OpenServo.com Forum Index OpenServo.com
Discussion of the OpenServo project
 
 FAQFAQ   SearchSearch   MemberlistMemberlist   UsergroupsUsergroups   RegisterRegister 
 ProfileProfile   Log in to check your private messagesLog in to check your private messages   Log inLog in 

Open-Encoder-MG995
Goto page Previous  1, 2, 3, 4, 5, 6, 7, 8  Next
 
Post new topic   Reply to topic    OpenServo.com Forum Index -> Hardware
View previous topic :: View next topic  
Author Message
Nocturnal



Joined: 06 Dec 2009
Posts: 3

PostPosted: Sun Dec 06, 2009 2:45 pm    Post subject: Reply with quote

ginge wrote:
As for the I2C reading of the chip... this may be a problem. I got it into my head we would be using the SPI functionality of the chip.


Sorry to stick my largely uninformed head in, but your using the AS5046 right? could you not use the AS5045, which seems to use half duplex SPI? 18bits to a message.
Back to top
View user's profile Send private message
jharvey
co-admin


Joined: 15 Mar 2009
Posts: 362
Location: Maine USA

PostPosted: Mon Dec 07, 2009 12:38 am    Post subject: Reply with quote

I'm not familiar with the SSI bus. Seems like it would use the same number of wires, ect. I don't see why it couldn't be done. Are there some benefits to using this bus?

I see both chips cost the same, both are 12 bit when data is received via digital interface. They look nearly the same to me. About the only real difference I see is that the 5046 offers an analog output is so chosen, while the 5045 offers a PWM output.
Back to top
View user's profile Send private message Visit poster's website
Nocturnal



Joined: 06 Dec 2009
Posts: 3

PostPosted: Mon Dec 07, 2009 1:37 am    Post subject: Reply with quote

SPI can be handled by the hardware in the AVR. Even if that won't work, its far simpler to bit-bang, since you can just clock the data out, you don't have to worry about addressing, start conditions, ack's etc etc etc

Pity about not having an analog output. The max clock frequency for the SSI port is 1mhz, which if my math is right, means you can theoretically poll it about 55k times a second, the maximum sampling rate of the chip is 10k, so you should be able to read the position as fast as the chip is sampling it. Whats the current sampling rate with the ADC in the AVR?
Back to top
View user's profile Send private message
jharvey
co-admin


Joined: 15 Mar 2009
Posts: 362
Location: Maine USA

PostPosted: Mon Dec 07, 2009 10:40 am    Post subject: Reply with quote

There is a chance the OE board I've drawn up can be used by both SSI and I2C. You might want to check into that. If you leave out the I2C resistors, they are just wires.

Is SSI the same as SPI? I see SPI with MOSI, and such signals, they have a different name in SSI. I haven't looked to see if they contain the same signals.

I don't how many A/D samples per time it current handles.
Back to top
View user's profile Send private message Visit poster's website
Nocturnal



Joined: 06 Dec 2009
Posts: 3

PostPosted: Mon Dec 07, 2009 12:08 pm    Post subject: Reply with quote

No, but it really depends on the specific implimentation of SPI or SII. In this instance, from looking at the datasheet, it looks like the SSI interface should be compatible with the SPI interface on the AVR, though the 18bit message might cause some issues since I think the SPI hardware will only do 8bit, but as long as CS is low, you should be able to read it in 8bit chunks. The datasheet is not forthcoming on what will happen if you try to read past the 18th bit though, and you have to pulse the CS line between messages.
Back to top
View user's profile Send private message
robotjay
co-admin


Joined: 01 Aug 2006
Posts: 225
Location: Nebraska, USA

PostPosted: Mon Dec 14, 2009 6:51 am    Post subject: Reply with quote

Nocturnal wrote:
Pity about not having an analog output.

