Author Topic: Speed modification  (Read 1632 times)

0 Members and 1 Guest are viewing this topic.

Offline davelectronicTopic starter

  • Frequent Contributor
  • **
  • Posts: 772
  • Country: gb
  • Life's too short.
Speed modification
« on: October 09, 2019, 07:32:33 pm »
Hello out there to the community.
Think I've chosen the correct section to post this in.
I'm trying to find a way to slow down a hand held belt sander. I think it's a universal series motor of 800 watts, on the tool handle is a small control wheel that goes from 1 - 6 for speed. Although 1 on the wheel is reasonably slow, I could do with Slowing the belt down a little more. It's for standing flat some heatsinks I have, these had a raised portion which I cut off. So need to flat the face off. I don't have access to a lathe unfortunately,  Slowing the sanding belt speed will give me better control for final finished surface of these heatsinks.

Been thinking about DC of 50 - 60 Volts, but the tools speed control wheel has got me wondering,  I don't want to destroy the tool before I've even got started. My supply mains voltage is 240 Volts AC at 60hz single phase. Any thoughts how I can slow this tools belt just a little more would be of great help.
Thanks for reading, any ideas appreciated.
I've attached a picture of this belt sander, it was cheap as you can see.
 

Offline davelectronicTopic starter

  • Frequent Contributor
  • **
  • Posts: 772
  • Country: gb
  • Life's too short.
Re: Speed modification
« Reply #1 on: October 11, 2019, 01:34:14 am »
So no replys  :( Anyway I'd forgot that i purchased a reasonably high powered boost converter some time ago now. It was for another project that i never got around to. Cautiously i wired up this boost converter, the power source being a server power supply converted for bench use. So 12 Volts source, and 460 watts. I have bigger higher powered server psu, but though that was powerful enough with the lower voltage. Starting at 24 Volts, nothing, 30 Volts, nothing, 40 Volts gets it going. But this was to slow, pressed on up to 50 Volts now, and sort of ok. But still a shade to slow, i know as the voltage climbs my expectations on current begin to diminish.

I used two multimeters, one for voltage the other for current beingdrawn from the converter. I know i should have had a current reading from the boost converters input. But i didn't measure that  ??? But when i set it up again, before project box time, i will test the current at the input to the boost converter. What i ended up with was 60 Volts, drawing 2.5 - 3 Amps from the boost converter. The heatsink was always touchable, so below 60° C i guess, a pro would have set up a thermal measurement reading. The 40mm fan spun up a few times, but not that much. I was aware of tool temperature, as the fan in the belt sander was running much slower at 60 Volts. But all seemed fine, the sander got warm at best. The process of facing off large processor heatsink with a belt sander does not achieve the tolerance needed for cpu cooling. But my objective is to passively cool a large transformer, buy bonding the heatsink to the transformers outer surfaces. I've found one thermally conductive epoxy, but not sure if its suitable for the purpose. If not i will create some fine copper particles and mix with suitable epoxy at time of bonding.
Ignoring the location   ;D the image shows the knocked up assembly, its 60 Volts, and no one was going to come in contact, other than me, i know not best practice...
But acceptable results i think. The boost converter is largely over rated in its listed advertising. This one came off Amazon, claiming 1500 watts, err no, definitely not. There is a really good you tuber that tests buck and boost converters, most of those available from Amazon and Ebay.
I'm only posting the results, as its a viable option, staying with in reasonable parameters.
Thanks for reading and looking in. Hope its useful to someone.
 

Offline Nominal Animal

  • Super Contributor
  • ***
  • Posts: 7192
  • Country: fi
    • My home page and email address
Re: Speed modification
« Reply #2 on: October 11, 2019, 08:14:17 am »
I've modded some PC heatsinks  -- cheap aluminium ones, mostly originally older style processor heatsinks; like Arctic Alpine M1 --, but only by hand: using a hacksaw for the bulk removal, then various grades of sandpaper-on-mirror, with 400 and higher grades with mineral oil as a lubricant.  The mirror surface is absolutely flat, so ensures the flatness of the heatsink as well.

