Official Luthiers Forum!

Recently learned about these Plek machines.

As primarily a guitar repairer does this mean i am beginning to be replaced by a machine?!


Does anyone have more info on them?

http://www.plek.com/en_US/home/


anyone experienced a finished product?



Slightly scared,
Jack!

_________________

Stringed instrument Set up and Repair based in Manchester - UK
http://www.jacksinstrumenservices.com

You're plenty safe. Those things are big bucks and they basically just level the frets.

_________________
http://www.PeakeGuitars.com

And I'm told they're subject to use error, as well.

_________________
Bob Garrish
Former Canonized Purveyor of Fine CNC Luthier Services

At $500,000 each I doubt if the neighborhood repair shop is going to be putting one in anytime soon.... Actually I am not sure that they are 1/2M but I have heard this - anyone know?

I still think too that a human can be just as accurate leveling and dressing frets if they know what they are doing.

_________________
Ann Arbor Guitars

$80K is the last I was quoted, but that's a couple of years old. Still not in my budget though.

They can be optioned up to cut nuts and do inlays too. According to one factory that I visited, it doesn't save them any time on fretjobs, but they like the diagnostics end of it. They can call up a serial number at any time and get the exact relief and fret specs. Valuable info if you get a neck back to work on.

If your repair shop is in a busy center, the Plek might make you money vs putting you out of business. That was Joe Glaser's take on it. I haven't talked to Mike Lull about his.

I saw this thing a few years ago and immediately wondered about it. You can scan the frets under tension and identify issues, that's for sure. I've used laser scanning and reverse engineering as a matter of course. The technology is sound. What I DON'T get is how the frets are "adjusted" once a problem is detected. It's hard to tell form the web site.

I'm guessing it goes like this:

1. Scan the frets under string tension.
2. Identify problem frets.
3. Software profiles course of action.
4. Remove strings and relieve stress.
5. Rescan frets without string tension.
6. Software plots tool path applying former profile to new scan.

That might work but by the time the technology is developed to the point where it actually WORKS the cost of the procedure and equipment puts the price
out of reach when you consider having a luthier set up a guitar.

It looks to me that they are trying harder to sell the Plek machines than the service.

_________________
I read Emerson on the can. A foolish consistency is the hobgoblin of little minds...true...but a consistent reading of Emerson has its uses nevertheless.

StuMusic

I've re-dressed PLENTY of Plekked instruments and have gotten superior results nearly every time. Sure the laser scanning portion is accurate, but the grinding and polishing aspect of it is still mechanical and subject to machine tolerances, wear, and the guitar still has to be jigged up and tightened down by a human.
A skilled human who takes his time, checking and rechecking as he goes can do some pretty tightly toleranced work IMHO.
The Plekked instruments I've seen have been mostly Gibsons, and the Plekked instruments are still better than the guitars that have had the 10 minute grind-polish-setup that they usually get, but I think a human can still do a better job.

I've plekked a couple hundred guitars.

The machine is ridiculously accurate. However it, like anything else, has to be kept in tip top shape, and the user has to know how to operate it correctly to get the best results. Is it more accurate than a human? Yes. Does that extra accuracy actually mean anything? Not necessarily. If a experienced person does a perfect fret level, crown, polish, setup while the plek will still be more accurate your at a point where the difference between the two is not noticeable. They are awesome machines, but a good luthier can do work that will feel/perform just as good. Likewise if the person doing the plek job is half assin' it then the results will suffer.

_________________
John Mayes
http://www.mayesluthier.com

I don't agree that a plek is more accurate than a human.

The methods that I have learned from David Collins deal with tolerances the thickness of marker ink... this is sub .001". Plek does not claim this level of accuracy in their promotional materials and anything mechanical unless new or perfectly maintained is not going to operate at sub .001" tolerances.

Granted not only with the plek but with manual methods the human does have to perform the tasks flawlessly AND really understand what they are doing and why to obtain this level of accuracy.

What a plek does do very well IMHO is these tasks with very good accuracy over and over again never complaining of tennis elbow or impediments to a satisfying.... well never mind.... It's a cool machine if you can justify the cost by keeping fret jobs stacked up and ready to go and have an operator who enjoys using the technology and won't get bored. Methinks that the guys who recover their costs the best must be promoting the machine pretty well and getting lots of business. One has to sell a LOT of guitars or do an awful lot of repairs to justify the six figure price for a new one with lots of bells and whistles.

A Plek is no different than any other computer based, mechanical system and process - it's accuracy is dependent on any and all things that impact mechanical things. A human can do a more accurate job though.

