DIY PCB

[strange island]

Making this a thread for visibility...

[forest thorn]

from what i can find, lekkers are 100-900G roughly

[blissful bluff]

This is not the place for DIY talk. #🔧│keyboard_modding should be

[hasty jolt]

so in general HE sensors can be calibrated just check the datasheet on how. for the signal being noisy well thats what the MCU will have to handle and somehow smooth it out

[hasty jolt]

not sure what keyboard modding has to do with completely custom PCBs

[blissful bluff]

It's a mod?

[hasty jolt]

no they said custom pcb thats not a mod

[blissful bluff]

Custom PCB sounds very much like a mod

[hasty jolt]

so you want people talking about custom music instrument PCBs in keyboard modding?

[cursive basin]

🧠💥

[hasty jolt]

eitherway id lvoe the topic to stay here if anything as we dont have any channel for anything outside of keyboards really.

[strange island]

Nope, not a mod, a completely custom design based around Lekker Switches.

[strange island]

I ordered a PCB with the DRV5053VAQDBZ, which has a sensitivity of -90mV/mT. The sensor output is useful in the first 2 mm of switch travel, but afterwards the sensor output bottoms out. I'm guessing I need the DRV5053RAQDBZ with a sensitivity of -45mV/mT, but since a run of PCB's costs me 100 euros to make, I'd like to not waste that money 😛

[forest thorn]

can you put them further away?

[strange island]

I think you're referring to the I2C type sensors (Allegro A1454) but I'm using Analog sensors

[strange island]

Yes, but it's not ideal

[forest thorn]

yeah makes sense

[strange island]

I think you're right, this datasheet (that I got from Wooting some time ago) mentions 102G at the top of the travel to 905G at the bottom of the travel

[hasty jolt]

ah yeah i checked the datasheet now

i dont really see anything purely spec wise that would scream throws off sensitivity to me though

the only issue i might see is the saturation

although we also recommended the 5053 in the past

the BSAT i can find for all 5053 lies at 90-730G

the drv5055 doesnt seem to be as limited found the range and it would be easily able to sense our switches and stronger ones even

(the BMAX aka absolute maximum flux that you shouldnt go over for the DRV5053 is unlimited so it shouldnt receive damage or such from too high flux)

[hasty jolt]

thats not a datasheet for our switch

but the G range sounds about right

[hasty jolt]

you need to check what TI calls BSAT which is the amount of magnetic flux needed to saturate the sensor aka it wont change the reading if you apply more

[strange island]

Yeah I did that, but wrong 🤦 I ordered the DRV5053VA because they have a Bsat of 9 mT (900G, which matches the datasheet for the switch), but now I'm only getting half of the range, with the sensor output bottoming out after only 2 mm of travel. I'm probably going to need double that Bsat, so either the DRV5053RA or the DRV5053EA...

[hasty jolt]

9mT is 90G

the drv5053 listed 1mT as 10G if i saw that right

unless theres a documentation issue on TIs side i wouldnt see any 5053 variant that could do 900G (or 90mT)

yeah checked with a conversion table an 1 mT is indeed 10G not 100G

so the drv5055 is the better option as they offer sensing ranges of 169/176mT at 5/3.3V respectively

although raising the magnetic further away of course also lessens the gauss acting on the sensor

idk how much distance you can create but id try moving the switch further away and see how far you need to move to gain back the sensing range

and ofc if you dont loose the start then

[strange island]

For using the PCBs I already have, this is ideal, but I'm trying to make a compact design (hence using Lekker switches in the first place) so adding 4mm or more of empty space is not ideal...

[strange island]

I don't think this is true. The DRV5053VA is too sensitive, so going for a higher sensitivity will make the problem worse.

[hasty jolt]

sensing range means how much flux it can take before saturating

looks quite usable to me for the drv5055

[strange island]

Yeah seems you're right, doubting between DRV5055A3 and DRV5055A4. The switch goes from 10mT to 90mT, so 88mT range of the A3 would only miss out on 2 mT of travel but use as much of the output range as possible, the A4 has more than enough range, but that also means less accuracy.

[hasty jolt]

keep in mind that those 20G can also still be worked around by raising the switch and pcb thickness

[strange island]

The datasheet for the switch I posted here mentions the flux density is a north pole, can anyone confirm that? Otherwise I could use the DRV5056, which is unipolar, so a bigger output range, but only responds to s-poles. It would eliminate the need for analog amplification of the signal.