Hi Boden,
In China of course. Easy to find here, there may be more sellers for the same item:
https://www.ebay.de/itm/35489151606.../Cpx45IfHMCpWn3eZwiPwEb7Bxm5|tkp:BFBM8vmTzMti
Hope the link works, otherwise just search for TD8620 Gauss Meter.
All the best
Mattes
In China of course. Easy to find here, there may be more sellers for the same item:
https://www.ebay.de/itm/35489151606.../Cpx45IfHMCpWn3eZwiPwEb7Bxm5|tkp:BFBM8vmTzMti
Hope the link works, otherwise just search for TD8620 Gauss Meter.
All the best
Mattes
If you use a sine motion, you don't have to integrate, just use voltage. We do use integration, but it's always a constant zeroing problem due to leakage and offset drift.
A constant speed motor coupled to a linear motion via an offset pin would work for a VC in a gap.
I do agree with others, this micro-gauss to tesla stuff is a PITA unit wise, and really weird. I work in one building at the micro-gauss level, then go to my other office and work 25 tesla. It gets rather confusing.
John
Yes, that´s an in direct way, not sure how would you calibrate it.
Integrating is the old Physics Lab method, for sure whatever Maxwell and others usd (mechanical integration, no Electronics way back then)
As of Op Amp drifting, yes, it´s real problem, and complicated if you want to keep readings available for a long time (they will drift away) but for single measurements I had no trouble.
Mind you, my earliest integrating meters (we are talking early 70´s) used humble bipolar input LM741, go figure.
The "universal" workhorse back in the day, and the only one widely available, the Industry Standard.
Usable because to begin with you can balance output to zero, it has 2 pins for that,and then using a largish capacitor, think around 10uF, so input bias current will charge or discharge it, no doubt about that, but will do it slowly.
My test box had a momentary pushbutton switch which shorted the cap (so 0V by definition), I first zeroed OpAmp output with a trimmer pot.
For measuring I inserted coil in speaker gap, momentarily shorted cap (to start from zero), pulled coil so its current carged cap and a second switch disconnected it from Op Amp output so I made a kind of sample and hold circuit.
Cap voltage could be read at leisure, since self discharge was low, specially at such low voltges (a couple V) and meter input impedance was high (1M to 10M).
Values DID drift, but sllloooowwwwllllyyyy 😉
Long term "memory" consisted of writing read value in a piece of paper, of course 😉
Then the Math: integrated Volts and considering Cap value became Maxwells (Integrators measure total flux) which divided by gap surface gave me Gauss.
Easier done than said, quite straightforward.
Integrating is the old Physics Lab method, for sure whatever Maxwell and others usd (mechanical integration, no Electronics way back then)
As of Op Amp drifting, yes, it´s real problem, and complicated if you want to keep readings available for a long time (they will drift away) but for single measurements I had no trouble.
Mind you, my earliest integrating meters (we are talking early 70´s) used humble bipolar input LM741, go figure.
The "universal" workhorse back in the day, and the only one widely available, the Industry Standard.
Usable because to begin with you can balance output to zero, it has 2 pins for that,and then using a largish capacitor, think around 10uF, so input bias current will charge or discharge it, no doubt about that, but will do it slowly.
My test box had a momentary pushbutton switch which shorted the cap (so 0V by definition), I first zeroed OpAmp output with a trimmer pot.
For measuring I inserted coil in speaker gap, momentarily shorted cap (to start from zero), pulled coil so its current carged cap and a second switch disconnected it from Op Amp output so I made a kind of sample and hold circuit.
Cap voltage could be read at leisure, since self discharge was low, specially at such low voltges (a couple V) and meter input impedance was high (1M to 10M).
Values DID drift, but sllloooowwwwllllyyyy 😉
Long term "memory" consisted of writing read value in a piece of paper, of course 😉
Then the Math: integrated Volts and considering Cap value became Maxwells (Integrators measure total flux) which divided by gap surface gave me Gauss.
Easier done than said, quite straightforward.
HT-20 for example, you dont need to spent a fortune and this measures to 2 Tesla, its price is about 150 dollars.
If you want good one, then DX-150 but it costs 5-6x more.
In case if your magnetic field is low, you can try DIY.
http://rakarskiy.narod.ru/publ/iz_seti/prostoj_gaussmetr/2-1-0-15
(online translation translated it into much better technical English than to my minority language but I completely understood this text and everything necessary is described there and you only need arduino and a few more common parts).
