Adler 104 stitch length - adjustment manual

[Pterytus]

Great, thanx! Maybe the solution is somewhere in there.

[Bert51]

I have pulled the information that I needed as a word doc, but it is to big to post here.

If you don't mind giving me your email in a private message, I will email it to you.

Bert.

Edited December 9, 2019 by Bert51

[Pterytus]

So I went through all the adjustment rules, but there seems to be missing something.

The description tells you to adjust the eccentric on the main shaft (and I've done that).
But that alone CAN NOT be the solution for uneven F/R-feed as it only synchronizes the needle position to the feed dog position.

So let's remove the needle. The machine still transports the fabric (correct or incorrect). So the needle timing does not affect the feed kinematic in itself and thus can not cure the problem at hand.

Next would be feed dog lift. Not relevant in this case, as it also does not affect the (in-plane) F/R feed dog movement which is the main problem (you might be able to deribelately shorten F feed to R feed length by a prematurely diving feed dog - but that's cheating).

The key problem is visible with the machine in F when the feed dog is in its in-plane dead center nearest the back of the machine. When changing to R, the feed dog moves to the front of the machine but changes direction shortly before the stitch lenght regulator reaches it's end stop and moves a little to the back of the machine again. THAT is the "missing" R feed - which has nothing to do with needle timing or feed dog lift timing. It's a result of the internal stitch length regulator kinematic (slider position, pivot points etc.) and I'm not quite sure it can be adjusted - or maybe I am missing something...

 

 

 

[Bert51]

Have you checked the position of the needle to the hook.

The eye should be no greater the an eight of and below the hook and about 1/32below at it's highest.

Here is a pic from the manual, but I do not have the gauge, I guessed.

Bert.

 

[Constabulary]

How much difference is it between F & R? Is the difference in all lever positions or just in the longest stitch length?

[Pterytus]

@Bert51

Yes, needle timing is checked and ok (I built myself something like this gauge). The machine sews ok with different needle and thread sizes, only F and R stitch holes do not match.

@Constabulary

I'd have to measure this to have the exact figures, but it's something like 10% off, I'd guess. And yes, it's in all lever positions. Otherwise I would not really mind (or would not have found out in the first place as 9mm is not my regular go to stitch length ;-)

[Bert51]

I need to reset mine as I am just enough to be a pain. It's about 8mm forward and 7 in reverse.

Mostly I am lucky enough that I can turn what I am sew and back stitch that way. I had to learn that when I use the 132K6.

From memory, it was a pain to get it to be spot on in forward and reverse, it appeared to be as the instruction said, but I have to keep fiddling to get it just right. Adjust it, stand it up try it, lay it down and do it again, I had to do it a few times. I don't know why, but it did.

Bert.

 

Edited December 11, 2019 by Bert51

[Pterytus]

I checked the stitch lenghts F and R for different settings of the stitch length regulator and measured 10 stitch lenghts each. First test was full stitch length, the others were changed by full knureld knob revolutions, with zero (centered regulator) being 13 revolutions in.

And the result is really interesting:

0: F97mm/10, R81mm/10 (-17%)

5: F77mm/10, R62mm/10 (-20%)

7.5: F52mm/10, R52mm/10 (+/-0%)

10: F33mm/10, R42mm/10 (+27%)

12: F14/10, R21/10 (+50%)

So with my current setup there is ONE stitch length at ~5mm in which F and R stitch lenghts are equal.

Longer stitches result in reduced R stitches, shorter stitches result in increased R stitches.

 

Would be really interesting to compare it to an equal machine (sister 104-4 from the same production year). Or maybe there is something to adjust other than what I already tried.

[Constabulary]

The 104 / 105 came with 3 (three) different stitch length levers so maybe they changed them for a reason. The last one is the simplest and close to what they have used on the class 4 model. Maybe your SLL has been replaced once...

Let me see if I have pictures...

EDIT:

So I would guess they "simplified" the SLL for a reason. But I only can guess...

I think in a time line yours is the middle model. The latest model has 2 limiters - just as the class 4 and you have to adjust F & R manually for either direction by setting the limiters.

So if you are "good with metal" you can probably reproduce the latest model SLL and solve the problem that way.

You know what I mean? Could work but again I only can guess - just my 2 cents.

Edited December 11, 2019 by Constabulary

[Pterytus]

I think I'm "OK" with metal ;-)

So I COULD in fact do something like that. But it would only partly solve my problem: On such a machine (late type or modified mine) you'd have to adjust R stitch length each time you change F stitch length, if you want it to be equal. Does not seem that practical - but in fact "solves" the problem for the company because noone can complain about, uh, what-I-do. So this could be (part of) the reason.

So a real solution would be to design something like a curve which protudes more or less outward on the R (top) side of the SLL frame and limits the SLL in the correct angle to match F stitch length.

With my data it would mean plus-material directly above 0 (center). Level-material at 5.2mm F stitch length and minus-material above 5.2mm stitch length. That way I'd only have to "flick the switch" in order to get correct (matching) R stitch length.

I believe the problem is the pivot point position of the SLL relative to the direct line between the centers of the main shaft and lever connection of the fork (below the bedplate). If the pivot point would be in line with those two centers, F and R would be exactly the same (simple geometry). On my (the) Adler 104, the pivot point is significantly off this line to one side - maybe there is a good reason for that, but this seems to result in the observed behaviour (there's a draft attached, don't know if that helps to explain what I mean :-)

It's the same on my Haid&Neu, but on a garment-sewing household machine noone really cares about unequal stitch length (or even recognizes it). If you work with leather it does matter.

So if all that is true, it is not a matter of adjustment...