If you have access to a woodworking bandsaw, you can use that instead of a hacksaw for the bulk shaping of aluminium heatsinks.
 
The following users thanked this post: davelectronic

Offline mikerj

  • Super Contributor
  • ***
  • Posts: 3382
  • Country: gb
Re: Speed modification
« Reply #3 on: October 11, 2019, 08:43:09 am »
Agreed, a belt sander isn't going to give you a very flat surface and the finish won't be great either.  Wet and dry paper on a piece of glass is the way to go, I usually use WD40 or similar to lubricate the paper.
 

Offline Nominal Animal

  • Super Contributor
  • ***
  • Posts: 7192
  • Country: fi
    • My home page and email address
Re: Speed modification
« Reply #4 on: October 11, 2019, 09:07:54 am »
(Yup; I've only used mineral oil ("parafin oil") as a lubricant, as that's what I have (I oil my wood work surfaces with it, so always have some), and didn't have anything better like WD40 at hand.)
 
The following users thanked this post: davelectronic

Offline davelectronicTopic starter

  • Frequent Contributor
  • **
  • Posts: 772
  • Country: gb
  • Life's too short.
Re: Speed modification
« Reply #5 on: October 11, 2019, 07:16:55 pm »
I did take the bulk off with a hacksaw, used wd40 as a cutting lube for that. Its quite time consuming process. Yes i wil use a true flat surface and fine grades of wet and dry to get a better finish. Its a result to have got that sander to run slower, you can always remove material, but there's no putting it back. And a flat out belt sander is very hard to control at speed. I'm guessing here, but the motor in the belt sander must be some kind of universal series wound motor.
Just one small observation,  the variable 1 - 6 wheel does nothing when its run at 60 Volts dc off that boost converter. If i had to hazard a guess, its only functional useful run on AC. Stil not sure how DC voltage passes across this though. But it does nothing to speed control on lower voltage DC.
 

Offline davelectronicTopic starter

  • Frequent Contributor
  • **
  • Posts: 772
  • Country: gb
  • Life's too short.
Re: Speed modification
« Reply #6 on: October 12, 2019, 12:23:03 am »
Just some idea of the heatsinks that I've acquired, there quite substantial in size. This is post hacksaw stage. I think they where only a couple of quid each, dell and Lenovo. Make good cheap heat removal solutions.
 

Offline davelectronicTopic starter

  • Frequent Contributor
  • **
  • Posts: 772
  • Country: gb
  • Life's too short.
Re: Speed modification
« Reply #7 on: October 12, 2019, 12:24:13 am »
Normal side up.
Thanks for the input.
 

Offline Nominal Animal

  • Super Contributor
  • ***
  • Posts: 7192
  • Country: fi
    • My home page and email address
Re: Speed modification
« Reply #8 on: October 12, 2019, 06:14:08 am »
Looks quite similar to what I've done.  I started with 80 grit sheet, using a figure-8 pattern (supposed to give an even wear) to get all marks off, followed by 120 and 240, then switching to wet-and-dry (400, 600, 800 and 1200, I think); with the progressive grits just smoothing the scratches left by the previous grit.  Wet-and-dry works much better with a lubricant, to move the debris/swarf out, so that the abrasive particles in the paper make contact with the surface, with room for the removed material to move out.  Make sure you get good quality sandpaper, intended for metals; you are sanding off aluminium oxide, which is quite tough, after all.  Spray adhesive works well to stick the paper to your glass surface, although I've also resorted to just taping all four sides with packing tape.

I do not lap the surface.  (Lapping is using a flat surface with a lapping compound, a paste that contains abrasive particles, to get a really flat, shiny surface.)  I use small amounts of thermal paste to fill the tiny voids between the mating surfaces.  The idea is not to have a layer between the surfaces, just to fill any voids in between.  Good mechanical contact -- that is, having positive force pushing the heatsink to the hot surface -- is quite important for good thermal convection; screws with stiff springs is a good idea (as the springs keep relatively constant tension, even when ambient temperature changes and the heatsink expands and contracts).
 
