BD's Mechanical Disk Recorder Build

[Gus]

if you need my own plugin for Audicity who creates 90 deg. out of phase sine wave just let me know

[Fela Borbone]

To generate sine waves with a microcontroller you just need to read a table of values in the program and load them in the port, spaced with atimer, who sets the frequency. For two sines at 90 deg, poke in one port at n index an then poke at another port at n plus offset needed for the 90 deg. For d/a conversion the infamous r-2r ladder can suffice. Other way of cotrollig frquency is manipulaing the clock. I can ptovide table of values if you need. BTW, nice project!

[EpicenterBryan]

This is presumably too fine at this point, but simply put: obtaining smooth motor operation without auditory interference via the belt-to-cutting-platter is certainly possible with square wave use.

Ah, but let me make sure you are clear on this. The micro-stepping drive will be outputting what becomes closer and closer to a sine wave (current wise) as you increase the micro-stepping steps per rev resolution. So the drive is really taking care of generating the two out of phase sine wave drive current profiles. The beauty is you just send the drive digital pulses at the rate you want to spin the motor.

What the drive is really doing for you is varying the ratio of current between the coils, making a small change for each step pulse you send it... Want smoother movement? Increase the number of steps per rev on the drive. By the time you get to 50,000 steps per rev the sine wave looks as good as what your DAW can generate.

gr_FAD_LVSA06.gif

Bryan

[Bahndahn]

Want smoother movement? Increase the number of steps per rev on the drive. By the time you get to 50,000 steps per rev the sine wave looks as good as what your DAW can generate.

Ah hah! I see I see.... Far higher resolution than the [sine wave] output I am capable of generating via my arduino uno!

Thanks for the clarification.

[EpicenterBryan]

To generate sine waves with a microcontroller you just need to read a table of values in the program and load them in the port, spaced with atimer, who sets the frequency. For two sines at 90 deg, poke in one port at n index an then poke at another port at n plus offset needed for the 90 deg. For d/a conversion the infamous r-2r ladder can suffice. Other way of cotrollig frquency is manipulaing the clock. I can ptovide table of values if you need. BTW, nice project!

Fela, you nailed it.
20 years ago, I did a similar thing to generate a sine wave from a ROM look up table - but all in hardware with no processor. Back then, it was harder to do this. I keep these boards around to remind myself about how awesome the tools and toys we have now are, and how much easier it is now to bring ideas to reality!

Below is a photo of the empty prototype board and final version of the 8 layer board (and back then 8 layers was really expensive too).

IMG_3397.JPG

This was basically a voltage controlled oscillator hybrid, with analog on the front and back end, and a state machine and ROM table look-up in the center. The state machine did several tasks, but basically read the data from the ROMs, and sent it to a 12 bit DAC. This had to operate between 0.1hz to 100khz as I recall but also had some special conditions needed about starting and stopping on the low end per the customers requirements. Basically, this was a temperature stable DC controlled wheel speed simulator for an aircraft manufacturer to test brake designs.

So my point here, is these micro-stepping drives have simple state machines built in to generate "modified sine waves" just like what you mention. Making one yourself is totally possible. Buying one is a better option if time, or the value of your time is a bigger issue.

I do truly appreciate all of your posts Fela. You seam to have more time than money though, and I do appreciate that some times re-inventing the wheel is very interesting also. What I was trying to do by steering Bennett in the other direction was to try to save him time because of his school / build schedule. But Fela, what you mentioned is totally doable!

Bryan

[EpicenterBryan]

Ah hah! I see I see.... Far higher resolution than the [sine wave] output I am capable of generating via my arduino uno!

It depends. Your arduino can send anything you want OUT digitally. As Fela noted, you could send it to a DAC to generate the sine wave and the other one 90 degrees out of phase.

You can not send analog values directly out from an arduino. You can send PWM out, and when using an external filter, generate a voltage....


But, you can output digitally to an external DAC which can generate the analog signal you need just from your digital data in any precision you want. For a stepper, you have two phases to mess with, that's where your 90 degree stuff comes from. But you can do that too if you want.

I think you understand which direction you need to go for your project.
Go with a micro-stepping drive, send it a pulse train and it will take care of the pattern and phase stuff.
Call if you need clarification.

B

[Bahndahn]

It depends. Your arduino can send anything you want OUT digitally. As Fela noted, you could send it to a DAC to generate the sine wave and the other one 90 degrees out of phase.

You can not send analog values directly out from an arduino. You can send PWM out, and when using an external filter, generate a voltage....


But, you can output digitally to an external DAC which can generate the analog signal you need just from your digital data in any precision you want. For a stepper, you have two phases to mess with, that's where your 90 degree stuff comes from. But you can do that too if you want.

For the sake of this continuum, here is a boundary I have found according to the arduino documentation:

Screen Shot 2015-03-15 at 10.44.57 AM.png

It looks like the highest trustworthy scheduling with a normal arduino yields 333333.33333333Hz. If I understand the lookup-table technique correctly, a sine wave would have many 'values' to look up per cycle, requiring a scheduling frequency accordingly high.

I may be missing something here - also I have come across techniques where people clock their arduinos at a higher frequency with an external crystal.

