PDA

View Full Version : Making Dovetail Straight Edges



epicfail48
11-26-2017, 02:44 AM
First off, this aint a how-to, unless you want to learn how to do everything the wrong way! This is just my chronicling my first attempt at it!

Decided i needed a few dovetail straight-edges around the shop for a few projects down the road, as some point im planning on doing a complete tear-down on both my mini mill and lathe, all the little odds and ends. New bearings, delrin feed nuts, and most importantly, scraping the ways in so the movement on the table/cross slide is actually smooth. My opinion is that both the machines could work pretty well for the kind of work that i do, but 90% of the problems could be improved with better fit and finish of the moving parts.

So, making a few straight edges serves a few purposes. The first is, well, i need a straight edge so reference the dovetails off of. The second, and arguably more important reason, is scraping is a bit of a new skill for me, and id rather not learn on my tools. Much cheaper to kill a piece of cast iron after all. I figure, scrap in 2 or 5 straight edges to get a feel for it, get the process down, then move on to the expensive stuff.

Started off sourcing some cast iron, which turned out to be quite the task. Nowhere local to me carried it, somehow, and very few places online seemed to carry it. Fortunately, one of the forum members here came through in a big way, so i got a pretty massive piece of dura-bar to work with. Started off by hacking off a few pieces of roughly the right size. Funnily enough, hacking off a few pieces ended up being the hardest part, so far. My cheapie little portaband didnt want to cut without the blade wandering, so my pieces came out pretty wonky. Nothing that didnt mill off, milling just took a bit of time:
https://i.imgur.com/Crmzcpem.jpg

Up top is a rough cut piece off the saw, down below is a piece ive already taken down to dimensions of about 1x3/4x9ish. Really i wasnt going for any exact dimension, just wanted to get it as square and flat as i could. Thanks to the wandering of my bandsaw blade though, i had to cut the pieces close to 3/8" oversize, so i ended up with a fair bit of material to hog off with my little mini-mill. A 2 inch carbide face mill did a pretty fair job of it, though it did take a bit of time, the iron milled pretty nicely but i was still limited to about .030" DOC before the mill started making some pretty funny noises.

Cutting the angle for the dovetail proved to be a bit of a challenge. I had originally assumed i would need something along the lines of a sine vise, but in a post here i made a while back a few people made the very valid point that there was no reason that the straight edge angle needed to be exactly 60 degrees, in fact it needed to be less and wasnt critical. It still took some doing, figuring out how to cut the angle, but i managed to work something out that worked with my present tools. I ended up cutting a few wedges of ~55 or so degrees from some aluminium, then superglued them to the cast iron so i could grip the bars in my vise at the proper(ish) angle:
https://i.imgur.com/awGZQalm.jpg

Youll have to excuse the glare in the photo, it was kinda hard to snap. The setup was a little dicey, but it ended up working better than logic and reason say it shouldve. It did make me realise i need to invest a little in better workholding setups, a sine vise wouldve been really nice to have, even if it wasnt really needed, and some machinist jacks have made their way onto my wishlist. I ended up having to improvise a way to support the right side of the stock with a few 1-2-3 blocks and a bolt. Somehow it worked, and i got a pretty nice looking and fairly flat surface, but it really falls under "redneck engineering" and its not a setup id like to repeat.

Anyway, heres where the straight-edge blanks sit now:
https://i.imgur.com/WxjSdINm.jpg

I ended up maging the 2 different sizes, one is 3/4"x1", the other 1/2"x3/4". Bandsaw blade wandered a bit more than i though when i was roughing the blanks, but it ended up being for the better. The smaller one seems like itll be a little easier to maneuver. Now im just waiting for the Hi-Spot blue i ordered in to arrive before i can start the actual scraping. As it turns out nowhere local to me carries it. In the meantime though, i think I might start by 'lapping' the blanks with a piece of sandpaper, my mill didnt make a perfectly flat surface and im thinking starting with the surface a little closer to flat will make the scraping go by quicker.

Ill update once i can actually start working on these a bit more. In the meantime, id more than welcome any advise anybody cares to lend!

Danl
11-26-2017, 01:55 PM
Excellent work. Can you post a photo of an end view of those?

Dan L

J Tiers
11-26-2017, 02:10 PM
If you want an angle reference, the best idea is to make a short one that is not a lot of trouble to make. a couple inches is as much as you will need.

As for the dovetails themselves, it is really not critical what angle they are at exactly. Close to nominal angle will help if you ever need to replace the slide.

You can simply scrape the longest dovetailed part of the system (usually the slideway) to correct geometry, as far as flatness, direction, and distance between, etc. Then the shorter part (saddle or slide, usually) can be scraped to match more easily. (if you scraped the slide dovetail first, you would need to scrape the whole long slideway to match the slide).

I usually scrape the flat bottom of the slide to a granite flat (so that the two sections match) , then use that part to align the two surfaces on the slideway so they are in one plane. Then the dovetails on the slideway, for flatness, parallel, and alignment to external parts. Finally the dovetails on the slide are scraped to the dovetails on the slideway. Often with lesser equipment, one of the slide dovetails will have a gib of the multiple screw type (not tapered), so that side of the slide does not require scraping at all.

epicfail48
11-26-2017, 02:46 PM
Excellent work. Can you post a photo of an end view of those?

