A Straddle Knurler for the Sherline Lathe

Not open for further replies.


H-M Supporter - Diamond Member
Staff member
H-M Lifetime Diamond Member
Dec 20, 2012
A Straddle Knurler for the
Sherline Lathe

The Sherline lathe is far more capable than most people know. I have explored the limits of this amazing lathe for the last 30 years and while I’m sure of what it can do, I am not so sure of what it can’t do because I’m still building tools for it. With the proper turning tools, it will take 0.120” off the diameter in mild steel or 0.200” in aluminum in a single roughing pass, then it will take off 0.0001” accurately using just the hand wheels. It will part most materials from the rear in sizes as large as the chuck will hold, at several times normal turning speeds and with zero chatter. It cuts class 3 threads easily and will bore with incredible precision with a good operator behind it. It will also knurl as well or better than many lathes that are much larger with the proper tool … this tool.

I’ve shared pics of this knurler on HM before but never fully discussed its features or how it is built so I thought I would do that in case somebody was looking for ideas on how to build a better knurling tool. I hope this helps.

This tool was built to maximize rigidity, both in the tool itself and the way it is mounted. I have never seen a discussion about the need for, or the effects of, rigidity in knurling tools. However, the negative experiences I had with Sherline’s proprietary straddle knurler and a commonly available import straddle knurler suggested to me that rigidity might be really important. At the time I was contemplating building this knurling tool I had about 10 years of experience with these knurlers so I wasn’t exactly a beginner at knurling but for the life of me, I couldn’t get a really good knurl from either of them. They produced functional knurls but they had a smeared, sort of out of focus appearance to them, like this:

This is a knurl in 6061-T6 aluminum, made with an import straddle knurler. It is a full-depth knurl made in a single pass with 30 tpi Circular Pitch diamond knurls. It should be crisp and defined; it is not, at least not to me. The Sherline tool makes knurls that are not much better so I really needed to fix this.

After thinking it over, I determined that what these two tools had in common was movement. The Sherline tool is actually meant to move in use; they recommend you not lock it down to the cross slide. I suspect this was to improve tracking but whatever their reasons are, a moving mount does not work well. The import knurler moved excessively at the hinge joints, which was sort of expected given the overall design and quality of the tool. Both knurlers were also quite loud; they made a roaring sound that I’m pretty sure was due to excessive vibration as the tool is traversed down the work piece.

I decided that I needed a knurling tool that was built to maximize rigidity, both in the tool itself and the way it was mounted. If I was right about the movement and vibration adversely affecting knurl quality then things should improve with rigidity, right? To find out I built my straddle knurler and this is what happened:

first one.JPG
This test knurl in 6061-T6 aluminum is the very first one ever made with my straddle knurler. It was done in a single pass using 30 tpi Circular Pitch knurls without calculating the diameter of the work. Other than cleaning with Dawn soap and drying it with compressed air, this is just as it came off the lathe. This is what I was after – a crisp, clean, consistent pattern with no smearing and minimal flaking, along with clean starts and stops – and this is typical of what my knurler produces. Same operator, same lathe, just different tool.

Screen Shot 04-24-20 at 08.04 PM.JPG
This ornamental “square knurl” was made with my straddle knurler. It was originally described by GH Thomas of the UK and is made with straight Circular Pitch knurls to produce a depth of pattern that leaves a small flat on top. It is then reworked with a tool shaped like a threading tool across the knurl to produce the final pattern. Again, this was done in a single pass without calculating the work diameter. No smearing, mistracking or pattern variation can be tolerated in a design pattern like this so a good tool and good technique are important. The ability to make knurls like this was one of the reasons I needed a good tool.

Design considerations
When we traverse the tool down the work to extend the knurl there is a lot of lateral pressure on the arms of the tool. In fact, anything with a hinge, joint or arms that can move, will move. I suspected that this movement caused vibration that produced variations in the consistency of the knurled pattern to produce that out of focus appearance. So, for my project knurler I didn’t want to just restrict arm movement; I wanted to eliminate it to the extent possible. This became the key building parameter.

I also wondered if the way a knurling tool is mounted might have something to do with tool performance. We have all heard that straddle or scissors knurlers “absorb” all the load and that all the lathe has to do is provide the torque to turn the work. In my view, this is nonsense. The load has to go somewhere, and that somewhere is into the structure of the lathe. I reasoned that if this is so then rear mounting a knurling tool would be the best way to transfer these forces to the lathe because I feel that this is the most rigid way to mount a tool. But I didn’t want to just rear mount it; I wanted to mount it rigidly enough to eliminate the possibility of the tool moving. This was parameter #2.

