Low Cost Encoders

[mman1506]

Recently a number of encoders have popped up on ebay like this and this with excellent prices (<15$)! Considering the cost of some name brand rotary encoders they look like a excellent option for applications that do not require perfectly calibrated sensors. They are also available for purchase in small quantities unlike a number of other encoders. Does anyone have any experience with them?

[techhelpbb]

Not yet but I soon will .
I will let you know when they arrive.

[mman1506]

Originally Posted by techhelpbb

Not yet but I soon will .
I will let you know when they arrive.

Great, I would grab a few for myself to review but It will be a while until I get purchasing access. Most of the items I order from china take about 2-3 weeks to arrive.

[techhelpbb]

Got the encoders today.

5 wires each
A & B
Vcc / System ground
Shield ground

Color code in Mandarin Chinese.
Actually not a problem.
http://mandarin.about.com/od/vocabulary/a/colors.htm

Black = system ground
Red = Vcc
Shield ground is black higher gauge or bare
White, Yellow, Green are A & B

Will power them up tomorrow.
These are likely too big for AndyMark gear boxes without shaft couplers and spacers.

[Tungrus]

Originally Posted by mman1506

Great, I would grab a few for myself to review but It will be a while until I get purchasing access. Most of the items I order from china take about 2-3 weeks to arrive.

I would rather pay little more and buy from someone who has designed and made the encoders (or anything or from any country for that matter) as long as they own the intellectual properties for the product. It's unfortunate that many companies steal designs or copy products from others and make quick buck.

[techhelpbb]

Originally Posted by Tungrus

I would rather pay little more and buy from someone who has designed and made the encoders (or anything or from any country for that matter) as long as they own the intellectual properties for the product. It's unfortunate that many companies steal designs or copy products from others and make quick buck.

No offense but optical encoders have been around for decades.

As much as I too would like to take the 'high road' here:

First let us start with a basic question: who invented the first shaft encoder?
You are tossing around the suggestion that they stole it.
So let us demonstrate from whom.
Consider this a lesson for the students.
Hint: Look up Murray Hill, NJ....Dr. John Northrup Shive....Bell Telephone Laboratories.

I take some issue with the use of non-American parts to operate U.S. FIRST as well.

Then again ask National Instrument to catalog where their parts and assembly come from.
The $2,000+ 8 slot cRIO myself and another mentor bought was shipped to us from Turkey.
That doesn't sound like 'Made in America' to me.

So why is it okay to use labor unknown from one country but not another?
Can you prove that abuses of labor are any less in one place than the other?
Have you personally checked?

Keep in mind when I bid on the 2015 control system I was going to do *all* the assembly in the United States.
I flatly stated in writing that the additional cost was irrelevant to the potential quality control issues.
Never mind the potential language barrier (my Chinese is laughable).
I did not suggest these other countries were thieves without evidence.
Here's your chance show your evidence.

I note that in the topic about the RoboRio you never asked where that product came from but did admire the cost savings.

Further *who am I* to decide that teams with deep pockets are more entitled to working encoders than teams just managing to show up? Why should those with less in their purse not buy parts probably made by those with less in their purse?

Now I have encoders to test you seem relatively new here you might want to take note:
Of all the people you could try to insert your politics with >I am the last person on Earth< to do it with.
If I think your case is questionable I will poke holes in it.
It is not personal nor is it about whom you are with politically or otherwise.
It is merely my scientific method which is quite incompatible with politics.
Do not assume that as a member of Team 11 they control me either or that I speak for them.
I am me and my choices are my own.

Sorry if that's a little rough but when you end on an angry emoticon you should recognize the types of response you might get back.

[techhelpbb]

Have not forgotten about this.
Dealing with a Linux bug that has some urgency.
Will have an answer before Sunday.

The A&B outputs appear to be open-collector.
So I will set these up and turn them with a motor then sample the output quality and provide it.

I will also work up a circuit to divide the output down to encoder resolutions more common to FIRST.