_________________
Ann Arbor Guitars

Plek is accurate to .0004 of an inch on the cutter, or .01mm. The sensor finger is .0002. Sorry but we are not that accurate.

And of course, as you state, the machine has to be kept in top shape to be that accurate. Same can be said for a human.

These accuracies are stated in their literature.

Don't get me wrong I don't feel the need to have one at all. I do a perfectly great job by hand, and it feels and plays as good as any plek can do, but I'm not four ten thousandths of an inch accurate. Nor is there any need to be.

_________________
John Mayes
http://www.mayesluthier.com

Well I still beg to differ John and I don't believe everything that I read - I'm sure you don't either so no insult intended.

The CNC guys that I know, and I have one in the room with me at the moment, tell me that the thickness of my marking pen's ink or his bluing is a closer tolerance than any mechanical CNC machine could ever match - reliably, repeatedly, and over time. Granted humans have bad days too and need maintenance... from time to time.

Some very demanding players require the benefit of considering all of the variables when setting up their guitars. Not saying that sub .001" results are always required but I find something of value in knowing that how I learned to do fret dresses, manually will provide me with these results on demand.

Wow I'm arguing against CNC technology for a change...

_________________
Ann Arbor Guitars

Well I think your wrong and you think likewise... so be it. I've done hundreds by hand and hundreds on a plek. I believe what I know and so do you... I'm sure we can both live with it.

_________________
John Mayes
http://www.mayesluthier.com

No prob John - here's a link to a very good discussion of plek vs. highly skilled Master Luthier and a far better defense of a skilled artisan than I provided. It's great reading and as always my money is on David Collins, besides he's not outsourcing to foreign lands.... : http://www.tdpri.com/forum/tele-technical/230744-plek-machines-zombie-thread.html

_________________
Ann Arbor Guitars

Interesting discussion. I still hold my same views.

_________________
John Mayes
http://www.mayesluthier.com

Humans can rub two surfaces together randomly better than most machines, but that doesn't extend to being able to work in three or even two dimensions accurately. The only thing it proves is that we're better at finding blue spots and dull spots than most machines. David's argument is absolutely correct in that tolerances cascade from the weakest link: a human being can't even draw a straight line or a round circle to tolerances you can't find on a measuring tape.

_________________
Bob Garrish
Former Canonized Purveyor of Fine CNC Luthier Services

We have to remember that the most accurate machine on the planet evolved from human hands. (and brains)

True.....but WHY did they evolve?

When that question is honestly answered, the person asking must conclude that machines and technology are a means to an end in a practical world. Of course....there is also the mystical world in which many luthiers reside. In this world, "craftsmanship" is a mystical, virtous spirit which takes up residence in the hands. But I live in the practical world. In my world, "craftsmanship" resides in my head and heart and not in my hands...or in my tools.

I don't understand why value is attached to the idea of an instrument being "handmade" when accuracy can be measured, and therefore quantified, except as a rationalization for those who can't or won't embrace newer technologies. I have EVERY respect for the craftsman that does his or her BEST with what they have, and there some marvelous examples of people like this on this board...but I can't even define "handmade".

Can you?

What does "handmade" mean when vitually all of us would cease to build guitars if we couldn't obtain the tools and materials made within a practical, industrialized world?

Cnc machines are now invented. They won't ever be uninvented. They will become more prolific, easier to use, more versatile, and gain more applications as time goes on. If they don't have value simply as a tool they will become extinct....as poor tools do. Plek is a threat to luthiers if it functions consistently well and its cost can be justified within its market. The jury is still out on the Plek machines I think. The thing only exists because someone had a vision and followed it through. That person may turn out to be an intelligent knucklehead though.

_________________
I read Emerson on the can. A foolish consistency is the hobgoblin of little minds...true...but a consistent reading of Emerson has its uses nevertheless.

StuMusic

I dunno. All I know is that there's about ten guitars down at Guitar Center right now that have been plekked already and could use a fret dress.

_________________
Old growth, shmold growth!

If indeed the plek machine is supposed to level the frets, and even I can detect high frets with my straightedges on specific frets that produce fret buzz, then I think it says at least SOMETHING about the Plek process...

My point is simply that, no matter what any machine looks like on paper, it MUST produce real world results.

I know that some high powered guitar techs with great reputations and a large clientele use the Plek machine. I also know they had phenomenal reputations BEFORE the Plek machines. I think the Plek process is benefiting from THEIR reputation, not the other way around.

