Garrett ACE 250 calibration advice needed

[rheslip]

I picked up a non-working ACE 250 very cheap a couple of weeks ago. It had multiple issues - bad solder joint on the coil connector and caps C38 and C39 on the attached schematic were bad. C38 measured 100pf or so and C38 about 2.3nf. I replaced the caps with the values in the schematic and after that the signals that control the analog switches looked much closer to what I expected. TX signal is a clean sinewave 8Vpp 6650hz which seems correct. The differentiator and integrator signals from the opamps seem reasonable - slow blips as you sweep metal thru the coil.

I set the pots to align KT12 to the zero crossings of the RX signal at KT5 and KT13 to the peaks of the RX signal but it didn't work well set like that - no ferrous/non-ferrous tones and little to no discrimination.

Then I set pot R48 to its original setting of 3.15k and pot R67 to about 11k. This more or less aligns test point KT12 on the schematic so it switches near the zero crossings of the TX signal at KT3 (not shown on schematic) and test point KT13 switches close to the peaks.

Its working with these settings - detects reasonably well, generally distinguishes ferrous and non-ferrous and the discrimination seems to work. If I increase the gain past half it starts beeping with no targets present which doesn't seem right - ground balance issues?

Can anybody offer suggestions on how to set up the timing of the analog switch control signals at KT12 and KT13?

Thanks to whomever posted this schematic - it was a huge help!

[Altra]

I believe R48 adjusts the ground balance and R67 is for the target ID. You can try setting both trimmers to center. Then take a piece of ferrite and adjust R48 until you get the lowest signal on KT14 test point. Or the lowest audio response. Then use a US nickel and adjust R67 until it is matched on the LCD ID.. Verify with a US quarter. This is how I would approach it.

[Carl-NC]

I set the pots to align KT12 to the zero crossings of the RX signal at KT5 and KT13 to the peaks of the RX signal but it didn't work well set like that - no ferrous/non-ferrous tones and little to no discrimination.

Were you using the residual (no target) RX signal? The phase of that can be all over the place. You want to introduce a ferrite target that produces a really strong RX signal (but not saturated) and calibrate to that. In fact, you can do it "by ear" without an oscope by bobbing the ferrite while adjusting R48 and listening for a minimal audio response. This is done in all-metal mode.

Altra is right, R67 is used to adjust the target phase. I would use a US nickel and US half dollar for this, and you may need to do the R48 + R67 cal 2 or 3 times to get both settings dialed in.

If I increase the gain past half it starts beeping with no targets present which doesn't seem right - ground balance issues?

Not necessarily, could be EMI.

[Altra]

I have an early ACE250 and it only has one trimpot. rheslip can you post a photo of the your pcb? The voltage regulator marked on the schematic "LD1807" is a LM2733. Here is a picture my pcb plus a schematic that seems to match

[rheslip]

Thanks for the advice - very helpful! I'll give that a shot and report back.

Attached is a photo of my PCB before I replaced the two caps - really weird that two SMT caps would fail. As you can see my board has two pots as per the schematic which is a later rev of the Ace 250 I believe.

[rheslip]

Were you using the residual (no target) RX signal?

I was. Did not realize it needed a target for this to work properly.

Thanks Carl!

[Ruben2000]

Thanks to whomever posted this schematic - it was a huge help!

I'm glad it was helpful to you!

[ivconic]

Here is a picture of a Chinese copy ACE250, which works 100% identically to the original. A few weeks ago I got it, more out of curiosity.
I didn't need to do anything extra, the copy works perfectly fine.

[ivconic]

"Decoding" some SMD components was quite a challenge:

[ivconic]

Sprint Layout file (not 100% sure if I done it corrctly):

[ArchibaldSTM]

Sometimes resetting to factory settings helps. Turn on the device (with the coil), lift the coil up. Then press the power button and hold it for 5-6 seconds. When you hear a signal, release it.

[Ruben2000]

Sometimes resetting to factory settings helps. Turn on the device (with the coil), lift the coil up. Then press the power button and hold it for 5-6 seconds. When you hear a signal, release it.

Hi ArchibaldSTM!
What do you mean by lifting the resistance?
I tried to repair my Ace 250, but it seems the microcontroller is failing, so I gave up.
Everything I'm talking about is in this thread I created:
https://www.geotech1.com/forums/foru...-ace-250/page2
I'm thinking of using the rod for my SMW, but I'm embarrassed to leave it like this. I got it for a very low price.
That's why I'm asking you this question to see if there's any hope.

[Ruben2000]

Hi ivcónic, thanks so much for sharing your files!
Do you have a schematic of the microcontroller part?
I'd like to compare it with mine.

[rheslip]

"Decoding" some SMD components was quite a challenge:

Thanks for posting your schematic! C30 and C29 on your schematic are the caps I took out because they didn't match the schematic I originally posted. Mine measure 155p and 2.2n measured out of circuit which is very close to what you have. Based on that I put the original caps back and used the suggestions above to set the ground balance and target pots. Its working very well now!

The original problem with the unit was bad solder joints on the coil connector. I'm not sure why it took so much more fiddling to get it working but it was kind of fun to debug and I learned some new things. Thanks for all the help folks!

[ArchibaldSTM]

Hi ArchibaldSTM!
What do you mean by lifting the resistance?

Google translator translated incorrectly. I talked about this: https://www.youtube.com/live/ty1n3B7ndLo