A secondary reason for rear mounting it was to enhance visual access to the wheels so I could see the pattern developing as I dialed it in. You see, there is more to the depth of the pattern than just appearance. Form-Rol, a maker of very good knurls, says that knurling to greater than 90% depth is not good for a knurling wheel so being able to dial in your depth accurately is important for knurl life. Rear mounting the tool greatly enhances your ability to visually dial in the desired depth of pattern in real time, a definite advantage.

To test these notions, I set out to build a knurler as a test bed that had two key design features:

This knurler design has cheek plates to limit arm movement, which is good all by itself, but I took it further and made it so there is near-zero clearance between the arms and the cheek plates. If you push or pull on the arms laterally there is literally NO perceptible movement or play, yet the arms move freely and fully throughout their 1/8” – 2-1/8” range.​

I also built it to mount to the rear of the Sherline cross slide such that one cheek plate contacts and registers on the edge of the cross slide; once it is locked down, the tool is automatically oriented perpendicular to the work and it cannot move under load. I used screws that pass through both cheek plates to lock the tool to a relatively large aluminum mounting block that is then locked down on the cross slide with two screws into elongated steel T-nuts.​

This arrangement is really solid; the only movement on the tool are the wheels in their slots at the end of the arms. I left 1/32” of space on each side of the wheels and found that this is all that is necessary to allow them to track and synchronize.

As I showed you, this tool works. It has never failed to track that I can recall, and produces crisp, clean knurls in mild steel, medium carbon and stainless steels, brass, aluminum and Delrin. I can easily see and adjust the depth of the knurling pattern, and because there is no loss of tension as the knurl is made the pattern remains consistent all the way down the knurl. This tool is also pretty quiet, with no roaring as my other tools had. It has the same 1/8 – 2-1/8” knurling range as my Eagle Rock K1-44 and Aloris AXA-19 knurlers but I believe my knurler is more rigid than they are despite their size. It disassembles completely for cleaning in under a minute due to its simple design. And it was made mostly from my scrap bin so it cost almost nothing to make.

All tool builds teach but this tool taught me more lessons than any other tool I’ve built to date. I had never made a knurling tool before and had to figure out how to build it as I went along. In the process, mistakes were made and lessons were learned. I’ll share some of these lessons as I show you how I built this tool in case you’re interested in building one, too. And yes, this tool was built entirely with the Sherline lathe and mill.

I would say that this is an intermediate level tool to make – not hard, but you have to work carefully. If you feel the need to use larger knurls, upscale the dimensions of the tool to suit but I think this tool will work as configured on lathes up to and including 10”.

For transparency and to give credit where it is due, I should tell you that this is not my original design. I saw this picture of a knurler on UK-based Chris Heapy’s now defunct blog. He did not give any construction notes or build details but this single picture was enough to figure out how to build it and incorporate the design parameters I had. By the way, this cheek plate design is not Mr. Heapy’s, either. It has been around for decades and is popular in the UK where I think it was developed. Brilliant!

Materials List (dimensions in the text):

  • ¼” thick mild steel plate for the cheek plates
  • ½” square mild steel keystock for the arms
  • ¾” 12L14 hex stock for the tensioning nut on top
  • ¼” OD round 12L14 rod for the tensioning screw
  • O-1 tool steel for the half-moon pivots, ½” OD
  • O-1 tool steel for the arm pivots, 5/16” OD
  • 3/16” drill blanks for the wheel pins
  • Stainless washers and screws
  • Stainless spring
  • Delrin for a washer
  • 6061-T6 aluminum rod, 3/8” OD for spacers
  • 6061-T6 block for the mount
  • 3/16” reamer, #14 pre-pre-reamer drill, #13 pre-reamer drill.
  • 1/4” reamer, C-pre-pre-reamer drill, D-pre-reamer drill.
  • ½” reamer, 12mm pre-pre-reamer drill, 31/64” pre-reamer drill.
If you choose to make this tool you can substitute a piece of steel square stock for the mounting block to allow you to use this tool in a QCTP holder. However, if it is at all possible to rear mount it I strongly encourage you to do so; a solid mount is important.

