Plek

[FRANK]

 has anyone ever toyed with going out and Having a plek machine do the fret work on a new construction instrument ? Ive been thinking of this, 2010 I was at the Plek Booth at the NAMM show the guy there explained on my best day doing frets Id be 80% accurate I said no way Joe his response was yes it is, and his Machine would smoke me and so Now I Have some beautiful instruments coming to completion and I would only want the best for the customers that play them has anyone PLEK ed before ?

 

[Jeffrey L. Suits]

Jeffery explain Plz

David Collins' post covered anything I need to say, other than I use the strings-on method of leveling, so that the neck is in real time tension.

[John]

When I install frets I bevel the ends and put the strings on and play it. If there are problems I address those problems by dressing out specific areas. Sometimes there is more dressing than others. I like to leave as much fret as I can. I like technology but CNC machines are slow, accurate but slow.

I worked with an old school builder years ago he would freehand pickup and trem routes on this huge and fabulous pin router, he followed pencil lines with perfect speed and accuracy. I asked him why he didn't put the template on the bottom he said it took too long.

 I read recently that some of the bigger manufacturers in North America are using rotary cutters in place of CNC machines. Don't get me wrong I think they are cool, but firstly are they really faster than a good guitar tech or builder and secondly are they  cost effective in both time and money savings.

I don't know where you draw the line between craftmanship and technology but when solo acts and duos were using sequencers was it really any different than using backing tracks? And is using a CNC machine any different than just buying a body or neck? 

Back to the PLEK there is a small margin for error in fret dressing, meaning that the strings need to clear the frets below them. If they do that. And the frets aren't perfect. Is it a problem?

[David Collins]

To begin with, I don't use files for leveling much anymore. You can eyeball or check them with a straight edge, but when it comes down to fractions of a thousandth I just don't think you'll find many you can trust for perfect accuracy (if such a thing exists). In addition to inconsistencies in straightness, they just aren't easily available in lengths to cover an entire board, which I believe to be important. 

 

I buy drop cuts of 2"x1" aluminum rectangle tubing at my local metal supplier, and then cut them to various lengths and true them up on a granite surface plate. The surface plate was trued and certified to .000025" (25 millionths of an inch) shortly before I bought it, and though it's been a few years without being re-certified it's been lightly used, and I'm pretty sure should still be good enough. I lap all my leveling bars and check them on this, and am pretty confident they are at least well within tolerances of .0001". I use 6 different lengths from 6.5" to 23", each length with four different sanding grits for a total of 12 bars. 

 

My process starts usually with a bar just longer than the entire board with 220 grit paper. This is the first referencing stage to see how good or bad it is. From here I can adjust the truss rod to strategically minimize material removal, go to a heavier grit or bring in the files if much material needs to be removed from any area, or file/level away a rise in the extension to get it out of the way before starting the true leveling of the main section, etc. 

 

Once I've hit all or nearly all peaks of the frets, I will recrown them all to ensure that in subsequent lapping sessions the leveling bar is hitting a minimal and approximately equal area on each fret. After the first crowning I will come in and mark the tops again, and come back with a finer grit to do what will hopefully be the final leveling. At this point I hope to see all the peaks hit within one or two strokes of the bar. If it takes more than this, then I will sometimes have to recrown, remark, and relevel again. 

 

Typically after two or three leveling/crowning sessions I will find each peak hit with a single stroke of the bar over each area as I move from treble to bass, though sometimes it can take four or five sessions before I'm completely satisfied with the results. This may sound like a lot of sanding and material removal, but in these last refinement sessions we are really only talking about evening out errors in the tenths of thousandths at most. 

 

At this stage I will begin the polishing, usually with 320 grit sandpaper, first on the sides, then across the top. Then, after the polishing is begun and all the leveling and crowning scratches removed, I will do one final reference check. At this point I use a leveling bar set up with 3M 30 micron precision lapping film to eliminate any false readings that may be possible from irregularities in traditional sandpaper. This is simply a final reference check, and does not remove any significant material or flatten the tops in any significant way. All I'm looking for is to make sure that each peak is hit all the way along each fret. 

