Schmidt trigger for the next hardware rev? #113
Replies: 4 comments 4 replies
-
Hi Chris, I did see a slow rise of the encoder signals but I dealt with it in 2 ways:
If you think 17HC14 will benefit your project I have no objections though.
What problem would that solve for the NanoEls H2? I don't think I saw reports saying they're losing encoder pulses or something like that. Or is this to use 1 encoder both with Nano and some other device e.g. TouchDRO? Thanks! |
Beta Was this translation helpful? Give feedback.
-
Oh that was purely to adapt an esp32 to the h2.
…On Thu., Mar. 30, 2023, 02:03 Maxim Kachurovskiy, ***@***.***> wrote:
Hi Chris, I did see a slow rise of the encoder signals but I dealt with it
in 2 ways:
1. Added 1kOhm pull-ups on the A and B lines, this helped a lot
2. Added a logic change where if the encoder pulses are coming fast
enough, I stop reading the B line because spindle can't reverse the
rotation direction due to the mechanical inertia -
https://github.com/kachurovskiy/nanoels/blob/main/h2/h2.ino#L442
3. If you see high-frequency noise, inline ferrite bead is the usual
way of dealing with it
4. For ESD protection, normally 6V ESD diodes to ground are used
though NanoEls doesn't prescribe them in the published schematics
<https://github.com/kachurovskiy/nanoels/tree/main/h2> to keep things
simple for those building it at home
If you think 17HC14 will benefit your project I have no objections though.
That brings up another idea: how about designing an adapter board that
does a pin-translation to the nano
What problem would that solve for the NanoEls H2? I don't think I saw
reports saying they're losing encoder pulses or something like that. Or is
this to use 1 encoder both with Nano and some other device e.g. TouchDRO?
Thanks!
—
Reply to this email directly, view it on GitHub
<#113 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ABMZQGR3HCJ6F3YSNWWO2HLW6UV25ANCNFSM6AAAAAAWMH2IYE>
.
You are receiving this because you authored the thread.Message ID:
***@***.***>
|
Beta Was this translation helpful? Give feedback.
-
It could also allow additional periferals without needing to change the
main h2 board, such as physical end stop switches or output to a touchdro
or something like that, along with level shifters to make an esp
voltage-compatible.
…On Thu., Mar. 30, 2023, 06:51 Chris Chatelain, ***@***.***> wrote:
Oh that was purely to adapt an esp32 to the h2.
On Thu., Mar. 30, 2023, 02:03 Maxim Kachurovskiy, <
***@***.***> wrote:
> Hi Chris, I did see a slow rise of the encoder signals but I dealt with
> it in 2 ways:
>
> 1. Added 1kOhm pull-ups on the A and B lines, this helped a lot
> 2. Added a logic change where if the encoder pulses are coming fast
> enough, I stop reading the B line because spindle can't reverse the
> rotation direction due to the mechanical inertia -
> https://github.com/kachurovskiy/nanoels/blob/main/h2/h2.ino#L442
> 3. If you see high-frequency noise, inline ferrite bead is the usual
> way of dealing with it
> 4. For ESD protection, normally 6V ESD diodes to ground are used
> though NanoEls doesn't prescribe them in the published schematics
> <https://github.com/kachurovskiy/nanoels/tree/main/h2> to keep things
> simple for those building it at home
>
> If you think 17HC14 will benefit your project I have no objections though.
>
> That brings up another idea: how about designing an adapter board that
> does a pin-translation to the nano
>
> What problem would that solve for the NanoEls H2? I don't think I saw
> reports saying they're losing encoder pulses or something like that. Or is
> this to use 1 encoder both with Nano and some other device e.g. TouchDRO?
> Thanks!
>
> —
> Reply to this email directly, view it on GitHub
> <#113 (comment)>,
> or unsubscribe
> <https://github.com/notifications/unsubscribe-auth/ABMZQGR3HCJ6F3YSNWWO2HLW6UV25ANCNFSM6AAAAAAWMH2IYE>
> .
> You are receiving this because you authored the thread.Message ID:
> ***@***.***>
>
|
Beta Was this translation helpful? Give feedback.
-
I'm sorry I'm not very clear, I'm basically thinking out loud. I enjoy modularity and being able to piece gear together with whatever I have on hand, so I was thinking about 2 things at the same time and asked them in the same discussion. Firstly, I think schmitt triggers are the bees knees, but yes your solution for interferance works fine. Secondly, making the next version (h3?) somewhat modular so you can use adapter boards to add different MCUs or other filtering circuitry or whatever the user may feel the need for is a kind of design I like. For example, a header to plug in LCDs, but also a header for an I2C lcd as well. I have an I2C version of the 2004, 4 wires, really easy. Saves some pins if someone wants to use an esp8266 or something. I first started down the world of stepper control via the 3d printing world, and the early 3d printing firmwares tended to attempt to allow people to piece stuff together with whatever they could find, which is definitely not a problem here for this project because all of these components are readily available. Just a friendly design discussion, no need to actually solve problems :) Feel free to close it. I did whip up a home etched version of the H2 board, and I've gotta say, it works really well. There are a few tracks that are a little close together that didn't etch well, I had to manually repair those, but it wasn't a big deal. Have you considered selling parted out kits for people to solder together? I think they could sell well. |
Beta Was this translation helpful? Give feedback.
-
I saw your historical repo of the ESP version, and I was also working on a TouchDRO at the same time. I noticed TouchDRO's esp implementation uses a Schmitt trigger for signal inputs to help clean things up and give a nice square edge, especially if the edge is noisy. I find it works really well with long wires with high capacitance where the pulses have a slower rise and get more sawtooth shaped, and the slight hysteresis helps a lot in noisy environments. it seems to be more esd resistant than the ESP32 seems to be also. Have you considered using one, such as the 17HC14? Could combine it with a simple transistor for level shifting too, to support encoder voltages up higher than 5v.
https://www.ti.com/lit/ml/scea046a/scea046a.pdf?ts=1680067926446
That brings up another idea: how about designing an adapter board that does a pin-translation to the nano, includes a schmitt trigger/level shifting circuit, and plugs straight in the same space on the H2. A simple define at the top of the pin configs could switch pins. Thoughts?
Beta Was this translation helpful? Give feedback.
All reactions