Knurls: I really like the quality of Form-Rol knurls so I ordered directly from them. I chose to go with their EQ series (Sherline uses the EP series) because these knurls are ¼” wide vs 3/16” wide in the EP series. It is best to use either beveled edges or convex wheels for traversing along the axis of the work so I had Form-Rol bevel the edges of all my knurls. While this leaves only 1/8” of pattern surface to work with it works fine for traversing. I ordered all Circular Pitch knurls because most of my work will be on non-nominal stock:
  • EQS 230 – straight knurl pair, 30 tpi
  • EQR/EQL 225 pair – 25 tpi diamond
  • EQR/EQL 230 pair – 30 tpi diamond
All these knurls are ½” OD, ¼” wide, 3/16” center hole. All knurls are fine pitch in order to minimize the mistracking that is common with coarser knurls. Finer pitch ones will tend to fall into the pattern and track more easily; this is especially true with straight knurls.

Knurl pin detail: These are cut from a Precision Twist Drill 3/16” HSS drill blank which is already somewhat hardened. Instead of a set screw to secure it, I used an idea from Guy Lautard’s first Machinist’s Bedside Reader and cut a groove on one side of the pin that allows insertion of a flat washer with a flat ground into one side of it; this retention washer is held down by a single 6-32 SHCS. This is secure and allows quick tear-down of the knurls for cleaning or changing wheels. The pins and knurls are pulled, cleaned and oiled after every knurling session and this arrangement makes it quick and easy to do.

Since the retention system keeps the pins from rotating, no sleeves were required to house them. The screws have never loosened in use and the pins have shown very little wear in the last 20 years. They have seen at least a hundred knurls so I guess drill blank is not a bad choice for pins, in a hobby shop anyway.

pin1.jpg pin2.jpg
Most experienced hobby guys should be able to make this knurler just from the two pics shown below but I’ll detail how I built it for our newer guys.

assembled2.jpg apart2.jpg

The Build:
PLEASE NOTE: If two or more pieces must be identical, the only way to achieve this is to machine them together using the same process at the same time.
This applies to all parts made for this tool – plates, arms, etc.

I squared a 6” long piece of ½” square MS keystock for the arms first, then cut it in half and milled the ends square as a pair; this assures all dimensions are the same. At finished length, each arm is just under 3” long.

Then I cut and squared the ¼” thick mild steel plate for the side plates. The mill scale was flycut off so final thickness is just under ¼”. The plates are 2-1/2” high and 2” wide. Once they were surfaced, the plates were mostly machined as a pair – squaring, drilling, etc.

I used the reinforced abrasive wheel from my Dremel in my drill press that has a Phase II X-Y table on it. With the blank in the vise, 9/16” long pieces were cut off using the movements of the X-Y table to move the piece into the wheel. Cuts of 0.050” deep per pass were made to keep temps down but lots of diluted soluble oil are still required to cool the steel or it will anneal. I basically dribbled this onto the wheel as I fed the steel into the wheel. Deburr with a diamond card or stone.

Then I cut a 0.040” deep slot into one end of each pin. The slots are just outside the width of the arms. This will allow the use of a #6 stainless washers with a flat filed on one edge to lock the pin in place while in use as previously illustrated.

The main tensioning nut and bolt are two pieces. I made the tensioning nut from ¾” OD 12L14 hex stock, tapped ¼-28. The stem of the nut (under the hex) is long enough to enter the space between the plates without the head of the nut hitting the plates at max travel. My nut is 1-5/16” long from the top of the hex to where it meets the upper half-moon pivot. The length under the hex is ¾” long. None of this is critical. I cross-drilled the hex head for a stainless steel T-bar but have never had to use it. If you build this tool, I suggest making a largish knurled knob instead of a hex; that will be more than sufficient to apply the pressure needed for almost any material we are likely to work with in a hobby shop.

I made the bolt that actuates the arms from ¼” OD 12L14 rod, threaded ¼-20 to go into the lower arm half-moon pivot. The other end is threaded ¼-28 to accept the tensioning nut. Note that all the travel as the top knob is adjusted occurs in this upper threaded section; the bottom just threads into the lower half-moon pivot and doesn’t move. I chose a finer pitch thread up top for more precise adjustability. The length of this bolt is enough to allow the arms to travel all the way in and out to over 2”; mine is 3-1/4” long but this is not critical. Threads are screw cut on the lathe to allow for the best fit (no play).