 

Then I go to final polishing, starting with 400 grit paper, then of to my buffing arbor set up with 3M radial bristle polishing wheels. When I was first working to refine these methods I completed a number of fret dresses and went back to check them with the lapping film bar to ensure that no errors could be introduced in the final polishing, and never found the slightest problem. 

 

Now there are a number of other little details to the process. Leveling in fall away at the extension is of course quite simple. For very minor fall away I may use a long bar, 15-18", with one or two pieces of tape at one end to create a very shallow angle. Mark the frets, and level until they are touching the area you wish for the fall away to begin. For more dramatic fall away I can use a shorter bar, more layers of tape at the end, or both. This can be done early or late in the leveling job, though I usually prefer to start it off early. 

For introducing variable relief between treble and bass sides, this can be a bit more tricky. The basic idea is to level the treble side straight, then gradually induce some degree of backbow as you move toward leveling the bass side. I do this differently on different necks. For a bolt on neck I may clamp the heel tightly in a vise and wedge a padded block under the peghead, tighten the truss rod to straighten out the center, then push upward slightly on the back of the neck to induce backbow as I move to level the bass side. For acoustics I may rest the center of the neck in a cradle and apply slight pressure up or down on the peghead to change the neck's shape. 

 

There's a lot of slight of hand in this technique, and feel and experience can be critical in getting consistent results, but it's certainly doable with practice. It might sound a bit haphazard, but like any skill you can get pretty consistent and accurate with practice. I started out using a neck-jig type setup modified to apply true string tension to the neck, and with rods and gauges to observe just how much backbow I was inducing on the bass side. After a while though, through comparing results of doing it by gauge or by feel, I just decided that the jig/gauge setup was overkill and really didn't offer any improvements once you got the feel for doing it by hand. 

 

Now what advantages do I think the Plek really has to offer? Well, there are a few. First, it is true that some necks can take a different form under true string tension that is not easily simulated with forces perpendicular to the neck. Some necks and truss rod systems can be more active in some areas of the neck, and create curves or twists which only exist when they are adjusted under string tension. I find this phenomenon fairly rare to occur to any significant degree, but it can happen. Or ribs may twist, tops may hump, and ramp up the extension on an acoustic which disappears once string tension is off the neck and body. These can be largely accounted for in hand leveling, but it can be tricky and would certainly be easier with a Plek. I wouldn't mind having one with a special program for Pete Seeger longneck banjos for example. 

 

On the other hand, how much do these generally minor long range anomalies effect final results in playability? Even if Plek'd, do they stand only to change with season or string gauge anyway, or are they essentially far outweighed by normal neck movement in actual playing? I can't say for sure, but I'm not too worried about differences here between hand lapping and Plek machining in most cases. 

 

Short range tolerances however can be easily seen to be of utmost importance. If adjacent frets are out of level, and if string buzz on a fret immediately after the one being played is a limiting factor to setup, then any error between those two frets will have to be multiplied by eighteen at the saddle height. If fret 5 is one thou higher than fret 4, then the saddle will have to be .018" higher to get the same clearance over fret 5 when playing fret 4. Same rule applies to any adjacent frets across the board. Half a thousandth error means .009" increase in minimum saddle height. By the time we get down to tenths of thousandths, then it probably gets to be insignificant, as the limitations of the overall plane of the fret surfaces and buzz on more distant areas from that being played will probably be the more limiting factor.  

 

In these short range tolerances especially, I find it hard to see how the Plek could beat the precise tolerances of careful hand lapping. A cutting head on a series of rails and bearings, raised and lowered by gears or lead screws, continually moving up and down guided by sensors to level a series of points simply is not capable of achieving the level of precision that a traditional lapping methods can achieve. The technology and construction may be advanced enough to be as good in practical application, but when you really want to split hairs on precision I see no way it can match the standards of a good old analog leveling. 

