Shield TLE5206

[GardenRobot]

---

This is a companion discussion topic for the original entry at http://maslowcommunitygarden.org/Shield-TLE5206.html

[jwolter]

[2cents]

Hi jwolter!

I would like to take the opportunity to announce shields are available now!

Thank you

[bar]

@2cents Are you planning to make more of these?

Since they aren’t available I made a pull request to take it down, but if you are planning to bring them back we’ll delete my pull request

[dlang]

leave them up as the design data is there, just mark them as no longer
available.

David Lang

[2cents]

Yes, absolutely! I’ve been very busy lately and was planning to work on inventory over the weekend, they are back in stock now. Sorry for the delay.

Gab

[dlang]

could you make a version of these that will work with the Due (including an
eeprom and level converters between the existing shield connections and the
arduino)

something like these chips:

0.5US $ |3.3V 5V TXS0108E 8 Channel Logic Level Converter Convert TTL Bi...

ideally, have a pingout that matches a due, and a place where a ESP32 could be
soldered on.

but even if you don’t make it have a nice pinout for any particular processor,
just having a board that we know will work well at 3.3v would be very valuable.

David Lang

[Joshua]

I thought I would provide a link in this thread, just to make sure it is well-publicized.

@Joshua

Hey, I wanted to pull a recommendation from another post into this thread.

There was a recommendation to put a 0.01 uf capacitor on the receiver end of the encoder wires. This reduced the high frequency electrical noise coming from the wires, and reduced/eliminated the occurrence of false positive and negative pulse-detections. Can capacitors be added to the board, to serve this purpose? We need to be careful to appropriately size the capacitors, so they don’t cause missed pulses.

Here is a link to the thread:

• 2nd Generation MaslowCNC Controller on GRBL

Also, this one:

@Joshua

dlang:

you would have to have a false pulse on one wire at the same time as a real
pulse on the other wire (or a second false pulse on the other wire) before your
count would be off.

That is not true.

• You could have two simultaneous false-pulses.
• You could have two false-pulses that happen in coordination:
  • one high followed by the other high
  • one low followed by the other high
  • etc.

I am not so sure that this is such an infrequent thing. The processor operates at a 16 megahertz rate. That means there is the potential for 16X10^6 false positives/negatives every second. The probabilities don’t have to be very high for this to happen on a regular basis. It is not a question of -if- it happens, but how frequently.

Further, the external circumstances that cause false-positives and false-negatives happen very similarly for both encoder lines. In other words, both encoder wires are subject to the same EMF noise that causes voltage deviations. If there is a positive voltage deviation on one line, the other line is subject to the same environment, and likely experiences the same positive voltage deviation. So, if one encoder line experiences a false-positive pulse, it is very likely the other encoder line will, too, because they are run in parallel and receive the same ambient noise.

dlang:

Adding a cap to the wire will reduce false pulses but it will also slow down
real pulses, so I’m not sure how much it will help vs hurt. I would shield the
wires first.

I agree with as a potential issue. However, if we appropriately size the capacitors, this won’t happen.

dlang:

I would shield the wires first.

I like the idea of shielding the wires, and I think it is a very good step. However, capacitors cost something like $.01, which is a really cheap and easy modification. I would like to do both, because I see this as an unknown. I would really like to see some community members do some work to analyze this; both the effect of the capacitor and the shielded wire.

[2cents]

Hi David!

I just found out about the Due hardware/firmware initiative, it looks great!, I would love to help in any way I can, however it will take me some testing time on a breadboard before commit to produce them. Unfortunately I’ve been very busy lately and I won’t have time to put my head into it until I complete the kits/parts restocking process that I’m in right now. As soon as I have everything back on tracks I will give it a try.

Gab

[2cents]

Hi Joshua!

I don’t think this is big issue, but I will hook up my oscilloscope and test it as soon as I have time, the induced noise is likely from the DC motor power wires and probably isolating them would be the best solution. What do you think of a Schmitt Trigger for getting rid of higher frequency noise?

[Joshua]

@2cents

What do you think of a Schmitt Trigger for getting rid of higher frequency noise?

I think the capacitor is simpler. As long as there is some kind of information that justifies the solution over alternatives.

