AVR+DSS Frequency Mod SD2100V2

[Finderskeepers]

Hi all,

I'm in the process of drawing the circuit board of this project, but I presume that the sinusoidal out terminal and ground terminal is connected in lieu of the crystal on the sd2000 board.
Just want to verify this is correct.
When its complete and working, I intend to publish pcb artwork for a doublesided holethru version for this forum.

thanks and regards

Martin

[jtz54]

If you check the detector mods forum and look forGP extreme broken parts, there is a lot of info on common failures and replacement parts. Considering your level of expertise should be fairly easy for. Good luck and thanks for your efforts in this forum.

[mario]

Is your 2100 still broken? I fixed my friend's GP Extreme with the help of Woody and the guys at Detectormods forum. It's hard when you don't have a schematic and everything is covered with paint.

Yeah, it's still broken.

With everything else done, I'll probably try to get it working again one of these days. It's been hunting me.

[jtz54]

Is your 2100 still broken? I fixed my friend's GP Extreme with the help of Woody and the guys at Detectormods forum. It's hard when you don't have a schematic and everything is covered with paint.

[mario]

G'day Mario,

Good to see your still kicking around!

Now, square waves or sine waves..... It matters not. As long as it is stable. The sine waves from the oscillator are converted to square waves before feeding the clock signal to the 4040. I have actually built a "vco" type cct with 10 coarse frequency settings from 1 to 4 mhz and a fine tune for tuning out emi. It all fits on a board 36mm X 18mm. The output is extremely stable with only a small amount of slow frequency drift due to temperature. It uses 29 components and the output is a square wave.

Cheers Mick

Yeah, it's been years.

After my machine broke, I kind of got bummed out and stopped my minelab projects. Around that time, I also moved to a new house, so I replaced those projects with remodeling work.

So yeah, I've been busy, but still feeling bad about my old SD2100...

[Mechanic]

G'day Mario,

Good to see your still kicking around!

Now, square waves or sine waves..... It matters not. As long as it is stable. The sine waves from the oscillator are converted to square waves before feeding the clock signal to the 4040. I have actually built a "vco" type cct with 10 coarse frequency settings from 1 to 4 mhz and a fine tune for tuning out emi. It all fits on a board 36mm X 18mm. The output is extremely stable with only a small amount of slow frequency drift due to temperature. It uses 29 components and the output is a square wave.

Cheers Mick

[mario]

Thanks Mario and Mick. Going to try one of the function generator modules at first. If that doesn't work too well I'll build your project (I'm a lazy guy). Just picked up Carl's book, maybe I'll learn a few things. It's forums like Geotech and guys like yourselves that make this hobby interesting and successful. Much appreciated.

Always glad to help,

Good luck

[jtz54]

Thanks Mario and Mick. Going to try one of the function generator modules at first. If that doesn't work too well I'll build your project (I'm a lazy guy). Just picked up Carl's book, maybe I'll learn a few things. It's forums like Geotech and guys like yourselves that make this hobby interesting and successful. Much appreciated.

[mario]

Hey Mario - was reading up on your project. I'd like to play around with the variable frequency concept on my 2100.
Where did you inject the signal on the 2100 board? A pic would be great. Thanks.

John

Hi Mick, Hi guys, long time no see.

John, download the files included with the project. One of them is an image with the correct connecting points.

I'm not sure what type of oscillator you are trying to use, but the signal has to be a very pure sine wave. Otherwise, it might be **very** problematic.

In short, the frequency:

1) has to very pure for timing purposes
2) has to be within a flexible range to avoid interfere

failure of the above 2 will lead to noise, which will prevent you from achieving any noticeable depth.

[Mechanic]

G'day John,

Is yours a 2100(original green) or a 2100v2? Cause they could be different.

Basically though, you have the large eeprom, the part next to that, moving toward the rear endcap edge of the board, you have a so16 package, which is a 74hc4040. Pin 10 is the clock signal for the 4040. If you lift that pin from its pad, you can feed signal into that, but make sure the voltage levels are correct. To be safe use a 1nf cap between your oscillator and the 4040.

Cheers Mick

Ps this thread will also be of interest to you.... http://www.geotech1.com/forums/showt...highlight=2100

[jtz54]

Hey Mario - was reading up on your project. I'd like to play around with the variable frequency concept on my 2100.
Where did you inject the signal on the 2100 board? A pic would be great. Thanks.

John

[mario]

I need to make an update.

Work on the DSS project has been completed. I feel I have attained a good level of polish, and don't want to include unnecessary bloat. Last thing I want is redundant features. Currently, the program is very easy to use and powerful. This last update gives the user far more control over EMI, making these detectors substantially more quiet.

The archive included in this posting has all the necessary information to get a dss board up and running should any one here decide to undertake such a project. It includes the necessary schematics, binary, and detailed documentation.

I made last minute revisions to the schematic.

It now leaves open the possibility of future upgrades, like the current ground balancing project I'm undertaking. Having said that, I have made good progress with the X9118 chip, and I'm very impressed with the results.

I will be starting a new thread devoted to that second project since I plan on making these two projects independent of each other. However, I will give those who have the DSS boards the option of upgrading to ground balancing.

Since I have not finalized the hardware, I have not made the schematics for the ground balancing public. I'm currently debating if to implement something more meaningful like automatic ground balancing, which will obviously require additional hardware other than the two X9118 RDACs.

At 20 mips, we stand a chance at implementing some form of fft. I plan on doing something similar to what Garrett did with the emi handling in the Infinium, but with ground balancing, obviously.

If I don't, I will at least make the manual ground balancing public, which will be useful on its own right.

All in all, I'll try to make fewer, but more substantial posts on my next mod, so expect fewer posts when I do start making them.

It's a great time to own these older analog minelabs. Now that more and more people are becoming more comfortable with these microprocessors, these older minelabs will only get more functional.

[mario]

I thought about the AD5231.

Instead, I ordered the X9118 (X9118TV14IZ) (http://www.intersil.com/products/dev...=&js=&pn=X9118) because using I2C will simplify everything.

It seems like SPI (the interface the AD5231 uses) will require us to use 3 pins for SPI + 1 pins for addressing( total of 4 pins). Because we need 2x5231 in order to address Channel 1 and Channel 2, the pin count becomes much more manageable if we use the X9118 instead. For instance, with the X9118, we simply use the same 2 (I2C) pins to address up 4 devices .

Also, you can directly order from the manufacturer (clicking on the small shopping cart to the right) instead of relying on distributors.

I have them on order and will let you know how it goes...

[Jiminy]

What about the AD5231? It has 1024 taps and 100k?

[mario]

Well, this little add on should also apply to the SD2000. If I'm not mistaken, Minelab also released them with manual ground balancing through the use of those 50k-100k pots.

Anyway, it seems like the AD5242 only has 256 positions, so the resolution will not be fine enough for our purposes...

I'm going to have to find another part...