LM394 on my Minelab PCB

[gef12]

Ye well I said can't garantee it .. tis 2100 or 2200 .. dont know

[gef12]

But you would be correct Mario.. pin 1 on all 3 IC is should read pin 6.

[mario]

Fets update

Well, after hours of research, I found some interesting fets.

The TN2404KL is definetely an improvement over the VN2410L N mosfet. However, I found one with more interesting specs:

specs for
IRFU214PBF
(IRFD214, or IRF614S)
V = 250
Resistance = 2
Input Capacitance, Ciss =140 pF
Output Capacitance, Coss= 42 pF
Reverse Transfer Capacitance, Crss =9.6 pF
http://www.mouser.com/catalog/specsheets/91269irf.pdf
http://www.vishay.com/docs/91130/sihfd214.pdf

The IRFU214PBFhas half the resistance of the TN2404KL, lower Ciss and Crss....(not sure about Coss).



-------------

Also, how about an update to the P Fet?

the TP0610KL is definetely an improvement over the stock ZVP4105A P mosfet. Resistance and capasitance figures are half the ZVP4105A.

V = 60
Resistance = 6
Input Capacitance, Ciss =23 pF
Output Capacitance, Coss= 10 pF
Reverse Transfer Capacitance, Crss =5 pF

http://www.vishay.com/docs/71411/tp0610k.pdf


It occurred to me why not use components used in high performance audio receivers? The P Fet will be connected to the LT1080, so why not use P Fet 2SJ133?



I noticed that when it comes to these fets, the capacitance varies much so with the resitance. In other words, in one hand you have fets with low capacitance figures, but the resistance will be high. Or on the other, the fets will have higher capacitance figures, but the resistance will be low.

In our application, it seems we want fets with lower capacitance figures, rather than ones with a lower resistance.

My question is, in what applications do these two variations come into play? Hitachi and Toshiba make some great fets for their audio equipment, yet we don't want them in our detectors because even though they're super low resistance, they're higher capacitance specs.

Why can't we use those fets in our detectors, when they're used successfully in audio equipment?

[mario]

I have read that the 2200 is a little more stable than the 2100s. I think that the schematic in question is for the 2200. Also, there's v2 in addition to the original 2100 and 2200s, so who knows, that schematic might be correct, just not for the 2100v2. Some of the components on the schematic look familiar, others, I just wasn't able to find on my pcb.

[mario]

back to the drawing board with the N fet

I got my chips today.


N channel fets not compatible:
TN2404KL : asides from not being available in the right package (Only available in TO92), prongs (D,S,G) do not align. I had to twist prongs G and S to get the fet to work. And when it did work, it only worked with my DD coil. My home made coil has a resistance of .6 OHM and quickly got this fet hot. The DD coil also got this fet hot.

IRFU214PBF: much more compatible and did not get hot when I hooked up the stock DD coil. However, my custom coil did not work AND also got this fet hot.

Fortunately , my P channel fet was compatible:
TP0610KL

=========

The N fet used on the SD2100v2 is not the same as the SD2000. When I removed it, it reads as *21*9*1.... unfortunately, I scratched too many digits to properly identify it.

Oh well, at least I didn't damage it like i did with the P fet ZVP4105A (Which is in deed used on the 2100v2)

-----------

Not sure I'm going to replace other chips until i get a right N fet.

[Mechanic]

Hi Mario,

If the pins don't line up you need to check the datasheet and be absolutely sure that you get the pins right. If it goes in backwards(D and S mixed up) it will get hot and only work with a DD. If you choose to go with surface mount packages just solder some small leads to the board and attach the part on top. It can be a bit tedious but it works.

With the 4066's, if you go with a lower resistance chip you need to be aware that the gain of the integrator will be affected and the machine could potentially become noisier. Also if you change the 4066 you will then need to reset the earth field trimmers to cancel a moving magnetic field. But by all means fiddle and play but make sure you are careful when removing components as you may need to put them back if it all goes up the creek! Also before assuming an ic is something check the pinout to make sure the power pins line up.

Cheers Mick

[mario]

hi Mick,

The pins S and G didn't align properly so I twisted them in accordance with the data sheet. It did work but got hot really quick with my dd coil after testing it for 30 mins. The original fet comes in package to220, a much bigger package than to92.

I'm not saying it doesn't work in other versions, I'm saying it doesn't work on mine. I now know the SD2100 is a different animal from the 2000.

My guess is that the N fet is not compatible enough. The transmit cct of the 2100 might use higher amperage components and thus must be overwhelming my replacement fets?

Unfortunately, there's no way of knowing the requirements of the original part since I don't know the part number. Still, the best I can do, since it's a N fet after all, is experimenting with higher amperage parts.

