Leg vise toe-in

[Gene Pavlovsky]

I was reading about leg vises, and found out several sources mentioning that a leg vise (as well as other vises) should have a small amount of built-in toe-in, so that the top of the vise contacts the workpiece first.

Benchcrafted, who manufactures a commercial parallel guide for leg vises (as well as vise hardware), explains toe-in in their blog post: http://benchcrafted.blogspot.com/2015/03/what-is-toe-in.html
(http://benchcrafted.blogspot.com/2015/03/what-is-toe-in.html)According to them, toe-in is somehow built in the Crisscross (their parallel guide) design.

I also found this workbench build log: http://kesterhouse.com/etct/bench.html
He says "the leg vise screw base needs to be installed so the threaded rod has a very slight angle up as it exits the bench leg".
I don't quite understand it, to me it seems it would provide the opposite of toe-in.

Paul Sellers talks about toe-in too: https://paulsellers.com/2018/08/vise-toe-in-out/
(https://paulsellers.com/2018/08/vise-toe-in-out/)Paul says that if a particular vise doesn't have toe-in, it can be retrofitted by simply planing the vise jaw, removing more material at it's bottom.

Most of the workbench build logs and videos (and Chris Schwarz's book) don't mention anything about toe-in or how to achieve it.

I still can't decide what kind of parallel guide to use for my upcoming workbench/leg vise built.

The most low-tech solution, but seemingly effective and reliable, seems to be using a foot-operated wedge (with a corresponding slope in the bottom of the vise chop). The other tried-and-true option is the parallel guide with rows of holes and a pin. I've also seen a ratcheting foot-operated guide, and read about (but can't find at the moment) a second screw in the bottom, with a foot-operated "handle". These options provide either infinitely variable, or "stepped" adjustment which would allow toe-in.

However, the convenience of "automatic" parallel guide which doesn't require adjustment, appears very attractive. I've also read many people saying that once they have used a bench with a St. Peter's Cross/Croix St. Pierre/Crisscross guide, they can't imagine going back to the pin-style parallel guide. I've found out about a couple of styles of automatic parallel guide besides the cross. One option uses a heavy-duty chain and a couple of sprockets. Derek Cohen wrote an article about this one. The chain doesn't really appeal to me, feels a bit over-engineered and looks too gadget-y to have something like that on a woodworking bench. Another design uses a guide rod in a linear bearing. Multiple people who tried that one, couldn't get it to work well (besides, the parts are rather expensive).

So I'm thinking about either going with one of the "manual" options, or a handmade cross-style guide. Here's a quite nice video on making a cross-style guide: https://www.youtube.com/watch?v=yaDZlUIkpgU
(https://www.youtube.com/watch?v=yaDZlUIkpgU)The question is, how do I make one that is "designed with toe-in"?

Other recommendations/advice?

Gene

[Jim Koepke]

Benchcrafted, who manufactures a commercial parallel guide for leg vises (as well as vise hardware), explains toe-in in their blog post: http://benchcrafted.blogspot.com/2015/03/what-is-toe-in.html
According to them, toe-in is somehow built in the Crisscross (their parallel guide) design.

It may be the vise screw is putting more pressure on the top of the crisscross and the tow in is caused by the natural tendency of vises to rack.

It could also be the pivot point is not at exact center.

jtk

[ken hatch]

Gene,

Yes slight toe-in is needed but my question is why the overthinking. The Benchcrafted system works very well, the cost is reasonable and it works better, not only speed wise but better holding power than any other vise I've used, and with no monkey motion. In addition it is a very easy install, much easier than a parallel guide or I expect any of the other parallel guide/pin workarounds. I recently posted a short video of the Benchcrafted system in use.

In case you missed it here it is again. The bench is not finished and is in the back garden while I fit the vise and slab to the base units. The vise hardware is Benchcrafted "Classic" screw with the 14" Crisscross:


https://www.facebook.com/kenneth.hatch.52/videos/2664438680305565/?t=0

BTW, of the things about a workbench that will/can drive you to barking at the moon crazy the vise is usually number one and vises are much easier to install during the build rather than as a retrofit. So much so I've built new benches just because it was such a PITA to change the vise.

ken

[Joshua Lucas]

The Benchcrafted crisscross definitely wins in terms of style and utility. Unfortunately my bench design doesn't have room for even the 14" version between the wooden screw and the through mortise for the long stretcher. I was planning to use Will Myers' ratcheting design (https://eclecticmechanicals.com/2018/10/28/ratcheting-parallel-guide/) but ended up ordering Jim Ritter's chain guide. My build isn't done yet so I can't speak to its utility though. At first I too was hesitant to use the chain because of its looks, but I realized recently that most of the mechanism isn't easily visible. Most of the time all you can see is the "button" that connects the chain to the chop. On Derek's build even that isn't visible because it's under the garter. The actual chain mechanism is only visible from the end of the bench.