I'll say what I said earlier. Based on my observations in a busy guitar shop, the Plek machine does a better job than the average 10 minute Gibson fret dress and setup, but a careful technician who check and rechecks as he goes can do a better job ultimately. If that weren't true, I should be blown away by how phenomenal the Plekked Gibsons hanging on the wall at Guitar Center are, and I shouldn't be able to detect high frets in front of specific frets that are producing buzzing.

_________________
Old growth, shmold growth!

Very interesting thread...I often wondered about the Plek machine and its value.

For my day job I work as a manufacturing engineer for a company that makes large gears. We have many large grinding machines that kind of work similar to a Plek machine (other than they cost between $1.5 - $2.5M each). The grinding machines probe the gear teeth, determine the error in the part, grind the gear teeth, and then remeasures the part to ensure accuracy. If the part still has form error the machine will redress the grinding wheel, regrind the gear teeth, and remeasure. It is an amazingly productive and accurate process.. However to ultimately determine the part accuracy we need to take the gear off the grinder and send it to a separate CNC coordinate measuring machine since measuring parts on the machine that produced the part can also lead to a certain amount of error. It isn't uncommon for use to achieve tolerances better .0004". The grinding process and coordinate measuring machine are both held at 72 degrees since temperature variation will affect the part quality. Achieving accuracy at that level on every part is extremely difficult.

When I read that the Plek machine is .0002" accurate on the probing system and .0004" accurate on the grinding process I am impressed. However when I think of the effort and control it takes to truly produce a part with accuracies between .0002 and .0004" I am not convinced that it can be done on a Plek machine since creating part with .0004 accuracy is not the same thing as creating a machine with .0004 accuracy. But lets face it, at the end of the day is it essential that a guitar is perfectly created to .0001" precision or is the "feel" of the guitar more of an art than a science? I would say that the feel of a guitar is more or an art and if the operator of the Plek machine doesn't understand the art then the machine itself isn't much of a threat.

_________________
Brad
Avon, OH

I, a while back, sent plek an e-mail asking them about the accuracy of the machine in a non-based manner. Here is a snippet of what they sent back:

"Hello John,

Those numbers are correct and they do already represent the repeatability. The absolute resolution of our sensor module is even higher: 0.001 mm = 0.000039''. Of course the resolution needs be higher than the repeat accuracy. A professional guitar player will start noting fret height deviations of .001'' - so PLEK is on the safe side! And yes, PLEK is very different from a typical CNC machine."

And then I got an e-mail from Joe Glaser who some of you may know is the US rep for Plek. Here is his thoughts on my questioning of the accuracy.

"Hello John,

Not that you need me to weigh in on this subject but here is my point of view:

There is a lot of discussion about the accuracy, resolution and repeatability of the PLEK and a lot of this discussion happens in the theoretical realm of the Internet but we, as PLEK, have to or get to live, test and execute in a real world of measurement. Sometimes this is gratifying, sometimes cruel. It is not without issues and it is all about overcoming problems. Seen over fifteen years, things do improve constantly, backslide a bit here and there and then improve again but all of this takes place in a calibrated and verified arena and this is the key; we're not talking paper accuracy, we measure, and measure again and like anything you test more than once and with different tools, you prove or disprove its validity.

These claims of excellence and maximum accuracy are obviously hard to accept sight unseen because virtually no one else has the ability to measure and draw conclusions except by the sort of faith that preceded the PLEK in my world: I wanted to be excellent, I felt that I was excellent and it was my word against the world. Still, absence of evidence is NOT the same as being correct or good. There is safety in non-verification but, for me, this had always been a world of bluff and rumor, not fact, and that made me uncomfortable. One thing that convinced me to consider an expensive PLEK was scanning my own work, seeing that it was mostly very good but not perfect and seeing how it measured up to others and much more importantly, to the best work possible both in quality and consistency. In the long run, I have come to value the saved data, saved time and saved arm wear and tear but the eye opener was the visibility.

In fact, the noticeable and acceptable fret deviation that we have derived from testing/playing and use as a standard limit for the PLEK is about +/- 0.02 mm though it may tighten for very, very low action settings. This is more like 0.0008 in. unless I am mistaken. This means that at a low action, people can detect 0.0010 in. or we would not have made this our go/no-go limit and it has been my experience that they can, in fact, detect 0.0010 in. That is why our typical actual results of about +/- 0.005 mm or 0.00020 inches yield excellent playability as long as the other setup is correct, a neck stays fairly well adjusted and wear is limited. (Making this tolerance a little more complicated is the need for the relief and total fretplane surface to also be ideal, not just local levelness, but that is the general spec.)

