Forum

Welcome Guest 

Show/Hide Header

Welcome Guest, posting in this forum requires registration.





Pages: [1]
Author Topic: New features for hardware
sandy
Administrator
Posts: 1427
Permalink
sandy
Post New features for hardware
on: July 8, 2013, 15:55
Quote

Hallo everyone, I'm revisiting the hardware design of the polarshield soon, and am planning on removing the XBee socket / opamp / voltage regulator. I don't think anybody uses it, and it doubles the complexity of the board. Am I right?

I will still break the pins out, so you can install such a thing using your own breakouts. Are there any strong feelings against that?

Are there any other things that any of you would like to see in the Polarshield? Things moved, orientation of cables, things like that. Types of connectors.

The impetus for this is that the screens I'm using are no longer available and I've got to switch to different ones anyway, so I'm having a rejig.

I'll also be writing this up for the main blog.

Cheers!!
sn

kongorilla
Pro
Posts: 385
Permalink
kongorilla
Post Re: New features for hardware
on: July 8, 2013, 22:25
Quote

I've never settled on a good way to attach my polorgraphSD to my drawing board. Too often you'll find it dangling by its stepper cables, like this:

Image

My natural impulse is to put it up on top of the board, but the cables come out the bottom, so I'd have to turn the polargraphSD on its side. Also, my servo extension cord isn't long enough to allow the polargraphSD to be perched on top (I could buy or make a longer one).

I've been meaning to make a mount that fits along the side of the drawing board, low enough to be out of the way of the counterweights, but then I'd have to deal with the stepper cables coming out the front and needing to be longer. Not a big thing, but it's so much easier to leave the poor thing dangling. 😉

I'm not sure what I'd suggest you do, because you can't anticipate everyone's desired orientation, but with the present arrangement there doesn't seem to be a convenient, default one.

kongorilla
Pro
Posts: 385
Permalink
kongorilla
Post Re: New features for hardware
on: July 8, 2013, 22:44
Quote

I just remembered that it's kind of a pain when running in tethered mode if the polorgraphSD is on top of the board. Keeping the long USB cable out of the way of weights and servo cable...yeah...that's why I was thinking of a side mounted arrangement.

Just thinking out loud. I'll leave now.

sandy
Administrator
Posts: 1427
Permalink
sandy
Post Re: New features for hardware
on: July 8, 2013, 22:49
Quote

I know what you mean. I designed the case and the cables to be hung up on a hook, with everything hanging out the bottom (so there was no twisting on the cable connectors), but I'll admit that's not what my own machines are like (more like dangling, or precariously balanced), and when I _have_ set up machines like that, the routing is never very convenient, even though it might look neat by the end.

The first machines were mounted on the top edge of the board (http://www.flickr.com/photos/euphy/6785955898/in/set-72157626497662024) but that's a bit bespoke. Mounting stuff on walls is always a bit fraught (paintwork etc).

I wonder if I could hang it off one of the motor brackets or something like that. That's not a bad idea. Is it?

Changing the orientation of the LCD is not out of the question either.

cheers!!
sn

Kelticfox
Newbie
Posts: 18
Permalink
Post Re: New features for hardware
on: August 8, 2013, 22:15
Quote

Probably a bit late to the party, but I have a question:

If the PolarshieldSD is run unconnected (via the SD basically), could it be mounted at the top? My plan was to have the Polarshield at the top in the middle with the stepper cables being directed around the back, which would keep everything out of the way of the weights.

I'm just planning ahead as mine I have to be able to draw at least A1 (I work on A1 or A2 doing architecture vector drawings). If I have to run tethered so be it, but I would prefer to stick the whole drawing board and Polarshield in the hallway out of the way (with the kids under VERY strict instructions of touch it and someone looses a finger!). The more space efficient (with cables out of the way) the better!

kongorilla
Pro
Posts: 385
Permalink
kongorilla
Post Re: New features for hardware
on: August 9, 2013, 01:30
Quote

Yes, Kelticfox, up top is fine. You'll still have to deal with routing the power cord so it doesn't interfere with the weights (not a big hassle), and, assuming you're using a servo, the servo cable (is it long enough, and is it out of the way of the weights?).

Each position has its advantages and complications, so do whatever works for you. You can always alter it later if your situation changes.

Kelticfox
Newbie
Posts: 18
Permalink
Post Re: New features for hardware
on: August 9, 2013, 17:13
Quote

Cool. I have a load of servo extension cables anyway (I have several R/C planes), so the servo shouldn't be an issue. I was thinking of routing that down and under.

