Tesoro - Make it deeper

[getsa]

Same applies to me - I bought few, tried them, but the result was awful... probably also fake copies or renumbered ICs.
––––––––––––––––––––––––––––––––––––––––

Regarding increasing the depth - it depends on the types of the ICs, but sometime connecting pull out resistors to the outputs and (-V) may doubled the depth.

[dbanner]

I think it is time for a true op amp tester, not the one with the blinky led's, but one in which the op amp can be put through its paces and various results viewed on the oscilloscope.
This sort of rigorous testing would tell the fake from the genuine or even help to identify or verify the op amp by comparing with original one.

Here is a true tester circuit. It can be made from a smaller PCB layout. I'm sure there are other circuits. Even one based around an arduino can be designed.

snoa637.pdf

One sure giveaway would be measuring the slew rate. But also to determine whether you have a jfet or cmos or bipolar op amp. Otherwise buy from only reputable sources, even so, testing is always a good idea when your circuit is under performing.

[Carl-NC]

Once upon a time, I replaced the TLC2262 chip with TL074, the sensitivity has almost doubled. Try it, it's not difficult.

That doesn't make sense, the 2262 is the better opamp. Even discounting which opamp is better, where it is used in the design is not demanding and I can't imagine any scenario where depth would be cut in half based on the opamp used. Did you also do something else?

[Carl-NC]

Yes, both Bandido and DeLeon use 5.6/6.2mH coils.
Lowering the LTX implies making a new search head. Phew!
Just what I want to avoid!

You can probably find a different concentric design and adapt it to the Bandido, it's not all that hard.

[Artyom]

That doesn't make sense, the 2262 is the better opamp. Even discounting which opamp is better, where it is used in the design is not demanding and I can't imagine any scenario where depth would be cut in half based on the opamp used. Did you also do something else?

I saw in the documentation that 2262 is of higher quality, but in fact it worked better with 074. I tried in different devices and won 074 everywhere. but I will make a reservation that the chips ( 2262, 347, 074) were bought in China, from different manufacturers and sellers, and they were all different. probably unidentified fakes. and I've been thinking about the chip tester myself for a long time))

[Carl-NC]

OK, it makes sense if the 2262 was a fake.

[ivconic]

OK, it makes sense if the 2262 was a fake.

I couldn't find a 2262 anywhere so I used a TLC27M2... is there a possible problem?

[Carl-NC]

That should work just fine.

[ivconic]

As far as I can see, the tester is of an older generation and requires a serious approach, an oscilloscope and certain "mathematics".
In principle it is a good approach. But it's too complicated for me and won't tell me much.
For the purpose of achieving better metal detector performance; i think a much more direct approach is to put the IC sockets on the pcb and test "in vivo" a series of opamps.
Carl, I see that you are in disbelief about the radical differences in the behavior of different opamps on a metal detector, especially on the "depth" issue.
I made my first SMW xx years ago. It worked solidly but not "sensationally".
I think I used a TL074 or 064 or 084 back then... not sure. It was a long time ago.
In those days, I came across a pile of pcbs from scrapped medical military instruments at a military junkyard.
I spotted a dozen LF347s in a dark brown ceramic package.
Until that moment, I didn't even know what the LF347 was, I just didn't have the opportunity to see and use that opamp.
Dialup internet... datasheet... I see it's a quad opamp.
Crazy idea; why don't I put it in SMW and see what happens?
So, same pcb, same coil... IC sockets. Same targets, same test conditions.
With the LF347 the SMW worked insanely well, unheard of, I couldn't believe what I was seeing!
I don't want to give wrong information, it was a long time ago, I can't remember the coins and items and the performance achieved, the only thing I clearly remember is that the difference in behavior was HUGE! So big it was hard to believe.
I even think I wrote that somewhere on this forum, it was a long time ago.
So yes, it's true. The morphology of SMW is nothing special. It is a "hybrid" between White's Classic and Fisher 1235.
But that rather rudimentary metal detector is still very popular today and makes a huge difference in real fields.
But not to digress; yes Artyom is right, I have the same experiences.

[walter]

https://www.tme.eu/at/details/tlc226...s-instruments/

lg walter

[ivconic]

Thanks walter. It is sometimes very complicated for me to buy material from abroad.
I did notice one detail though, browsing through the link you sent!
TLC2262 says explicitly that it is rail to rail, while TLC27M2 doesn't say that!?
Maybe that's the reason?

[ArchibaldSTM]

There are a lot of TLC2262 on Fisher F4\F2 boards. You can buy a broken one and unsolder it. There will be a guarantee of authenticity , and not a fake.

[Artyom]

I don't pretend to be a genius. I propose for discussion and consideration the proposed scheme for checking the gain of the chip at different frequencies and signal levels. U2 is a test chip, we connect the output to anything. it can be supplemented with both a divider and a load at will.

[dbanner]

Military grade op amps like the LF347 in the dark brown ceramic package that Ivca mentions above, have to go through very stringent testing. Extreme temp. range is not the only thing. If you check the wafer distribution of the production lot for example(some datasheets show this), you will notice variations. You will see "typical" and "Max" values in the datasheet for some specs.

There is a document, Military standard for testing microcircuits.

std883.pdf


I think just plugging in various chips and trial and error is good enough sometimes, unless your entire stash of chips are fake.

[Carl-NC]

Carl, I see that you are in disbelief about the radical differences in the behavior of different opamps on a metal detector, especially on the "depth" issue.

It depends on where the opamp is used. The 2262 is typically used in the post-demod gain stages where the signal bandwidth is under 100Hz. The most important spec here is usually input bias current or offset voltage where the 2262 is slightly better, and it has a rail-to-rail output as well. I've never seen any opamp make much of any difference in these stages.

[dbanner]

I looked at the photo of your bandido pcb, and those 27M2's all look suspect(fake). Aren't Texas Instruments DIL thru-hole chips normally with the notch? I have some and they all have the notch.