If you want good one, then DX-150 but it costs 5-6x more.
In case if your magnetic field is low, you can try DIY.
http://rakarskiy.narod.ru/publ/iz_seti/prostoj_gaussmetr/2-1-0-15
(online translation translated it into much better technical English than to my minority language but I completely understood this text and everything necessary is described there and you only need arduino and a few more common parts).
If you need it for production of identical units, just drive it at a known frequency in a gap of measured field strength.
put an amp on the coil out and adjust the gain to produce 1 v rms when it's immersed in a gap of 1 tesla.
To transfer that cal to a different size coil, make a flat gap field generator with, two pieces of iron, a neo, and two iron slugs. Glue the neos at one end of both irons in an elongated C, then glue the two slugs at the opposite end. The slugs are used to constrain the gap field a bit more .Any coil you drive at the slug end will produce voltage on the coil. For varying coil diameters, the output will be proportional to the number of turns traversing the field. I suspect you will find that it can be used to cal a good range of coil diameters.
John
I make them from scratch.
I buy full (6 meter long) cold rolled round bar steel in 12L14 alloy, which is magnetically acceptable and much easier for the Lathe operator, since is gives out short (couple mm long) swarf; actual low carbon steel (SAE1008) is very soft and gives you impossibly long swarf which clogs the lathe, sharp edges can cut opeerator´s fingers or worse, a mess.
I turn them to proper size, including a little plate thickness length stump which is used to "rivet" it to back plate.
Plates are hydraulic press punched (if 130 mm or less) or Oxy flame/water jet/Laser cut from 6.35 mm, 7.92 mm or 9.5mm cold rolled SAE 1008 steelplate ... what everybody else uses by the way.
Then backplate and polepiece are pressed together, again in a Hydraulic press, I spot (resistance) solder front plates to stamped steel frames/"bells" orin the heavier models bolt gthem to cast aluminum ones.
All steel parts are zinc plated, best finish in my view.
Paint takes up too much thickness and when scratched allows water + oxygen in, rust then travels under paint, while zinc "sacrifices itself" to protect steel.
I wind my own coils, but buy cones and spiders.
Oh, and magnets, of course.
As an example (just what I had nearby), here you have a turned 1.5" diameter polepiece, showing the mounting stub , (I like to drill a vent hole through them), a similar polepiece mounted on a 130 mm backplate (my Guitar Speaker standard) and a 19 mm polepiece mounted on a 76mm backplate for my high power 4" cone tweeters,which I use in biamped Bass and Keyboard cabinets.
Hi,
I have never seen ready-made loudspeaker motor parts on offer.
I think that those who make their own motors must decide about the design and the material of the motor and must then either make the parts themselves or have them made by a specialist.
I use pure iron for my motor parts (ALLIEDPUREIRON TM) and am very glad I have found a company which is willing to make the parts in small quantities. This is not cheap.
You may have more luck with standard sized magnets, but be careful with pre-magnetized units, especially Neodym. This can get dangerous. I use a custom-made large N52 magnet, as my design couldn´t make use of standard units.
This is expensive AND dangerous. Don´t ask me how I know.
Back to the gaussmeter, the TD 8620 standard probe just barely fits my 1,3mm x 10mm airgap. Any smaller wouldn´t work, but a 1mm airgap is an adventure on his own...
Good luck!
Mattes
I have never seen ready-made loudspeaker motor parts on offer.
I think that those who make their own motors must decide about the design and the material of the motor and must then either make the parts themselves or have them made by a specialist.
I use pure iron for my motor parts (ALLIEDPUREIRON TM) and am very glad I have found a company which is willing to make the parts in small quantities. This is not cheap.
You may have more luck with standard sized magnets, but be careful with pre-magnetized units, especially Neodym. This can get dangerous. I use a custom-made large N52 magnet, as my design couldn´t make use of standard units.
This is expensive AND dangerous. Don´t ask me how I know.
Back to the gaussmeter, the TD 8620 standard probe just barely fits my 1,3mm x 10mm airgap. Any smaller wouldn´t work, but a 1mm airgap is an adventure on his own...
Good luck!
Mattes
Attachments
Hi Boden,
Indeed. This motor was mainly designed with FEMM, but of course was as well evaluated in real life at various steps. Reality showed good accordance with FEMM and proved not only that FEMM is working as claimed, but also that my method how to work with FEMM is correct.
