Singer 108WSV36

[brongle]

So, I impulse bought this machine the other day. Saw it pop up on facebook, the guy happened to be passing through my area the next day, and the price/timing was just too good to pass up - hell, it's the first time I've had a machine dropped off at my shop, really can't beat that. I've been looking for a cheap cylinder arm machine for a while now and learned that those words don't usually go together, so when this one came up for a few hundred bucks, I jumped on it. I'm planning on using this as a binding machine for my bag projects. 

 

It's a 108WSV36, apparently came out of a shoe factory in MA quite a while ago. I've found the limited info that's out there on this model - seems one other forum member had one with this specific subclass, but it's very similar to the 108W20, and that's the manual/parts list I've been referencing so far. I just started getting into it yesterday, and so far, I've identified the two primary functional parts that are missing: the cloth timing belt and the feed regulating spindle.

Found the specs for the timing belt in another thread and was able to verify them against the pulleys in my machine, so I'm planning on 3D printing a replacement in TPU today, if not only to get things moving. I'm hoping a printed replacement can hold up for actual use too, as the cloth belts seem to be increasingly rare. I'll share that file here once I test it. If that doesn't hold up in the way I like, I'll probably end up making a jig to locate and crimp the staples for DIY belts that match the original style.

As for the feed regulating spindle, it seems like I'm going to need to get a little more creative. I've got a screwdriver jammed in there where it'd normally sit, and was able to verify that the feed mechanism works! Surprisingly, despite this machine's condition and the fact that it hasn't run in quite some number of years, I haven't found much that's frozen up with rust or varnish. If anyone here with a 108 class machine would be so kind as to share pictures of their feed regulating spindle with a scale reference in the photo (a ruler is fine), that'd help me out tremendously. Seems like it's a 1/4" OD, based on the hole it goes into, but I'd like some reference for OAL and point length so I can reproduce it. I'd love to see a pic of the inside of the threaded cap as well if possible. Below is the subassembly in question.

 

This particular machine seems to have originally been set up with a stationary binder of some kind,  and there's no way to connect a synchronized binder to the drive slot in the feed dog plate through the cover that's on there. This means I'll also need to either make a new cover plate with those mounting features or modify the existing one. I'm leaning towards a reproduction there in order to preserve the parts it came with. I've got a cheapo generic single fold right angle binder that I'm going to attempt to set up as a synchronized binder. I'd love to see pics of any binding attachments that other folks have for their 108 class machines, particularly showing the mounting and drive features. 

I'll keep updating this thread as I get deeper into the machine, and I'm happy to share any CAD that I come up with in the process if other folks are interested in printing or machining replacement parts for their machines. I'm trying to make as much as I can here, because that's the part of this kind of work that I enjoy the most and it's pretty clear that the fairly uncommon parts I'm missing will rapidly add up to more than the machine itself cost me. 

 

A semi-related question that's been bouncing around my head - this machine, as well as my other two, has a horizontal rotary hook. I like that from a design standpoint, and I'm really curious as to why oscillating hooks seem to be a little more common. I understand that on larger machines that have a lot of needle travel and really big bobbins, it becomes impractical to wrap that much thread all the way around a hook, but I see a ton of smaller machines that still have oscillating horizontal hooks. Is one more reliable than the other? Gentler on the thread? More tolerant of a wider range of thread sizes? Seems like a rotary hook would be much smoother at high speeds and I don't really see any obvious downsides.

