Does mirror finish = flat?

[Tullphan]

Perhaps silly questions, but I lapped (or tried to lap) a Big Typhoon.
I got a fairly good finish on it...not "mirror" mirror, but pretty good.
I don't know about this "hold a razor blade to it in the light" thing, but I tried it as well as a machinists' square.
When holding it up to the light, I see an even gap across the base a few thousandths wide. Is this an optical illusion? Is this normal?
Thanks.

 

[Zepper]

Flat is flat, shiny is shiny - not the same at all. Flat is what you want. Too shiny is not good can reflect IR waves back into the device. Flat is a macro condition, shiny is micro. Thought we'd gone over this one before.

.bh.

 

[PCTC2]

Flat = using a razor blade to measure flatness, not shinyness.

 

[NinjaJedi]

I posted this link on a couple other threads on another screen name. For those that do a lot of lapping it would be a great tool. The smallest one is about $22 shipped

http://www.grizzly.com/product...aspx?q=granite%20plate

The thing with using a razor blade is you are assuming that the razors edge is completely flat. I don't know the tolerances of a razors edge. Is a few thousandths going to matter? I do not know. I would think the square would have a higher tolerance than a razor.

 

[Tullphan]

Originally posted by: Zepper
Flat is flat, shiny is shiny - not the same at all. Flat is what you want. Too shiny is not good can reflect IR waves back into the device. Flat is a macro condition, shiny is micro. Thought we'd gone over this one before.

.bh.

I'll have to go back over my notes.

Originally posted by: PCTC2
Flat = using a razor blade to measure flatness, not shinyness.

As I stated in my original post, when holding it up to the light, I see an even gap across the base a few thousandths wide.

 

[Navid]

A mirror is not necessarily flat.

To see if a surface is flat, create a high-resolution grid, by drawing straight lines, and look at the image of the grid on the surface.
You can clearly see any curvature or uneven area.
http://i259.photobucket.com/al...id200/FlatnessGrid.jpg

 

[EarthwormJim]

Originally posted by: Tullphan
Perhaps silly questions, but I lapped (or tried to lap) a Big Typhoon.
I got a fairly good finish on it...not "mirror" mirror, but pretty good.
I don't know about this "hold a razor blade to it in the light" thing, but I tried it as well as a machinists' square.
When holding it up to the light, I see an even gap across the base a few thousandths wide. Is this an optical illusion? Is this normal?
Thanks.

It just means your base is not flat if there's a gap.

Try using a flat piece of glass when you sand, and also print out a grid like suggested.

Think about it, a car can have a mirror perfect shine, yet the body isn't flat.

 

[DerwenArtos12]

Smooth and flat don't entirely co-inside. I've found that the best I can do to get things flat is either use a milling machine or I have a glass desk that I lay the paper on to use. With a little water on the backside and some tape at the very edges it holds in place quite well, then I hold the heatsink by the base and start to sand. I also like to use a trick my dad taught me, draw an X from corner to corner and a + from side to side so you get a star on the base, that way you can tell where the high and low spots are and check your progress.

 

[wwswimming]

if you can find a copy of ANSI Y14.5M, that is a drafting spec that describes
various details of Geometric Dimensioning & Tolerancing, including Flatness.

it describes how flatness is measured, e.g. with a gauge tracing a profile
on the "DUT" (device under test - part being measured - the part you're
using to do the lapping & need to be flat, that is.)

if the top surface of the DUT is not parallel to the bottom surface, that
can cause a mis-read on the flat-ness.

 

[NinjaJedi]

Flatness - A measure of the dimensional deviations from a mathematical plane over a given surface area. The flatness is the smallest spatial separation between two parallel planes that can contain all the measurement points for a surface area between them.

In the manufacture of precision parts and assemblies, especially where parts will be required to be connected across a surface area in an air-tight or liquid-tight manner, flatness is a critical quality of the manufactured surfaces. such surfaces are usually machined or ground to achieve the required degree of flatness. High-definition metrology, such as digital holographic interferometry, of such a surface to confirm and ensure that the required degree of flatness has been achieved is a key step in such manufacturing processes. Flatness may be defined in terms of least squares fit to a plane ("statistical flatness"), worst-case or overall flatness (the distance between the two closest parallel planes within which the surface barely will fit, or other mathematical definitions that fit the intended use of the manufactured part.

