206RB Reverse stitch length

[Dave4]

 I've learned from Gregg's posts that RB1 and RB2 have a stitch length lever and push-up reverse.  RB3-5 have a SL dial, and push-down reverse. RB5 has a dial lock tab.  RB4 came in two series, Japanese TH series and a later apparently mixed SH series, some Japanese, some Chinese.

My Machine is a 206RB4-TH series.  My reverse SL was shorter than forward.  With the dial set on 10, the forward stitch measured 10mm, the backstitch measured 8mm.  This 20% disparity persisted across the range of dial settings. 

The Seiko chapter 8-13 Feed Regulator Adjustment procedure had no effect on my machine because my RB4 regulator pinch bolt is on the regulator itself rather than on an adjacent linkage, therefore the relational arrangement of regulator to rod cannot be changed.  If you have an RB-4 or 5 with the later dual parts at the regulator rod, then adjustment of fwd vs rev SL is very simple, just loosen the pinch bolt on the linkage arm, slightly alter the relationship between the linkage and the regulator, tighten the screw, and see what effect you get.  See drawings below.  

Some RB4's and apparently all RB5's have a slotted feed rock shaft crank (FRSC), which would allow possible adjustment of feed regulation (and therefore fwd vs rev SL) at that location.  If you have an RB4 that does not have the later dual part configuration, but does have the later FRSC, it is probably possible to adjust fwd vs rev SL at that point, but also probably possible to wreck something, so proceed with caution.  My FRCS isn't the slotted variety, therefore I was unable to try this.

So I still needed to find a way to make this adjustment on my machine.  

It has been suggested here that altering the inner presser foot timing can change fwd vs rev SL.  According to the manual, this timing can be adjusted by advancing or retarding the lifting cam on the arm shaft.  I tried that, with almost imperceptible effect on SL.  Only a small range of adjustment can be made on this cam before the machine becomes inoperable, and it's a bear to attempt to make a tiny adjustment and then later put it back where it was.  Be sure to make index marks with a sharpie before you start.  I only recommend trying this adjustment if your SL difference fwd vs rev is very small.  In fact, I don't really recommend it all because it attempts to later fwd vs rev SL by moving the machine out of a normal feet timing paradigm, which I suspect will impact machine performance in some way that will show up later.  

It has also been suggested that altering the relative lift heights of the feet can change inner foot timing enough to change rev vs fwd SL.  This is done by loosing the clamp screw on the lifting rock shaft (the slotted 6-sided hex nut on the upper rear of the machine) and advancing or retarding the shaft.  This makes the inner foot lift height different than the outer lift height, and can therefore change inner foot timing, therefore theoretically changing fwd vs rev SL.  I tried it, and it had no effect on my SL.  I was okay with this outcome because I wasn't sure what the long-term of effect of sewing with irregular feet lift heights would be anyway.  So I returned the feet to the equal lift height setting.  BTW - the Navy manual Chpt 4 procedure for this adjustment is a neat trick and the easiest way to equalize the ift height of each foot.  Pasted here:

a. Turn the balance wheel toward the
operator and observe the action of the alternating
pressers. If they do not lift equally, proceed to
step b.
b. Turn the balance wheel until the foot
that lifts too high is just above the throat plate.
c. Loosen the presser-lifting link crank
pinch screw (the same slotted hex nut I referenced above). The foot should
snap down; if it does not, push it down.
d. Tighten the presser-lifting link crank
pinch screw.
e. Repeat step a. If necessary repeat steps
b through d until the feet lift to the same height.

Continuing, I also tried to equalized fwd and rev SL by changing the feed dog position, but that didn't work.  I also tried changing the position of the reverse crank relative to the reverse lever, but that won't work because Screw 7 has a pin that engages a hole of the reverse lever shaft, permitting no adjustment.

I couldn't find any other method for fixing this problem, so i started looking at how the engagement between the feed regulator part and the SL dial rod worked.  

It's pretty simple.  The regulator has a vertical V notch valley on its forward face that receives the rounded nose of the SL dial rod.  This V notch has curved walls.  It's a really shallow valley, perhaps it's more like two rounded lobes separated by a V.  The reverse lever connects to an internal crank that engages the fork on the bottom of the feed regulator.  A spring keeps constant downward pressure on the crank, which keeps constant forward pressure on the feed regulator, which means that the feed regulator is always trying to rock forward.  The only thing that holds it back is the rod of the SL dial hitting the upper lobe of the V notch valley.  When you turn the SL dial CCW, backing the rod out, the feed regulator rolls forward slightly (with the SL dial rod landing higher on the upper side of the V notch).  This creates a longer feed action, therefore a longer stitch.  When you press the reverse lever, this forces the regulator to rock all the way backwards until until the SL dial rod hits the other side (the lower lobe) of the V notch valley.  In a perfect world, the two lobes of the V notch valley are of proportional  dimensions that makes the reverse feed action the same as the forward feed action.  Mine weren't.  The lower lobe of the V notch valley on my regulator was too tall.  When the reverse lever was pressed, the amount of distance that the regulator traveled was not enough to make the reverse feed action of the machine correspond perfectly to the forward action.  It was 20% less travel than what was needed.  