[Constabulary]

you still can find some 108w / 108K parts f.i. at college sewing in the UK. But you have to enter the parts number or "search word" for certain parts to find them cause they do not have all the "108 parts" listed in the right machine section.

https://www.college-sewing.co.uk/catalogsearch/result/index/?cat=9&q=108w

https://www.college-sewing.co.uk/catalogsearch/result/index/?cat=9&q=108k

They have the feed dog that can govern the binder attachment plate:

https://www.college-sewing.co.uk/224704c-feed-dog-singer-108w.html

and the swinging attachment plate:

https://www.college-sewing.co.uk/224738-binder-swing-base-singer-108w.html

Certain Singer feed regulating spindles also can still be found but they most likely need a modification in length (at least) and the stitch length number on the knob may not match but you at least can "get something".

https://parts.jacksew.com/feed-reg-spindle-generic-202310/

https://www.ronsew.com/feed-regulating-spindle-224122-for-singer-112w115/

 

Edited Sunday at 03:35 PM by Constabulary

[Constabulary]

I have a 111 feed spindle somewhere so I can compare the length and tip shape with my 108w20 machine.... stay tuned

EDIT:

Top: 111 feed spindle (224122 on spindle shaft)

Center: 107w feed spindle

Bottom: 108w feed spindle (208540 on spindle shaft)

Edited Sunday at 04:11 PM by Constabulary

[Constabulary]

here: 108w on top and 111 below - tip shape is different but I think you can machine a 111 spindle to make it work for the 108w.

The knobs seem to be the same.

So I think this spindle is a good candidate for reworking

https://www.ronsew.com/feed-regulating-spindle-224122-for-singer-112w115/

But in the end decision is yours - I only can give "part hints".

Edited Sunday at 04:21 PM by Constabulary

[Constabulary]

108w20: Total length from inside of the knob to spindle tip is approx 116mm and the machined tip is approx 13.23mm. But I´m not a machinist - the measures are close but probably but not 100% correct.

EDIT:
cheapest new hook I found was on Amazon:

https://www.amazon.com/460243-HSH-8-20-DURKOPP-Binding-Machine/dp/B0F7K6LCRD

Edited Sunday at 04:47 PM by Constabulary

[friquant]

48 minutes ago, brongle said:

A semi-related question that's been bouncing around my head - this machine, as well as my other two, has a horizontal rotary hook. I like that from a design standpoint, and I'm really curious as to why oscillating hooks seem to be a little more common. I understand that on larger machines that have a lot of needle travel and really big bobbins, it becomes impractical to wrap that much thread all the way around a hook, but I see a ton of smaller machines that still have oscillating horizontal hooks. Is one more reliable than the other? Gentler on the thread? More tolerant of a wider range of thread sizes? Seems like a rotary hook would be much smoother at high speeds and I don't really see any obvious downsides.

This topic interests me as well. I've been cataloguing the traits of the hook and bobbin case systems that I have, noting the differences that I see.

Questions to Consider

1. Is hook to needle distance adjustable?
2. Is hook to needle distance reliable?
3. Is there a needle guard?
4. What is required to change to a different needle size and maintain the same hook to needle distance?
5. What is required to change to a different size thread and maintain quiet operation?
6. Is a latch opener required?
7. Is there a range of acceptable hook timing?
8. Effects of changing stitch length and forward/reverse

Hook to Needle Distance

Reliable hook to needle distance is paramount for reliable stitches (for not skipping, that is.) 

The two machines I have with horizontal hook axis, oscillating hooks (Singer 31-15, Singer 237) have the downside that any axial play of the hook in the hook race constitutes variable hook-to-needle distance. In the case of the Singer 31-15, hook to needle distance is NOT adjustable.

The one machine I have with a horizontal hook axis, (double) rotary hook (Singer 96-87) has considerable axial play of the bobbin case in the hook race, but this axial play has no bearing on hook to needle distance, as the hook point is attached solidly to the outer part of the hook assembly. On this machine the hook to needle distance is adjustable.

The two machines I have with vertical hook axis, (double) rotary hook (Singer 144W103, Jianglong 341) include a needle guard which can be set to deflect the needle around the hook, and is also useful for stuffing the hook. In the case of the Jianglong 341 the needle guard has an adjustable screw in it for making precise adjustments. Hook to needle distance on these two machines is adjustable.

Noise

The angular freeplay of the hook for horizontal hook axis, oscillating hooks (Singer 31-15, Singer 237) means there is an additional ticking sound with every change of direction of the hook. Supposedly the amount of freeplay can be adjusted but I have not managed that yet. Also, this amount of freeplay needs to be large enough that the thread can slip by. One needs a large enough gap for the largest thread sizes that they intend to use.