[Prashun Patel]

I used a leg vise for a few years with a manual 'wedge' (mine is actually a bunch of 1/4" mdf slices bolted with a wing nut; use as many as I need (1/4" increments is nice because it's easy to calculate).

I recently greatly increased its holding power by lining the vise jaw (which gives you some grip).

I never angled my screw, but would if I were installing fresh. I would also put it as HIGH as possible. The leg vise holds best near the screw, and it means less stooping over. I'd rather take holding power at the sacrifice of the perceived loss of holding depth, which has turned out to never be an issue in my use.

Also, if you are a handle guy (vs wheel) make it long. The extra leverage is nice.

The parallel guides are IMHO over-rated - or should I say, the manual guides are underrated. They are not hard to use. Further, I have found that I don't need to readjust it all that often in practice.

I guess because I don't own one, it's puzzled me how over engineered those parallel guides have been over the years.

[Gene Pavlovsky]

I'm sure Benchcrafted make good products (and they do look well-made), but "cost is reasonable" is up for debate (especially since I'm in EU and it's an imported product). 129 EUR (that's 142 USD) for basically two metal plates pinned in the middle (+some nuts and bolts and liner plates for the bottom of the mortise)... A hand-made version could be done for 10% of the price (or free, if suitable second-hand material is salvaged), which would also be a fun little project. That's why I would like to understand better, how such a toe-in could be figured in the design.
As suggested, I was also thinking that if the two bars are joined some distance off-center, and then installed with their longer "legs" closer to the ground, then this would naturally push the bottom of the chop outwards a bit. However, when I picture this in my head, the amount of toe-in would then depend on how wide the vise is opened - at the widest opening of the vise the toe-in would be more exaggerated. Another option I'm thinking: if the two bars are joined in the middle, but the ends (or one end) are sticking outwards a bit on the bottom, that would accomplish the same effect, and the amount of toe-in would be the same regardless of the vise opening.

Prashun, a parallel guide is a proven solution and quite simple to make, I'm considering it as one of main options. I don't understand how lining the vise jaw gives toe-in? Well, the workpiece will be contacted by the lining earlier, but more pressure would be applied to the bottom side of the workpiece than it's top - opposite of toe-in?

[Prashun Patel]

You are right Gene; I was wrong about the liner; it increases grip but is the opposite of toeing in. I edited my post.

[Joshua Lucas]

The Benchcrafted install instructions (https://www.benchcrafted.com/download-files/CrisscrossGlide_Instructions-BC0419.pdf) (PDF warning) cover how to adjust the toe-in of the vise on pages 10-11. Basically it's as simple as shimming the bearing plates that the bottom ends of the crisscross ride on.

I've played around in CAD with homemade crisscross designs and found that this method is indeed the best for adjusting the toe-in. You want the arms of the crisscross to be the same length and the center pivot to be centered on both arms. At first I thought that making the lower ends of the crisscross longer (i.e. having the pivot above the center of the arms) would be the best way to produce toe-in. It does produce toe-in, but the problem is that the amount varies (sometimes dramatically) as you move the chop in and out. So the shim method works much better.

If you want I can send you the Fusion360 CAD file I was using.

[ken hatch]

The Benchcrafted install instructions (https://www.benchcrafted.com/download-files/CrisscrossGlide_Instructions-BC0419.pdf) (PDF warning) cover how to adjust the toe-in of the vise on pages 10-11. Basically it's as simple as shimming the bearing plates that the bottom ends of the crisscross ride on.

I've played around in CAD with homemade crisscross designs and found that this method is indeed the best for adjusting the toe-in. You want the arms of the crisscross to be the same length and the center pivot to be centered on both arms. At first I thought that making the lower ends of the crisscross longer (i.e. having the pivot above the center of the arms) would be the best way to produce toe-in. It does produce toe-in, but the problem is that the amount varies (sometimes dramatically) as you move the chop in and out. So the shim method works much better.

If you want I can send you the Fusion360 CAD file I was using.

