Getting into it:

Click on any image for a larger view.

"Hard Numbers"

Allright, now I've taken some basic measurements of the HF8x12 lathe's out-of-the-crate setup and accuracy. I used 4 different dial indicators that I use in my regular work: a Starrett Last Word 0.0001 resolution, a Brown & Sharpe BesTest 0.0005 resolution, a Starrett Last Word 0.001 resolution, and a Grizzly large range 0.001 resolution Chinese DTI. I started with the coarsest resolution DTI for each measurement and worked my way up in resolution as it became evident that it was warranted. I used a combination of "Rolie's Dad's Method" and the broader lathe measurement regimen authored by J. Latta in Model Engineer July 21, 1955.



Before taking measurements I cleaned the lathe thoroughly with WD40 as a solvent, mostly for the convenience of the spray can; I had a gallon of kerosene on hand but chose to try the WD40 first. It worked very well and it was a simple job to clean off the "clear" grease, using less than 1/2 can of "odorless" WD40 and perhaps 1/3 roll of paper towels. I dissassembled a minimum of parts for the initial cleaning because I didn't want to disturb any factory setup before measuring it. What I did do was slide the tailstock off the bed and backoff both the cross-slide and compound feed screws so that I could inspect them for lubrication and apply it if necessary. Everything else was left in the as-received condition. I lubricated the two screws with white lithium grease and everything else including the ways with Mobil 1 10W-30 synthetic motor oil. The carriage, cross-slide, and compound rest assembies are liberally sprinkled with 10 brass oil ports so it is a reassuring and simple exercise to oil up all the moving parts of these assemblies, including the carriage longitudinal travel shaft, bearings, and gears. I chose, by personal preference, to use white lithium grease on the two screws as mentioned above, so did not use the oil ports for those.

First I took some basic overall dimensions off the lathe:

Overall Length: 34-1/2" (from belt/gear cover outside to end of bed)

Bed Length from Headstock to End of Bed: 24-3/4"

Distance Between Centers: 14-1/8" (Yes Virginia, this HF 8x12 is actually an 8x14 lathe)

Width of Bedways: 5-1/4" (Broader and more stable than the 9x)

Distance from Face of Chuck to Tailstock Center: 13"

Maximum Depth at Motor Housing: 17-13/16"

Overall Height at Switch Housing: 17-1/4"

Maximum Handwheel Extension at Front of Lathe from C/L of Bed: 11"

Spindle Centerline above Bench Top: 11-1/4"

Bed Height above Bench Top: 7-1/4"

Mounting Dimension, C-C: Tailstock single bolt: 3-1/4" from end of bed; Headstock double bolts (inline): 21-13/16" and 25-7/8" from Tailstock bolt, 5-1/4" then to outside of gear cover.

Weight: 272 lbs. crated; 242 lbs uncrated.

For a principal DIMENSIONS diagram click here.

All that remained at this point were some electrical measurements I wanted to include, mounting the two belts: one vee and one timing, adjusted the idler pulley for the vee belt, and finding the right mounting bits and pieces for the DTI's. Oh, and also a double edged old-fashioned razor blade. The latter was the toughest chore in the whole bunch. Luckily my dear wife hit the drugstores and came home with a pack of 25 genuine Walgreen's SS DE razor blades. (She even knew it was for a machine shop project; I couldn't hide it because I haven't shaved in 40+ years.)

Why electrical measurements? I wanted to know that the 3-prong grounding cord was correctly wired and that I was, indeed, protected by grounding of the lathe assembly, housings, ways, and anything else that could conduct electricity. My shop uses 100% GFI (Ground Fault Interrupter) 120 volt receptacles that are fed from the shop's main 200 amp board in addition to a large surge supressor across both sides of the 120/240 line. Nevertheless, it is imperative that non-double-insulated tools have their grounds correctly wired internally.

The HF8x12 electrics are CSA certified, not UL, not that it really matters in a discrete motor driven machine tool like this. I found that the ground was correctly wired and all conductive parts of the machine were grounded. The motor was also properly grounded. My one complaint about the electrics is that the builder uses a "flying" terminal strip for all connections not made on the front panel switches. By "flying" I mean it is not fastened down to any surface but is supported by the wires leading to it, and is free to move around as wires are moved or the electrical box back cover is fastened/unfastened. This strip is relatively light weight, in my opinion, with very small screws that clamp on tiny pin type terminations of the conductors going into it. Everything is recessed into insulated wells, so the device is not necessarily unsafe, I just would prefer a more substantial connecting means and may change my unit later for my own satisfaction. I pulled on it and twisted it but nothing came loose or seemed to suffer. Actually there is a second item, the "missing switch" that I will get into later on.