I think my mine problem is going to be speed. I'll probably be using it at the slowest speeds to get the highest detail.

kongorilla
Pro
Posts: 385
Permalink
kongorilla
Post Re: New features for hardware
on: August 9, 2013, 20:34
Quote

Be sure to play around with speeds before you settle at a slow one. I've done the same drawing with fast and slow settings, and have gotten sharper, better drawings done at the 'fast' end of the scale. I don't know why -- it's certainly counterintuitive. Maybe I'm setting acceleration unwisely for slow drawings? Maybe when the gondola is traveling slowly there's more time/opportunity for wobble? Maybe beaded cord slack is taken up more quickly when using fast speeds? I dunno.

So, what I'm saying is, test your assumptions. These machines are simple, but they can defy expectations sometimes. Have fun!

Kelticfox
Newbie
Posts: 18
Permalink
Post Re: New features for hardware
on: August 9, 2013, 21:47
Quote

Well it was your CAD head drawing that convinced me to buy one of these.

Like I've said to Sandy in emails, I cannae draw for toffee. But I can do computer art easily. Vectors come easily to me (after years of me being a graphic designer).

Although I really need to find the software needed so I can have a tinker...

sandy
Administrator
Posts: 1427
Permalink
sandy
Post Re: New features for hardware
on: August 9, 2013, 22:35
Quote

Kong is right, though I reckon deceleration speed might even have more to do with the quality of lines, than actual top speed. When there's a fast, sudden move to a corner, I think the momentum of the gondola carries the pen tip deeper into its "proper" place. Whereas a dead slow, painstaking move will end up stopped a small distance away from the intended corner.

I haven't ever done any kind of proper testing of that, and that might just be a naive conception. I also think the initial max speeds (and particularly acceleration) are very conservative. They were low by default so that your machine doesn't pull itself to pieces when you have first built it, and give it a test, and I just never changed them. My common speed is 2000 for max speed and acceleration.

When drawing with felt-tip pens, on fairly absorbant paper, its quite easy to move the pen so fast the line fades out, but that's really the only negative aspect of fast moves that I've noticed. There is, in principle a hard limit to how fast the arduino can practically spit out steps though, opinions vary about this, but I've not gone higher than 4000 odd. Of course there are practical limits that depend on motors, current, voltage etc too.

sn

kongorilla
Pro
Posts: 385
Permalink
kongorilla
Post Re: New features for hardware
on: August 9, 2013, 23:34
Quote

Hey Sandy, I'd appreciate some discussion of acceleration settings, either in a dedicated thread or in the controller instructions (assuming you haven't added the info since I last looked). It's one of the settings I've only experimented with, increasing values until I start experiencing problems and then I back down a little.

I typically run with a max speed of 3600-3800, sometimes higher, but suspect that I never get near those speeds due to the drawings being made of such short segments and my use of acceleration around 800-1000. So, I'm guessing, it might be better for me to lower the max speed and raise acceleration. Does it sound like I'm understanding this right?

Is there an explanation of what the acceleration value actually represents that my mathematically challenged brain could grasp? Perhaps using a metaphor involving a hamster on a wheel?

sandy
Administrator
Posts: 1427
Permalink
sandy
Post Re: New features for hardware
on: August 9, 2013, 23:59
Quote

I don't think I know much more about it than you do, ha! But aye, I assume I almost never hit the max speed either, and I have taken to bumping the acceleration right up, 4000+.
I do think that acceleration needs to be tuned to the friction and weight of the gondola too though - I don't reckon it's a very fine piece of tuning, but I think it's the particular deceleration (and nothing else) that leads to judders at the ends of lines.

The acceleration number is acceleration in steps per second, per second. Which I thought I understood, but now my brain isn't working so I can't remember how to work out how long it'll take to reach top speed. Perhaps I must load a new hamster, this one is tired.

sn

kongorilla
Pro
Posts: 385
Permalink
kongorilla
Post Re: New features for hardware
on: August 10, 2013, 06:02
Quote

The acceleration number is acceleration in steps per second, per second.

Oh yeah, sorry, now I can remember you telling me that before. Since there's nothing tangible like "hamster steps per revolution of the wheel", my brain just lets the information traipse through without leaving any permanent impressions in its cement. Maybe if I say it over and over again, something'll stick and make sense...

...."acceleration in steps per second, per second"
"acceleration in steps per second, per second"
"acceleration in steps per second, per second"...........

So, "steps per second" is initially zero? Then you increment "steps per second" by the value "acceleration" once per second until the "max speed" value is achieved or the impending end of the line segment triggers deceleration? (By the same method as acceleration, just negative values?)

(My hamster is now panting heavily).

sandy
Administrator
Posts: 1427
Permalink
sandy
Post Re: New features for hardware
on: August 10, 2013, 10:49
Quote

Er yes that sounds right to me I think, the speed is steps per second, and is zero while the motor is stopped. And the acceleration is a statement of how rapidly that speed is changing.

I think your description is good, except that the calculation is based on milli or microseconds rather than seconds (although the value uses full seconds as the time base because that makes sense to us). Essentially the core of the movement engine is a loop in which current speed, max speed, acceleration and steps to go is constantly evaluated. The loop itself runs at a fairly high frequency (at least a few kHz - thousand times a second), and each run through decides to step or not step the motor based whether a step is due to make it move at a certain speed.