Dan L

Can do, it'll just have to be later tonight. Dunno why I didn't have one to begin with, should've occurred to me that would be nice to have

If you want an angle reference, the best idea is to make a short one that is not a lot of trouble to make. a couple inches is as much as you will need.

As for the dovetails themselves, it is really not critical what angle they are at exactly. Close to nominal angle will help if you ever need to replace the slide.

You can simply scrape the longest dovetailed part of the system (usually the slideway) to correct geometry, as far as flatness, direction, and distance between, etc. Then the shorter part (saddle or slide, usually) can be scraped to match more easily. (if you scraped the slide dovetail first, you would need to scrape the whole long slideway to match the slide).

I usually scrape the flat bottom of the slide to a granite flat (so that the two sections match) , then use that part to align the two surfaces on the slideway so they are in one plane. Then the dovetails on the slideway, for flatness, parallel, and alignment to external parts. Finally the dovetails on the slide are scraped to the dovetails on the slideway. Often with lesser equipment, one of the slide dovetails will have a gib of the multiple screw type (not tapered), so that side of the slide does not require scraping at all.

Thanks for the tips on order of operations. I haven't come up with an exact plan of attack yet, I'm still trying to sponge up all the information I can. So far the 2 links below have been my Bible for this, both are for mini Mills rather than lathes, but the concept is about the same. Planning on doing the mill anyways, so it sure doesn't hurt to plan

http://www.modelenginemaker.com/index.php/topic,3306.0/all.html
http://www.cnczone.com/forums/benchtop-machines/157385-cnc.html

epicfail48
11-27-2017, 02:20 AM
https://i.imgur.com/2zeUWD0m.jpg

And a quick end-on shot. Hi-Spot should be hear tomorrow according to UPS, so i should be able to get to scraping!

boslab
11-27-2017, 02:59 AM
Might be a bit soon after cutting, I've found stuff cut long and slim moves about a bit as the stress settles down, I was told that sticking in the oven at a low heat helps, I can see the logic but haven't tried it.
Otherwise good job, hand held bandsaws tend to be a bit challenging, sometimes I've cut curved once and every time I used it after it wanted to cut the same, the set must have been buggered
Mark

Mcgyver
11-27-2017, 07:48 AM
Might be a bit soon after cutting, I've found stuff cut long and slim moves about a bit as the stress settles down, I was told that sticking in the oven at a low heat helps, I can see the logic but haven't tried it.


I don't think its a bad thing to do, but its advantages seem an unresolved subject imo. Laslo who used to post a lot here claims latest engineering info suggests it is (or sited a reference that claims) an old wives tale that cast iron creeps. Offsetting that are armies of apprentices and machinists who hot cold cycled parts to work out the stresses. Of course we know the tales of manufacturers burying castings to remove stress - removing stresses is definitely a real thing to achieve stability for subsequent ops, but the need to do so after after all cutting and work is done is according to one source, disputed.

I'm not sure which is right. but I find the idea compelling that if you finished up cutting a piece its not going change shape (plastic deformation) on its own....except for this thing called creep. But apparently creep only happens with lots of force (but less than that required for plastic deformation) The other belief is that ringing it for awhile with a hammer removes stresses. Could be, but whether those stresses will change shape or not is not clear.

It can't hurt, but in my experience a good piece of cast has shown to be quite stable.

EddyCurr
11-27-2017, 01:25 PM
Funnily enough, hacking off a few pieces ended up being the hardest part, so far. My cheapie little portaband didnt want to cut without the blade wandering, so my pieces came out pretty wonky.

... Thanks to the wandering of my bandsaw blade though, I had to cut the pieces close to 3/8" oversize, so i ended up with a fair bit of material to hog off with my little mini-mill.Until a purpose-made horizontal or vertical bandsaw enters the picture, have you considered creating a bench w/ a table your portaband can be mounted to in a way that it functions as a vertical bandsaw? There are several links and videos with examples of how others have gone this route.

If you have access to a welder, fast-forward through this video to see the end result: M^3 Portable Bandsaw Table Build (https://www.youtube.com/watch?v=9s_H9xgcXuk)

Below are a few commercially available products for additional inspiration

Making a Table for my Portable Bandsaw (https://www.youtube.com/watch?v=9s_H9xgcXuk)

Milwaukee Horizontal Table 48-08-0260 (https://www.homedepot.com/p/Milwaukee-Portable-Band-Saw-Table-48-08-0260/100059645)

SWAG Vertical Table (Versions 1-4) (http://www.swagoffroad.com/SWAG-Portaband-Tables-Accessories_c_35.html)

A feature I do not recall seeing on portable band saws like my Milwaukee is an extendable blade guide for the purpose of closing down the unsupported length of blade across the throat gap to the minimum necessary for the material to be cut - like on purpose-built saws. If you incorporated something like this in your modifications, it would help further improve cut accuracy.


You'll have to excuse the glare in the photo, it was kinda hard to snap.Consider buying or building a tripod or other mount that will stabilize the camera/phone well enough to permit long exposures and higher f-stops in ambiant light for well-lit, sharp images with good depth of field.

