pic: $14 1000ppr, 16384-position encoder

[asid61]

loading photo ...

[mman1506]

I'm excited for this and have been thinking of doing something like it myself for a while now. With the Canadian dollar grayhills are $75 each which is way to much to justify over a CTRE encoder but I still love the bearing supported format for ease of use especially as WCD encoders.

[asid61]

Originally Posted by mman1506

I'm excited for this and have been thinking of doing something like it myself for a while now. With the Canadian dollar grayhills are $75 each which is way to much to justify over a CTRE encoder but I still love the bearing supported format for ease of use especially as WCD encoders.

If you have a 3D printer, it's possible to make one of these into a shaft-style format (see my AEAT-6600 encoder).

[Ether]

Originally Posted by asid61

-Max 28,000 rpm (stick it on the back of a 775pro!)
-Can be run at this speed with Talon SRX or roboRIO encoder readings
-4000cpr (aka 1000 ppr) incremental with index pulse
-14-bit (16384 position) absolute accessed via SPI

Is the 28,000 rpm a mechanical limit or an electronic limit?

What's the speed limit for absolute angle?

How are you defining "cpr" and "ppr"? It seems to be the opposite of how US Digital defines these acronyms.

[mman1506]

Originally Posted by asid61

If you have a 3D printer, it's possible to make one of these into a shaft-style format (see my AEAT-6600 encoder).

Yeah I saw that, very cool.

I've been thinking of doing something similar with a AS5304/6 encoder to make a 1/2 hex through axle encoder.

[asid61]

Originally Posted by Ether

Is the 28,000 rpm a mechanical limit or an electronic limit?

What's the speed limit for absolute angle?

How are you defining "cpr" and "ppr"? It seems to be the opposite of how US Digital defines these acronyms.

Whoops. Honestly, the CPR/PPR gets me every time. I edited the description to match it.
There's no mechanical limit that I can think of, as it's just operating off a contactless magnetic field anyway, and due to the magnetic field operating limits I wouldn't think induced current is a problem. It does run an internal computer for the incremental conversion so the 28k rpm may come from that as an electronic limit.

AFAIK the 28k rpm also applies to absolute (datasheet doesn't single out either mode of operation), but it could be higher than that.

Originally Posted by mman1506

Yeah I saw that, very cool.

I've been thinking of doing something similar with a AS5304/6 encoder to make a 1/2 hex through axle encoder.

PM me about that if you'd like. I designed one for an AS5304 a few months ago but never got around to building it because of lack of good 3d printing resources.

[marshall]

Schematics? Oshpark link? More details?

[AdamHeard]

Originally Posted by Ether

Is the 28,000 rpm a mechanical limit or an electronic limit?

What's the speed limit for absolute angle?

How are you defining "cpr" and "ppr"? It seems to be the opposite of how US Digital defines these acronyms.

And US digital calls it opposite of several other vendors too

[Ether]

US Digital defines "CPR" as "cycles per revolution". It's the same value whether you're talking about one channel or two. And there's no ambiguity concerning rising and falling edges.

If "CPR" is defined as "counts per revolution", the number is ambiguous unless you specify whether it's one channel or two, and whether you're counting only rising (or falling) edges or both.

[asid61]

Originally Posted by marshall

Schematics? Oshpark link? More details?

I'm getting them up on grabcad as we speak (well I'll get around to it soon). Come back tomorrow.
Here's an Oshpark link for now: https://oshpark.com/shared_projects/zTrcd4cp
Just an AS5047P chip and 0805 leds, capacitors, and resistor. The "incremental" led is nonfunctional so I'm leaving it off for now.
I have no more details to give... I'll see if I can get an oscilloscope on it soon.

Originally Posted by AdamHeard

And US digital calls it opposite of several other vendors too

Cripes...

Originally Posted by Ether

US Digital defines "CPR" as "cycles per revolution". It's the same value whether you're talking about one channel or two. And there's no ambiguity concerning rising and falling edges.

If "CPR" is defined as "counts per revolution", the number is ambiguous unless you specify whether it's one channel or two, and whether you're counting only rising (or falling) edges or both.

Cycles, ok. That makes a lot more sense than the "counts" I was thinking of in my head. Thank you!

[asid61]

Eagle files and CAD are posted here.

The folder MVRE-1.2 has everything you need.
As a side note, MVREX is a hex shaft encoder using the AS5304 if anybody is interested.

[marshall]

Originally Posted by asid61

I'm getting them up on grabcad as we speak (well I'll get around to it soon). Come back tomorrow.
Here's an Oshpark link for now: https://oshpark.com/shared_projects/zTrcd4cp
Just an AS5047P chip and 0805 leds, capacitors, and resistor. The "incremental" led is nonfunctional so I'm leaving it off for now.
I have no more details to give... I'll see if I can get an oscilloscope on it soon.

Thanks! Can you give values for the caps and resistor?

Suggestions:
-Some of the vias are a bit close to pads and should be moved. I'm not sure if you auto routed this or not but the routing is not the best.
-Add some mounting holes to aid in stability for a 3d printed case.
-If it were me, I wouldn't break out all of the pins unless you need to.
-I'd also flip the solder points to be single side instead of thru-hole. Can then put it in a case and solder it without having to deal with both sides
-Make it square and/or panelize it. Round is the worst shape for per-inch pricing the way Oshpark does it.

Don't take any of those as complaints, these things are MFD.

[Gdeaver]

AMS has evaluation boards for most of their magnetic encoder boards. Digikey link.
http://www.digikey.com/product-searc...2557?k=as5047p
We were going to use the AS5600 evaluation board for our swerve but 2016 was not a swerve year.

[asid61]

Originally Posted by marshall

Thanks! Can you give values for the caps and resistor?

Suggestions:
-Some of the vias are a bit close to pads and should be moved. I'm not sure if you auto routed this or not but the routing is not the best.
-Add some mounting holes to aid in stability for a 3d printed case.
-If it were me, I wouldn't break out all of the pins unless you need to.
-I'd also flip the solder points to be single side instead of thru-hole. Can then put it in a case and solder it without having to deal with both sides
-Make it square and/or panelize it. Round is the worst shape for per-inch pricing the way Oshpark does it.

Don't take any of those as complaints, these things are MFD.

1. The vias are on the pads to make it easier to fit stuff. I figured that they're all part of the same signal path anyway, so unless it starts messing with my soldering it's not a problem (I think).
I hate the autorouter. It never gives me what I want...
2. It's meant to snap into the case, and mounting holes cost money. Seriously though, I have made some with mounting holes and some without and I've found that if anything the ones that just snap in are easier to deal with albeit definitely more sketchy. The headers align it with the casing to some extent, or you can glue it in.
3. The only pin I may not need to break out is the 3.3v pin, as the 5v pin is suitable. But, I wanted a power supply pin available on SPI interface. Everything else is unfortunately necessary.
4. Hm I'm look into making squares, I might be able to get it a bit more compact that way too.