The acceleration calculator is constantly changing that particular speed over the course of any given second.

For a fixed loop frequency of 4000 Hz, running at 2000Hz will mean every alternate loop does not result in a step. You can see where the hard limit creeps in here - once the motor is running at 4000Hz, every single run through the loop results in a step, and there's nowhere to put the extra steps to make it go faster.

But, I think that means that with an acceleration of 3000 and a max speed of 3000 it would take only one second to reach top speed. That intuitively sounds too fast - in real life, accelerations have always seemed more protracted, and top speeds rare, but perhaps I just run with lower numbers in reality than I think I do!

sn

gontiki
Newbie
Posts: 24
Permalink
gontiki
Post Re: New features for hardware
on: August 11, 2013, 15:23
Quote

Okay, this is maybe a little bit provocation, but why not drop the whole Arduino all together and go for a Rasperry Pi (see topic "Raspberry Pi Powered Polargraph")?
We would get a much cheaper hardware and a unified software.

sandy
Administrator
Posts: 1427
Permalink
sandy
Post Re: New features for hardware
on: August 11, 2013, 18:49
Quote

Arduino is attractive for a couple of reasons, but simplicity is the main one. No OS or resource sharing to worry about, but of course that goes both ways - doing complex, memory-hungry stuff is complex, or impossible in arduino.

It wouldn't end up that much cheaper because you'd still need an addon board with motor drivers and probably some power handling hardware, but that would be a condensed version of a polarshield I guess. Could use a I2C/SPI touchscreen for standalone work. But it probably would be a little cheaper if pi supplies stay reliable.

sn

gontiki
Newbie
Posts: 24
Permalink
gontiki
Post Re: New features for hardware
on: August 11, 2013, 21:33
Quote

It is true, any other configuration comes to a similar price (which is anyway not a big issue because this is probably the cheapest drawing machine one can have).

I just got the diffuse opinion that having Rasperry Pi with a shield for the stepper and servo (same functionality like the EiBotBoard) would really simplify the Polargraph. Right now, you have to maintain two SW systems simultaneously. Then, part of the image generation is in the server, part in the controller, which makes it difficult to hack, i.e. to add new drawing algorithms.

Anyway, this thread is about HW, and not SW, and this thoughts are more concerns about SW. So never mind. 😉

As for the polarshield, I also would wish to have more soldering options for different types of the stepper connectors. USB and power connector should be sidewise (and not toughing the bottom side of the polarshield). I also have the polarshield mounted on top of the machine because that gave me the shortest cable to the stepper.

sandy
Administrator
Posts: 1427
Permalink
sandy
Post Re: New features for hardware
on: August 11, 2013, 22:06
Quote

I've moved the 2.54mm pin headers further out from the 3.5 pin headers, so there's room to actually fit both now. Not much else has changed in v1.4, except that there's no longer a 3.3v voltage regulator because there is no longer an XBee socket on board. Nobody seemed to use it, and it increases the cost and build time by a wee bit.

On the other other hand, I have added space for a third stepper driver. This will be unpopulated until I figure out what to do with it, but I have a few schemes that might benefit from an extra motor.

The revision is mainly to get the new screens fitting inside the board footprint again.

sn

Kelticfox
Newbie
Posts: 18
Permalink
Post Re: New features for hardware
on: August 17, 2013, 10:14
Quote

As you mentioned it, any update on the new screens?

I'm holding off ordering one until you say go 😛

sandy
Administrator
Posts: 1427
Permalink
sandy
Post Re: New features for hardware
on: August 17, 2013, 14:32
Quote

The plot thickened slightly with the new screens:
The 2.2" touch screen supply dried up.
I didn't really want to move to the 2.4" screens because the different resolution means maintaining and testing on two code bases, but I didn't have a choice.
So I started making plans for the revision to the board to move the connector, and then my supplier announced a new 2.2" touchscreen, with the orientation changed, and _just_ perfect, so I put the shield redesign on hold and ordered a pile of those 2.2" screens.
But then when I got them I found that the item description had been misdescribed and it was a screen only, not a touchscreen! Boo!
So they are currently on the way back to be exchanged for 2.4" touchscreens, and we're back to where we started.
I've got a handful of 2.4" touchscreens for now, but don't have the v1.4 polarshields yet.
The v1.4 shields will have no xbee port, but will have space for a third stepper driver. I have half a dozen v1.3 polarshields leftover, and am still using that stock for vitamin kits, along with a the 2.4" screens. I hope to get some preliminary v1.4 boards back from the makers around next weekend, but I wouldn't expect to have anything to release until the weekend after that.

sn

Pages: [1]
Mingle Forum by cartpauj
Version: 1.0.34 ; Page loaded in: 0.027 seconds.