[Tungrus]

Yes, I do believe that there are companies who steal intellectual properties. My post was related to intellectual properties and not where it is made. It doesn’t matter who steals from whom, whether its US based company who steals from non US company or the other way around. The company that designs products invests lot into their design and testing. More over these companies spend lot in making sure that their product stands by their specifications and meets industry standards. I was a Design and Application engineer and I know what goes into a product development.

It’s naïve to think that this does not happen. Recent dispute between Apple and Samsung is one example. And just don’t forget the lady who got electrocuted with fake iPhone charger!

Before you vent your frustration, read the post and understand, my message did not mention anything about labor abuses. What “politics” are you talking about? My point is just plain and simple; before I put my money, I will do my best to make sure it does not go to thieves. It is not always easy or simple, but when someone is selling for half price, it raises a red flag!

This is for those who steal!

[techhelpbb]

Originally Posted by Tungrus

Yes, I do believe that there are companies who steal intellectual properties. My post was related to intellectual properties and not where it is made. It doesn’t matter who steals from whom, whether its US based company who steals from non US company or the other way around. The company that designs products invests lot into their design and testing. More over these companies spend lot in making sure that their product stands by their specifications and meets industry standards. I was a Design and Application engineer and I know what goes into a product development.

It’s naïve to think that this does not happen. Recent dispute between Apple and Samsung is one example. And just don’t forget the lady who got electrocuted with fake iPhone charger!

Before you vent your frustration, read the post and understand, my message did not mention anything about labor abuses. What “politics” are you talking about? My point is just plain and simple; before I put my money, I will do my best to make sure it does not go to thieves. It is not always easy or simple, but when someone is selling for half price, it raises a red flag!

This is for those who steal!

Your argument is still without demonstrable evidence.
I will dismantle these encoders, post pictures and then you can make your case.

It is round, can like, has a 6mm shaft is not theft of anything unless you want to show me the design patent on that form. Surely quadrature encoding itself is not subject to any existing utility patent I am aware of. Prior art would be acceptable as evidence but for something as simple as the way that picture looks it really is pressing on the point of being a bit silly.

Once again without that evidence and without any patience to get that evidence you assume these are stolen designs. It's just not acceptable to me.

I too am an engineer. I have 2 patents pending. I too would be *quite* annoyed if someone anywhere just simply took my work. However I would not assume they did so.

All technology tends to be derivative. You can't blame a tire manufacturer cause they didn't invent the wheel. You can't blame V-Tech because they didn't invent the telephone. I think it would be very ungracious professionalism if one FIRST team went at another for shooting frisbee because they did to. People do arrive at similar solutions from different places all the time. It is not magic that so many teams use wheels to shoot frisbee. Should I be annoyed at that?

As far as Apple and Samsung is concerned that is an example of business usurping engineering and a patent system that is utterly broken. We've got patent trolls sitting on mountains of valid ideas that are trapped because they have *no intention* to use those ideas as anything but weapons against innovation. So even mentioning that is political.

Apple did not invent the GUI and neither did Xerox. That idea was military technology long before they used it.
Yet for a *very* long time Apple ran around accusing Microsoft of stealing their idea.
Really?

STEM means scientific method should have a very firm grip on this community.
If your concern is that this is stolen: wait for me to do the work for you to prove the point.
If your concern is that this is unreliable: wait for me to do the work to demonstrate that.
Doing this as you are is political plain and simple. You are not following scientific method.
You are trying to 'poison the well' with suspicion and allegation.
That sort of business is natural for humans but science has to be above it.

[techhelpbb]

Finally all the parts arrived:

So I have a 5V 4096 position per revolution stepper (it is geared down) with ULN2003 to be driven by an Arduino or the like.

A dual LS7366R quadrature encoder PCB to monitor the encoder outputs.
U.S. Digital sells this component I bought the PCB from SuperDroidRobots.com.

A ServoCity set screw hub for the stepper.
A ServoCity set screw hub for the encoder.
They are different size bores. The same 6-32 tapped bolt pattern.

The Chinese encoders to be tested.

Now I need to make a mount and wire this up.

Then I will have set up a realistic environment turning the encoder back and forth for an extended period of time.