The solution: remove needle and throat plate.  Remove the cast feed regulating base from the back of the machine, pulling the regulator out with it.  It slides off the linking arm and pops right out.  Take the regulator piece off the casting and grind down the lower lobe of the V-notch.   This is basically grinding down a stop to allow more travel.  Both lobes of the V notch are stops.  The lower lobe was simply a stop that was too tall and therefore needed to be ground down.  

I used a small sanding drum on a drill clamped in a vise, run at low speed.  I oiled the drum for wet sanding.  The critical aspects of the mod:  Take a little at a time, test, repeat if necessary.  Preserve the curvature of the face, so that rev SL will = fwd SL at all SL settings. Don't take too much off, else your rev SL will then exceed fwd, requiring a grind on the upper face of the V-notch to correct it.

When you put the feed regulator back on the cast base, use the set screw on the bottom of the base to lock the regulator rod temporarily while you tighten the magic split retention screw into the right end of the rod to hold the assy together.  Procedure: gently bend the magic split open with a wedge, just enough so that it is almost too wide to be introduced into the receiving threads of the rod, tighten the bottom set screw to hold the rod still (the rod has a flat face for this), then insert the retention screw and run it in until nicely tight, then back it out 1/3 turn or so until the regulator spins just freely on the rod, then loosen the set screw, then place the assy rod-down on a flat hard wood block and strike the head of the magic split retention screw a good whack or two with a small hammer.  The whack causes the screw split to spread so that even though the screw is slightly loose, it won't back out in normal operation.  So you whack it hard enough to make magic, but not hard like you want to break something.       

Get a helper for reassembly.  Remove the throat plate!  Remove the needle!  The unregulated machine has maximum feed action, which will drive your feed dog into the limits of your throat plate or your needle into the limits of your hook if you don't remove them.  Procedure: You have to insert the assembled base/regulator into the window, slide the linkage parts together, roll the machine slightly to find optimal fit for continued insertion, and you have to actuate the reverse lever up and down while poking at the nib that must fit and slide into the fork of the regulator as you slide the parts home.  And you have to hold your tongue just right.  The helper should not be a child unless the child is fully acclimated to an environment of choice profanities.  The helper should not be your spouse or anyone you may wish to avoid offending.  Probably should not be opposite gender, because then one gender will blame the other.  If you have a monkey, and the monkey only does what you say half of the time, and the other half it does the complete opposite, then this will be a 6 second job and the nib will literally jump into the fork.  The way i did it was to sit in a chair behind the machine and wear a headlamp and drink beer and stack layers of 3x3 plywood squares under the reverse lever to hold the lever at various positions and then got lucky.  One time i got it in, then bumped the plywood stack and it fell right back out.  It was great.      

Test the result.  If you didn't grind enough off the lower side of the V notch, take it apart (easy), grind more (easy), put it back together (yah).  The grinding bit is an art, but hey, we're artisans right?  Mine turned out perfect on the 2nd attempt.  I sewed some cardboard without thread and measured stitch lengths and drank beer.  Then I got up the next day and sewed more cardboard and marveled at how easy it was to fix the problem and how well it worked now.  

If your rev SL is longer than your fwd SL, you'd grind the top rather than bottom side of the V notch valley. 

When you finish a mod that alters feed action, you reassemble and roll the machine by hand with a needle installed and you go slow and feel for binding.  Watch the extremes of feed dog motion within the throat plate slot.  Make sure it doesn't collide.  Make sure the needle doesn't hit anything in the bobbin area.  If you are reasonably careful, there is no way your mod will cause the machine to exceed feed travel limitations, but the way to test that is by hand, not by motor.  

If you have an RB1 or 2, and you have the unequal SL problems, I suspect you can solve it by rigging a simple external stop on one end of the SL slot.  I've never seen one of these machines in person, but the parts diagram suggests that an external adjustable stop would be the simple fix for the problem.

If you have an RB4 or 5 that has either the adjustable feed regulator (pinch screw on adjacent linkage rather than right on the regulator), or that has the slotted FRSC, then you don't need this mod.

If you have an RB3 or 4 without adjustable regulator or slotted FRSC, then you can try to equalize rev SL to fwd SL by tweaking feet timing.  If that doesn't work, you can buy a new regulator and see if it works better, or you can grind the one you have.  I chose the latter option because there is no assurance that a new regulator wouldn't have the same proportional dimensions as the existing regulator.  I suspect the regulator parts all have the same dimensions, and the common problem of unequal fwd vs rev SL is caused not by variations in regulator dimensions, but rather a variation in cumulative tolerance stack between machines.