A Range of Acceptable Hook Timing

With a (double) rotary hook, if the hook timing is advanced enough the hook may grab the thread again on its second time around. With any hook, if the hook timing is retarded enough the thread will make a snapping sound as the thread is force-dragged around the bobbin case. Ideally there is a range in between these two extremes where the thread gracefully rounds the bobbin case and the hook does not grab the thread on its second time around. Even better when the factory specification is somewhere in this ideal range.

My machines that have oscillating hooks have fixed hook timing, so I have not experimented with those.

Effects of changing stitch length and forward/reverse

With a vertical hook axis, as the stitch length or direction is changed the hook to needle distance also changes. I wrote about this in Hook to Needle Distance: A Compromise.

With either a vertical or a horizontal hook axis, hook to needle distance is unchanged but the effective needle bar height would still vary with changes in stitch length or direction.

[AlZilla]

In case you don't know, the "SV" means it was a Special Variant or close enough. Even if you find something about the 108W36, there's still something different. 

Is it a true Unison Feed? As in the feed dogs drop and the foot and needle walk? A lot of times on here we see Binding Machines where the feed dog runs back and forth but doesn't drop. 

Yep, I'd have been on that one in a heartbeat.

[Constabulary]

real triple feed but feed dog goes back and forth only on 108w cylinder machines. But a different triple feed system compared with f.i. 111w or 153w.

Extract from the 108w manual:

 

This is my 108w20 binding edge of a WWII radio back pad.

https://www.youtube.com/watch?v=oY5KjWLJREo

 

Regarding SV machines: 

 

Edited Sunday at 05:43 PM by Constabulary

[AlZilla]

1 hour ago, Constabulary said:

real triple feed but feed dog goes back and forth

Yeah, that does qualify as "real" or "true". I guess the word I should have used was "conventional".

Tangent: would the binding type of feed (flat, non-dropping) be problematic for conventional kinds of sewing? Either cloth or leather? Just curious, really. I always look past binding machines, even if they're a steal.

[brongle]

Wow, this is great info, thanks everyone! 

6 hours ago, Constabulary said:

you still can find some 108w / 108K parts f.i. at college sewing in the UK. But you have to enter the parts number or "search word" for certain parts to find them cause they do not have all the "108 parts" listed in the right machine section.

https://www.college-sewing.co.uk/catalogsearch/result/index/?cat=9&q=108w

https://www.college-sewing.co.uk/catalogsearch/result/index/?cat=9&q=108k

They have the feed dog that can govern the binder attachment plate:

https://www.college-sewing.co.uk/224704c-feed-dog-singer-108w.html

and the swinging attachment plate:

https://www.college-sewing.co.uk/224738-binder-swing-base-singer-108w.html

Certain Singer feed regulating spindles also can still be found but they most likely need a modification in length (at least) and the stitch length number on the knob may not match but you at least can "get something".

https://parts.jacksew.com/feed-reg-spindle-generic-202310/

https://www.ronsew.com/feed-regulating-spindle-224122-for-singer-112w115/

 

Good to know! Looks like my feed dog has the same features as the linked one, so it should able to drive a swinging attachment plate. 

5 hours ago, Constabulary said:

here: 108w on top and 111 below - tip shape is different but I think you can machine a 111 spindle to make it work for the 108w.

The knobs seem to be the same.

So I think this spindle is a good candidate for reworking

https://www.ronsew.com/feed-regulating-spindle-224122-for-singer-112w115/

But in the end decision is yours - I only can give "part hints".

 

That slot in the side is quite interesting. I don't see it anywhere in the parts diagram or the machine manual, and the closer I look at this figure, the more I realize just how many artistic liberties were taken in the manual 😅

Any idea what the slot engages with, if anything? I couldn't feel much in there with the screwdriver, and the below is definitely a bit simplified. One thought that comes to mind is that it might act as a sort of flexure brake, expanding slightly against the inside of the driveshaft when you tighten it to keep it from shaking loose and reducing the stitch length as you sew.