 

[Tullphan]

Originally posted by: NinjaJedi
Flatness - A measure of the dimensional deviations from a mathematical plane over a given surface area. The flatness is the smallest spatial separation between two parallel planes that can contain all the measurement points for a surface area between them.

In the manufacture of precision parts and assemblies, especially where parts will be required to be connected across a surface area in an air-tight or liquid-tight manner, flatness is a critical quality of the manufactured surfaces. such surfaces are usually machined or ground to achieve the required degree of flatness. High-definition metrology, such as digital holographic interferometry, of such a surface to confirm and ensure that the required degree of flatness has been achieved is a key step in such manufacturing processes. Flatness may be defined in terms of least squares fit to a plane ("statistical flatness"), worst-case or overall flatness (the distance between the two closest parallel planes within which the surface barely will fit, or other mathematical definitions that fit the intended use of the manufactured part.

:Q
Alrightythen!

 

[ColdFusion718]

Originally posted by: DerwenArtos12
Smooth and flat don't entirely co-inside. I've found that the best I can do to get things flat is either use a milling machine or I have a glass desk that I lay the paper on to use. With a little water on the backside and some tape at the very edges it holds in place quite well, then I hold the heatsink by the base and start to sand. I also like to use a trick my dad taught me, draw an X from corner to corner and a + from side to side so you get a star on the base, that way you can tell where the high and low spots are and check your progress.

What's "co-inside?" Did you mean "coincide?"

 

[DerwenArtos12]

Originally posted by: ColdFusion718

Originally posted by: DerwenArtos12
Smooth and flat don't entirely coincide. I've found that the best I can do to get things flat is either use a milling machine or I have a glass desk that I lay the paper on to use. With a little water on the backside and some tape at the very edges it holds in place quite well, then I hold the heatsink by the base and start to sand. I also like to use a trick my dad taught me, draw an X from corner to corner and a + from side to side so you get a star on the base, that way you can tell where the high and low spots are and check your progress.

What's "co-inside?" Did you mean "coincide?"

Teh grammer natzis strik agin!

 

[NinjaJedi]

Originally posted by: DerwenArtos12

Originally posted by: ColdFusion718

Originally posted by: DerwenArtos12
Smooth and flat don't entirely coincide. I've found that the best I can do to get things flat is either use a milling machine or I have a glass desk that I lay the paper on to use. With a little water on the backside and some tape at the very edges it holds in place quite well, then I hold the heatsink by the base and start to sand. I also like to use a trick my dad taught me, draw an X from corner to corner and a + from side to side so you get a star on the base, that way you can tell where the high and low spots are and check your progress.

What's "co-inside?" Did you mean "coincide?"

Teh grammer natzis strik agin!

My spelling and grammar are horrible that is why I tend to write my post in a email, use the spell check and then C&P it.
I think the thing people need to remember is that English (or should I say Engrish) is not every ones first language.

 

[Tullphan]

Originally posted by: NinjaJedi
I think the thing people need to remember is that English (or should I say Engrish) is not every ones first language.

Heck, I had to take 10th grad English when I was a senior. Kinda embarrassing sittin' in there w/the soph's. Needless to say I was englished out that year! I did love to terrorize my English teacher by saying "ain't" alot, tho...gave her chills! :laugh:

 

[DerwenArtos12]

Originally posted by: NinjaJedi

Originally posted by: DerwenArtos12

Originally posted by: ColdFusion718

Originally posted by: DerwenArtos12
Smooth and flat don't entirely coincide. I've found that the best I can do to get things flat is either use a milling machine or I have a glass desk that I lay the paper on to use. With a little water on the backside and some tape at the very edges it holds in place quite well, then I hold the heatsink by the base and start to sand. I also like to use a trick my dad taught me, draw an X from corner to corner and a + from side to side so you get a star on the base, that way you can tell where the high and low spots are and check your progress.

What's "co-inside?" Did you mean "coincide?"

Teh grammer natzis strik agin!

My spelling and grammar are horrible that is why I tend to write my post in a email, use the spell check and then C&P it.
I think the thing people need to remember is that English (or should I say Engrish) is not every ones first language.