The following users thanked this post: davelectronic

Offline Nominal Animal

  • Super Contributor
  • ***
  • Posts: 7192
  • Country: fi
    • My home page and email address
Re: Speed modification
« Reply #9 on: October 12, 2019, 07:16:17 am »
Funnily enough, der8auer showed that going above 80 grit isn't really necessary, if you use thermal paste (to fill in the scratches).  See results at 9:50,


Flatness is much more important than surface smoothness here.
 
The following users thanked this post: davelectronic, tooki

Offline davelectronicTopic starter

  • Frequent Contributor
  • **
  • Posts: 772
  • Country: gb
  • Life's too short.
Re: Speed modification
« Reply #10 on: October 12, 2019, 04:53:17 pm »
Yes final finish could be lapped, it is time consuming though. In an ideal world I'd have put them in a four jaw chuck and face them off.
Thanks for the input.
 

Offline mikerj

  • Super Contributor
  • ***
  • Posts: 3382
  • Country: gb
Re: Speed modification
« Reply #11 on: October 12, 2019, 07:58:50 pm »
Use a hand file to remove the saw cut marks before moving to wet and dry paper.  Chalk the file to discourage aluminium sticking in the teeth and draw file (i.e. hold both ends of the file and apply is sideways across the work) to get the best finish.

Sorry if this is frustrating, I know this isn't addressing you original question but given the application I just don't think a belt sander is good solution.
« Last Edit: October 12, 2019, 08:01:52 pm by mikerj »
 
The following users thanked this post: davelectronic

Offline rstofer

  • Super Contributor
  • ***
  • Posts: 9963
  • Country: us
Re: Speed modification
« Reply #12 on: October 13, 2019, 12:31:16 am »
Hello out there to the community.
Think I've chosen the correct section to post this in.
I'm trying to find a way to slow down a hand held belt sander. I think it's a universal series motor of 800 watts, on the tool handle is a small control wheel that goes from 1 - 6 for speed. Although 1 on the wheel is reasonably slow, I could do with Slowing the belt down a little more. It's for standing flat some heatsinks I have, these had a raised portion which I cut off. So need to flat the face off. I don't have access to a lathe unfortunately,  Slowing the sanding belt speed will give me better control for final finished surface of these heatsinks.


Google for "universal motor speed control" and there are hundreds of projects but you can also buy a speed control for < $20.
https://www.harborfreight.com/router-speed-control-43060.html
 
The following users thanked this post: davelectronic

Offline davelectronicTopic starter

  • Frequent Contributor
  • **
  • Posts: 772
  • Country: gb
  • Life's too short.
Re: Speed modification
« Reply #13 on: October 13, 2019, 02:39:25 am »
Thanks for the tips, for now I'm going to project box that boost converter. It seemed to cope with it fairly well, i think it got up to about 180 watts. And i was using serious pressure at times. But sure will look for universal motor speed controller. With a file i find it hard to keep a real flat profile on such a large flat area, its ok with file until you grt close, with about 1mm ish. Carrying on with a file zfter thst risks uneven material removal.
But thanks again for all your tips and ideas.
 

Offline Nominal Animal

  • Super Contributor
  • ***
  • Posts: 7192
  • Country: fi
    • My home page and email address
Re: Speed modification
« Reply #14 on: October 13, 2019, 12:52:28 pm »
If I were to do more than a couple of more of those heatsink mods, I would probably construct a horizontal disc sander, from wood/plywood, a large lazy susan bearing, a ~100 RPM 12V geared DC motor, and an adjustable DC power source.

Typical sheets of sandpaper are about 230mm on the shorter edge here, so I'd make the rotating square 230×230mm in size.  I'd get a meter long piece of glue board (solid wood, glued together; very flat) 250mm wide. For paper holders, I'd use aluminium L profile I already have, with a bead of silicone on the inside for grip, bolted to the side of the rotating board (using thumb screws or something, for changing the sand paper easier). I've access to Ikea, so I'd get a set of four 300×300mm mirror tiles for 7€.  The 6" lazy susan bearing I'd get from eBay (like this 6" one), and the motor from Banggood (like this or the 110 RPM here; a 6mm flange is also useful).