Joshua,

Good post and good info. One of the things I've learned the hard way over the years is in the end it is better and cheaper to buy vs. make. I can not count the times I've said "self, I can make that cheaper", and end up spending twice as much for something that works half as well.

ken

[Gene Pavlovsky]

Prashun, name's Gene, not Greg :) By the way forgot to say thanks for the advice on the screw location (making it not too low).

Joshua, a CAD file would be useful, do share it please. Just recently I was looking into starting to learn Fusion 360.
Yes I had read the toe-in section in the Crisscross instructions. I wonder if those bearing plates cover the whole length of the mortises, or only the parts where the bottom parts of the cross slide? I'm guessing it must be the latter, otherwise shimming the bearing plate in the chop wouldn't have any effect (besides when the vise is open for very thin stock). It's not clear from the drawings in the instructions, though. I was hoping someone who has a Crisscross installed would chime in with some actual measurements and answers. I also wonder if location of the pins for the top is making the top touch the floor of the mortise (while the bottom is offset by the thickness of the bearing plates)?

Ken, you have a good point, nevertheless I'm really curious about this topic and would like to find out the answer :)

[Prashun Patel]

Sorry Gene! Boy I am really putting my "TOE IN" my mouth today.

[mike stenson]

Ken, you have a good point, nevertheless I'm really curious about this topic and would like to find out the answer :)

There can only be one answer then, go make it! Include the price is costs for you to perform the engineering work and the manufacturing work as well. Your time is worth money (I know mine is :) )

[Joshua Lucas]

This is my first time sharing Fusion360 files, so lets see if this works: https://a360.co/2UoY0oX
(https://a360.co/2UoY0oX)
It's a parametric model so when you open it in the desktop application you can play around with the various dimensions. (Under the Modify menu, select "Change Parameters". I keep it pinned to the toolbar but I'm not sure if that's the default.) I was playing with the idea of making the crisscross out of wood (hickory), so the design is slightly different from the Benchcrafted version. The specific parameters that adjust toe-in are slide_Offset (equivalent to shimming the bearing plates) and crisscross_centerOffset (making the pivot point of the crisscross off-center, not recommended). crisscross_length does about what you'd expect, and the rest of the parameters mostly just adjust the thickness of the components.

[Gene Pavlovsky]

Prashun, that's ok, you're funny :)

Mike, if it was as simple as "time is money", I would probably just buy a good quality commercial workbench. As long as what you're doing gives you fun, pleasure, sense of doing something useful, or learning/practicing skills, "time is money" goes out the window to some extent. I was just thinking about this a few days ago. I bought a 20-year old bench grinder that looked like new, however it didn't start without giving it some helping spin. I first replaced the start capacitor, which had leaked and was no longer working. Surprisingly, that didn't help. I proceeded to disassemble the motor, and found a break in the joint where one of the wires connects to the starting winding. I've re-soldered the joint, all that is left is to seal the insulation that I've had to cut (waiting for some high-temp silicone sealant to use for that) and put it back together. I've already spent 5 hours or more fixing up that grinder, and I'm quite mindful that it might have been cheaper to bin it and buy a new one. However once it will be back together and working (hopefully), I will have a warm and fuzzy feeling, and a bit more understanding of how an inside of an induction motor looks like. And next time I have something not working, I'll have more confidence to try to fix it.

Joshua, thanks for sharing. I've opened it in the web viewer (I still have to register to download the desktop one) and was able to understand how it works.
Interesting design with the 3 bars instead of 2. I wonder if a wooden one would be strong enough. It might bend enough for that to be a problem.
I see in your design the square sliding blocks at the bottom have offset holes, indeed being the equivalent of shimming the bearing plates. I'm now convinced that's how the Benchcrafted Crisscross achieves the toe-in. Cool! It seems that I should just build one and see how it works. Shimming the bearing plate, if needed, seems like an easy adjustment that wouldn't require reworking the bench leg or the vise chop. What thickness of bearing plates to start with? Probably 1/16"-1/8" or 1.5-3 mm would work? Unless I'm mistaken, two 1/16" plates would give 1/8" of toe-in, and two 1/8" will give 1/4".
By the way, do you have any recommendations on how to start learning Fusion 360? Some books, tutorials, video courses?

