Gaussmeter Kit Group-Buy

[StarryNight]

ok, well... it seems to be working today (I think). I separated the 3 pins on the sensor slightly incase there was a short, now when i turn it on it seems to zero but jumps around between 0 and 4 when idle. When I put a magnet to the sensor it climbs up and wavers + or - a few gauss. Is that common for you guys? I'm finding the most consistent reading i get is when the sensor is touching the magnet or pole piece. It still jumps around slightly but seems pretty stable. I'm not sure what I did but it's better today than yesterday. The polarity reads correct too. If I touch an A5 magnet to the sensor I'm getting 500 to 600 gauss +/-. If I test a humbucker with an A5 I'm getting 350-450 gauss (not the same magnet). Does that seem somewhat normal?

[David King]

Are you testing Fender rod magnets or Gibson bar magnets? Those figure would be very weak for the former and about right for the latter.
Fender A2 .188 x .75" =700-900G
Fender A5 .188 x .75 =1050-1300G

never mind -I see you are looking at HB magnets.

[David Schwab]

David, as much as I like this idea I'd be a bit concerned about someone changing probes with the power on and frying the sensor.

From what I can tell the sensor is most sensitive to getting the (+) supply voltage to the wrong pin, specifically to the sensor output.
If you go ahead with this plan I'd put the (+)Vs to the Tip and not the ring or ground lug. That way the Vs is the last connection made as the plug is inserted and Vs won't rake across the output and ground terminals on the way in.

I had the same thought. That way if you accidentally pull the plug, you remove the power before anything else.

If you have the room then a mini DIN or even an XLR jack would be safer. I'd personally want to use a small Deans polarized 3 pin plug from an RC model shop (used to connect servos) as these are so cheap and widely available.

That was my first idea, but I wasn't seeing any sockets and plugs that I liked.

My sensor has come back from the dead but it didn't get Vs to the wrong pin I'd only flipped the ground and the output connections.
I discovered that some of my out-of-range readings were because the heat-shrink tubing I installed over the sensor leads had caused an intermittent short. Be sure to check for this possibility before tossing the hat in.

I haven't even touched the sensor yet! I'm going to get the case set up, and then solder on a three pin header to the sensor pads, and then finish assembling the unit (PIC and LCD). Then I'll work on the connector and probe.

[RedHouse]

...That was my first idea, but I wasn't seeing any sockets and plugs that I liked....

+1

But more to the point you can do what PC-Card does, connect the power before the data lines, like wise disconnect the data before the power.

It's not apparent when looking at the pins on a PC Card, the 2nd gen PCMCIA spec designated make-before-make, and break-before-break contacts. That is the way hot-plugging works. When you shove a PC-Card (PCMCIA) into it's socket, the data lines connect BEFORE the power lines. What does that have to do with us?...

If you use a "stereo" 1/4" jack (or even 1/8" 3.5mm jack) you CAN'T ensure the power pins make contact before the data pins. This leaves only the one SURE method of connecting/disconnecting probes ...TURN OFF THE DAR_N POWER! then plug-n-play stuff.

[Joe Gwinn]

Also, can anyone point me to a place where I can source a trannie socket suitable for plugging the allegro into? (for easy sensor changing)

Any socket intended to hold a transistor in a TO-92 package will work. These sockets have three socket pins in a line. I use MIL-MAX 0.100" spaced single in line socket strips, available from the usual large suspects, but there are lots of suitable socket types.

[Joe Gwinn]

Easy Joe, I'd leave that for the experienced folks only, inexperienced experimentation with the input line to the PIC might ruin the PIC also. It's cheaper to get another A1302 than another PIC from Italy.

But how will they become experienced (=scarred for life)?

I recommend members build their probe (w/o the PIC installed) then install the PIC and test the unit. (static strap on of course)

That would work, and one can use the voltmeter on the sensor output and a magnet to determine if the sensor works.

Try not to touch or otherwise play with the sensor wiring (PIC input) as little as possible.

I think I'll add a pair of 1N4148 diodes wired such as to constrain the sensor-output input line to remain between 0 and 5 volts, at the very least for ESD protection. Will the PIC tolerate all voltages between zero and +5v at its pin 2 input? I assume it will, and that it has some built in protection diodes.