Frank Ford's site shows a camera-mount rig he built for his shop that can be easily moved from place to place around the work area. Compose the scene, set the timer and voila! - great pics.

One or two inexpensive flex lights on mag bases fitted with warm white LED bulbs from CREE (CREE features good Color Rendition Index numbers) will be a benefit. Use these and cover up the window.

Peter.
11-27-2017, 05:21 PM
Good job but beware that those straight edges will likely bend over time. I made just the same thing and it printed in nicely. Six months later it would spin like a top and I had to scrape a couple of tenths off the middle to get it to print again.

10KPete
11-27-2017, 05:36 PM
I'm most of the way through making a similar straight edge 14" long out of Durabar. The bar was cut a little over 4 years ago then milled off about 3/16" on all 4 sides. I let it sit for about 6 months then checked it. It had moved about .003" out of straight. So,not having any way to stress relief anneal the thing, I just put it in the oven at 500*F for an hour then let it cool overnight. I repeated that cycle four times, then cut it straight. A few months later it hadn't moved! So I machined the 60* angle and it still didn't move. Cue the scraping action.

That's where I am now; it's within about two tenths overall with good print on all sides. If I apply my self another few hours of work should finish it.

Then I can get on with straightening my Benchmaster mill!!

Persevere!!

Pete

epicfail48
11-27-2017, 06:21 PM
Might be a bit soon after cutting, I've found stuff cut long and slim moves about a bit as the stress settles down, I was told that sticking in the oven at a low heat helps, I can see the logic but haven't tried it.
Otherwise good job, hand held bandsaws tend to be a bit challenging, sometimes I've cut curved once and every time I used it after it wanted to cut the same, the set must have been buggered
Mark




Good job but beware that those straight edges will likely bend over time. I made just the same thing and it printed in nicely. Six months later it would spin like a top and I had to scrape a couple of tenths off the middle to get it to print again.


I'm most of the way through making a similar straight edge 14" long out of Durabar. The bar was cut a little over 4 years ago then milled off about 3/16" on all 4 sides. I let it sit for about 6 months then checked it. It had moved about .003" out of straight. So,not having any way to stress relief anneal the thing, I just put it in the oven at 500*F for an hour then let it cool overnight. I repeated that cycle four times, then cut it straight. A few months later it hadn't moved! So I machined the 60* angle and it still didn't move. Cue the scraping action.

That's where I am now; it's within about two tenths overall with good print on all sides. If I apply my self another few hours of work should finish it.

Then I can get on with straightening my Benchmaster mill!!

Persevere!!

Pete

I know it's a bit soon after the rough milling, and I'm expecting to see a little movement in these down the line. The way I see it though, is it either ends up as more scraping practice, or I get pleasantly surprised 6 months from now when they haven't moved. Before being used on anything critical I was already planning on spotting them, just o double check straightness


Until a purpose-made horizontal or vertical bandsaw enters the picture, have you considered creating a bench w/ a table your portaband can be mounted to in a way that it functions as a vertical bandsaw? There are several links and videos with examples of how others have gone this route.

If you have access to a welder, fast-forward through this video to see the end result: M^3 Portable Bandsaw Table Build (https://www.youtube.com/watch?v=9s_H9xgcXuk)

Below are a few commercially available products for additional inspiration

Making a Table for my Portable Bandsaw (https://www.youtube.com/watch?v=9s_H9xgcXuk)

Milwaukee Horizontal Table 48-08-0260 (https://www.homedepot.com/p/Milwaukee-Portable-Band-Saw-Table-48-08-0260/100059645)

SWAG Vertical Table (Versions 1-4) (http://www.swagoffroad.com/SWAG-Portaband-Tables-Accessories_c_35.html)

A feature I do not recall seeing on portable band saws like my Milwaukee is an extendable blade guide for the purpose of closing down the unsupported length of blade across the throat gap to the minimum necessary for the material to be cut - like on purpose-built saws. If you incorporated something like this in your modifications, it would help further improve cut accuracy.

Consider buying or building a tripod or other mount that will stabilize the camera/phone well enough to permit long exposures and higher f-stops in ambiant light for well-lit, sharp images with good depth of field.

Frank Ford's site shows a camera-mount rig he built for his shop that can be easily moved from place to place around the work area. Compose the scene, set the timer and voila! - great pics.

One or two inexpensive flex lights on mag bases fitted with warm white LED bulbs from CREE (CREE features good Color Rendition Index numbers) will be a benefit. Use these and cover up the window.

Some form of table for using my bandsaw horizontally would certainly be nice for most of the work I do, but in this particular instance wouldve been akin to tilting at windmills. I was using a high-tooth blade with some age to it to go through 2 inches of cast iron, on a saw that barely tracks the blade to begin with. Pretty perfect storm for blade deflection, though I'll admit I wasn't expecting nearly 1/4" of deflection. Think I'll try an angle grinder too.

I've also got a pretty good photography setup, just didn't seem worth setting it all up to take a few snaps of a metal bar. I'll save that for the end when they're all shiny and scraped in!