 

I could get in to concerns of deflection of the neck and distortion of the leveling bars under pressure, or how the Plek relies on the same series of rails and sensors to check its work as it uses for the cutting, and all sorts of other nitpicky details, but I think this is probably too long a post already. Point is though, I'm pretty sure that I can at least match the tolerances of the Plek with my methods no problem. It might take me an hour instead of 20 minutes, and I don't want to do this all day every day like the Plek does, but when needed I think hand lapping methods can easily stand up against Plek work. 

[Robert May]

David, that's a great post. Thanks for taking the time to go into such detail. I've never used "precision lapping film", but you have me interested in trying it. Does it come with an adhesive backing ? If not, how are you attaching it to the leveling beam ? How long is your final lapping film beam ?

I've been going up to 2,000 grit 3M black wet/dry on my 8" level beam, 'cause the 3M paper isn't long enough for a longer beam, unless I use double-stick tape, which I did for a while, then I installed spring clips inside the ends of the leveling beam, and then I discovered since my (home-made) beam has glass glued on the business edges, wet 3M black paper sticks to it very well, at least long enough for the job at hand.

 

[FRANK]

Incredible Man I feel I am a dam good fret artist also not ever having many problems after building some skill with most guitar necks Ive encountered .But 99% Accurate is 99% Accurate and if I am not 99%,and I can achieve this by cnc well then wouldn't my guitar seam to play better than most or is that just horse shit. each time I am asked to set the action on a Guitar everyone says give me the best results and they each make sure I hear them say this GIVE ME the Best ,now if I go CNC and pass down the cost didn't I just deliver the goods so to speak .My discussion is about a couple 30 acoustic guitars I built recently over the summer and giving it the true test with the Plek to find out if all around playability improves.its also about My potential customers as well Id like to see my customer play anywhere with ease .

[Mark Pollock]

Dude, if you can level and dress frets so that they have no noise, feel good, and make a customer happy, then what else could you possibly need? A PLEK can do more of them than you, faster than you, and without repetitive stress syndrome, but that's like saying that we need to eat out every night because Denny's has a more efficient kitchen that we have. If you can do the job just fine, then there's no need to mess with a machine.

 

I would also question the sanity of using a PLEK on anything that is brand new and not broken in, unless you are a factory. I don't know about you, but it takes more than one go for me to get a fret job just right, partially because I'm a bit of a noob, and partially because everything adjusts with new strings and time. I get it right at first, but then something settles a bit, and I get a buzz somewhere, which I tackle with some careful polishing, and everything turns out fine. I don't see going back to a PLEK over and over to get it dialed in.

 

And then I read Rusty's entirely too reasonable response and I feel humbled a bit. But screw it. If you have the time to do the jobs, go for it. If there is truly enough savings to go find somebody and see what happens, then go for it.

[Russell Vance]

Hi David, thanks for that, it's probably a landmark post on how to do a good fret-job and as as a Scientific Instrument Maker by trade I can see your practice very sound, comprehensive and appropriate - I did my 5 year apprenticeship hand lapping stuff and generally wearing out my knuckles and eyesight as well.

But, to compare and contrast machines to hand finishing is a futile exercise - automotive cylinder bores and crankshaft bearings used to be hand lapped.  Watches were hand made with hand polished pivots and gears. The end of these practices signalled the beginning of high reliability and high performance as the machines consistently finish to tolerances unheard of in human endeavor. 

The ability to consistently replicate a complex and asymmetrical fret profile can only be a bonus to the average guitar player and buyer - the real cost of a well executed hand fret job of this type is probably way beyond what most of us are charging - more than the market can bear for most musicians who already  think 100 bucks is too much. 

