Little Bee B Field / Current Probe

[Weston]

As a personal project during lockdown I have been developing a low cost B field / current probe based on a cool AMR magnetic sensor I found on Digikey. It's called the Little Bee field probe, get the pun?  The goal is serve as a low cost alternative to the $$$$ DC current probes sold by Tektronix and the $$$ I Prober 520. Similar to the I Prober 520 it allows me to probe the current in a PCB trace without having to modify the PCB.




I recently assembled my first prototype and it works a lot better than I hoped for. The -3dB is to 9MHz, which is high enough for most of my current sensing needs. The sensitivity at full bandwidth is a few 10's of mA. Based on some feedback I received I have been considering trying to fund a batch of units for sale via crowd funding before open sourcing the design.

To try decide if crowd funding this would be practical I put together a market survey. If anyone who is thinks this is interesting would fill it out it would be a great help!  https://forms.gle/1jCCtYXS3TVfTNUC9

[Twospoons]

Do you have geomagnetic field cancellation? Adjusting that is a PITA with the  I Prober.
Also , whats your noise floor?

[Steve Ward]

This looks like a great tool to have, even if the relative amplitude of the signal isn't too accurate.  A lot of use cases would allow for an extra "cored" sensor, usually at some other point in the circuit, to be used as a reference for the amplitude of the current.  One particular use case would be measuring switching losses, but im not sure the bandwidth is there for fast switches.

Im curious, what would be roughly the upper limit on current it can measure before saturating?  Assuming the wire size is comparable width as the sensor.

[station240]

This would have been really handy for debugging this PCB I designed.
Some sort of weird design issue, I 'solved' by removing that part of the circuit entirely.

[Weston]

Thanks to everyone who filled out the survey!

In the past week the sensor I use (HMC1041Z by Honeywell) has gone out of stock at every distributor and I am getting suspicious its getting discontinued. This is causing a lot of personal stress because that part basically has no substitutes and I have not had any luck getting in contact with anyone at Honeywell. I have 8 parts left for future testing. Best worst case is I have to buy the minimum order quantity of 3k units while only having 8 left for development, which I could possibly pull off. Worst worst case is that the project is dead in the water. I guess at least this has been a learning experience.

Do you have geomagnetic field cancellation? Adjusting that is a PITA with the  I Prober.
Also , whats your noise floor?

The buttons on the side control gain / bandwidth select and auto zero. All you have to do is press the button when the probe is a inch or so away from the wire / trace and it will take ~ 0.5 seconds to zero. Noise is 25mG p-p at full bandwidth of ~9MHz. With a maximally close wire that equates to ~20mA p-p (blue is my probe, orange is a CT-2 current transformer)

This looks like a great tool to have, even if the relative amplitude of the signal isn't too accurate.  A lot of use cases would allow for an extra "cored" sensor, usually at some other point in the circuit, to be used as a reference for the amplitude of the current.  One particular use case would be measuring switching losses, but im not sure the bandwidth is there for fast switches.

Im curious, what would be roughly the upper limit on current it can measure before saturating?  Assuming the wire size is comparable width as the sensor.

I have not looked at high currents much with the probe yet, largely due to my lack of a well behaved current source (at this point I would be a bit worried by all the EMI from a tesla coil). The sensor is specified at +-6 gauss though (and can probably go up to +-10 gauss with a small loss in linearity). With the wire geometry I have been testing with this would be +-5A full scale, a higher max range could be achieved with a larger spacing (field goes as 1/D).

This would have been really handy for debugging this PCB I designed.
Some sort of weird design issue, I 'solved' by removing that part of the circuit entirely.

A tool with these capabilities seems useful but given the relatively unproven use cases I fell the $800 cost of the iprober is hard to justify. This is what gave me the idea of creating this probe. If I bring it to market, I want it to be as low of cost as possible. It turns out that making anything hardware related is difficult, so it seems like "as low as possible" is going to be ~$120, but it's still a small fraction of the cost of the iprober with almost all of the same capabilities.

[johnf]

Weston
Field drops off at 1/D cubed not 1/D ie inverse cube law
even small distances cause massive field drop off

[johnf]

Weston although not as fast
https://www.nve.com/webstore/catalog/product_info.php?cPath=27_28&products_id=659
but it is in a different package
I have used NVE stuff before
But i used a temp compensated hall sensor for my magnetometer that goes 0.1 gauss to 40000 gauss

[Weston]

1/D^3 is for a magnetic dipole. In the near field most PCB traces or wires can be approximated as an infinitely long wire, for which the field falls off as 1/D.

I have looked through the entire magnetic sensor parts catalogue of Digikey, Mouser, and LCSC, as well as doing a lot of google searches. The hardest requirement to meet is the high bandwidth, Z axis sensing is very nice to have as the geometry of the probe gets pretty ugly otherwise. The NVE parts you linked are unipolar. The bipolar part only has 300KHz bandwidth.

Still working on getting in contact with the sales people at Honeywell. Other than that, the best alternatives seem to be the AFF756 ( https://www.sensitec.com/service-support/download/file-category/magnetic-field/lang/0 ) and the TMR2503 (https://www.aecsensors.com/components/com_virtuemart/shop_image/product/Magnetic-Tunnelling-Magnetoresistive-(TMR)-Linear-Sensors/pdfs/TMR2503-Datasheet-EN-V1.0.pdf ) .

Both seem like they will at least be somewhat lower bandwidth. Only the TMR2503 is Z axis sensing. I have samples of the TMR2503 arriving soon, still talking to the sales people for the AFF756.

[Weston]

As an update, I submitted the project to Crowd Supply (and have gone through 3 more revisions....) and now I have a pre-launch page!

https://www.crowdsupply.com/weston-braun/little-bee

[harrimansat]

As an update, I submitted the project to Crowd Supply (and have gone through 3 more revisions....) and now I have a pre-launch page!

https://www.crowdsupply.com/weston-braun/little-bee

Great!

[Weston]

Its been going a bit slower than I would like, but all the files are now on github and I am hoping to have the crowd supply campaign up in the next week or so https://github.com/westonb/little-bee-B1

Due to Crowd Supply's requirement for CE regulation, the initial campaign is only going to be USA only. I am still figuring out what would be required to get the probe CE certified, and if it comes to that I think I might want to make a full case for it and make it a more finished product.

[harrimansat]

I need one!!!   

[Mads Barnkob]

Its been going a bit slower than I would like, but all the files are now on github and I am hoping to have the crowd supply campaign up in the next week or so https://github.com/westonb/little-bee-B1

Due to Crowd Supply's requirement for CE regulation, the initial campaign is only going to be USA only. I am still figuring out what would be required to get the probe CE certified, and if it comes to that I think I might want to make a full case for it and make it a more finished product.

There is no such thing as CE certified.

You can put a CE mark on your product to claim that you upholding the standards and national laws. You have to self-sign a certificate of that and ship that along with the documentation for the product. Its called a EU declaration of conformity, and the requirements to the documentation has really gotten easier since standards around 2017-2018. What used to be specific requirements on standards to follow for specific documents, it now just says "bla bla bla schematics ... bla bla if needed".

Lets chat on IRC about it if you want to know more, and check this: https://europa.eu/youreurope/business/product-requirements/compliance/technical-documentation-conformity/index_en.htm

[Weston]

I probably spent a bit too much time optimizing the design, but the Little Bee is now live on crowd supply!

https://www.crowdsupply.com/weston-braun/little-bee

Edit: Updates related to the Crowd Funding of the project are going in this thread now: https://highvoltageforum.net/index.php?topic=1361.0