EddyCurr
11-27-2017, 08:29 PM
I missed the part about the high tooth count earlier.


I've also got a pretty good photography setup, just didn't seem worth setting it all up to take a few snaps of a metal bar. I'll save that for the end when they're all shiny and scraped in!The beauty of Mr Ford's rig is that once implemented, there is little to set up - it is just THERE, pretty much ready to use for project documentation whenever the desire occurs. A foundation for consistently great photos w/ minimal impact on momentum.

In case you haven't already seen:

"Sky Cam" Camera Arm by Frank Ford (http://www.frets.com/HomeShopTech/Shop/SkyCam/skycam.html)

"Mill Light Bar" by Frank Ford (http://www.frets.com/HomeShopTech/Tooling/MillLightBar/milllightbar.html)

"Ball Head" Camera Swivel Mount by Frank Ford (http://www.frets.com/HomeShopTech/Projects/BallHead/ballhead.html)

epicfail48
11-28-2017, 04:30 AM
Alright, the Hi-Spot came in today, so i actually managed to get some work done, and i gotta say, its going better than i thought it was going to. Started off by using a strip of sandpaper to roughly lap the straight edges a little closer to truly flat:
https://i.imgur.com/0uH0KBSm.jpg

You can see that theyre pretty shiny in the center, whereas the ends still have some pretty clear tool marks. Seems to be a pretty clear indication that my mill had some trouble with table sag at the far ends of its travel. Thats one of the biggest motivators for this whole series of projects, i dont like the fact that i have to tighten the gibs to the point where i can barely move the table and i still have dimension issues like this. Anyway, after the lapping, i went ahead and printed (pretty sure thats the right term) the parts:
https://i.imgur.com/cOQaMclm.jpg

Took a few tries to get the thickness of the dye right. First time i applied it way too heavy to the plate, wiped it off and tried again and it was way too thin for the roughing passes. Once i got that figured out everything went pretty smoothly, the dyechem stuff was easy to see where it was spotting, and the cast iron was almost entertaining to scrap. Print, scrape, clean, repeat. The first printing showed more of what i already knew, that the center was high by a decent bit. After 35 iterations, i ended up here:
https://i.imgur.com/eHqCZ3pm.jpg

Spots all along the length of the straight edge, though the density of the spots is hardly where it needs to. I was ling on the dye fairly thick for these few rouging iterations, just to get the process going a touch faster. Ill start thinning out the blue as i get closer to the end. I went ahead and stopped there for the night, ill consider that plenty roughed in and will spend the next few days bringing up the precision to where it should be.

Course, the problem there i where the precision should be, ive heard numbers anywhere from 30 to 55ppi for a reference edge. Anybody care to weigh in? I was thinking 35-40 would be a good target for me. I understand that more precision never hurts, but i feel like in my case id hit a point of diminishing returns, where the added effort for the extra precision would outweigh the actual benefit for me. After all, these are being made with a grade B surface plate as a master by a complete amateur and will be used to scrape in the ways of what is generally considered to be a Fischer Price mini mill. Seems like it would be a touch of overkill to go for laboratory grade AA for that

TGTool
11-28-2017, 10:43 AM
35 - 40 ppi should be plenty good. As you break up the high areas you have now the greater points per inch should come in fairly quickly. There are no big holes of non contact and the areas showing no ink at the moment are probably not far below the contacting surfaces.

epicfail48
11-28-2017, 04:17 PM
35 - 40 ppi should be plenty good. As you break up the high areas you have now the greater points per inch should come in fairly quickly. There are no big holes of non contact and the areas showing no ink at the moment are probably not far below the contacting surfaces.

Well that's certainly reassuring to hear, that it should take horribly long to get the point down. Getting to this point didn't take too much time, about 2 hours, but still, I couldn't really feel my arm at the end! Not hard work, but a little repetitive

10KPete
11-28-2017, 05:15 PM
IMHO 20 ppi is plenty for the type of work intended. More won't hurt (except your arm!). I notice that the edges of the part are not inked. It's a common problem to 'roll' the edges over. Use great care near the edges. I don't recommend worrying about the point density until you have very even point coverage over the whole surface.

Pete

Dan_the_Chemist
11-28-2017, 10:23 PM
You have inspired me.

I've had four 21" drops of durabar sitting around for a while waiting to be turned into straight edges... So, tonight I tried facing one off with the Battle Shaper. Wow, a 22" stroke is impressive. It took a long time, especially since I'm being cautious while learning. But now I know I can do it. So, one out of 16 faces done. :)

I'll start another thread when I get pics... I just wanted you to know that you are the inspiration !

Dan

J Tiers
11-28-2017, 10:43 PM
..... I don't recommend worrying about the point density until you have very even point coverage over the whole surface.

Pete

Repeat that 10 times before starting a session of scraping. And when you finish. It is great advice.

Anyone can get a high point count. Getting it evenly all over, getting a good percentage of coverage, and not having any "holidays", is much harder

epicfail48
11-29-2017, 12:15 AM
You have inspired me.