For the capacitor solution, one would have to demonstrate that the frequency of the filter were above the frequency of the pulses coming from the motors at max speed. Substantial margin would be desirable, in case the maximum observed pulse-frequency increases. If the capacitor is sized correctly, tour oscilloscope should show reduced noise, but the pulses should remain largely unchanged.

@2cents

the induced noise is likely from the DC motor power wires and probably isolating them would be the best solution

I agree about the DC motor part. However, capacitors cost about $.01 each, and can be a very simple improvement.

[dlang]

For the capacitor solution, one would have to demonstrate that the frequency
of the filter were above the frequency of the pulses coming from the motors at
max speed. Substantial margin would be desirable, in case the maximum
observed pulse-frequency increases.

so max speed currently is ~2.5K steps/sec (so each signal line being ~1.3KHz
signal)

If the capacitor is sized correctly, tour oscilloscope should show reduced
noise, but the pulses should remain largely unchanged.

Agreed.

This is an area that I know enough to ask about problems, but not enough to have
a feel for what solutions will work.

David Lang

[Bee]

Hi folks!

I’m experimenting with this and am contributing to this on this thread:

2nd Generation MaslowCNC Controller on GRBL

Just wanted to create a reference here in the forum to some work that has been going on with electronics and firmware around the MaslowCNC. This info may interest folks that are looking for a little more out of their machine and GRBL compatibility. May be a platform to build on for a 2nd generation or Pro-version. All the available information can be found at: MaslowDue

Let’s concentrate on one thread rather than fragmenting?

Thank you

[2cents]

First off, I want to thank you for your suggestion, inputs and exchange of ideas like this contribute to improve the project.

Here are my thoughts about this:

For a false pulse to occur (if they occur at all), it has to be an induced spike similar to a real pulse (amplitude/frequency), in which case it will be impossible to filter, the only solution to it would be isolating the wires

However, noise may be preventing the encoder pulses from being decoded at random times, and in this case a capacitor may help.

I think it’s worth testing it and if it helps on cleaning the square wave, I will incorporate your recommendation.

[dlang]

frequency is not a factor here, the ‘frequency’ of the pulses varies based on the speed you are moving, and the ‘frequency’ of the leading edge of the pulse is close to infinite.

the amplitude matters a lot, if you have a 5v signal, it typically takes about 3v to trigger it. If you have a 3.3v signal, it typically takes about 1.8v to trigger it. anything below that point doesn’t matter.

what matters in the timing, look at this truth table on this page and see what happens when a line goes high and then drops back down before the other line changes, you will see that the count goes up and then back down (no effect)

Understanding Quadrature Encoding

This is why I am saying that a faster cpu should have less problem with noise as any noise spikes are more likely to go away and be noticed.

Adding a capacitor to the input will slow down the rise of all pulses, which will distort the legitimate signal, but will also slow down the noise, potentially to the point where the noise never reaches the threshold to trigger a state change.

If you have noise on both wires, 3/4 of the time it’s the same as having noise on only one wire (as one or both of the lines are already high), In that other 1/4 of the time it’s an invalid transition in theory. In practice, it depends on the implementation, if the cpu processes state changes on each line individually, it processes one before the other and may throw the count off.

[2cents]

Hi David!

I was referring to the ‘noise’ frequency, not the pulses ‘period’ from the encoder.

@dlang

…the ‘frequency’ of the leading edge of the pulse is close to infinite.

Shouldn’t be close to zero?

@dlang

the amplitude matters a lot, if you have a 5v signal, it typically takes about 3v to trigger it. If you have a 3.3v signal, it typically takes about 1.8v to trigger it. anything below that point doesn’t matter.

I agree…for a ‘noise’ spike/burst to be picked up as an encoder pulse, it has to have amplitude enough to trigger a hi/lo, but also a negative spike may prevent a high to reach the trigger, so anything bellow that point may cause a problem too.

@dlang

This is why I am saying that a faster cpu should have less problem with noise as any noise spikes are more likely to go away and be noticed.

Shouldn’t a faster ‘cpu’ be picking up more noise?

Anyways… whatever it is, false pulses, missed triggers, attenuation, ground loop, or even harmonic frequencies from the power supply, until we test it we’re just shooting in the dark, I will check it out when I have a chance and post the results.

Thanks for the link

[Bee]

Has anyone here built/designed a motherbaord above a 386 or a RF engineer?

Thank you