I guess the micro-stepping drive would require that same high frequency as it is essentially doing the same thing.

[Bahndahn]

Whipped up an oil dashpot this weekend using a glass guitar slide, a couple springs [the upper spring removed in the photo], 3d printed parts, epoxy resin, and aluminium rod.

Here it is:

[dimi751]

Wow this dashpot looks great well done very neat !

[Fela Borbone]

BD ,you are buliding a clean and nice machine!

I do truly appreciate all of your posts Fela. You seam to have more time than mon

he he,thanks,you are rigth!

[Bahndahn]

Brief update of many little things:

Made the dashpot before, but hadn't an attachment plan set out. Spent some time working on that recently and will have photos once I get it setup.

Also, the halfed nut for pitch travel I milled needs attaching to the mobile suspension [x-axis]. The method I am working on will attach to these screws [photo], a simple spring loaded arm holding the halfed nut. Spring loaded to 'lift' it off the fixed position to quickly move it back to home position.

photo 3.JPG

I didnt previously put much care into motor spindles [pulleys for the belts], I thought: I will just 3D print them when that time comes. That didn't work so well for me when I tried it so I decided to learn the metal lathe. I made a quite heavy steel pulley for the pitch screw and the the stepper motor.

photo 2 (1).JPG

photo 4.JPG

Put together basic, untested, improvised cutterhead details using 3D printed rigid pushrods, steel [sewing needle] more flexible pushrods, nylon torque tube with 3D printed leaf/v-spring, 3dprinted stylus holder that currently is holding a carbide vinyl plotter blade etc.:

photo 5.JPG

I also filled my platter with 4 tubes of silicon caulking the other night, somthing some of my broke-ass DJ friends used to do. It truly has an effect I have found, ,my platter went for very light and 'ringing' to quite heavy and dampened.

I received the "BigEasyDriver" that is capable of 16microstep driving. I did some brief tests and deduced that I think its going to be fine as a drive, noise-wise. I am even confidently planning to mount the motor on the aluminum plate that is the base of my machine. It will be nice and compact that way. A previous thought had me planning to make a little 'motor island' that was totally decoupled from the lathe main body. If its noise, I will order a 128microstepping drive, they are inexpensive if you dig for them.

More to come!

[Bahndahn]

Photo update:

[Sillitoe]

Looks great Bahndahn! Are you going to anodize your whole build?

[Bahndahn]

Thanks Sillitoe!

I am thinking of anodizing all the aluminium which is everything except the 'pitch traveller' rods that are now steel, the motor pulleys, and some brass details I recently implemented. I still need to find a place to do it here in Montréal. Supposedly the engineering students at the university I study at take their parts to a place that charges them next to nothing to drop them into a batch, provided they are not picky about color. I want to keep it the natural color of aluminium.

I also bought some spray-on laquer for the steel and brass bits.

This week I will finish the build. Last week I worked from 9-5 monday to friday, this thing is coming together! This week I will finish up the details, throw together metal cutterhead details, and sand/polish the parts in prep for treatment.

[dimi751]

Love the motor case such neat work

Your doing such a great work keep it up !

[Bahndahn]

Heres another visual update from today, steel crossbars have replaced the aluminium ones, 3/8"- 24[threads per inch] lead-screw has replaced the 3/8"-16 one. Releasable half-nut is held up to the lead-screw by a spring, it is 'lifted' by that brass bar with the aluminium knurled handle sticking up out the back. I think I will shorten the brass rod a bit!

I have been doing 9am to 5pms on this build since last week monday, and am continuing this way till its done!

Tomorrow I'm going to make a center-disk stabilizer on a metal lathe. I wanted to do it in brass but the hunk I was looking [2" dia, solid rod, 6" length] at was $65?! Steel will do!

Stay tuned!

[Bahndahn]

I'm trying my hand at precision machining on the lathe. I am really enjoying this build!

Today I worked on the aluminium rigid portion of the push-rods. From these will extend a flexible sewing needle portion. Tomorrow I will make other cutterhead details.

These pushrods thread into the springs on the drivers I am using, they are held taut with the nylon nut.

I am measuring them along the way with a micrometer, I will be weighing them tomorrow.

[Bahndahn]

Today I set the sewing needles into the aluminium shafts using the mill as a vertical guide.

First the aluminium portions were set up in the vice, shortened to length, center-started, and drilled with the finest drill bit I had available.

mill.JPG

Next I mixed 5-minute epoxy and placed a blob on the face of the aluminium still in vice; submerged the sewing needles and vertically positioned them by closing the chuck jaws around it, unmoved from drilling. The results look positive!

Needle glued.JPG

MainSuspension.JPG

[Sillitoe]

Hi Bahndahn, clever way to insert the needle.
Looking good.

[Bahndahn]

My record cutting lathe is at the anodizing facility, so this week I went on down to the United States of America where I picked up this metal cutting lathe! I will clean this up over the next few weeks and begin experimenting with some special transducers or something!

This lathe is presumably also used by markrob mentioned in his account of cutting a new presto leadscrew.

I'm very much looking forward to cleaning this machine up!

photo (3).JPG