The goal was to have a brutally strong motor with saturated pure iron pole plate and core, in order to have the strongest control and a max Qe. Don´t judge this motor by size. Many expensive FR drivers have gigantic motors, but if you look at the Qe values of the final driver, the energy seems to have been lost somewhere underway...
Despite their aesthetics, these motors are a true form-follows-function design. The geometry is optimized to the very last detail; every change of dimensions or shape of a mm or so will deteriorate the result.
All the best
Mattes
Indeed. This motor was mainly designed with FEMM, but of course was as well evaluated in real life at various steps. Reality showed good accordance with FEMM and proved not only that FEMM is working as claimed, but also that my method how to work with FEMM is correct.
The goal was to have a brutally strong motor with saturated pure iron pole plate and core, in order to have the strongest control and a max Qe. Don´t judge this motor by size. Many expensive FR drivers have gigantic motors, but if you look at the Qe values of the final driver, the energy seems to have been lost somewhere underway...
Despite their aesthetics, these motors are a true form-follows-function design. The geometry is optimized to the very last detail; every change of dimensions or shape of a mm or so will deteriorate the result.
All the best
Mattes
Attachments
Very slick - is the left hand edge of the picture the centre-axis? 1.6T is pretty extreme - can it be even higher?
Hi Mark,
Yes, the left edge is the central axis, that´s how FEMM works with circular motors. Compare to the pictures above of the motor in reality. The central recess is a thread for the phaseplug.
1.6 T is extreme, yes. It is the simulated value - reality measurements show a value of 1.54 T. That´s close enough for me.
I wanted maximum motor control, as I have friction-free low-loss carbon arm front and rear suspensions which result in a very low mechanical damping, therefore the motor has to have control. It is extreme as this results in a very low Qts (0.09) driver which will not do any bass in "normal" applications.
With the given airgap dimensions (for a 40mm diameter voicecoil - chosen for various reasons - and a length of 10mm, to provide +-2.6mm linear travel of the 5mm coil, while having a width of 1.3mm as a compromise between flux and mechanical reliability) it cannot be any higher - as you can see, pure iron metal parts are already saturated at 2.14T, the maximum permeability of pure iron.
Whatever change in dimensions of metal parts or magnet you do, it will not get any better. In a simulation you can quickly explore gigantic magnets, field coils and changes in quality and shape of the core and plates. Nothing will give a better result, I have been there and have done it.
The only ways to have about 8% more flux in the airgap are to make the airgap more narrow (which means a higher risk of issues with production tolerances) or to use Permendur for the metal parts (or combine both, of course).
The use of Permendur would be extremely expensive in a small series, this motor already is a lot more expensive than many other drivers. It employs custom made pure iron (ALLIEDPUREIRON TM) parts in the complete motor and a custom made N52 magnet.
All the best
Mattes
Yes, the left edge is the central axis, that´s how FEMM works with circular motors. Compare to the pictures above of the motor in reality. The central recess is a thread for the phaseplug.
1.6 T is extreme, yes. It is the simulated value - reality measurements show a value of 1.54 T. That´s close enough for me.
I wanted maximum motor control, as I have friction-free low-loss carbon arm front and rear suspensions which result in a very low mechanical damping, therefore the motor has to have control. It is extreme as this results in a very low Qts (0.09) driver which will not do any bass in "normal" applications.
With the given airgap dimensions (for a 40mm diameter voicecoil - chosen for various reasons - and a length of 10mm, to provide +-2.6mm linear travel of the 5mm coil, while having a width of 1.3mm as a compromise between flux and mechanical reliability) it cannot be any higher - as you can see, pure iron metal parts are already saturated at 2.14T, the maximum permeability of pure iron.
Whatever change in dimensions of metal parts or magnet you do, it will not get any better. In a simulation you can quickly explore gigantic magnets, field coils and changes in quality and shape of the core and plates. Nothing will give a better result, I have been there and have done it.
The only ways to have about 8% more flux in the airgap are to make the airgap more narrow (which means a higher risk of issues with production tolerances) or to use Permendur for the metal parts (or combine both, of course).
The use of Permendur would be extremely expensive in a small series, this motor already is a lot more expensive than many other drivers. It employs custom made pure iron (ALLIEDPUREIRON TM) parts in the complete motor and a custom made N52 magnet.
All the best
Mattes