Machine the arms. As noted, the keystock was milled square before cutting them in half to form a pair. The arms actually have a lot of machine work done to them so before continuing on with construction, I’m going to take the time to give specific tips to help you. Forewarned is forearmed …

The arms are basically mirror images of each other. Every feature must be identically produced on each one. Except as noted, the ONLY way to do this is to machine them together as a pair using the same process at the same time. The arms you see in this tool are the second set; the first set taught me what I just told you.​

A spotting drill is more accurate than a center drill for starting holes. I used a 1/4” cobalt 120° spotting drill to suit my 118° screw machine drills. I used an edge finder and my X-Y hand wheels to position the drill for greater precision. When hole location matters, do not rely on a center punch or scribed marks.​

Reaming steel is not a problem but drilling in preparation for the reamer needs to be done correctly. Best results come from spotting and then going straight to a drill one size smaller than the pre-reamer drill, then use the pre-reamer drill, and then the reamer. This makes certain that the on-size pre-reamer drill only has to cut a clean-up hole so is more likely to be accurate. Avoid step drilling in small steps because small drills can wander and subsequent drills and your reamer will follow that wandering hole; instead, use the sequence above.​

When reaming use a good full-strength cutting fluid (not diluted coolant) and make the pass steady, without dwelling anywhere. Never bottom a reamer … ever. Speed is the slowest your drill press will go; if using the mill, try for about 100-200 rpm. The reason for the slow speed is so you can feel the tool cutting as you feed; high speeds eliminate much of the tactile feel manual guys need. Feed steadily so there is a slight resistance to the feed; doing this allows the reamer to cut continuously but without rushing the tool. For best accuracy and finish, make only one reaming pass and ream only on the way in, then shut off the machine before withdrawing the reamer. Withdrawing a reamer under power will dull the cutting edges prematurely. I had already learned this lesson before but am passing it on here – do not pull your cutting tools back while under power. This applies to reamers, lathe tools, boring bars and other tools meant to cut in one direction. Okay, moving on …​


With the arms vertically stacked in the vise, I pre-drilled (C, then D drills) the arm pivot holes (the holes the arms hinge on at the rear) centered between the top and bottom of the arm and ¼” from the rear of the arm, then reamed to ¼” ID.

The knurl wheel pin holes were spotted 3/16” back from the tip of each arm and vertically centered, then drilled and reamed to 3/16” ID. I also drilled and tapped for the 6-32 retention screws on the sides of the arms, vertically centered and 1/8” proximal to the edge of the pin holes; the screw that goes here will lock the knurl pins.

To cut the slots at the end of the arms where the knurls sit I band sawed the excess material from the slot, then clamped the two arms together and used a new 3/16” end mill to debulk and clean up the slot in both arms at once. I then followed this with a new on-size ¼” end mill to finalize the slot. The slot is ½” deep from the tip of the arm to the bottom of the slot and is obviously ¼” wide.

The wheel-end of the arms were then shaped for aesthetic appeal on a belt sander and then hand sanded to a nice finish.

Next, I drilled and reamed the half-moon pivot hole by aligning and clamping both arms in the vise and drilling and reaming a ½” hole at the seam where they meet so that each arm has half of an identical hole. This hole is located 1/8” off-center toward the tip of the arms to provide more of a mechanical advantage when the knurls are tightened. It works – I can knurl mild steel to full depth using hand pressure alone to turn the top knob.

With the arms clamped together, one on top of the other and both half-moon holes facing down, I drilled two adjacent ¼” holes (edges touching) to debulk the slot through which the tensioning bolt passes. This slot is centered over the half-moon holes. A 3/16” End Mill was used to clean up between the two holes, then a ¼” EM was used to finish the slot. Then working on each arm separately, I machined the front side of this ¼” slot; this has to be cut at a 5 degree angle to allow the arms to open up to max capacity without hitting the bolt. If you look closely at the picture of the upper arm above, you can just make out that the wall of the slot at the rear is vertical, while the front wall is canted at a slight angle towards the tip of the arm; that angle is 5°. I used a 5° angle block to raise the tip of the arm in the vise and machined into the top edge of the slot with a ¼” EM until the EM just meets the edge of the slot at the bottom. Repeat for the other arm, then deburr everything carefully.

Drill the plates for the arm pivots. Square and clamp the plates together and drill #18 holes through both plates. This is the tapping drill for the screws that pass through the arm pivots; you will only tap the inner plate. These holes are centered ¼” from the top and back edges of the plate at their upper (and lower) rear corners. Then separate the plates and drill the outer plate through those holes with a C-drill, then D-drill, then a ¼” reamer; this is the through hole for the arm pivots. Tap the inner plate for a 10-32 SHCS.