Other: Slow wet sanding was effective.  I only spent 15 minutes grinding, just going slow and careful.  It was easy.  The part appears to be case hardened, but light scratches are made fairly easily with the corner of a file, so the hardening is not prohibitively hard or deep.  

Wow, this is long.  I've learned a lot on this board.  Hopefully this post is a way of returning the favor that may help someone else.   Also hopefully, someone won't reply to tell me that all I needed to do was some simple thing that I overlooked before I broke out the grinder.  We'll see...

 

Edited December 3, 2016 by Dave4
fix photo, clarity

[MadHatter]

Wow!  

[Uwe]

Thanks for the detailed write up on this topic Dave! I've inspected the various reverse mechanisms on my machines on occasion.

The RB1 and RB2 mechanism is similar to the one on the Durkopp Adler 204/205 machines (also Pfaff 145/545/1245). On the DA machines the (plastic) stitch regulator cover plate is not flat but has built-in steps that compensate for asymmetries in the design.

 

One alternative to grinding down the high lobe on the non-adjustable RB designs is to build up (make higher) the low lobe, perhaps by attaching (glue/solder) a very thin metal strip/shim. This approach might reduce overall maximum stitch length but might be easier to reverse.

[Dave4]

Thanks Uwe.  You make a good point.  One could lightly scuff and degrease the surface of the "short" lobe and then JBWeld a strip of shim metal to build it up.  

Pro: this would simplify preservation of the original curvature of the lobe.

Con: it may be difficult to match shim thickness to exactly what is needed to equalize stitch length.  

Theoretically, if you got it close, you may be able to make up the remaining difference by tweaking feet timing as has been suggested elsewhere, but I must confess that I had no success in altering SL with either suggested method of feet timing adjustment.  

FYI for future readers, my reduction of lobe height was at least .065"; not a "small" adjustment.  So if you try the glued shim method, you may need some hefty shim if you have SL variation of 20% as I did.

Wow that OP was long!  I know that there are a lot of folks out there who own this class of machine, but I suspect not many of them keep a disobedient monkey, so perhaps I should have left that part out.  

Best to all.  DS

[TinkerTailor]

21 hours ago, Dave4 said:

Thanks Uwe.  You make a good point.  One could lightly scuff and degrease the surface of the "short" lobe and then JBWeld a strip of shim metal to build it up.  

Pro: this would simplify preservation of the original curvature of the lobe.

Con: it may be difficult to match shim thickness to exactly what is needed to equalize stitch length.  

Theoretically, if you got it close, you may be able to make up the remaining difference by tweaking feet timing as has been suggested elsewhere, but I must confess that I had no success in altering SL with either suggested method of feet timing adjustment.  

FYI for future readers, my reduction of lobe height was at least .065"; not a "small" adjustment.  So if you try the glued shim method, you may need some hefty shim if you have SL variation of 20% as I did.

Wow that OP was long!  I know that there are a lot of folks out there who own this class of machine, but I suspect not many of them keep a disobedient monkey, so perhaps I should have left that part out.  

Best to all.  DS

Or you can insert shims behind either the top or the bottom of the stitch adjuster cover plate to angle it and change the contact points. If reverse is too long, shim out the top of the plate, if it is too short, shim the bottom. I firmly believe that casting flaws and irregularities on the bodies of these machines change the angle the plate sits at from the originally designed angle and causes the inconsistency between machines. At least machines with this style of stitch adjust. It is more likely that the external body has casting flaws than the internal parts. In a fiber environment, where most of these machines live, this irregularity is not an issue so they do not correct it in manufacturing due to cost. They could easily machine this to be consistent but it would increase cost of manufacture.  We leather people are the only ones who care that the backstitch is in the same holes....and we are an insignificant portion of the machine sales..

[Dave4]

Excellent points Tink.  Thanks.  

RB1 & RB2 could easily be modified by shimming the cover plate.  This would be better than the magnet bar idea I suggested above, because once you got the correct shim in place, rev SL would = fwd SL across the full range of SL adjustment.  

 

[Gregg From Keystone Sewing]

On 12/2/2016 at 7:18 PM, MadHatter said:

Wow!  

That was my 1st impression as well!  I have to come back and study this when I have more time, seems interesting to say the least.

[Gregg From Keystone Sewing]

I've been able to read a bit of this, and only now clearly understand the issue.  Usually, I don't move on anything until someone brings the issue to my attention.  I don't recall many of the 206RBs, 3s and 4s, that have this issue, but I can see where this would be a problem as in this case.  Great fix, this is not easy at all to do, or to get correct without cutting, welding, and grinding many times until we get it correct.