Thread: Magnet charger

A while back I posted a gauss meter I made - I was intending to use it to compare magnet strengths before and after remagnetising when using a home brew magnetiser or magnet charger I was planning on making, to be used down the track to help a friend restore a vintage stationary engine - the engine project however fell through.


I do have though an analog multimeter that reads low about 18% on all ranges, this is due to a meter movement that has lost magnetism, with the movement removed and dismantled I went ahead with the magnet charger, using only materials I had on hand.


The photos just about show it all, the two central poles are mild steel 5inch long x 1.75 inches diam cut from a lenght of old line shafting, drilled and tapped for 8mm bolts - the plywood end pieces are glued on and approx 200 turns of wire wound onto each pole, the wire is pvc insulated 7 core mains cable, my late father ran power to the rear shed with 2 lengths of this cable which ran from the garage along the side fence - the back fence and into the rear shed. The coils/poles are bolted to a length of mild steel bar 1/2inch thick and 3 inches wide - the top two flat mild steel poles have been machined to fit closely on the multimeter magnet, before I had my cnc mill I would have clamped a length of flat bar to the lathe face plate, bored a hole and then cut the bar into two pieces, not a difficult job in the lathe but certainly easier to do in the cnc mill about 10 min for each pole


The contacts on the 3ph contactor have been connected in parallel - they are rated at 25amps ea so I should be able to switch 75amps through it, the lower left corner of the magnetiser has a socket that takes a connector from an old laptop power supply for 20V to engergise the 24V contactor via the push button. I have organised the coil connections so I can if needed connect the coils in either series or parallel, currently connected as series for 400 turns total. I marked the poles as N and S, in use a north pole always goes against a south pole - I check the polarity of the magnet with the gauss meter and place the magnet south pole on the magnetiser north pole - hit the button 4 or 5 times


I measure the coils resistance as 0.4ohms - with my old car battery which is rated at 660 CCA and has an open circuit voltage of 12.25volts the voltage across the coils drops to 10V when engergised, this gives around 25 amps through the coils, with 400 turns the amp/turns is 10,000, I was hoping for around 25,000. Options I have to increase this are to use my stick welder and fit a 200amp bridge rectifier for dc, these bridge rectifiers are readily available on Ebay or Aliexpress.


Another option is to use a bank of high voltage capacitors - I have a bunch of capacitors that total 12000mfd at 200V - something to be explored.


I have done some tests and initial results are encouraging - using my gauss meter for before/after comparison I was able to double the strength of an old horse shoe magnet that was once part of a telephone generator, I pulled the magnet from a magnetic door catch and was able to increase the strength around 10%, I was also able to reverse the poles by reversing the magnet. I'll do some more tests before I try the multimeter magnet, the movement is quite delicate and I really only want to have one go at increasing the magnet strength - I'm not keen on taking the movement apart multiple times.

20220325_102640.jpg 20220325_102655.jpg 20220325_102731.jpg 20220309_112751.jpg 20220309_170057.jpg 20220309_170206.jpg

That interesting. Alot of work and calculations gone into your setup.

So, your analogue multimeter is out of calibration because the magnet in the meter has lost some of its magnetism over the years?

How do you know exactly how much the original magnet needs to be increased in or to get the meter back into reasonable calibration?

Is it a trial and error thing, using another meter?

Cheers,

Simon

Sent from a galaxy far far away

This is interesting because at work in in labs with big magnets (I'm talking up to 6 tons) we always used analog meters because they would be more reliable than digital meters when working close to these magnets. A number of things on the digital meters would get screwed, not the least of which were the LCD displays.

Hi Bob,

Is also with those precision meters that with digital it was inherently not as accurate because the circuit would require a AD converter and some accuracy is lost in the conversion?

Or is that not an issue with a well designed circuit?

Simon

Sent from a galaxy far far away

Originally Posted by simonl

Hi Bob,
Is also with those precision meters that with digital it was inherently not as accurate because the circuit would require a AD converter and some accuracy is lost in the conversion?
Or is that not an issue with a well designed circuit?

We used several different types of systems.
The electrical signals (streams of ions) we were measuring were very small <10^-10 A

The system Is spent most time on had 9 DC amplifiers in thermally stabilised (Peltier cooled) magnetically shielded box about 1/2m from magnet.
Each of the 9 amplified DC signals, typically in the 10-0.001V range, were connected by shielded coax to individual DVMs
The limit on a single read was the thermal noise in the electrical resistance in the amps (10^-14A) but by integrating over long periods we could reach a max measurement precision of 0.001% but it required everything else to be lined up

Initially we used 6 place Solatron DVMs (Schlumberger) which were electrically/magnetically shielded and good to about 2 in the last decimal place and linear to about 0.002%
There were 9 of these DVMs in a bank, just to the right of the operator in the pic below.
The large black and green object was the 250kg magnet, the DACs in the Peltier cooled box were in the smaller back box furtherest to the RHS of the picture
Before every significant measurement the DC amplifiers underwent an (automated) basic calibration and once a month a full system calibration was undertaken

TIMS3.jpg

In later systems we used DC amplifiers and 7 place ADCs all enclosed inside a magnetically shielded and thermally stabilised enclosure.
The output of these went straight into the computer (no DVM).
These systems were theoretically about 10X better than using the Solatrons

We also has systems for measuring currents down to 10^-18 A using photomultiplier tubes but these could only operate below 10^-13 A and had at best a precision of 0.01%

Teh problem came when we had to use hand held DVMs ned rthe magnet - the displays would go all wonky (screwing with the liquid crystal polarisation) and even regular readings would be sus.

Very interesting Bob.

In a previous life (feels like many lifetimes ago) I studied a physics degree at RMIT.

Learning about thermal noise in sensitive circuits and the amplification of low output signals and boosting the S/N ratio was part of our learning.

Forgoten most of it now but I still have the interest and appreciation.

Simon

Sent from a galaxy far far away

Originally Posted by simonl

That interesting. Alot of work and calculations gone into your setup.

So, your analogue multimeter is out of calibration because the magnet in the meter has lost some of its magnetism over the years?

How do you know exactly how much the original magnet needs to be increased in or to get the meter back into reasonable calibration?

Is it a trial and error thing, using another meter?

Cheers,

Simon

Sent from a galaxy far far away

For me there is no way of knowing how much to strengthen the existing magnet - I'm assuming that when it is made it is taken to saturation and then aged with an AC magnetic field to bring it back to a specific point, online reading of magnet manufacturers web sites leads me to believe that this process can result in a 10% variation between magnets, to compensate for this some moving coil meters have an adjustable magnetic shunt placed across the poles so the full scale deflection can be set - this meter has no shunt but does have 4 internal adjustments (preset variable slider resistors) to calibrate the various scales/functions.
When I finally do the re-magnetisation it will be hit or miss - I'm hoping for close enough so I can use the internal adjustments - if I overshoot the mark too much and make the movement too sensitive, then I can place a shunt resistor across the meter in much the same way that a 1mA meter can be made to read 5amps full scale - if I undershoot too much then there is always the option of making a meter amplifier using a micropower opamp, to drive the movement - these things operate from a 3v supply and consume almost no current so a 3V coin type battery should power it for it's shelf life which is around 5 yrs.
My normal goto multimeter is a Fluke 87 - I don't use an analog meter that much but there are instances with AC where a digital meter can measure non existent voltages, so it is more than handy to have a descent analog meter available should I need it.