I've had four 21" drops of durabar sitting around for a while waiting to be turned into straight edges... So, tonight I tried facing one off with the Battle Shaper. Wow, a 22" stroke is impressive. It took a long time, especially since I'm being cautious while learning. But now I know I can do it. So, one out of 16 faces done. :)

I'll start another thread when I get pics... I just wanted you to know that you are the inspiration !

Dan

Just as long as you read my disclaimer! Id love to switch places with you, ive been following along those threads on your shaper and that is a beast of a machine. Im guessing youll have a much easier time of it than i did roughing in the blanks, if absolutely nothing else i imagine getting the angles cut in will go a lot easier for you. Looking forward to seeing how you go about it, and glad i could help with the push!


IMHO 20 ppi is plenty for the type of work intended. More won't hurt (except your arm!). I notice that the edges of the part are not inked. It's a common problem to 'roll' the edges over. Use great care near the edges. I don't recommend worrying about the point density until you have very even point coverage over the whole surface.

Pete

Repeat that 10 times before starting a session of scraping. And when you finish. It is great advice.

Anyone can get a high point count. Getting it evenly all over, getting a good percentage of coverage, and not having any "holidays", is much harder

Doesnt one kindve go with the other? Wouldnt getting the point count up also imply having a consistent surface? Im not trying to argue the point, i see where you guys are going, but at the same time it seems like if you have a straight edge with 40ppi, but only on a quarter of the surface, its not really 40ppi. Any way you slice it though, i was going to make sure the entire surface was consistent. The spotting shown was just the end of my roughing, trying to get the entire surface at least somewhat closer to flat. Still got a lot of work ahead of me, i know!

And youre right about the edges Pete, they are a touch low. The passes prior to that last one they were high because i was pretty skittish scraping close to the corners, may have overcompensated a touch. I dont know that theyre particularly rolled, but if it turns out that way i may just cheat and mill some off the side... I am curious about your recommendation for final point density, seems closer to what ive seen recommended for bearing surfaces. Im not trying to contradict anything, i swear, im just trying to sponge up enough knowledge as possible. The entire methodology is still pretty foreign to me at this point

boslab
11-29-2017, 02:46 AM
With a shaper that size you could even make a surface plate!
Mark

epicfail48
12-01-2017, 02:44 AM
Came home off a 12 hour shift, got bored and decided to do a few more hours of scraping:
https://i.imgur.com/fkUrxinm.jpg

Spotting is a little more consistent now, if a little sparse. Turns out i did round over the edges a little, i figure ill keep going until ive got the surface looking like an actual straight edge with consistent spotting over the entire surface up pretty close to the edges, then cheat a little if needed and mill a bit off the offending sides. I know, cant do that with the actual parts, but hey, thats why im practicing!

Only thing im having a bit of trouble with is breaking up those high spots. Im in 25 more iterations, and it seems like ill scrape off the high spots, then ink up again and see the exact same number of high spots, just in slightly different positions, scrape those, and somehow end up back where i started. I know im doing something, but the progress is annoyingly slow-going for someone who likes instant gratification.

Only other thing im having problems with is getting a good layer of the blue on the surface plate. Theres a very, very fine line between too thick and invisible ive notices. Doenst help that my brayer wasnt actually round, more some exotic form of polyhedral not usually found in nature that refused to roll properly. Went ahead and took it to the disc sander to try and true it up a bit, well see what benefit that has some other day.

God i wish i had a surface grinder! Cant feel my bloody arms...

Mcgyver
12-01-2017, 08:24 AM
Doenst help that my brayer wasnt actually round, more some exotic form of polyhedral not usually found in nature that refused to roll properly. Went ahead and took it to the disc sander to try and true it up a bit, well see what benefit that has some other day.

God i wish i had a surface grinder! Cant feel my bloody arms...

everybody should have a surface grinder.....all in good time :)

Forrest advocates the brayer, I always use a piece of folded over felt. Why I like it better is with the wiping action you can basically get the blue as thin as you'd like, the wiping removes contaminants (chips and such) from the mix and you can rub over the plate (without adding blue) to even things out between cycles. Doing so, you can can keep cycling until the blue is so thin its not visible.....and you add just a tiny bit of blue to the plate with a quick rub pressing slightly harder on the felt. Disadvantages is its messy on the fingers. I often scrape with latex gloves to avoid blue stained hands

J Tiers
12-01-2017, 09:05 AM
......

Doesnt one kindve go with the other? Wouldnt getting the point count up also imply having a consistent surface? Im not trying to argue the point, i see where you guys are going, but at the same time it seems like if you have a straight edge with 40ppi, but only on a quarter of the surface, its not really 40ppi......

No, as in your case, where it's good except for the edges. You want coverage that looks as if you sprayed the blue on through a piece of perf metal, an even, regular pattern of spots over the whole surface.

And that's a good mental picture also because most round hole perf metal has 25% to 50% open area, which is in the percent contact range.

Mcgyver
12-01-2017, 09:44 AM
Doesnt one kindve go with the other? Wouldnt getting the point count up also imply having a consistent surface? Im not trying to argue the point, i see where you guys are going, but at the same time it seems like if you have a straight edge with 40ppi, but only on a quarter of the surface, its not really 40ppi.

It is 40 ppi - remember that's points per square inch. Your question though seems more along the lines of "so what if its 40 ppi, 3/4 of it isn't in contact".