 

 

5 hours ago, Constabulary said:

108w20: Total length from inside of the knob to spindle tip is approx 116mm and the machined tip is approx 13.23mm. But I´m not a machinist - the measures are close but probably but not 100% correct.

EDIT:
cheapest new hook I found was on Amazon:

https://www.amazon.com/460243-HSH-8-20-DURKOPP-Binding-Machine/dp/B0F7K6LCRD

 

Those dimensions should work just fine for me, thanks! I can't imagine they were holding these to incredibly tight tolerances some 70-80 years ago, anyway. I'm not too concerned with getting the stitch length numbers dialed in just right (although it might be fun to add them later based on empirical measurements), but I can definitely get the taper plenty close enough to match the original range from what you provided. 

Fortunately, the hook is in good shape! But I'll bookmark that for later, just in case.

 

6 hours ago, friquant said:

This topic interests me as well. I've been cataloguing the traits of the hook and bobbin case systems that I have, noting the differences that I see.

All of this is really valuable info, thanks so much for sharing! I've come across a bunch of your posts in my time lurking here when I was first learning about industrial machines last year, and they were a tremendous help in developing a mental model of how these machines work.

5 hours ago, friquant said:

Hook to Needle Distance

Reliable hook to needle distance is paramount for reliable stitches (for not skipping, that is.) 

The two machines I have with horizontal hook axis, oscillating hooks (Singer 31-15, Singer 237) have the downside that any axial play of the hook in the hook race constitutes variable hook-to-needle distance. In the case of the Singer 31-15, hook to needle distance is NOT adjustable.

The one machine I have with a horizontal hook axis, (double) rotary hook (Singer 96-87) has considerable axial play of the bobbin case in the hook race, but this axial play has no bearing on hook to needle distance, as the hook point is attached solidly to the outer part of the hook assembly. On this machine the hook to needle distance is adjustable.

The two machines I have with vertical hook axis, (double) rotary hook (Singer 144W103, Jianglong 341) include a needle guard which can be set to deflect the needle around the hook, and is also useful for stuffing the hook. In the case of the Jianglong 341 the needle guard has an adjustable screw in it for making precise adjustments. Hook to needle distance on these two machines is adjustable.

That's fascinating, I never would've expected that axial play in a hook would be tolerable, or that the needle distance would be fixed. My DA class 272 constrains the bobbin case pretty tightly, but the hook point is also attached to the outer part of the assembly. Haven't played with my Chandsew 100rb enough to know offhand if it's similar in that regard. 

Stuffing the hook makes so much sense! I had no idea you could do that with the needle guard (or really what purpose the needle guard served in the first place). 

5 hours ago, friquant said:

A Range of Acceptable Hook Timing

With a (double) rotary hook, if the hook timing is advanced enough the hook may grab the thread again on its second time around. With any hook, if the hook timing is retarded enough the thread will make a snapping sound as the thread is force-dragged around the bobbin case. Ideally there is a range in between these two extremes where the thread gracefully rounds the bobbin case and the hook does not grab the thread on its second time around. Even better when the factory specification is somewhere in this ideal range.

My machines that have oscillating hooks have fixed hook timing, so I have not experimented with those.

Re: your (double) note there - do all rotary hooks rotate twice per stitch cycle? All three of my machines definitely do. Didn't realize you could end up grabbing the thread twice, but that makes sense. 

 

3 hours ago, AlZilla said:

Tangent: would the binding type of feed (flat, non-dropping) be problematic for conventional kinds of sewing? Either cloth or leather? Just curious, really. I always look past binding machines, even if they're a steal.

I'm wondering the same thing. I'd imagine it depends quite a lot on the specific materials you're working with, but I have to imagine that if it can feed the binding tape without causing problems, it should be able to feed normal fabric too... I'll definitely play around with that once I get this machine running!