designer.