There can also be a percentage of contact specification, however its usually a bit superfluous. Why? DOC. DOC is limited in scraping, by the force available and the tool geometry. DOC limitations determine how low the lows are. Once you've roughed and got coverage all over, you start working towards finishing and the 40ppi. With finish scraping its easy to have DOC that is a tenth or less. So.....if the entire surface has been touched by the scraper, and depth of cut is .0001, the entire surface is going to flat to within .0001" (which assumes a perfect surface plate, which of course doesn't exist....it to as a series of peaks and valleys within a tolerance). Its not too difficult to scrap something and barely discern any flicker on 10ths indicator needle as you draw it over the surface.

As there is no "flat" just a tolerance with a range of highs and lows, the 40 ppi surface can now be thought of a flat with the entire surface lying in a single plane, with a .0001 tolerance (again, ignoring surface plate error). Its not too difficult to scrap something and barely discern any flicker on 10ths indicator needle as you draw it over the surface.

epicfail48
12-01-2017, 04:05 PM
No, as in your case, where it's good except for the edges. You want coverage that looks as if you sprayed the blue on through a piece of perf metal, an even, regular pattern of spots over the whole surface.

And that's a good mental picture also because most round hole perf metal has 25% to 50% open area, which is in the percent contact range.


It is 40 ppi - remember that's points per square inch. Your question though seems more along the lines of "so what if its 40 ppi, 3/4 of it isn't in contact".

There can also be a percentage of contact specification, however its usually a bit superfluous. Why? DOC. DOC is limited in scraping, by the force available and the tool geometry. DOC limitations determine how low the lows are. Once you've roughed and got coverage all over, you start working towards finishing and the 40ppi. With finish scraping its easy to have DOC that is a tenth or less. So.....if the entire surface has been touched by the scraper, and depth of cut is .0001, the entire surface is going to flat to within .0001" (which assumes a perfect surface plate, which of course doesn't exist....it to as a series of peaks and valleys within a tolerance). Its not too difficult to scrap something and barely discern any flicker on 10ths indicator needle as you draw it over the surface.

As there is no "flat" just a tolerance with a range of highs and lows, the 40 ppi surface can now be thought of a flat with the entire surface lying in a single plane, with a .0001 tolerance (again, ignoring surface plate error). Its not too difficult to scrap something and barely discern any flicker on 10ths indicator needle as you draw it over the surface.

I may have misworded something, what I meant was if going for a 40ppi surface you had 1 square inch that was a perfect 40ppi but the adjacent square inch was 3ppi, then your surface wouldn't actually be 40ppi.

10KPete
12-01-2017, 06:00 PM
I may have misworded something, what I meant was if going for a 40ppi surface you had 1 square inch that was a perfect 40ppi but the adjacent square inch was 3ppi, then your surface wouldn't actually be 40ppi.

That's absolutely right!!

Pete

epicfail48
12-01-2017, 06:35 PM
That's absolutely right!!

Pete

Okay, good, I was having a hard time figuring out how to explain myself. I wasn't trying to disagree with either of you guys and didn't want it to come across that way, my phrasing was just bad.

My initial comment should've been that I was a little confused by the initial comments because, don't properly, bringing the point count up should result in a consistent surface unless you're only paying attention to one particular spot. Probably still not the best phrasing, but hopefully this and my previous comment can clear it up!

10KPete
12-01-2017, 07:02 PM
I understand. It takes a fair amount of 'interpretation' of the spotting but it should come fairly quickly as you progress. You have enough area spotted that the amount of blue you are putting on the reference surface should be pretty thin.

I don't use a roller except for roughing as I find it can leave the blue sorta thick without a lot of work. I use a cloth 'wad' or pad about the size of a quarter and spread a little very evenly. By rubbing in a circular motion all over and in all directions I can get the blue so thin it almost doesn't show.

Your work lighting must be just right to see the resulting spotting. I do the work on a bench with a big window behind it and find that natural light works well. Florescents, cool white type, don't seem to work as well as a more natural spectrum.

And don't rub the work on the plate very much. A couple of strokes along with the spin check is all you need. Any more and the blue will 'pile up' against the high spots.

When I get back on scraping again in a few days I'll post a couple of pics....

Pete

Mcgyver
12-01-2017, 08:46 PM
I may have misworded something, what I meant was if going for a 40ppi surface you had 1 square inch that was a perfect 40ppi but the adjacent square inch was 3ppi, then your surface wouldn't actually be 40ppi.

agreed, but stating a surface was scraped to 40 ppi means you can check any square inch on it and it'll be 40.

J Tiers
12-02-2017, 12:06 AM
basically, you do not want "holidays" in the scraping, whether you can count 40 ppi or not.

Ideally you would want your 40 ppi spotting result to look , just as I said, like you sprayed blue through a piece of perf metal with 40 evenly spaced holes in it per square inch. Right out to the edge, even from edge to edge, end to end.

