Porter-Cable 7428 speed

[the other pc]

The Porter-Cable 7428 rotary buffer is a fine machine and is currently available at very good prices.

MaximusZTS and Mike-in-Orange both have PC7428s and the question of their speed came up. The buffer is specified as being adjustable from 1000 RPM to 3000 RPM. The speed setting dial is marked from 1 to 6.

It would be easy to assume that speed varies linearly between the specified values with dial setting but electronic motor controls don’t always work that way. So we decided to have a look.

A sample set of two isn’t statistically significant but the data is still useful for these specific machines and may be useful for other owners of the 7428 if PC’s (Flex division) manufacturing is highly consistent.

MaximusZTS’s buffer is called machine “S” while Mike-in-Orange’s is “M”

We took two sets of readings on each. Along with the measurements the chart below also includes the average of measurements for each machine, the range (max-min) and the range expressed as a percentage of the average. Also included is the average, range, range expressed as a percentage of the average, standard deviation and standard deviation expressed as a percentage of the average (coefficient of variation) for both machines combined.

The data was all taken under no-load conditions.



That’s it for now, we’ll have a closer look at the data soon.


PC.

[MaximusZTS]

Thanks Paul for taking the time to do this!

Originally Posted by the other pc

The data was all taken under no-load conditions. PC.

Now it makes me wonder what the speeds recommended for the RB are? If they are speeds under load or not. I would imagine there wouold be a big difference. But at this point it may just be splitting hairs.

[the other pc]

I’m pretty sure the recommendations are no-load. There are too many variables that affect loading, pad type, size, downward pressure, product, product volume, etc to make under-load specifications meaningful. You could specify some particular drag torque but without a tachometer and a calibrated brake (or brake dynamometer) users wouldn’t be able to replicate the load anyway.

Not that no-load is all that precise either. All the same loading parameters will come into play and different buffers will load down differently.

It’s just to get you into the ballpark. Then you have to go by feel and experience.


PC.

[the other pc]

A look variation in the data table above shows that largest variation appears in the lower speed range while the variation in speed reduces in the upper range. I would say that overall speed repeatability looks very good considering the 7428 is a power tool, not a precision instrument.

If we plot the speed data on a linear graph we can get a feel for whether the speed response curve is linear or follows some other curvature.



Looking at the graphed data we can see that the response does indeed appear to be straight-line linear.

Based on our findings I would say that anybody could come up with a completely practical graph for estimating a 7428’s speed vs. dial position by merely laying down a ruler and drawing a straight line from 1000 RPM at setting 1 to 3000 RPM at setting 6 (which is equivalent to the mathematical expression: Speed = 400 x Dial setting + 600).




Well, that really should be good enough but…


PC.

[MaximusZTS]

Originally Posted by the other pc

PC.

I may print this out and tape it on my rotory! Great work Paul.

It would be interesting to see how this data would compare to other rotorys out there. Makita, etc... And to see if there may be some standars out there.

[Michael Stoops]

Fantastic write up Paul, thanks so much for taking the time to do this - and for bringing in the test equipment in the first place.

Shawn's numbers look to be spot on when looking at that last linear graph you did while mine seem a bit off. I wonder how much that difference is a function of the machine and how much is due to how careful we both were in duplicating the settings on each run. Either way, we both know our rotaries a whole lot better thanks to you.

[MaximusZTS]

Originally Posted by Mike-in-Orange

Shawn's numbers look to be spot on when looking at that last linear graph you did while mine seem a bit off.

Well M1 was the first run at the mesurements. That may have something to do with it.

[the other pc]

I’m guessing that temperature is part of the variation we see on a given machine, drift that happens as the motor and circuitry heats up. I’ll bet there’s hysteresis (the electronic equivalent of mechanical backlash) in the system too. The deviation of M1, point 3 could have been caused by overshooting the target position then coming back to it. Without more experiments it will remain a mystery.

Even though the straight line from 1k to 3k would probably be good enough for all practical purposes we can come up with curves that more accurately represent the real behavior we measured.

Using the “ruler method” we could draw a line through the actual data points rather than the specification end points. In fact, we could draw any number of lines through the data points and be “pretty close.”

There are mathematical techniques for doing just that. They’re called “curve fitting” and more than just drawing curves that are “close” they can be used to find a “best fit” curve that comes closest to representing the real data.

And, thanks to the marvels of modern computing, that’s not hard to do. So I did. The graphs show the averaged data for each machine and the two together along with the equation for the “best fit” line though the data points. The tables are based on those equations.

Originally Posted by MaximusZTS

I...It would be interesting to see how this data would compare to other rotorys out there. ..

I did it for my Metabo PE7175. The 7175’s response isn’t even close to a straight line. It did fit very nicely to a logarithmic curve.




PC.

[Michael Stoops]

So according to your "curve fitting" graph, I need to have my 7428 tweaked by Spinal Tap since it only goes up to "6" as it is.

Interesting that your Metabo is not (practically) linear like the 7428s are. I guess I just assumed all such devices would be, but we all know what happens when we "assume" something.

[Superior Shine]

After evaluating your charts and graphs and appreciating the time it took to put all that together, I can safely assume that you have already completed your Christmas shopping.