[Andrey Kharitonkin]

I did a mock build of leg wise with linear bearings. With parallel guides like that or cross or chain toe-in is indeed needed to grip better along the whole surface of holding piece. The amount of toe-in depends on flexibility of jaw (how thick and what wood) and also flex between jaw and parallel guide or whatever is trying to hold it parallel. And, also how much pressure you want to apply to a holding piece. The more force the more flex the more toe-in is needed. Leather liner increases the grip and requires less pressure. So, it is very hard to guess it upfront, unless you build a prototype using the same wood for jaw. The good thing is that you can adjust it later, in some cases at least. Something like 1-2% of inclination is generally suggested (1-2 cm over 100 cm length, that is).

Important thing that I learned with my prototyping is that the hole where the screw is going through the jaw has to be bigger than just the screw. That would allow some room for error and adjustments. Because it is very hard to make it precise, it has to be fixed later, during assembly. Otherwise, the screw might bind inside the leg nut or more with resistance. It should glide well when not applying pressure. Once you start to tighten it it will go out of perfect alignment but it should be enough to hold the piece before it binds in the thread. So, this two edge cases have to be tested and adjusted. Something like this is also suggested in leg vise hardware kits manuals.

You can also change the angle by planing inside the jaw to taper it to the lower end, as indeed suggested in leg vise hardware kits manuals, works even for other types of vises. Provided you can detach the jaw from hardware, means no permanent things like epoxy until you get it right.

With linear bearings it is a little more difficult because also the guiding rod has to be attached at the same angle from perpendicular. The more pressure the bigger inclination between rod and jaw has to be. And once done you have to live with that or through all of that again...

[Gene Pavlovsky]

Good points, Andrey. I will keep this in mind with regards to screw installation.

As for the linear bearing, I wonder if anyone had done any mechanical analysis of this solution. It seems to me that someone just had an idea to try it (because it kinda made sense to them), posted about it on the internet (and mentioned that it worked well for them), and then naturally a bunch of people repeated that (with many saying it didn't work well for them). Considering that the leg vise can develop a lot of force, any bending of the shaft and play in the linear bearing will result in the chop's bottom moving inside more than desired. So to me it seems that to counteract that (and to provide some amount of toe-in), the linear bearing and it's shaft shouldn't be installed exactly parallel, the shaft's far end should be slightly angled towards the floor. But which angle to use is a question, and once the angle is chosen, it would be difficult to adjust it later.

I've found about another version of the cross guide - X Link by Hovarter. It mentions more clearly how toe-in is achieved "This is easily accomplished with the included shims which are added behind the wear plate in the vise jaw.", supporting what was mentioned earlier by Joshua. Their product looks even more fancy than the one by Benchcrafted, incorporating a spherical bearing to allow for less-than perfect precision when drilling the pivot holes.

I consider my original question (how does Crisscross achieve toe-in) is resolved, but welcome any further discussion on the topic of how toe-in can be planned for, when installing some other kind of automatic parallel guide.

I am currently quite attracted to a ratchet system parallel guide. I found examples by Will Myers and Mike Davis.
Article by Will Myers: https://eclecticmechanicals.com/2018/10/28/ratcheting-parallel-guide/
(https://eclecticmechanicals.com/2018/10/28/ratcheting-parallel-guide/)Video of it in operation: https://www.youtube.com/watch?v=0cji8ymNjJQ

[Joshua Lucas]

Gene, I learned parametric CAD in school so I'm not the best person to ask for advice on how to learn it, but I think I've heard that Autodesk has a good series of YouTube videos on getting started with Fusion360. Fusion has a lot of features, most of which you'll never use for woodworking design, so try not to get too overwhelmed. If you can draw a sketch and extrude it you're good for 99% of furniture design.

In case you weren't aware, Fusion is free for hobbyists, you just have to sign up for a hobbyist license. Autodesk changes the process it seems like every few months so I can't tell you the exact process, but it wasn't hard when I did it.

I actually did do a simulation in Fusion of the wooden crisscross, but Fusion doesn't support simulation of wood so I tried it with a couple materials with similar mechanical properties, ABS plastic and MDF. ABS is less stiff than most wood, but it is less brittle, while MDF is almost the opposite - more brittle, but similar stiffness. Both seemed to perform acceptably with a clamping force of 500 lbs (IIRC), although the MDF was close to splitting where the dowels go through the arms. I think the ABS probably could have taken quite a bit more before the bending got unacceptable. I think a hard, strong wood like hickory or European hornbeam would work pretty well. Also, if I were really building it I would exclude the sliding blocks. They made simulating in Fusion easier but they actually decrease the overall strength of the mechanism.

I like the ratcheting design too, but I decided to go with the chain more for ease of construction than anything else. If someone made a kit for the ratchet I'd likely go with that. And the crisscross would probably be my first choice if my bench design were compatible with it.