Peter.
12-02-2017, 01:39 AM
Came home off a 12 hour shift, got bored and decided to do a few more hours of scraping:
https://i.imgur.com/fkUrxinm.jpg

Spotting is a little more consistent now, if a little sparse. Turns out i did round over the edges a little, i figure ill keep going until ive got the surface looking like an actual straight edge with consistent spotting over the entire surface up pretty close to the edges, then cheat a little if needed and mill a bit off the offending sides. I know, cant do that with the actual parts, but hey, thats why im practicing!

Only thing im having a bit of trouble with is breaking up those high spots. Im in 25 more iterations, and it seems like ill scrape off the high spots, then ink up again and see the exact same number of high spots, just in slightly different positions, scrape those, and somehow end up back where i started. I know im doing something, but the progress is annoyingly slow-going for someone who likes instant gratification.

Only other thing im having problems with is getting a good layer of the blue on the surface plate. Theres a very, very fine line between too thick and invisible ive notices. Doenst help that my brayer wasnt actually round, more some exotic form of polyhedral not usually found in nature that refused to roll properly. Went ahead and took it to the disc sander to try and true it up a bit, well see what benefit that has some other day.

God i wish i had a surface grinder! Cant feel my bloody arms...

If you're moving your islands of blue around but not increasing spots then you're not scraping deep enough and each scrape is too close to the next. Time to get rough on it.

epicfail48
12-02-2017, 04:10 AM
Little more word done tonight:
https://i.imgur.com/0KC0ET1m.jpg

Starting to look a little closer to how it should. Still some inconsistencies in the spotting, but im working on that, for not im happy the surface is looking better than it did. I can tell its getting closer to flat, theres actually a bit of stiction when trying to lift it off the surface plate. Pretty cool really. Some closer looks at the surface:
https://i.imgur.com/T65r0Cum.jpg
https://i.imgur.com/Yogf8DMm.jpg
https://i.imgur.com/cHkEknLm.jpg

The spotting is fairly consistently inconsistent, if that makes sense. Theres a part of me thats tempted to stop here, mill a touch off the edges, scrape in the angled face and call it done, but ill probably keep going for just a touch and get it just right. Dunno that ill be able to get the rolling on the edges to go fully away, so ill very likely end up cheating


If you're moving your islands of blue around but not increasing spots then you're not scraping deep enough and each scrape is too close to the next. Time to get rough on it.

Well blast it, where were you yesterday?!

Kidding aside, i did end up stumbling this being a problem. Little more oomph in the strokes and the work went a lot better a lot faster. Wish i had a narrower scraper though, the one im using feel a little too wide to hit the spots like i should. Probably just need practice though

Mcgyver
12-02-2017, 08:55 AM
It is looking like you are making good progress. One mistake that beginners make, heck maybe we all make sometimes, is not properly gauging the gradual transition from roughing to finishing. I've been at it awhile and still get suckered in.

First you are scraping regions until you get contact all over. This overlapping pass after overlapping pass of the region without regard to where the blue spot is. Then the ideal development of the surface is something like a couple PPI ALL over the surface, then 5, then 10 and so on. you don't stop to measure, its just a sense you have that you are trying bring the whole down somewhat uniformly. As you progress of course, there is a gentle and subtle shift in 1) how thick the blue is applied 2) how much force you put into the cut (DOC) and 3) the length of the cut. It can seem like its taking forever and you just want rush in and start knocking of single points....but it is by far the most time efficient way to come at it, that is bring it all down uniformly.

As for the edges, do your best. always come to the edge at the diagonal and try to keep the force on the flat plane not the edge by slight adjustment in the tool position as the cutting edge of the tool goes over the edge of the work. I don't get perfect edges either, but it won't matter for what you are doing.....when scraping a dovetail there has to be corner relief so end of the day the edge of your reference flat won't be touching anything

There are two kinds of people in the world, those who get "making things really flat is not trivial" and those who don't. This is the pain required to put yourself in the former :D :D

J Tiers
12-02-2017, 09:57 AM
When you get to where you are just moving spots around, start splitting them.

Do not remove the spot, cut it in two.

Also, you will probably see spots where there is a ring of blue around a bare or very light blue spot. Known as a bulls eye. Split the bare spot, leave the blue.

That should get you moving again.

epicfail48
12-05-2017, 05:23 AM
Another few days, some more work done:
https://i.imgur.com/CRzPNgnm.jpg
https://i.imgur.com/dJHJCsMm.jpg
https://i.imgur.com/m0JmXSTm.jpg
https://i.imgur.com/3nyDqeom.jpg

Managed to get the spotting pattern a little but more consistent and a little denser. Im really considering milling a bit off the back side and angle to get rid of the rounded corners and calling this face done, at least for now. The surface certainly seems flat enough for my uses as is. I may try to get my hands on a slighty narrower scraper, the one im using im having a hard time staying precise enough to only hit the high spots. Its a little too wide, i keep accidentally rocking it and scraping where i dont want to.

jr45acp
12-08-2017, 12:40 PM
Looks to me that you're on top of it Chris. I was concerned that scraper was a touch on the wide size. Someone told me how much fun scraping could be. To me it is/was work on steroids.

J Tiers
12-08-2017, 01:48 PM
As you get it to a better appearing coverage, start reducing the amount of blue on the reference surface. Use a thinner coating of blue. That may be harder to do with a "brayer" than with a "dauber" to spread out the blue, but if you just add less when you refresh, it should happen naturally.