Please correct me if I'm wrong, but the datasheet says that pin 12 is an analog output. Although we lose 2-bits of resolution, at least the soldering becomes simple, and we don't have to bit bang anything.
_________________
"Nothing is fool-proof; For we fools are ingenious and will find a way."
Back to top
View user's profile Send private message Send e-mail Visit poster's website
jharvey
co-admin


Joined: 15 Mar 2009
Posts: 362
Location: Maine USA

PostPosted: Mon Dec 14, 2009 9:38 am    Post subject: Reply with quote

The 5046 does have AN out, the 5045 does not. The OE board I drew up includes pads for the two caps that are called for by the datasheet for the AN output. So the OE can be used as a POT. I have only tested and worked with the I2C on OE.
Back to top
View user's profile Send private message Visit poster's website
robotjay
co-admin


Joined: 01 Aug 2006
Posts: 225
Location: Nebraska, USA

PostPosted: Mon Dec 14, 2009 3:44 pm    Post subject: Reply with quote

Quote:
The 5046 does have AN out, the 5045 does not.

I neglected to notice that Nocturnal is refering to the 5045 for it's SSI interface. You're absolutely right. My bad. I do notice in the 5045's datasheet that they have a PWM output module, which again could use the pot holes that are present on the OSv3.

On a different note: How's it coming with the new shaft drawings? Were you able to find the necessary info? Instead of trying to get the tolerance exact on the first try, could we create an array that includes different sizes near the calculated values, and just "engrave" the size on each part to tell them apart? I'm sure you know how to do this in SolidWorks, but if not, just let me know and I can make it happen. (Unless you hate the idea altogether, and you're confident about the right sizing. I would understand, and you've earned my trust if you think you can get it exact on the first try.)
_________________
"Nothing is fool-proof; For we fools are ingenious and will find a way."
Back to top
View user's profile Send private message Send e-mail Visit poster's website
jharvey
co-admin


Joined: 15 Mar 2009
Posts: 362
Location: Maine USA

PostPosted: Mon Dec 14, 2009 5:02 pm    Post subject: Reply with quote

Good point, I just got to start looking into that yesterday. Holidays you know. My initial math was way off so I didn't post a reply. I do know how to do that engraving in SW. Good note.

I like the idea of trying a series of samples. We can always make them a pinch small and glue them if worst comes to worst, that is if the press fit doesn't work out quite right. Perhaps I should just take a stab at it empiricly instead of predictively. Another good though.

Sorry I've been dragging along on this. I've been busy with other things including I2C comms, and other parts of the project I'm using these in.
Back to top
View user's profile Send private message Visit poster's website
jharvey
co-admin


Joined: 15 Mar 2009
Posts: 362
Location: Maine USA

PostPosted: Wed Dec 16, 2009 10:28 am    Post subject: Reply with quote

So about those crazy numbers I was getting. Here my basic math on the subject. For starters, I'll assume that we have one rib that interferes with the ID of the magnet. That one rib is .005 in wide, and .1 inches long. So it's surface area that interferes with the ID of the .0005 in^2. Both tensile and flexure strength are over 10,000 PSI, so I'll start by using 10k psi. That means 5 lbs of force for each rib that is .005 wide (10,000lbs/in^2*.0005in^2). If that rib goes to .015 wide, it's up to 15 lbs of force for each rib. If we use 4 ribs we could be talking about 60 lbs of force in the ID of the magnet. I don't know the friction coefficient for this type of insertion friction when being pressed in, but if I assume it's .5, we would need to press it in with 30 lbs of force.

The 10,000 psi is also likely on the low side. I see it only deflects up to 16% before fracture, and it's tensile and fracture strength is about 300,000 psi. I'm not sure what the stress strain curve looks like for this material, but that likely means this 300,000 PSI happens at around 5% deflection or less. Basically the material is nearly uncompress-able. If that's the case I would expect closer to that 300,000 not 10,000. So those insertion pressures could go as high as 900 lbs. I know we wouldn't expect that high a pressure. What would happen is it would fracture and break before going together. I think we could expect it to break even under a 60 lb force.

Perhaps if we make the rivet like stub small and use a less rigid material over the stub like shrink wrap, as a filler.

http://www.mcmaster.com/#7132k512/=4yd60k

This stuff is very ductile, and will form fairly easily. I think it will form and kind of flow in the nooks and crevices similar to a nylon insert in a screw. If we can't find a happy filler material, at least with it being a clearance fit, we can always go back to the super glue method.