[Gene Pavlovsky]

Ok Joshua I will try their YouTube videos. Yes I've heard about the hobbyist license, will apply for that one.

Interesting about the stress simulation, too bad they don't support wood. Perhaps they will add such support if enough of us woodworkers ask them for it :)

The ratchet option sounds attractive and I'm thinking more and more about it. One con I see (or just imagine?) in the ratchet design, is that if you have to work on multiple workpieces of the same thickness (e.g. plane multiple same-thickness boards on one or more edges), each time you open the vise, the ratchet might move by a tooth or more, in that case when you close the vise again, you have to release the ratchet pawl again to set it to the position (hopefully the same position). The traditional pinned parallel guide, on the other hand, provides repeatable clamping. The adjustment step (distance between adjacent settings) of the ratcheting guides that I've seen also seems to be a bit on the large side, compared to what the pinned guide offers (those are often made with 3 rows of holes). This may or may not be a problem.
Another issue is the lack of available kits, or standard parts that can be used to make it. Sawing the sawtooth rack by hand / using an angle grinder, plus hand filing, seems like a lot of work. Of course sometimes it's fun to create something that works, using only basic materials, but there is also a lot of fun when you can adapt existing hardware to implement an idea. What is easily available (and affordable), is rack-and-pinion hardware for sliding gates. I'm seeing a couple of options here.
Both options include a gear rack that is installed on the parallel guide (instead of the linear ratchet rack). A matching gear is attached to the bench's leg to mesh with the rack. A ratchet and pawl set (commercially available) is obtained, an axle joins the gear and ratchet together. A suitable pawl release lever is rigged.
The second option is simpler and requires less precision. Only the gear rack is needed. A small piece of it is cut off, and is attached upside down to the bottom of a "brake block" which is installed on the back of the leg (or in a mortise on the front of the leg, same as the pawl in Will Myers' ratcheting solution). The brake block can be lifted by a release lever (foot- and/or hand- operated) to adjust the vise opening, or locked in the down position, which locks the gear rack. This seems like something that can be made using standard parts, and it would also be not too hard to start offering kits based on this idea. What do you think?
But yeah, a simple pin in the hole sounds much simpler :)

[Andrey Kharitonkin]

Good points, Andrey. I will keep this in mind with regards to screw installation.

As for the linear bearing, I wonder if anyone had done any mechanical analysis of this solution. It seems to me that someone just had an idea to try it (because it kinda made sense to them), posted about it on the internet (and mentioned that it worked well for them), and then naturally a bunch of people repeated that (with many saying it didn't work well for them). Considering that the leg vise can develop a lot of force, any bending of the shaft and play in the linear bearing will result in the chop's bottom moving inside more than desired. So to me it seems that to counteract that (and to provide some amount of toe-in), the linear bearing and it's shaft shouldn't be installed exactly parallel, the shaft's far end should be slightly angled towards the floor. But which angle to use is a question, and once the angle is chosen, it would be difficult to adjust it later.


Yes, that's why prototyping is very important. In fact, it is very useful for any vise installation. I did prototype from pine also for Veritas Twin-Screw Vise and I'm very glad I did.

And yes, shaft is installed at an angle, to compensate for flex of the shaft and the jaw at the desired holding force. Toe-in is required only for that anyway. It also depends on how far the jaw is open as it flex more easily then. Cross guide is better in this regard as it doesn't flex and does not depend on how much you open it, and you can apply any force. I want to try linear bearings guide because I liked the idea and because I could completely detach it from the workbench and then use workbench in a English workbench way, by clamping to its side. There are no sticking out parts when it is unscrewed completely.

[Gene Pavlovsky]

Yes, that's why prototyping is very important. In fact, it is very useful for any vise installation. I did prototype from pine also for Veritas Twin-Screw Vise and I'm very glad I did.

...

I want to try linear bearings guide because I liked the idea and because I could completely detach it from the workbench and then use workbench in a English workbench way, by clamping to its side. There are no sticking out parts when it is unscrewed completely.

I think I will heed the advice and do a prototype first (especially if I try to experiment with the gear rack approach, which I haven't encountered attempted/documented so far).
Is there any other solution that does not permit to remove the vise, with nothing left to stick out? Sure, a cross guide needs a bit more effort to detach (removing one pin holding the cross to the bench's leg), and other solutions may also need some extra effort. The ratchet rack (or gear rack) solution seems as easy to remove as a linear bearing.