The blue tends to be deceptive as you refine the surface, filing in areas which are actually a slight bit lower. Using less allows concentration on the points that are actually in the plane of the "top surface", the highest points.

While that may seem as if it is needless refinement, it has an effect on how fast the surface wears down. If there are more "singleton high spots", the "surface" will wear faster until those are at the level of many others, but in a way that you have no control over. Better to take care of them in a controlled, measured way.

Can you do a closer-up shot of typical areas?

You can ignore this if you used a bit more in order to have it show up in pictures. I know that when the blue is still quite visible to the eye, it disappears in pics.

epicfail48
12-08-2017, 04:28 PM
Looks to me that you're on top of it Chris. I was concerned that scraper was a touch on the wide size. Someone told me how much fun scraping could be. To me it is/was work on steroids.

I dunno that I'd call it fun... It's definitely work, especially for me. I've never been too big on anything involving ultra-high attention to detail or repetitive movements, I just get distracted too easily!


As you get it to a better appearing coverage, start reducing the amount of blue on the reference surface. Use a thinner coating of blue. That may be harder to do with a "brayer" than with a "dauber" to spread out the blue, but if you just add less when you refresh, it should happen naturally.

The blue tends to be deceptive as you refine the surface, filing in areas which are actually a slight bit lower. Using less allows concentration on the points that are actually in the plane of the "top surface", the highest points.

While that may seem as if it is needless refinement, it has an effect on how fast the surface wears down. If there are more "singleton high spots", the "surface" will wear faster until those are at the level of many others, but in a way that you have no control over. Better to take care of them in a controlled, measured way.

Can you do a closer-up shot of typical areas?

You can ignore this if you used a bit more in order to have it show up in pictures. I know that when the blue is still quite visible to the eye, it disappears in pics.

It's a little of both. I did lay on the blue just a touch heavier so it would show up better in the pictures, but I certainly understand what you're saying about thinking the blue layer to show better the high spots. I've been trying as I go, but it seems like there's a very fine line between "impossible to see" and "too much".

Compounding that issue is the fact that I'm still having a bit off trouble exactly hitting the tiny high spots. It's mostly poor control that just needs practice on my part, but man does it get challenging hitting the exact right spot!

J Tiers
12-08-2017, 04:42 PM
You must be using Canode.... it is like that. Regular HiSpot Blue thins out perfectly as you go along. I use Canode also, due to the mess issue.

You can tin it and re-work it by a fine spray of Windex. Stays good for 2-3 minutes unless the air is very dry.

As for hitting the spots, take note of where it did scrape, and try to move that place on the blade to the next one. It's like fast draw pistol competition, you get to where you can hit what you look at, just by "pointing at it".

epicfail48
12-08-2017, 07:01 PM
You must be using Canode.... it is like that. Regular HiSpot Blue thins out perfectly as you go along. I use Canode also, due to the mess issue.

You can tin it and re-work it by a fine spray of Windex. Stays good for 2-3 minutes unless the air is very dry.

As for hitting the spots, take note of where it did scrape, and try to move that place on the blade to the next one. It's like fast draw pistol competition, you get to where you can hit what you look at, just by "pointing at it".

I'm using the Dyechem Hi-spot blue, I have no idea which category that falls under. I think the biggest part of my issue getting the layer thickness just right is my work area is about 50f right now, seems a little chilly for it to smooth out in a the sweet spot thickness. Dunno though, I'm still getting the hang of everything!

Thanks for the advise on hitting the spots. I've been trying to do that, but it's kindve a pain. Usually I end up rolling the blade just slightly when repositioning and scraping a little to the right of where I want to. Still got a lot of practice to go I'm afraid

J Tiers
12-08-2017, 07:34 PM
Exact location is not important until you get it all roughed-in and are starting to focus on individual spots instead of areas. But nothing says you cannot practice location while roughing.

Being right handed, I hold the scraper shank in my left hand and usually can get it placed back down in the same orientation. Right hand provides force, but not so much positioning, left hand provides positioning, but only "down" force.

If you are using hi-spot blue, you may have too much blue on the roller (brayer). I do not use one, so I cannot offer much advice on its use. Try not applying any more for a while, and letting the roller gradually lose blue as you re-work the spotting area on the flat.

You might be starting off with too much, and so it does not deplete very fast. Fir HiSpot, I usually apply a dab for every 3" square area, and then use a dauber to spread it out. Then the dauber gets used to re-spread what is left after a spotting, moving excess to where the workpiece removed some blue from the flat. I have found that to thin out at nearly the rate I'd like it to. But I do not use much Hi-Spot anymore.

epicfail48
12-20-2017, 08:36 PM
Alrighty, last update for this one. Did a little more work on the flat and scraped in the angled portion:
Flat:
https://i.imgur.com/m7OulqIm.jpg?1
Angle:
https://i.imgur.com/vPNsGMHm.jpg

A skilled hand could probably do better, but honestly for my uses i think thats more than enough. Thanks again for all the help and advise, i really do appreciate it! Now hopefully i can put the knowledge to use scraping in the ways on my mini lathe...