Is this a major advance in PI sensitivity ??

[moodz]

Thanks! Is this the best bang for the buck, or is there something cheaper that works in the 160 ksps range? Also important... not surface mount!

Regards,

-SB

Unfortunately there is not much that is not surface mount ... actually I cant think of any that are not also very expensive.

The most cost effective chips like the Wolfson range offer 192K sampling ( and $5 price range ) and come in TSSOP. You can get TSSOP and other similiar adapters from here ....

http://www.futurlec.com/SMD_Adapters.shtml

This makes the use of surface mount very easy and very cheap.

Soldering is faster than soldering a DIP

1. Borrow the iron ( clothes not solder ) off your missus.
2. Buy an Infrared heat lamp at the local hardware store ( $10 )
3. Apply solder flux paste / liquid to the adapter board and "tin" the pads with a small amount of solder on the end of a soldering iron.
4. Apply some more flux and place the chip so it sits on top of the tinned pads ( the flux will act as a "glue" to hold the chip in place.
5. Put the iron on maximum heat and place the board on the iron ( using the iron as a heated platform or work area ).
6. The infrared lamp radiates the top of the board and is used to bring the final temp up to soldering temperature. The chip will "align" when the solder reflows and grabs the chip pins.
7. Once the iron is set up and using pretinned boards you can solder 20 chips in 5 minutes using this method.

The use of solder flux cannot be stressed enough ... you can buy 10 ml tubes of the stuff on ebay for about $5. Gunk it on so that the work surfaces are submerged in it when soldering. You will find that you will have trouble bridging even a 0.1 mm gap with solder when using good flux.

Of course you can also just solder by hand ...search utube for solder and TQFP etc and you will see lots of examples ... no fine soldering iron required ....I have used a quarter inch tip and 80 watts to solder SSOP and TQFP ( the secret is to use lots of flux , little solder and stroke outward along the line of the pins ... does not matter if the tip covers several pins at once. clean the tip often ).

Wash the boards afterward in a glass container using boiling water and a drop of detergent this will remove flux and finger contaminants. Put boards out to dry on paper towel whilst still hot.



moodz.

[hdphilip]

Can't agree more

seems that any chip worth designing with is surface mount.
although, moodz's fpga project may be a bit of a challenge with my current PCB process

Philip

all the space we save by usuing smd, we can replace with li-po's...

[simonbaker]

Unfortunately there is not much that is not surface mount ... actually I cant think of any that are not also very expensive.

The most cost effective chips like the Wolfson range offer 192K sampling ( and $5 price range ) and come in TSSOP. You can get TSSOP and other similiar adapters from here ....

http://www.futurlec.com/SMD_Adapters.shtml

This makes the use of surface mount very easy and very cheap.

Soldering is faster than soldering a DIP

1. Borrow the iron ( clothes not solder ) off your missus.
2. Buy an Infrared heat lamp at the local hardware store ( $10 )
3. Apply solder flux paste / liquid to the adapter board and "tin" the pads with a small amount of solder on the end of a soldering iron.
4. Apply some more flux and place the chip so it sits on top of the tinned pads ( the flux will act as a "glue" to hold the chip in place.
5. Put the iron on maximum heat and place the board on the iron ( using the iron as a heated platform or work area ).
6. The infrared lamp radiates the top of the board and is used to bring the final temp up to soldering temperature. The chip will "align" when the solder reflows and grabs the chip pins.
7. Once the iron is set up and using pretinned boards you can solder 20 chips in 5 minutes using this method.

The use of solder flux cannot be stressed enough ... you can buy 10 ml tubes of the stuff on ebay for about $5. Gunk it on so that the work surfaces are submerged in it when soldering. You will find that you will have trouble bridging even a 0.1 mm gap with solder when using good flux.

Of course you can also just solder by hand ...search utube for solder and TQFP etc and you will see lots of examples ... no fine soldering iron required ....I have used a quarter inch tip and 80 watts to solder SSOP and TQFP ( the secret is to use lots of flux , little solder and stroke outward along the line of the pins ... does not matter if the tip covers several pins at once. clean the tip often ).

Wash the boards afterward in a glass container using boiling water and a drop of detergent this will remove flux and finger contaminants. Put boards out to dry on paper towel whilst still hot.



moodz.

Those adapters definitely would be in order for my kind of prototyping. I'll keep that in mind. It's true most of the best new parts are SMT these days.

I still think we should pass a federal law requiring all parts to have non-SMT versions available ...

-SB

[6666]

Thanks for the tips and tutorial Moodz
those adaptors would be handy

[straton]

May I ask, what are you talking about? An average/amateur PI machine would have over 10000X amplification to get descent results. The that1510 has been tested my me over 3 years ago and it has major flaws. You get a bit over average, sensitivity, but it's still far away from a good commercial PI detector.

[straton]

Speaking of using an A/D converter, I have used the AD1671 12bit complete, 1.25MSPS, A/D converter with good results, for discrimination purposes but sensitivity went down. If anyone is interested I have a few of them for sale. They are compete/monolithic!!!, easy to use, parallel interface(easy interface), unipolar/bipolar input,with internal reference, excellent choice for this kind of job. Most atmel and microhip micros are suitable.

[moodz]

May I ask, what are you talking about? An average/amateur PI machine would have over 10000X amplification to get descent results. The that1510 has been tested my me over 3 years ago and it has major flaws. You get a bit over average, sensitivity, but it's still far away from a good commercial PI detector.

May I ask the same question .... I have only ever used the THAT1510 as an amplifier ... I have never thought of using it as a metal detector ...

Since it was designed to be an amplifier I would have to agree with you that it is still far away from a good commercial PI detector

[Davor]

THAT1510 is a nice thing when applied where it belongs, and that would be a balanced input. It works best at higher gains, and the only drawback I can think of are the "miller" capacitors that take care of the stability, but in turn also reduce gain at higher frequencies. Guess they'd be just fine for up to 100kHz bandwidth.
It is quite possible to make such preamp on your own, and optimise it for your purposes. It comprises a CFIA amplifier that is made from two opamps and two BJTs in current feedback configuration. All the feedback components are of very low resistance, thus introducing a bare minimum of additional noise.

The cornerstone in audio preamps was the Graeme Cohen pre, and THAT1510 is a near replica of that particular design. They both use CFIA frontend, but Cohen's pre had a cross coupled differential amp balanced output, while THAT has something else.

It is funny, but CFIA is a nice example of mind over electronics, where an intelligent design outperforms most of the off the shelf components. Yes, our detectors are using low impedance sources, and sure they can benefit from low noise pre.

[6666]

Moodz would you mind telling me where you are getting the SF18G diodes ?
thanks
6666

[mickstv]

I can help with that one Element14.




Cheers
Mick

[6666]

Hi Mick thanks
I had previously tried there, but must have typed it wrong
cheers

[mickstv]

Hi Mick thanks
I had previously tried there, but must have typed it wrong
cheers

No worries

[moodz]

Moodz would you mind telling me where you are getting the SF18G diodes ?
thanks
6666

Mick is right ( thanks Mick ) ...element14 have them ... I tried 1n4004s and there is a distinct speed / recovery difference using the SFs.

[mickstv]

No worries moodz. From memory I initially used some UF4005,6 which came out of an old smps. Definitely the 1n400 series are too slow for the job.

[Davor]

Yes, 1N4001 is a varicap in disguise.