I'm going to try 240 5-9 amp, then 300V 5-9amp fets. Eventually on to a 9amp, 400V.
Will I risk damaging AD797 Chip if I use +300v fets? I'm thinking I just need one that can handle larger amperage so i should start at 200-240 v range.

I know there has to be a better replacement than the stock N fet. The original P fet was made back in 1994!

Also , I got my oscilloscope, it's an older textronik 7330, but so far i'm happy with it. I like the fact that I can upgrade it with modules. This thing weighs a good deal and not very portable though. Still, since I will mainly be using it for coils and my experimenting, it's more than I ideal.

[Mechanic]

Hi Mario,

Try using a new fet as the one you put in may have fried. The front end fet arrangement has changed very little since the 2000. Only a small amount of current goes through the front end fets. The purpose of the n channel fet is to block the flyback pulse of 180v from getting to the front end. Changing the voltage rating higher than 200v will be of no benefit to you. Also don't go attempting to raise the flyback voltage as it will not be of benefit either also you could blow something up and without really knowing whats what, that would not be cool!

Anyway that huge n fet you got there, will be G,S,D Whereas the 2404kl is S,G,D. Don't forget that the view on the datasheet is looking from the top down. Higher amp parts have much higher capacitance. This will effect coil settling time. The n fet needs to be no more than 600ma, Vdss 200v and Vgs >10v. Have a look for fets with resistance and capacitance ratings better than the 2404kl that meet the above criteria.

Your journey has just begun!

Cheers Mick

[mario]

You're right, I wasn't viewing it from the top. The TN2404KL fet is the most compatible of all the fets I have tried.

After soldering a second fet, I got it working with my 11 mono coil. Still, this fet gets really, really hot to the touch. I'm trying to replace it with a to220 package.... at least I now have some specs to go by, thanks for the ones you listed.

[Mechanic]

Hi Mario,

It should not get hot at all. Remove the fet and make sure you have a square wave on the gate pin measuring 10v. By putting the other fet in wrong you may have fried the 4093 that drives the fet. There are some real good replacement ones that have protected inputs.
http://au.element14.com/jsp/search/p...sp?SKU=9666508 This one has triple diode protection and is the one I use in mine. It seems fairly indestructible!

Just before you go changing it, lets check the pinout, the cct around it looks identical to the 2000. Pin 14 and 13 should have +5v present, pins 5, 6, 7 should be at your - most supply(batt -) Your n fet gate should be connected to pin 11 and your p gate to pin 3. You may also have to replace the p fet again, but try the driver first, providing you are not getting 10v or more to the gate pins. When they fry ou might only get about 1v.

Make sure you practice on something else before you try removing smd ic's!

Cheers Mick

[mario]

My 4093 is dead...long live the 4093!

I have a replacement ic coming in the next few days. Having said that, I dug up an old post where Woody suggests creating a sub board for the driver and fets. what kind of advantages would that bring?

I know coil feedback currently resonates accross the entire main circuit via ground. will there be substantial increase in sensitivity?

[Mechanic]

Hi Mario,

Don't worry about the extra driver cct, if you use the 4093 that is rated to drive 8.8ma it will do its job just fine.

Mick

[mario]

Here you go Mick. I should have all chips sometime next week

I'll update the photo soon to include grounds and voltages

[Mechanic]

Hi Mario,

Was the coil connected when you did these checks? Thats the only way I can see that the coil gnd is hooked up to the drain pin of the n fet. I have corrected your drawing. The 100ohm resistor can be reduced to 10 ohm when you put the new fet in.

Cheers Mick

[mario]

Sorry about that Mick, I rushed my last picture. If you can reloaded the updated picture, I would like to ask you if you can tell me if my proposed pin arrangement of the 4093 IC is correct. I based it on the diagram to the right of the photo, and not ground, vcc paths. However, I did my best to trace the flow of continuity as correctly as possible.

I am getting 11.28V on pins 1,2,5,6,7,10, and 13.

The odd thing is that detector still works like nothing has happened. My guess is that it is unprotected from a voltage overflow until I replace the fet driver.

Also, You pointed out about the N fet being incorrectly connected based on my previous inattentive error. The wire that is connected to "S" of the N fet does indeed go through a diode.

After studying the matter, I think my P Fet is the one that is connected incorrectly. Shouldn't the N fet "D" be directly connected to P fet "S"...Right now the order of "DGS" on my P fet is the way I soldered the pins to my PCB, but after tracing the flow of components, the way the P fet is connected doesn't seem right....

If I were to re-solder the 4093 IC, I don't think it'll last long if my P fet remains soldered in my current configuration ? no? Do you think that this is what fried my chip, (assuming my chip is fried)?

If photo didn't update, you can click here.