Thoughts?
Back to top
View user's profile Send private message Visit poster's website
jharvey
co-admin


Joined: 15 Mar 2009
Posts: 362
Location: Maine USA

PostPosted: Thu Dec 17, 2009 2:42 am    Post subject: Reply with quote

I took a stab at it. I made some triangle protrusions, and added some holes such that glue can flow if it doesn't press in. If it turns out that they press in to hard, one can trim off the triangles. I put the files, including the Solidworks models in this directory.

http://jaredharvey.com/openservo/open_encoder/

This link shows how I might recommend placing the parts in an array such that "support" materails is minimized. I think this may help keep the cost down. However, it also may create shear planes that may weaken the part in the side to side direction.

http://jaredharvey.com/openservo/open_encoder/shaft_array.jpg

If support material is basically free, perhaps we will get better strength by placing the parts on their sides. Instead of up and down. I'm not sure if that would cause the center of the part to be filled with support material or not. I'm aiming to have the center hollow. I think that will help the shaft deflect, preventing stress concentrations.

Thoughts?
Back to top
View user's profile Send private message Visit poster's website
robotjay
co-admin


Joined: 01 Aug 2006
Posts: 225
Location: Nebraska, USA

PostPosted: Thu Dec 17, 2009 4:35 pm    Post subject: Reply with quote

I am more than comfortable with a slip fit and glue. We could try a filler like the shrink tubing you linked to, but I don't think it's necessary. It'll make assembly of these overly complicated, and I think super glue will perform as we need it to. I like the triangular notches you put on there, and trimming them off should be easy enough. Thanks for doing the leg work to make it happen. Are we agreed that we can go ahead and make an order?

-Jay
_________________
"Nothing is fool-proof; For we fools are ingenious and will find a way."
Back to top
View user's profile Send private message Send e-mail Visit poster's website
jharvey
co-admin


Joined: 15 Mar 2009
Posts: 362
Location: Maine USA

PostPosted: Thu Dec 17, 2009 9:37 pm    Post subject: Reply with quote

It just occured to me I didn't post a link to the PDF of the drawing for this version of the shaft. The PDF should be viewable to just about anyone.

http://jaredharvey.com/openservo/open_encoder/POT_shaft_B_SLA.PDF

I say we give other folks a couple days to chime in on this before we purchase. There might be some more thoughts we can add to it if we give it a pinch more time to stew.
Back to top
View user's profile Send private message Visit poster's website
jharvey
co-admin


Joined: 15 Mar 2009
Posts: 362
Location: Maine USA

PostPosted: Sun Dec 27, 2009 9:08 pm    Post subject: Reply with quote

I got a chance to do some stress analysison this. The physical properties were very close to acrylite, so I used that as my base for the physical properties.

Here's a picture showing the stress should be about half of the expected yield. I couldn't change the exaggeration, so try to look at the color, and the scale on the drawing.



Here's one showing the displacement is about .003in.



Shouldn't be long to until I place an order for those buggers.
Back to top
View user's profile Send private message Visit poster's website
jharvey
co-admin


Joined: 15 Mar 2009
Posts: 362
Location: Maine USA

PostPosted: Tue Dec 29, 2009 11:35 am    Post subject: Reply with quote

I'm ready to place a purchase for the SLA shafts. I've uploaded the newest solidworks files to this link.

http://www.jaredharvey.com/openservo/open_encoder/encoder_pot_2009-12-29.zip

I've also uploaded a step file of the array to this link.

http://www.jaredharvey.com/openservo/open_encoder/shaft_array_2009-12-29_qty110.STEP

The solid model should allow anyone with solidworks to confirm / deny tolerance build up, ect. I used COSMOxpress, witch is eh as a FEA tool. The full blown COSMO is better but still eh. A better stress analysis tool would have been CAELinux, but time constraints had me go the easy route. COSMO should be good enough for this kind of FEA. I've included the drawings and COSMO files with above noted zip file.

The array includes 110, that's 100 to match the 100 magnets, and 10 to play with.

I'll likely order them tonight. For those that want to pitch in, we can deal with that later down the road.

Any thoughts, concerns, ect before I pull the trigger?
Back to top
View user's profile Send private message Visit poster's website
Display posts from previous:   
Post new topic   Reply to topic    OpenServo.com Forum Index -> Hardware All times are GMT
Goto page Previous  1, 2, 3, 4, 5, 6, 7, 8  Next
Page 4 of 8

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum


Powered by phpBB © 2001, 2005 phpBB Group