I just quickly hooked one up here. It's not pretty, but it's working fairly well, and I'm able to trigger the closed pedal notes just fine. I couldn't position the sensor quite right, but I'll make it more optimal when I get more time. The way you orient the sensor and the way the magnet moves are very important to get a linear response.
By the way, the sensor I'm using is DRV5056A1QLPGM, and it provides a lot greater voltage swing than the SS49E given the same magnetic field. The reason SS49E is listed in the manual is because the first generation of eDRUMin devices couldn't supply enough current for the DRV5056A1QLPGM. I'll experiment more with my SS49E, but I need to get some more powerful magnets. I'll also update the recommended sensor in the manual.
Rob, thanks so much for digging deeper into this issue. Your observations and comments got me digging deeper into the sensor set up, and I think you might have made a key observation about the response linearity of my current set up, i.e specifically the magnet approaching the sensor on the arc of the pedal's travel. I was also experimenting with differing diameters of some fairly powerful Nd magnets. Unfortunately the sensor stopped working in the middle of these experiments. Is it possible to damage a Hall effect sensor with too strong of a magnetic field? Anyway, I saw some things that led me to believe that the sensor set up might be the root of the "soft close" problem. I will be ordering some DRV5056A1QLPGM sensors today and revamping the configuration.
However, the other problem remains - that of having to return to an almost full opening in order to trigger the next closure sound. I did experiment with your suggestion of kicking up the xTalk control. That didn't really affect anything, so I took the hat cymbals out of the picture completely. Still need an almost full opening.
So, a lot of work to do here, but I'm feeling pretty confident we'll get the issues beat down. I'll give you an update in a week or so after I receive the new sensors. Really appreciate your accessibility- thanks again.
I don't think you can damage the sensor as you are only ever applying 3.3V to it. The DRV5056A1QLPGM datasheet has lots of great information about how the magnet should move in relation to the sensor, so definitely worth having a look at. When I was playing around with the SS49E I noticed that I had to move the magnet a lot to get the voltage to drop off nicely which is probably why you are have troubles getting the pedal to return to open quickly enough. You shouldn't have that problem with the DRV5056A1QLPGM sensor.
So I improved my installation a fair bit. I was able to use a much smaller magnet. One thing I noticed is that it's easy for the magnet to get too close to the sensor. I needed to make sure the magnet never got closer than about 1cm from the sensor. I also noticed some display bugs. The splash control isn't updated and even the hihat position isn't being updated as often as it should be. I'll see if I can fix those this weekend.
In my design, also using the SS49E I have the project box mounted to the rear of the pedal rather than under. I mounted a pair of hard drive magnets (they look like boomerangs) south side down to the end of the rod. To give an idea of how strong these magnets are - they're hard to pry off a steel door. The magnet travels straight down until it's in front of the sensor. Because of the sensor being inside the project box it can only get to within 3mm of the sensor. The sensor is mounted vertically with the beveled side towards the magnet:
Here's what it looks like on the hi-hat stand. I used a Small Rig clamp to affix the project box to the stirrup. I trimmed the metal strip after this photo was taken. The magnet holds itself to the end of the rod and I used the white material (InstaMorph plastic) to keep it aligned.
I think I'm converted. The hall sensor works as good as the FSR in my VH-13. And if I'm 100% honest, the TI sensor is better. Here's my effective voltage range after calibration and a few manual tweaks.
image.png (3.95 KiB) Viewed 12598 times
I'm going to fab up one of these and just stick it down under my pedal with some doubled sided foam tape.
Hey, dsteinschneider- thanks for your input. Nice design for the sensor mechanism. It brought to light a potentially important point that I hadn't thought about before. In your application, the magnetic field approaches the active sensor surface from a right angle, as opposed to my setup where the field approaches on the same plane as the sensor surface, albeit on a slight arc. Intuitively, one could see how that could have a significant affect on both response time and linearity. I'm going to run some comparison experiments using an o-scope. Glad you brought this point up. Does your set up give you all of the response characteristics you desire? Is it extremely playable?
Rob- thanks for the further information, especially about the distance between sensor and magnet.
Your effective voltage range on the TI sensor is indeed impressive, more than 2x than I've got with the SE49 in current configuration.
Love the concept and simplicity of your pcb design. Are you able to etch your own boards, or do you send the design out to one of the small quantity fabbers? If so, and you have a MPR (min purchase req.), I would be happy to purchase a few from you.
Also, I had noticed the same firmware bugs you described, but didn't bother bringing them up, as I had larger issues to contend with. Glad you got pro-active with them.
This has been a very interesting DYI experience. I would like to think that after its all said and done, and a little more data collected, a small but concise guideline could be created (possibly as an addendum to the manual) for hi-hat DIY'ers that would like to take the hall effect route.
d simplicity of your pcb design. Are you able to etch your own boards, or do you send the design out to one of the small quantity fabbers? If so, and you have a MPR (min purchase req.), I would be happy to purchase a few from you.
I don't etch boards. I can get that PCB manufactured from JLCPCB. I'll order enough parts for make 5 units initially. I can print a simple enclosure for it as well. I think if it's going to be a product, I would also need to design some sort of adapter for a hihat stand to mount a cymbal.
Translation