I made the arm pivots from O-1 steel but did not bother to harden them. I cut them to 0.2495” OD X 0.727” L with a 0.05” thick shoulder. These are shoulder pins so they determine the distance between the plates. They pass through the outer plate through that reamed hole, then through the arms and butt up against the inner surface of the inner plate. Hence, the length must be accurate. Since your arm thickness will vary from mine you will need to tailor the pin length. The goal is to cut the pins so that the distance between the plates is the same as the thickness of the arms. Drill a close fit clearance hole through each pin for a 10-32 socket head cap screw and allow for that 0.05” thick shoulder for the head of the screw to bear on before parting them off. Deburr the pins.

I made a 0.050” thick Delrin washer to go under the head of the tensioning knob – this allows for a very smooth action that helps tune in the pattern and it doesn’t back out until I want it to.

I made the half-moon arm pivots from ½”OD O-1 steel round stock, not hardened. I cut two pieces and faced each to 0.485” long, the width of my arms. Clamp them in the vise and cut the tops off of both with an end mill until you have a remnant ~ 5/16” thick to create two half-moon shaped pivots. O-1 work hardens easily. I dulled a good Niagara Cutter HSS end mill cutting these pivots down to size so you may wish to use a carbide end mill.

O-1 also work hardens readily when drilled so use a lot of lube and use a speed that allows you to feed fast enough to keep the drill cutting continuously. I drilled the top pivot with an “F” drill for clearance and the bottom pivot was drilled with a #1 drill, two steps larger than recommended by the tap charts, and tapped with a ¼-20 tap. I went with the larger drill because I’ve learned that using the recommended tap drill found in the typical tap charts can snap a small tap in O-1 steel because the material is often work hardened from drilling. My tap cut just fine and there have been no issues with the thread to date. Besides, I cut the threaded rod to a very close fit so it works fine.

At this point the knurler is essentially done.

I marked the vertical centerline of the inner side plate that contacts the mounting base; this will later align with the spindle centerline marked on the mounting base as below. Then I clamped the two plates together and drilled two #11 clearance holes for the 10-32 mounting screws ¼” below the centerline mark (screw centers were 3/8” and 1-1/4” from the rear edge of the plates).These holes pass the screws that mount the knurler to the base.

I squared the 6061 aluminum block and brought it to its final shape and dimensions (non-critical). I machined steps on both ends for the lockdown screws and drilled a clearance hole for those 10-32 lockdown screws on each end such that they align with the T-slots in the cross slide when the knurler is attached.

A line was marked 0.9415” from the bottom of the base (the spindle centerline height of my specific lathe) and the centerline mark on the side plate (from bullet 1 above) was aligned with this mark. Clamp the plate to the mounting block and use a transfer punch to transfer the mounting hole locations through the plate and into the block. Holes were drilled on these marks with a #17 drill and tapped with a 10-32 roll tap. Roll taps form the threads; they do not cut. The resulting threads are stronger and more precise so I recommend you use roll taps in aluminum (and brass) when possible.

In a trial fit, I realized that if I want to cut to a 2” OD I needed to scallop the front of the knurler side plates with a radius of about 1-1/16”; this is a precision radius marked by scribing around the base of a pill bottle. I debulked the scallop with a hacksaw, clamped the plates together and used a Dremel sanding drum and file to finish. This produces that concave radius you see at the front of the plates.

Two aluminum spacers were needed between the plates so the arms would not be squeezed and trapped as the screws that mount the tool to the mounting block were tightened. I made these 0.001” longer than the arm width to allow for compression of the sleeves.

Once the tool was assembled, the arms were really tight inside the tool. To get them to move freely but with the absolute minimum of side clearance, I drawfiled both plates. DO NOT file the arms! This is not as tedious as you might think because everything is machined pretty accurately. I just carefully drawfiled the inside of each plate just about equally with a double-cut bastard file until the arms moved freely throughout their range. Try to minimize filing the area where the aluminum spacers and arm pivots sit. This took all of about 10 minutes to do … file, reassemble, check the fit, disassemble, file and so on until the arms move freely throughout their range but there is no lateral movement at all. When I was done, the thin layer of oil I use to prevent rust is enough to dampen arm movement so it is silky smooth, with no tight or loose spots.