I will be able to tell if the encoder is missing pulses because I will mount 2 limits at either end of the rotation (so it won't rotate all the way around).

I will program the stepper to rotate till the encoder hits a point just before the limits on either extreme.

If it misses (by any serious margin) it will smack the limit and an error can be recorded.

Leave that run for a good long time and eventually it'll exceed the entire service lifetime FIRST will likely ever ask of it.

Plus, of course, it gives me lots of time to record samples of the output waveform.

As the entire motion is under Arduino control various speeds of rotation can be tested.

It would be extremely unlikely the LS7366R would roll over or miss a pulse so it would be safe to say any error like that is the encoder or an electronic issue.

[mman1506]

Originally Posted by techhelpbb

Finally all the parts arrived:

So I have a 5V 4096 position per revolution stepper (it is geared down) with ULN2003 to be driven by an Arduino or the like.

A dual LS7366R quadrature encoder PCB to monitor the encoder outputs.
U.S. Digital sells this component I bought the PCB from SuperDroidRobots.com.

A ServoCity set screw hub for the stepper.
A ServoCity set screw hub for the encoder.
They are different size bores. The same 6-32 tapped bolt pattern.

The Chinese encoders to be tested.

Now I need to make a mount and wire this up.

Then I will have set up a realistic environment turning the encoder back and forth for an extended period of time.

I will be able to tell if the encoder is missing pulses because I will mount 2 limits at either end of the rotation (so it won't rotate all the way around).

I will program the stepper to rotate till the encoder hits a point just before the limits on either extreme.

If it misses (by any serious margin) it will smack the limit and an error can be recorded.

Leave that run for a good long time and eventually it'll exceed the entire service lifetime FIRST will likely ever ask of it.

Plus, of course, it gives me lots of time to record samples of the output waveform.

As the entire motion is under Arduino control various speeds of rotation can be tested.

It would be extremely unlikely the LS7366R would roll over or miss a pulse so it would be safe to say any error like that is the encoder or an electronic issue.

Wow, it looks like a very thorough test. My test would be to see if it reasonably well, drop it out a window and see if it still works. : But I think your test is much better.

[Phalanx]

Can you also report back on how much signal jitter there is or isn't when it's stationary. Sure a capacitor could easily address that, just want to know how stable the signal is.

[techhelpbb]

Originally Posted by Phalanx

Can you also report back on how much signal jitter there is or isn't when it's stationary. Sure a capacitor could easily address that, just want to know how stable the signal is.

I was planning on posting the actual signal as sampled by my oscilloscope.
So hopefully you'll be able to get that from there.

Just to manage expectations I wouldn't expect this done this weekend.
With NASDAQ doing a dance yesterday I am a bit preoccupied.
However I will work on it this weekend even if I don't finish it.

[mman1506]

Originally Posted by Tungrus

I would rather pay little more and buy from someone who has designed and made the encoders (or anything or from any country for that matter) as long as they own the intellectual properties for the product. It's unfortunate that many companies steal designs or copy products from others and make quick buck.

If you google "rotary encoder" you'll find 100's of companies that produce encoders that look nearly identical to these ones. Who copied who? No one really knows.You pay a premium to buy from American companies like Industrial Encoders because they don't just sell encoders they sell a product that comes with technical support and the guarantee that you'll get a perfectly calibrated encoder that will last years. This is great for a company like HAAS but not necessary for most teams. The end product is almost entirely different.

[AdamHeard]

Originally Posted by mman1506

If you google "rotary encoder" you'll find 100's of companies that produce encoders that look nearly identical to these ones. Who copied who? No one really knows.You pay a premium to buy from American companies like Industrial Encoders because they don't just sell encoders they sell a product that comes with technical support and the guarantee that you'll get a perfectly calibrated encoder that will last years. This is great for a company like HAAS but not necessary for most teams. The end product is almost entirely different.

I wouldn't be so quick to say a lower quality encoder is sufficient for our use.

If the lower quality means an increased failure rate, we'd pass on that. The encoders we get from us digital aren't crazy expensive, and we'll always pay to not lose matches due to preventable failure.