English is my first language, I speak and write spanish and german and am learning conversational chinese(guan or mandarin) and arabic and can read about a half dozen ancient european languages that are exticnt as spoken languages.

Quite frankly english is the most confusing language I know of. Too many rules with too many exceptions.

 

[Tullphan]

Originally posted by: DerwenArtos12

Quite frankly english is the most confusing language I know of. Too many rules with too many exceptions.

I hate rules.

So, back to my HSF, should I get me some 600 grit, make some marks on the bottom & go a few rounds with it & get rid of its' shiny-ness & make sure it's flat?

 

[DerwenArtos12]

Originally posted by: Tullphan

Originally posted by: DerwenArtos12

Quite frankly english is the most confusing language I know of. Too many rules with too many exceptions.

I hate rules.

So, back to my HSF, should I get me some 600 grit, make some marks on the bottom & go a few rounds with it & get rid of its' shiny-ness & make sure it's flat?

if it's not flat, you'll probably need to go to 400 grit to take off the high spots and work your way back up, depending on how far off flat it really is.

 

[NinjaJedi]

IMO you can get the HSF as flat as flat can be but if the IHS is not just as flat it is a waste of time.

 

[DerwenArtos12]

Originally posted by: NinjaJedi
IMO you can get the HSF as flat as flat can be but if the IHS is not just as flat it is a waste of time.

so he should measure the convexity of the IHS and match the heatsink to it?

 

[NinjaJedi]

Originally posted by: DerwenArtos12

Originally posted by: NinjaJedi
IMO you can get the HSF as flat as flat can be but if the IHS is not just as flat it is a waste of time.

so he should measure the convexity of the IHS and match the heatsink to it?

Lapping is meant to join two mating surfaces. So getting both surfaces as close to the same flatness would yield the best outcome. If both surfaces are not flat and you lap only one you are only half way there. If one surface was convex the other would need to be concave to the same tolerance. For the mating surfaces of the HSF and IHS there should be some degree of roughness or unevenness to allow for the TC. If two surfaces are too flat they can be made to stick together.

 

[DerwenArtos12]

Originally posted by: NinjaJedi

Originally posted by: DerwenArtos12

Originally posted by: NinjaJedi
IMO you can get the HSF as flat as flat can be but if the IHS is not just as flat it is a waste of time.

so he should measure the convexity of the IHS and match the heatsink to it?

Lapping is meant to join two mating surfaces. So getting both surfaces as close to the same flatness would yield the best outcome. If both surfaces are not flat and you lap only one you are only half way there. If one surface was convex the other would need to be concave to the same tolerance. For the mating surfaces of the HSF and IHS there should be some degree of roughness or unevenness to allow for the TC. If two surfaces are too flat they can be made to stick together.

By TC do you mean TIM? and what's wrong with lapping both surfaces to a pure flatness and smoothness. Theoretically if they were both perfectly flat and perfectly smooth no TIM would even be needed as it's sole purpose is to fill the gaps that would otherwise be filled with causing insulation between the two surfaces.

BTW my previous post was pure sarcasm. Lighten up a little, if you're not having fun you're wasting time, right?

 

[NinjaJedi]

Originally posted by: DerwenArtos12

Originally posted by: NinjaJedi

Originally posted by: DerwenArtos12

Originally posted by: NinjaJedi
IMO you can get the HSF as flat as flat can be but if the IHS is not just as flat it is a waste of time.

so he should measure the convexity of the IHS and match the heatsink to it?

Lapping is meant to join two mating surfaces. So getting both surfaces as close to the same flatness would yield the best outcome. If both surfaces are not flat and you lap only one you are only half way there. If one surface was convex the other would need to be concave to the same tolerance. For the mating surfaces of the HSF and IHS there should be some degree of roughness or unevenness to allow for the TC. If two surfaces are too flat they can be made to stick together.

By TC do you mean TIM? and what's wrong with lapping both surfaces to a pure flatness and smoothness. Theoretically if they were both perfectly flat and perfectly smooth no TIM would even be needed as it's sole purpose is to fill the gaps that would otherwise be filled with causing insulation between the two surfaces.

BTW my previous post was pure sarcasm. Lighten up a little, if you're not having fun you're wasting time, right?

I understood the sarcasm. I am completely lightened up explaining that was fun lol. Yes you are correct about not needing the TIM if the surfaces were completely flat. I just don't know how well it would transfer the heat without it. It might work better I do not know.

"damn it Jim I'm a machinist not a mechanical engineer"

 

[DerwenArtos12]

Originally posted by: NinjaJedi

Originally posted by: DerwenArtos12

Originally posted by: NinjaJedi

Originally posted by: DerwenArtos12

Originally posted by: NinjaJedi
IMO you can get the HSF as flat as flat can be but if the IHS is not just as flat it is a waste of time.

so he should measure the convexity of the IHS and match the heatsink to it?

Lapping is meant to join two mating surfaces. So getting both surfaces as close to the same flatness would yield the best outcome. If both surfaces are not flat and you lap only one you are only half way there. If one surface was convex the other would need to be concave to the same tolerance. For the mating surfaces of the HSF and IHS there should be some degree of roughness or unevenness to allow for the TC. If two surfaces are too flat they can be made to stick together.

By TC do you mean TIM? and what's wrong with lapping both surfaces to a pure flatness and smoothness. Theoretically if they were both perfectly flat and perfectly smooth no TIM would even be needed as it's sole purpose is to fill the gaps that would otherwise be filled with causing insulation between the two surfaces.

BTW my previous post was pure sarcasm. Lighten up a little, if you're not having fun you're wasting time, right?

I understood the sarcasm. I am completely lightened up explaining that was fun lol. Yes you are correct about not needing the TIM if the surfaces were completely flat. I just don't know how well it would transfer the heat without it. It might work better I do not know.

"damn it Jim I'm a machinist not a mechanical engineer"

I spent some time in the early 80's designing lasers for the USDOD so I'm fairly familiar with heat transfer and with creating convex and concave surfaces. With some of the higher power yag lasers we were building we had to dissipate upwards of 1000W of heat from the body of the projector and the lenses were machined in house(some by me) with tolerances in the range of 1/100th of a thousandth of an inch across an inch and a half lens. with anywhere from 3-7 lenses per laser depending on it's desired range. We knocked out 6 city blocks in tempe one night cutting a vault in half.

 

[NinjaJedi]

Originally posted by: DerwenArtos12

Originally posted by: NinjaJedi

Originally posted by: DerwenArtos12

Originally posted by: NinjaJedi

Originally posted by: DerwenArtos12

Originally posted by: NinjaJedi
IMO you can get the HSF as flat as flat can be but if the IHS is not just as flat it is a waste of time.

so he should measure the convexity of the IHS and match the heatsink to it?

Lapping is meant to join two mating surfaces. So getting both surfaces as close to the same flatness would yield the best outcome. If both surfaces are not flat and you lap only one you are only half way there. If one surface was convex the other would need to be concave to the same tolerance. For the mating surfaces of the HSF and IHS there should be some degree of roughness or unevenness to allow for the TC. If two surfaces are too flat they can be made to stick together.

By TC do you mean TIM? and what's wrong with lapping both surfaces to a pure flatness and smoothness. Theoretically if they were both perfectly flat and perfectly smooth no TIM would even be needed as it's sole purpose is to fill the gaps that would otherwise be filled with causing insulation between the two surfaces.

BTW my previous post was pure sarcasm. Lighten up a little, if you're not having fun you're wasting time, right?

I understood the sarcasm. I am completely lightened up explaining that was fun lol. Yes you are correct about not needing the TIM if the surfaces were completely flat. I just don't know how well it would transfer the heat without it. It might work better I do not know.

"damn it Jim I'm a machinist not a mechanical engineer"

I spent some time in the early 80's designing lasers for the USDOD so I'm fairly familiar with heat transfer and with creating convex and concave surfaces. With some of the higher power yag lasers we were building we had to dissipate upwards of 1000W of heat from the body of the projector and the lenses were machined in house(some by me) with tolerances in the range of 1/100th of a thousandth of an inch across an inch and a half lens. with anywhere from 3-7 lenses per laser depending on it's desired range. We knocked out 6 city blocks in tempe one night cutting a vault in half.

I spent most of my machinist career machining golf putters.