For an ELECTRICAL diagram click here.

But I get ahead of myself: before doing any measurements I bolted the lathe down firmly to my bench's 1-3/4" thick wood/metal faced top surface using 1/2" diameter bolts and a "strong-back" on the bottom of the bench top to ensure stiffness. (for more click on BENCH in the menu bar above.) I then leveled the lathe in two directions, cross bed at the spindle end and at the tailstock end, and longitudinally on the front bedway. I used my 12" South Bend precision level and was able to get the lathe level in the above directions within 0.001" by adjusting the leveling legs I welded up for the bench (an hour exercise in bending, wrenching, measuring, loosening, tightening, measuring, cussing, and repeating the whole thing until I was satisfied). I then left the unit for 24 hours to let it "settle in" and repeated the leveling process again. (For more click on LEVELING in the menu bar above)

Confident at this point that any discrepancies in the lathe's accuracy would be the result of the builder's efforts, not my own sloppiness, I proceeded to measure AS-RECEIVED spindle runout, spindle parallelism to the bedways (vertical and horizontal) at two locations, alignment of the spindle and the tailstock, perpendicularity of the cross-slide movement relative to the spindle axis, tailstock ram parallelism to the bedways (vertical and horizontal), backlash, and runout of the included 3-jaw chuck.

The Razor Blade Test

This is an old machinist's stunt. As I am old at least I qualify halfway. Seriously, the razor blade test is a valid and real method for checking alignment of very small diameters like the points of the centers in a lathe. The most difficult part is that you need a very thin and hard material. A piece of flat full hard shim stock would do just as well but the razor blade is an eye catcher; plus its nice and flat (shim stock can have a curl to it). The typical DE razor blade is 0.003-0.004" thick and quite hard while being very light in weight (no pendulum effect). A single edge razor blade is heavier with the edge stiffener and about 0.009-0.010" thick (the blade not the stiffener); a machinist's 6" pocket scale even thicker and unusable.

The thinner/lighter the material the more accurate is the indication of misalignment. The heavier the material the more the weight of the stock can tend to bring it vertical and influence the calculation. It's not easy to get it dead center; it will rotate on the sharp centers and the heavier side will be down and acting as a pendulum, straightening the angle. You want it to assume an angle that is representative of the misalignment; you want it hard so that the points don't dig in and keep it from assuming an angle.

It is purely an exercise in geometry/trigonometry. The second chore is to determine the angle of the shim/razor blade in the horizontal and vertical planes relative to the lathe's spindle axis. I determined that in my 8x the horizontal was essentially 0 degrees and the vertical was less than 3 degrees. You know the thickness of the razor blade at 0.0035" (Walgreen's SE SS) so using a right triangle solution with the angle <3 degrees and hypotenuse 0.0035" known we can solve for the vertical misalignment as sin A x h = a (misalignment). The solution for my 8x is <0.0002" (actually <0.00018" but 5 decimal places is a bit much).

So, as is graphically intuitive, the bigger the angle the bigger the misalignment.

The Numbers:

My findings are as follows:

(All measurements involving Tailstock were repeated with ram fully retracted and ram extended 2-1/2". Measurement values are variance from 0.0000".)

Spindle Runout: <0.00015" ("<" means "less than")

Spindle // to Ways Vertical: 0.0005" ("//" means "parallel") at 10" from front bearing

Spindle // to Ways Horizontal: <0.0001" (I couldn't measure any finer) at 10" from front bearing

Spindle to Tailstock O: <0.0002" ("O" means "Center Alignment" here; there should be a dot in the middle of the "O" but I can't get my computer to do it)

Cross-slide travel _|_ to Spindle: <0.0003" ("_|_" means "perpendicular")

Tailstock Ram // to Ways Vertical: <0.0002" (w/Ram extended: <0.0003")

Tailstock Ram // to Ways Horizontal: <0.001" (w/Ram extended: <0.001")

Relative Change in Ram with Clamp Tightened: None measurable but some rotation, perhaps 2 degrees towards the rear)

Backlash: Backlash was measured with a DTI; I am not a fanatic in chasing down backlash on a machine tool; a careful machinist always approaches his setting from the same direction anyway, so backlash is immaterial except as an indicator of wear in an older tool.