Also, the ability to graphically/computer read a neck under true string tension is a remarkable time saver:  I may just be unlucky but I've had more than my share of necks that do not conform under string tension to what they looked like on the bench - I've done too many fret jobs and touch ups twice to fix up wobbly necks.  To have a machine read and compensate for this is to me, anyway, a must have.

But, respectfully, these points are mute - at one end of the spectrum we have the professional guitar makers who need speed, accuracy and consistency along with low labor cost to compete with the rest of the world - I note that China and Korea and Mexico aren't awash with Luthiers but they surely put out a lot of reasonable guitars using machine based manufacture (include a lot of the world's Fender guitars here). Plek in this case is essential and 100K is not a lot of money in that context.  

At the other end of the spectrum we have our hand made beauties and our labors of love finished by hand with precision hand tools  - no need for a Plek at this end of the spectrum.  The discussion is more about dispelling the myths and preconceived dislike of these dreaded machines rather than discounting hand finishing and our role in the soul of a finished guitar.

Philosophically yours,

Rusty. 

 

[FRANK]

that was heavy Rusty thanks .

[David Collins]

As much as I tend to argue the often underestimated potential and merits of hand work, I actually do appreciate tools like the Plek machine as a wonderful advancement, and agree with much of what you say. 

 

There are many qualities that make the Plek machine a fantastic tool. First of course is the ability to scan under actual tension, compensate for changes in position once strings are removed, and machine the frets accordingly. There is indeed the occasional neck that frustrates me, maybe a thin jazz bass or banjo neck, or one that develops twists or humps not reproducible without string and truss rod tensions. I find it rare, but once in a while I do curse having to go back for touchup work that relies much on memory and approximation. Still, it does seem a rather uncommon phenomenon in my experience, and certainly not the norm. 

 

Beyond that, it saves time when doing fret work in high volume, never gets tired, never comes in to work with a hangover, never asks for a raise, or decides to go off and start its own shop leaving you to have to train another employee on a long learning curve. 

 

On the other hand I do agree very much with Hesh in terms of economics of the tool, and that you have to have a business plan specifically planned around this investment for it to be worth while. I was actually looking in to getting one for a time, and still wouldn't rule out the option at some point in the future. Ultimately though, I just decided that for the time being it just did not make sense as a wise or necessary investment for my repair business. It probably could if I wanted to focus more on that particular aspect of service, but for the market I'm currently focused on it just didn't make fiscal sense. 

 

I will admit that it was the development of the Plek that really pushed me to try and refine my methods. Even so, I still offer a basic fret dress, similar to what is written above but simplified with less time on repetitive checks, and a bit more tolerance for minor deviations. Even with a price increase of only around 50% for premium precision leveling, most clients still tend to choose the basic dress. That said, I automatically do the premium dress on full fret jobs, as I want my reputation in that market to remain as high as possible. 

 

So I may get one eventually if business scale and plans allow for it, but for the time being I find it more suitable simply to do the work by hand, and when I come across a neck that seems to be acting particularly squirrely, I may simply refer it out to a shop with a Plek if I feel they can achieve results I cannot. 

 

On a side note, though I never completed my machinist apprenticeship I did spend a few months lapping plates before I left the trade, and still enjoy playing around a bit with rebuilding old machines (have a well worn Hardinge horizontal waiting for my next project). I would never claim to have mastered the art, but it gave me a good start on understanding it.

 

I also must admit that economics aside, a part of me does enjoy the more traditional face of skilled trades. When working alone in the shop it's kind of fun to imagine yourself as somewhat like a character in one of Guy Lautard's stories (if you're familiar with his gunsmith prose). A gun maker may be able to bore a rifle barrel to incredible accuracy with great efficiency (and profitability) with modern tooling, but a gunsmith probably receives a much greater sense of contentment upon honing a barrel to just as good of accuracy by careful hands on methods, even though it may take a bit more time. Call me a romantic...