I should mention that the reason for drawfiling the plates instead of the arms is that the plates, while machined flat, do not remain flat when screws are tightened in various places. It is easier and more accurate to remove any high spots on the plates; filing the arms will result in excessive play in areas of the plates that are low.

Drawfiling, and hand filing in general, is a skill that hobby guys would do well to learn. It allows you to create fits that are otherwise very difficult to achieve. Quite often, a stroke or two will remove 0.0001” of material in the exact spot you need to get the fit you want.


End Notes:

I realize that a blow-by-blow narrative like this (instead of a drawing) seems cumbersome but it is a complicated build and I thought a narrative would enhance the chances for a successful build by a new guy. Also, I suck at CAD; someday I need to fix that, too.

I really like this little knurler. It has exceeded all my expectations in terms of performance and has also proven, to me at least, that rigidity in the tool and the mount has a significant impact on knurl quality.

I should make it clear that a rigid mount may not suit all applications. Some knurls are made by using straight or angled knurl wheels and turning to tool to achieve a specific angled pattern. My tool will work for this if I modify the mount but I prefer to mount it as described and use wheels with the angles I need already cut into the wheel. This has worked for me for 20 years and I don’t see a need to change because I’m my only customer.

I will be happy to discuss any questions or criticisms you might have.

I also thought I would share some tips on knurling for our new guys. As always, this is my way of knurling, not THE way. I will include this here and invite others to share their thoughts and wisdom with us as well so feel free to add on to this thread. Yes, I’m thinking of you, @darkzero!

Knurling Ain’t Rocket Science
The knurling process most of us use is more accurately called Form Knurling. It is a forming process that uses pressure to displace the work material into a reverse pattern to that found on our wheels. There are other kinds of knurling processes (cut knurling, bump knurling, etc) but as they are not commonly found in a hobby shop I’m not going to go into that here. For users of small lathes, I highly recommend you stick with a straddle or scissors knurling tool.

There are basically two kinds of form knurling wheels – Diametral Pitch (DP) and Circular Pitch (CP).
DP knurls are classified by the number of teeth per inch of diameter and are designed to produce accurate tracking on commonly used fractional sized round rod. There are only four pitches available: 64, 96, 128 and 160 and they are suitable for blanks from 3/32” to 1”. DP knurls are commonly used in production shops and if you do a lot of knurling on nominal stock, buy DP knurls.​
Circular Pitch knurls are classified by teeth per inch of the circumference of the knurl. This is an older style of knurl and will work with just about any round stock. Since most of my knurling is done on whatever work piece I happen to need a knurl on, the diameter may or may not fall on a 32nd or 64th increment. With CP knurls, the precise diameter is not critical and, in my opinion, CP knurls are the way to go in a hobby shop because most of our work will be made on non-nominal diameters.​

Regardless of the pitch type you use, if you plan to traverse down the work then it is best to use wheels with beveled edges or use convex knurls. Most knurl makers will grind bevels on your wheels on request. Accu-Track is a well-known maker of convex knurls but as I have no personal experience with them I will leave it to others to comment on them. Wheels with straight edges will work but not nearly as well.

If you’re the math-loving calculating type then there are calculators, formulas and pitch factors that you can find on line that tell you what the theoretical work piece diameter should be to allow your wheels to track. I’ve been down this road before and it isn’t that much fun … calculate, turn the diameter, mistrack, turn down 0.005” more, mistrack and so on. This doesn’t work for me. I just want to turn the work piece to whatever diameter I need and knurl it so I can move on.

It turns out that if you use CP knurls and enough pressure you do not need to calculate the work piece diameter at all. You simply crank up the pressure until your wheels track, then adjust that pressure to give you the depth of pattern you need and go. Yeah, it’s that simple. No, it’s not rocket science. And the same pressure recommendation applies to DP knurls, too. Take 0.005” of skin off, increase pressure until it tracks and go.

Of course, there is a bit more to knurling than just cranking a knob. I’ll run through a typical knurling session to give you an idea of what I mean. I am assuming you’re using CP knurls. Before knurling, I suggest you take the time to cover up whatever you can because you’re about to make a big mess. It also isn’t a bad idea to use a splatter shield between you and the work.
Always use a live center if the work piece is more than about 3-4 diameters out from the chuck; the tool puts a lot of stress on the work piece and it will move if not adequately supported. Turn the work piece to whatever diameter you need and put a 30 - 45° chamfer on the tailstock end of the work. This acts as an on-ramp for your wheels and only needs to be about 1/16” or so wide.
  1. Set up the knurler and put some lubricating oil on your pins. Make sure that your wheels can reach the end of your run without running into the chuck. Don’t laugh … check! If your stop point is close to the chuck, set up a carriage stop.
  2. Adjust the center of the knurler as close to center height as you can. Be sure to visually sight the center of the knurl pins so they are over the center of the work; this is important.
  3. With the lathe off, bring the wheels onto the end of the work so that about ½ of the pattern width of the wheel is just past your chamfer, then apply moderate pressure with the tensioning knob. Turn the chuck by hand and watch the pattern that forms as you make one revolution. It will likely mistrack on this first revolution if you don’t have enough pressure dialed in. Mistracking is when the pattern made by the two wheels produces a double pattern on the first revolution. If you go more than one revolution, you’ll have even more than a double pattern. Conventional practice if you see this is to stop and turn the diameter down about 0.005” and try again, and again, and so on. Please don’t do this. Instead, just increase the pressure on your tensioning knob until the wheels do track. In most cases, we are talking about less than a ¼ turn on your knob to get the wheels to track. The mistracking that was there before will be obliterated by the tracking pattern and will not be seen.
  4. Once your wheels are tracking, increase the tension until you get the depth of pattern you require. Try not to go beyond a 90% pattern if you can. A 90% pattern will leave a tiny flat on top of each peak. Since most knurls made to full depth are too sharp for our hands, we file those peaks off anyway so save your knurls and go for less than a full depth pattern. I recommend that you go for your desired depth of pattern in one go if possible. For some tools and some lathes and some materials this may not be possible, especially when knurling harder materials. We may have no choice but to make more than one pass but if your set up can handle it try doing it in one go instead of taking multiple passes to get there. The reason is work hardening and the consequence is flaking. As we knurl, the work surface is displaced into our pattern under tremendous pressure and this work hardens the surface layer of most materials. This hardened top layer fractures into flakes and tiny debris particles that are knocked off by the wheels and vibration, producing what is known as flaking. Repeated passes mash these flakes and debris into our pattern, causing a poor finish. If you must make more than one pass, clean the knurl and wheels as well as you can before the next pass so you aren’t mashing previously made flakes into your knurl. Now imagine if we also have a tool with arms or joints that move. In addition to flaking we have wheels that act almost like a vibrating saw, shearing off even more debris that is then mushed into the pattern. This combination of shearing and flaking is what I think produces that out of focus thing I mentioned before. Our goal is to minimize this and one way to do that is to try to do it all in a single pass. If you must make more than one pass, make the first one as deep as conditions allow.
  5. Once your wheels are tracking and you’ve set the depth of pattern you need, lock down the jam nut on your tensioning knob if you have one and back the tool off the work. Apply some lube to the work; this reduces friction and will also lube your pins. Knurling creates a LOT of heat – do not forgo the lubricant! Knurling is a high pressure operation so I prefer to use a sulfur-bearing lubricant with Extreme Pressure (EP) additives in it for all my knurling other than plastics. The brown pipe threading oil from the hardware store works fine. I find that this stuff reduces flaking even with aluminum and brass; try it and see if it works for you.
  6. Set your speed – Knurling Delrin causes a lot of heat to build up and this will literally melt the plastic, causing smeared knurls that are just plain ugly if you go too fast. So, when knurling Delrin I go as slow as my lathe will reliably turn and I make 2-3 passes at a fast feed rate. This keeps temps down and melting to a minimum. I typically run at about 200 rpm for everything else, including aluminum and brass. This is about feed rates and avoiding heat buildup that can seize the pins in the wheels more than anything else. I have seen recommendations to knurl at turning speeds; please don’t do that. You need to feed at an adequate rate to keep up with the knurl and too much speed makes keeping up with your feed difficult. In addition, the more speed you use, the more the wheels repeatedly roll over the work. This increases the amount of work hardening and flaking you produce so try to keep your speeds reasonable so you can keep the feed rate up to move those wheels along.
  7. Turn the lathe on and move the wheels onto the work at a steady rate. Don’t worry, the wheels will climb the on-ramp and track. I prefer manual feed over power feed so I don’t rush the ability of the wheels to produce the desired pattern. You are looking at the pattern just behind the trailing edge of the wheel to be sure that little flat you created when you set your depth of pattern is forming consistently; this feed rate will be faster than you might think. You should feel a slight resistance to feed when you have it right. Try to maintain a steady feed rate so the pattern forms properly; it isn’t hard to do. Of course, if you have variable power feed then you can find the exact speed you need and it should go well. Just be sure to find this speed on a practice piece first.
  8. Be aware of your stop point. When you reach it, conventional practice is to back the tool out with the lathe still running and the wheels still in contact with the work. There is nothing really wrong with this but what you’re doing is mashing all the flakes and debris you just made into your pattern. A better option is to shut off the lathe and back off the tensioning knob completely and clear your wheels of the work before moving the tool back and off the work. Just doing this alone will improve the quality of your knurls; try it, you’ll see.
  9. If you must make another pass, I suggest you clear the wheels and back off, then clean the wheels and the knurl before making another pass. The goal is to avoid pressing flakes and debris into the knurl so get rid of whatever you can before the second pass. Do not forget to re-oil your knurl pins before the next pass.
You should have an acceptable knurl at this point, with a crisp pattern. If you need to file it to remove any sharp peaks, now is the time. Then use a synthetic or brass brush and knock off most of the debris and blow the rest clean. Put some lightly moistened newsprint or something behind the lathe to catch the mess before using lowish pressure compressed air to clear the knurl. Don’t forget your eye protection before doing this.

Remove the knurling tool and clean up the ways before finishing any work on the work piece.

Hobby guys love precision. If there is a way to calculate something you can bet we’re going to do it, then we’ll try to turn that calculated number to within a tenth because that’s what the formula calls for. If this is you, go for it because you’ll get no argument from me. All I’m saying is that if your calculations fail, and they often/usually will, then do what the manufacturers of our wheels recommend and crank up the pressure!

April 2020

left.jpg right.jpg rear.jpg

Last edited:


H-M Supporter - Diamond Member
Staff member
H-M Lifetime Diamond Member
Dec 20, 2012
I wrote the original version of this article 15 years ago and basically re-wrote it so I could present it here, as promised. As usual, I cannot get the post to look like the original so I'm attaching a pdf of the article as written.

A lot of us have small lathes but that doesn't mean they can't knurl well. In my opinion, the biggest impediment to good knurling are the tools we use. The typical straddle knurlers allow too much movement and I know from experience that this is just not good. If you can, try to modify them to improve their rigidity and I think you'll find this improves the quality of your knurls. Or you can build a better knurler from scratch that meets your criteria of what a knurler should be.

I hope this article provides some food for thought. If you would like to reproduce this tool, please feel free to do so.

If there are questions or criticisms, lay it on me!



  • Making a Straddle2 - Copy.pdf
    1 MB · Views: 152


H-M Supporter - Gold Member
H-M Supporter Gold Member
Jan 11, 2018
Thank you Mike, the information in your article is very helpful to us beginners. I really appreciate your continued sharing.



In Memory
Rest In Peace
Aug 23, 2014
Very well written and executed as always Mikey. I quickly came to the same conclusion as you when I did my first knurls, they looked out of focus due to side movement of the tool. You did something to fix that. Beautiful craftsmanship too. Thank you for the extraordinary help you give the forum.



H-M Supporter - Gold Member
H-M Supporter Gold Member
Jun 12, 2017
Nice! I might have to make one. I need to get some good knurling wheels anyway. The Chinesium that came with the QCTP didn't hold up well to SS303 in the camjack. It worked, but I'm not sure the wheels have much life left. I was a bit... Aggressive with it.. :grin:


H-M Supporter - Diamond Member
Staff member
H-M Lifetime Diamond Member
Dec 20, 2012
Thank you for all the kind comments, guys. I meant to get this one done a while ago but for one reason or another, it took this long.

Knurling is one of those operations that is very much a tool-dependent thing. Good tool, good knurl; junk tool, junk knurl. Like @royesses said, you just gotta do something about it, so I did.

I have to admit that tool building is probably one of the best aspects of this hobby for me. I really enjoy thinking things through and then testing whatever theories I come up with to see if the reality meets the fantasy. When a tool works, as this knurler did, it motivates me to look for the next thing, and then the next. With each tool, I learn and in doing so I become a better hobby guy and that is the real gain - the lessons. As humiliating as they can be, these lessons are indelible and are the real benefit to any tool building project.

I am so grateful that HM exists so I can share them.
Not open for further replies.
It can take up to an hour for ads to appear on the page. See our code implementation guide for more details. If you already have Auto ad code on your pages there's no need to replace it with this code