Hmm. How closely does the Allegro sensor output voltage approach the Gnd and Vcc rails over the magnetic field range? I'll have to check this.

[Joe Gwinn]

I had the same thought. That way if you accidentally pull the plug, you remove the power before anything else.

I have to agree with RedHouse here - hot swapping sensors is likely to lead to tears, and stereo plugs are not a reliable way to sequence power.

[David Schwab]

I have to agree with RedHouse here - hot swapping sensors is likely to lead to tears, and stereo plugs are not a reliable way to sequence power.

I have no plans on hot swapping them. I was just saying that if the plug was accidentally pulled it would remove power from the sensor if the tip was used for (+). I would power down the unit before changing probes.

Ideally I wanted to find a connector like the one on my Behringer mixer that connects the power brick to the mixer. It has three pins and a screw collar to secure it.

But I think the 1/4" phone jack will work well if I don't find anything I like better.

[realtonecustompickup]

Are you testing Fender rod magnets or Gibson bar magnets? Those figure would be very weak for the former and about right for the latter.
Fender A2 .188 x .75" =700-900G
Fender A5 .188 x .75 =1050-1300G

never mind -I see you are looking at HB magnets.

I repair the pickups to more than 700.

But I never knew a gauss.

How levels gauss affect sound ?

low levels = warm ?

high levels = bright ?

thank you

[Joe Gwinn]

I have no plans on hot swapping them. I was just saying that if the plug was accidentally pulled it would remove power from the sensor if the tip was used for (+). I would power down the unit before changing probes.

Many plug-jack combinations usually achieve this sequencing, but not all, and not always.

Ideally I wanted to find a connector like the one on my Behringer mixer that connects the power brick to the mixer. It has three pins and a screw collar to secure it.

Philmore distributes such connectors.

General: LKG Home Page

Multipin connectors: http://www.philmore-datak.com/mc/Page%2054.pdf

I think these may also be a kind of DIN connector. (DIN has many connector standards, not just the few kinds that are loosely called DIN connectors.)

But I think the 1/4" phone jack will work well if I don't find anything I like better.

The miniature phone plugs and jacks will work as well as 1/4-inch, and are mechanically more suitable.

[52 Bill]

In order to change the input calibration to probe #1 or #2, the meter has to be powered down anyway.

Radio Shack has a 4-pin, "CB Microphone" connector set that has a locking collar.

[David Schwab]

Radio Shack has a 4-pin, "CB Microphone" connector set that has a locking collar.

That's the kind of thing I was looking for. I even looked on their web site... I never thought about CB Microphone plugs.

[RedHouse]

But how will they become experienced (=scarred for life)?...

Well yes, there is that!

...I think I'll add a pair of 1N4148 diodes wired such as to constrain the sensor-output input line to remain between 0 and 5 volts, at the very least for ESD protection. ...

I have a similar plan, back-to-back Zeners (aka Clamping Diodes)

BTW, did you build yours yet Joe?

[Joe Gwinn]

BTW, did you build yours yet Joe?

Partially. I have built all but the sensor probe assembly, but have not put the circuitry into the newly purchased box yet.

The only problem I had was self-inflicted - I didn't initially notice that the soldering iron temperature was set a bit low, so soldering was difficult at first.

As for back-to-back zeners, that may not work, as the PIC is rated for one schottky diode drop below Vdd to one drop above Vcc. I would use two schottky diodes from Vcc to Vd, with center to the PIC input, and a series resistor at the PIC input.

[RedHouse]

Partially. I have built all but the sensor probe assembly, but have not put the circuitry into the newly purchased box yet.

The only problem I had was self-inflicted - I didn't initially notice that the soldering iron temperature was set a bit low, so soldering was difficult at first.

As for back-to-back zeners, that may not work, as the PIC is rated for one schottky diode drop below Vdd to one drop above Vcc. I would use two schottky diodes from Vcc to Vd, with center to the PIC input, and a series resistor at the PIC input.

Good to hear Joe.

My only "problem" was also self-inflicted. When I hooked up my first A1302 sensor I was looking at the data sheet (page-3) instead of Elepro's diagram, and read the pin config for the LH package (SOT) instead of the UA package (3-pin SIP) which we have....Doh!

Anyway, check out page 45-48 of the PIC16F883 data sheet, shows back-to-back zeners protecting the I/O pins.