 

I would only differ slightly in what you have said, in that I think discussions like this are not only good to dispel myths or undeserved disdain of technology, but also the inverse of dispelling myths that equal results can not be achieved by more traditional means.

 

Thanks for the great discussion here. This certainly seems to be one of the more sophisticated luthier salons. ;)

 

 

Randy, I buy the lapping abrasive in rolls (minimum four rolls at a time smallest quantity I could find), and it does not come with an adhesive backing. For the adhesive I use 3M 9447 transfer tape. I used to use 3M 467, but find the 9447 to be a bit more sturdy, less like to lift when pulling off the backing. Plus it's 1mil thick as opposed to the 467 being 2mil. Applying it to the abrasive can be tricky, and I ended up getting some rollers pulled from an old laminator to lay it up. A bit of water between the layers helps prevent air pocket which can't be squeezed out, and it bonds very well to the aluminum bars. 

 

I don't know how necessary the precision lapping film really is, as a good wet/dry paper is likely to be pretty darn close, and any rogue grits are likely to be knocked down in the first pass or two. Still, now that I have a box of the stuff I might as well keep using them. They do have to be adhered well to the bars though. If there are any loose areas this can lead to false indicators, hitting low areas which are not truly level. 

[Ned Knepp]

I haven't tried or even see more than a video of a Plek but I do know about computers. One of the first things I learned is that a computer is only as good as the people who designed it, the people who build it and the parts used. I also learned that a good computer is only as useful as the code that is being run on it and  the person who is using that code. Lowering the "quality" of any of these and the system isn't as "good" as it could be.

 

Somewhere, I read some remarks about the definition of "quality" with the author commenting that "quality" in a high production environment is much different then the definition of "quality" in a low production environment. In the first, quality must incorporate larger margins in tolerances. Gibson, in the 70's equated "quality" with lower warranty returns and failed to maintain other "qualities" in pursuit of that goal.  In a small shop, " quality" is usually defined as producing the best playing, best sounding, best looking instrument possible. The point is that the definition of "quality" depends on who is defining it.  

 

CNC processes are helping large manufactures bring their definition of "quality" closer to the small shop definition because it helps them do "quality" work at the faster pace their volume demands. The years of repetition a person needs to invest to do that work are almost eliminated and the skill set a "builder" needs shifts to a new arena as a CNC operator. That doesn't necessarily mean that the operator doesn't need luthier skills, only that they don't necessarily need to prefect them to the same "tolerance" to be nearly as effective.  Again, the better the operator understands the "skill" being duplicated by the CNC machine, the better work they can get from the machine ( assuming they have the CNC skills as well). 

 

The point is that small shops don't necessarily have to adopt CNC processes to do the same level of work but they will need to invest a lot more time in learning to do it as well or better. Small shops can invest more time in getting it right that a large manufacture can not afford. Soon or later that time need to do the best job gets shorter as the person gets better at doing it.

The great thing about technology is that it always gets cheaper. Hopefully, machines like a Plek will be refined and drop in price to the point where almost any small shop can have access to the technology if they want it. As it is, there seems to be some concern about the expertise of the Plek operators,  When I think about it, who better to get the best out of such a device than the people who are best at doing it by hand? 

 

In the future it may be possible to fab metal parts on a "printer" in your shops. If and when that happens, a CNC like Plek will seem mundane.

 

 

[Bob Steinarsson]

Today, I could for the first time analyze the frets of a new guitar that I believe is pleked (a 2011 Martin D28). I was surprised to hear clicks from my little Stewmac straightedge due to several high or partially high frets along the neck. I'm not talking about very uneven frets, but personally, I don't like to give back a guitar with frets "clicking", at least not from those on the neck end of the 12th fret.

[FRANK]

yep It also could be the operator , Martin has a quota ,I am wondering if $440 is worth the Guy from San Fran to Plek a guitar  I hear he is one of the Best.