Cross-slide Backlash: <0.005" (As-received; this is adjustable with a split "nut" on the screw and I may play with it later or I may not bother)

Carriage Longitudinal Travel Backlash: <0.036"

Compound Backlash: <0.002"

3-Jaw Chuck Runout: <0.013" at 6" from jaws

Spindle runout is quite good and is no doubt the result of this lathe's use of "Grade D" bearings for the spindle (bearing precision is graded by letter, "A" being the least precision). Parallelism and perpendicularity are the result of a number of surfaces being hand-scraped, a fact I discovered with further disassembly after the measuring phase. For example the carriage, the tailstock base, and the compound slide have all been hand-scraped, and the removable items are serialized with a number matching the lathe s/n. These are all signs of a careful and precision-enhancing manufacturing process.

Speeds and Feeds

The lathe includes a tensioning pulley that is postioned such that it can be used with the drive set up for either belt arrangement. Incidentally, for the uninitiated, this pulley should bear on the back/outside side of the vee-belt, not the inner narrower edge. It is possible to do either with the positioning of the pulley but you want to assign any wear from the pulley to the outside of the belt not the inside. The geometry of the drive arrangement is such that having the pulley bear on the outside of the belt also increases the "wrap" of the belt on the step-pulleys; this is a good thing to also minimize slip.

The 6 speeds possible with the the 8x are: 125, 210, 420, 620, 1000, and 2000 RPM. The lower speed range is valuable for turning tough materials like nickel alloys, high series aluminum alloys, and larger diameters of any material (remember SFM (Surface Feet per Minute) is the criteria for picking a spindle speed.) The higher spindle speeds are useful for smaller diameters, surface finish requirements on materials like aluminum, machining some plastics, buffing, etc.

The 8x lathe offers 24 thread pitches with its included threading gear assortment. The various pitches are enabled by mounting various combinations of 4 gears to the "banjo" that drives the leadscrew. The 24 pitches are divided into 12 Imperial and 12 Metric as follows: 8, 9, 10, 11, 12, 14, 16, 18, 20, 24, 32, and 40 Imp.; 0.4, 0.5, 0.6, 0.7, 0.8, 1, 1.25, 1.5, 1.75, 2, 2,5, and 3 Metric. There are other pitches possible if you want to calculate them mathematically. This standard range is graphically shown on a large chart on the front of the belt/gear cover. Some study will be required to successfully interpret the Chinese graphical method. A little thought the first time around and you will have no trouble later.

Included in the standard gear assortment with the lathe are:
30T,35T,40T,45T,48T,50T,60T,66T,68T,(2)70T,72T,75T,80T,90T, and 100T gears, plus the 40T spindle and intermediate gears (1 each). 11 are shipped in the toolbox, 5 plus a spacer are in place on the banjo when received: 1st step:35T/75T,2nd step:90T/30T, and 3rd step:spacer/100T; these give the 0.005"/rev. feed rate.

Lastly, the 8x offers 2 specific power feed carriage travel speeds for the 12TPI leadscrew (leadscrew RPM translated into carriage speed). They are 0.005 in./spindle rev. and 0.010 in./rev. As received the banjo is set for the slower 0.005 in./rev. speed. Obviously this two speed selection is just a matter of convenience and round numbers. Any threading gear selection will give you a certain leadscrew speed since the whole idea is that the carriage traverses a fixed distance relative to the spindle rotation to give a certain number of threads (Imp. or Metric). This is leadscrew speed by another name and vice versa.

Bottom Line

All in all I am quite satisfied with the inherent accuracy of my HF 8x12 lathe as well as its as-received setup and obvious factory tuning, but I am disappointed in the 3-jaw chuck. If I can rig a suitable dust collection system I may try to grind the jaws in place to improve the runout number; or I may get another chuck to replace this one. Doing the latter will give me more peace of mind, as long as it truly is more accurate; I am not a big fan of grinding on the lathe.

If you are interested in a little history of the HF8x12, its relationship to the Lathemaster 8x14 lathe, and some other commentary please go to Page 3 of this review. My modifications for the 8x are in the 8x Mods page and are applicable to both the Harbor Freight and LatheMaster nameplated 8x14 lathe.

Return to top