Here's an OpenSCAD sketch:
Code: [Select]
module Lprofile(a,b,t,length=170,d=4) {
    difference() {
        cube([a,length,b], center=true);
        translate([t,0,-t]) cube([a-t,length+2,b-t], center=true);
        translate([1-a/2,a-length/2,-b/5]) rotate([0,90,0]) cylinder(h=3, r=d/2, center=true);  // Bolt hole
        translate([1-a/2,length/2-a,-b/5]) rotate([0,90,0]) cylinder(h=3, r=d/2, center=true);  // Bolt hole
    }
}

rotate([0,0,$t*360]) {
    color([0.7,0.6,0.3]) translate([0,0,78]) cube([230,230,18], center=true);  // Rotating 230x230x18mm base
    color([0.9,0.9,0.9]) translate([0,0,89]) cube([230,230,4], center=true); // 230x230x4mm mirror
    // Paper brackets out of L profile:
    color([0.3,0.5,0.4]) translate([-107,0,84]) rotate([0,0,0]) Lprofile(20,20,1,170);
    color([0.3,0.5,0.4]) translate([107,0,84]) rotate([0,0,180]) Lprofile(20,20,1,170);
    color([0.3,0.5,0.4]) translate([0,-107,84]) rotate([0,0,90]) Lprofile(20,20,1,170);
    color([0.3,0.5,0.4]) translate([0,107,84]) rotate([0,0,-90]) Lprofile(20,20,1,170);
}
   
// Static base
color([0.5,0.4,0.3]) difference() {
    translate([-30,0,50]) cube([290,230,18], center=true);
    translate([0,0,50]) cylinder(h=20,r=6,center=true);
}

// Supports and hand guard
color([0.6,0.5,0.4]) translate([-184,0,50]) cube([18,230,100], center=true);
color([0.6,0.5,0.4]) translate([124,0,30]) cube([18,230,60], center=true);
color([0.7,0.6,0.5]) translate([-153,0,109]) cube([100,230,18], center=true);

// Motor
color([0.4,0.5,0.6]) union() {
    translate([0,0,60]) cylinder(h=40,r=3,center=true);
    translate([-10,0,27]) cube([43,32,28],center=true);
    translate([-50,0,27]) rotate([0,90,0]) cylinder(h=40,r=13,center=true);
}
Save the above as sander.scad, and open it in OpenSCAD.  I did not include the lazy susan bearing in the model.  Press F5 to compile and display it.  Enable animation (View > Animate).  Set Steps: to 2/3 of FPS, to see how 100 RPM looks in real life. My laptop can do about 25 FPS in real time, so FPS:25 and Steps:15 is pretty realistic.

EDITED: Aw crap, the gearbox would not survive many abrupt starts and stops.  The momentum in the plates is so high it'd strip the gears.  Instead, a friction drive with a soft rubber wheel (say Tamiya 70192 or similar, or just a rubber O-ring on top of suitable wheel) makes a lot more sense.  A 30mm diameter wheel making a 220mm diameter circle makes a ratio of about 1:7.33, so something like 600-800 RPM for such a smaller wheel should work (perhaps something like this).
So, ignore the motor part in this post; that needs further thinking.

It is unsafe as heck, and I've no idea whether it'd work in practice; I'm only saying that is what I would try to build, if I needed machine help in sanding heatsinks.
Knowing what I know now, I'd probably sand the surfaces down to 160 or 240 grit, no further.  I would use another mirror as a reference surface, using some ink (like metrologists use on surface plates, when checking if something is flat or not), to check the flatness of the results.

The attachments contain screenshots from OpenSCAD for those who don't already have it installed.
« Last Edit: October 13, 2019, 01:52:35 pm by Nominal Animal »
 

Offline davelectronicTopic starter

  • Frequent Contributor
  • **
  • Posts: 772
  • Country: gb
  • Life's too short.
Re: Speed modification
« Reply #15 on: October 15, 2019, 06:34:28 pm »
Looks like a good idea, but I've no access to cnc, and my code skills mmm, i don't have any at the moment.  I will make it a priority, as microcontrollers are off limits until i learn it. But thank you for the suggestions.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf