Harbor Freight Calipers inside and out

[rgsparber]

This is a restart of a thread about repeatability and accuacy of the HF caliper.

Two articles are presently available that describe my best understanding of these calipers. The first

http://rick.sparber.org/Articles/PAC/PAC3.pdf

deals with just the caliper. I felt compelled to retract some of my conjecture related to extrapolted accuracy presented in versions 1 and 2 but added more related to repeatability and how it interacts with accuracy at values near zero.

These retractions may later be reversed when the work from ckelloug is done. He has Jo blocks and the background to clearly see the absolute accuracy of these calipers in the range of 0 to 4".

The second article is a 28 page monster which was coauthored by Larry Gill of the Yahoo group "shumatech" and myself.

http://rick.sparber.org/Articles/JB/JB4.pdf

It does talk a lot about the caliper but more from an internal standpoint. The shumatech DRO directly accesses the caliper's internal count. There is a display mode that permits the user to see this count.

One thing that should generate a nice little "fire storm" is the section on repeatability. The experiment suggests that if you use this internal number you can get a short term (i.e. less than 10 minute span) repeatability of ± 2 tenths rather than the adverstised ±5 tenths.

I welcome your questions and comments. No reason to stop at the present versions.

[Swarf&Sparks]

Rick, even if that is so (resolution)
They only display to 0.0005" or 0.01 mm
How does that help the user?
I don't wish to denigrate your work.
Maybe I'm just having a blonde night?

[lazlo]

The experiment suggests that if you use this internal number you can get a short term (i.e. less than 10 minute span) repeatability of ± 2 tenths rather than the adverstised ±5 tenths.

That's interesting Rick -- I don't have time right now to read the latest version of your document, but how far are you moving the caliper when you're measuring +/- 2 tenths repeatability?

Are you still seeing +/- 1 thou accuracy during the same time-frame/conditions?

[rgsparber]

Rick, even if that is so (resolution)
They only display to 0.0005" or 0.01 mm
How does that help the user?
I don't wish to denigrate your work.
Maybe I'm just having a blonde night?

I was refering to the internal number. A single count here ideally equals 0.5 tenths. The Shumatech software filters this number to eliminate jitter so I am really looking at the average value.

If you don't have a Shumatech DRO, then you are right. It does not help you. Sorry for the confusion.

[rgsparber]

That's interesting Rick -- I don't have time right now to read the latest version of your document, but how far are you moving the caliper when you're measuring +/- 2 tenths repeatability?

Are you still seeing +/- 1 thou accuracy during the same time-frame/conditions?

My repeatability test uses a recently factory rebuilt Starrett Last Word finger DTI to sense touch down on the Z axis. I move the quill up about 0.2" and then lower it until the DTI reads zero as viewed through my Optivisor. I have an ultra fine feed on my RF30 that permits me to move a lever 2" and get a 2 thou down feed (it is just a dumb 15" long rod threaded 1/4-20 on the end that screws into the Z axis collar).

At touch down I read the internal number in HEX on the DRO. This operation was repeated 10 times in less than 10 minutes. Past experience has shown me that I can hit TD within ± 1 tenths with the DTI. About 5 hours later I did one more TD and saw a drift of 2 tenths. No surprise there.

The overall accuracy is still, at best, ± 1 thou. As you get above 100 mm of movement, the error goes up.

[Swarf&Sparks]

Fair enuff Rick.
I was not aware of your DRO application.
I just use em as calipers and height gauge.
Cheers,
Lin.

[lazlo]

I move the quill up about 0.2" and then lower it until the DTI reads zero as viewed through my Optivisor.

At touch down I read the internal number in HEX on the DRO. This operation was repeated 10 times in less than 10 minutes.
The overall accuracy is still, at best, ± 1 thou.

Ah, got it Rick -- interesting results.

[IOWOLF]

Ummm, perhaps I am missing something, But ......IT IS A HARBOR FREIGHT CALIPER, WHAT DO YOU EXPECT?

Tenths? yea right on a Cheap DRO?

[rgsparber]

Ummm, perhaps I am missing something, But ......IT IS A HARBOR FREIGHT CALIPER, WHAT DO YOU EXPECT?

Tenths? yea right on a Cheap DRO?

It is not what I'm expecting, it is what I saw. The subject is short term repeatability, not accuracy.

I see a short term repeatability of ± 2 tenths. This repeatability is good for at least 10 minutes.

[Carld]

So Rick, are you making turbine parts for the military or rocket parts for NASA? What kind of machines are you making these parts on?