Secondly, it is true that our measuring resolution is approx 0.001 mm and it is through repeated scans of the same neck or test object that we see how accurate and stable this number is. In my experience, the deviation in repetitive measurements of the same object is never the absolute mechanical limit but is still less than 0.005 mm and often well less than that. The difference between measuring accuracy and cutting accuracy is explained by the minute rise and fall of frets during cutting, wood expansion from work and heat, cutting module and axis gear backlash/play and fixturing flexibility. These factors, all added together, still produce less deviation than our target tolerance of approx +/- 0.020 mm or 0.00079 in. on about 98% of what I run and always 100% before the job is done.

The other key is that we know this because we scan all work before and after each process and collect and compare that data. If a fret dress is graded as achieving PLEK quality then it is measured as achieving that quality and the actual results are graphed for viewing. Likewise, unsatisfactory results would be spotted and marked as needing further work. That before/after check establishes the accuracy of everything except the measuring finger itself and that part is proved by re-measuring the same fret or other object at different times and looking at the results.

I know these are just details but it is how we know what we know because we are not just promising a certain accuracy spec based on component part specs, we are able to check this ourselves and do it on every instrument. That is what make us different. That proves our claims and, assuming operator competence, assures quality control of the PLEK work. You are more than welcome to visit and see this in action or to do so at a NAMM show.

Thanks, Joe"

In the end I still believe as I did before. I doubt a human can not only level across such a long surface to those specs but to do so further in the arc that is needed for relief (which changes depending on string gauge, and action among other things). Again as I said before is there really a need to be THAT accurate? I don't think so. It's nice to know it's capable of that but that level of absolute accuracy is not an absolute must. Can a expertly trained, and skilled tech do a job that feels as good as the Plek? Yes.

_________________
John Mayes
http://www.mayesluthier.com

That is some rather amazing stuff and I'm sure it took a lot of thought and research to develop the Plek machine/process.
Very interesting, Brad and John.

John,
Thanks for taking the time to put the details in on the Plek's accuracy. Yesterday after I posted I was reading the link in one of the prior posts that talked alot about the Plek process and the machine accuracy. It is clear to me that the advantage of the Plek is really in the control software that allows various relief values for each string. I believe this to be the real brilliance of the machine. It would not surprise me if the machine is able to repeat accuracy of .0008" which is probably pretty good for guitars. While the probe on the Plek is .001mm resolution and in the post you menionted that Joe Glaser (I believe that was who wrote it) said he often gets results that show less than .005mm repeatability. In my professional experience we often make parts that are checked on the same machine that makes the part and get number less than .001mm. The same parts when checked on a separate CMM machine or usually in the .003-.004mm range. When chasing microns (.001mm) it is amazing how many things can affect the results. For most CMM's that are accurate to the .001mm range they typically have air bearings that everything rides on, the entire machine is on air springs to eliminate any vibration that could affect the machine, and the room the machine is located is air conditioned to + or - .5 degrees and humidity controlled.

The reason I challange the accuracy Plek machine isn't to try to disprove that engineers that designed it. What really interests me is at what point does the accuracy of a fret level affect the playability of the guitar?

All that beind said the Plek machines is an amazing machine. I wish some time to use the machine since I am sure it would expand my ideas of how to improve my setups.

_________________
Brad
Avon, OH

For all the talk of accuracy and repeatability and whatnot, and for all the millions of dollars spent researching and building the Plek process, there's STILL ten guitars down at GC that have been Plekked, that STILL need fret dresses.
I've also re-dressed MANY Plekked guitars and cleaned up notes that were buzzing due to high frets for customers.
What counts is what the customers actually get, NOT what a spec sheet or lab test shows, nor what people put on the internet.
To my mind, the variable that the Plek process doesn't seem to be able to take into account (based on Plekked guitars that I've seen) is the minor imperfections in neck straightness that all Gibsons seem to have, which shows up as an S-curve in the neck. What does the spec sheet say about that?
And even though the Plek guys might be able to come up with some explanatory theory about why their machine works around this variable, how come I'm able to re-dress and improve so many Plekked guitars?
I'm basing my thoughts on a specific company, since I know no one else who has Plekked guitars.
The only Plekked guitars I've ever seen have been Gibson factory Plek jobs.

_________________
Old growth, shmold growth!

I firmly believe it's not the Plek's fault in your case but rather the operator or some other circumstance. The Plek WILL account for the "s curve" in the neck. If it can be dressed out it will do it. If it cannot it will tell you it is beyond range. The Plek is a tool. It still has to be used correctly, and maintained correctly to be as good as it can be. If it is not then the results will be sub-par. If you're a repairman you should know by now not to think anything that comes out of Gibson's shop to be consistent!

_________________
John Mayes
http://www.mayesluthier.com