Regulating is too touchy

[BobG]

I’m fairly new to this and have worked on maybe 30 or so vintage and new movements. I recently went through a NH35 movement that I picked up with low amplitude. 140-150 degrees and impossible to regulate. I did find that the train bridge was not flat and when the screw nearest the barrel was tightened the wheels no longer spun freely so I replaced that with a spare and all good. 
I cleaned it and lubricated it, including the balance cap jewels and it came right up at 220. I could get the beat error down below 0.5 but I’m chasing the rate all over. It will be down -15 and I’ll just touch the regulator arm and it will be up +15. This is a practice movement for me as I want to learn and improve making positional error corrections on the Etachron system. 
The amplitude has fallen off a bit as well, 200. I’ll let it run overnight but is there anything I’m missing that can cause the rate to change so significantly with a very very small movement of the regulator? Could the hairspring be damaged and very particular to both the  regulator and stud position? 
Could the regulator pin positions off? Either too open or closed?  

[HectorLooi]

The terminal curve of your hairspring might be distorted. If the curve is perfect, the hairspring should remain in the centre if the gap of the regulator pins when the regulator arm is moved along the entire curve. And adjusting the rate should not affect the beat error.

However, reshaping the terminal curve is not easy stuff. Maybe at this beginner's level, getting the watch running is already a commendable achievement. 

[Nucejoe]

5 hours ago, BobG said:

It will be down -15 and I’ll just touch the regulator arm and it will be up +15. 

Concur with hector,

 A kink or twist  in hairspring, exactly at regulator gap  can cause this, so move the regulator arm to exposed hairspring ( H/S ) there and check with high maginfication. 

Regulator slot not wide enough or nearly/ fully  closed, the gap just limits breathing of H/S.

Damaged regulator gap,  a close up high maginfication of the gap helps. 

Is regulator arm fitted fully in the setting groove?  

Good luck.

 

[Louis1969]

The timegrapher displays significant beat error variations in the dial-up position. When the watch is lying flat (face up), the beat error is much higher than in other positions. Beat error readings are inconsistent, with especially large gaps in the face-up position.  Please help !

 

[LittleWatchShop]

16 minutes ago, Louis1969 said:

The timegrapher displays significant beat error variations in the dial-up position. When the watch is lying flat (face up), the beat error is much higher than in other positions. Beat error readings are inconsistent, with especially large gaps in the face-up position.  Please help !

 

Maybe show us 1) the watch, 2) timegrapher readings in DU, DD, PU, and PD.

[Birbdad]

I agree with hector it's probably the regulator curve, it almost always is on these movements.

So i work on these movements a lot and iv'e managed to fairly consistently get them running at like 50 degrees more amplitude than that with deltas in the 3 range and on the wrist deviations of sub 1 second a day.

They ALWAYS require work to get there though. The main thing is shaping the regulator curve and this is really finnicky and definitely something you practice on a movement you don't care about when you're new (I'm still new) but if the hairspring is flat and the coils are evenly spaced and the regulator curve is properly shaped and it's pretty wild how accurate these movements can get. 

But, it does sound like your regulator curve maybe needs a bit of reshaping. You can easily mess up the watch learning to do this so warning if you don't want to live with your current results but I"m not an expert, it's just speculation but there's a pretty easy way to check. 

This is a good video showing how to see if your terminal curve is properly shaped. Just make sure the regulator pins are open when you test this.  

He takes off the balance wheel to shape it. I do it with the movement disassembled but the balance on the mainplate. I use a homemade tool from a sharpened dental pic to do the adjustments (Tiny tiny TINY adjustments) and i use the regulator pins themselves to sorta brace the spring against to bend it very tiny amounts and just keep checking it by moving the regulator arm down it till it stops moving the hairspring.

Taking off the balance wheel over and over again is a good way to slip and destroy your hairspring and will make the process way slower. I try to avoid removing the balance wheel from the cock it on these movements due to how difficult it is to get the hairspring stud back into the balance cock. It's super easy to slip and twist your spring then you got way worse problems. 

Since these are mass manufactured and unadjusted the regulator curve is NEVER perfect but once you learn how to reshape them it's pretty easy to do and you can get REALLY low deltas with really low positional error and pretty remarkable accuracy. 

Definitely with it powered down and the balance on the movement with both balance jewels in place look across the spring to see if it's perfectly flat, then check the regulator curve how he does it in the video. I would bet money even if it's not your main issue the regulator curve is not ideal also. 

Edited 5 hours ago by Birbdad

[BobG]

Yes! The terminal curve shape is key. I struggle not only with shape but also the angle it leaves the stud and the the height it leaves the stud. Since on the NH movements the spring is attached to the stud by friction fit and set by machine it can be off and hard to get right. I find moving the stud angle really affects the coil spacing. 

I understand how to shape a terminal curve but when you install the spring and cock into the movement if gte spring is tilted up or down how do I get the spring flat? 

The sling on the left is after shaping and I thought was good but it is way off and required more work. The one on the right is new. 
 

 

[Birbdad]

16 minutes ago, BobG said:

Yes! The terminal curve shape is key. I struggle not only with shape but also the angle it leaves the stud and the the height it leaves the stud. Since on the NH movements the spring is attached to the stud by friction fit and set by machine it can be off and hard to get right. I find moving the stud angle really affects the coil spacing. 

I understand how to shape a terminal curve but when you install the spring and cock into the movement if gte spring is tilted up or down how do I get the spring flat? 

The sling on the left is after shaping and I thought was good but it is way off and required more work. The one on the right is new. 
 

 

So if anybody overrides what i say, listen to them haha. I'm not an expert but i've just been in this boat a dozen times as i used to mess up hairsprings bad, but the good part of that was that i got to learn how to fix them. 

What made my life SO much easier was learning how to adjust the hairsprings with the balance installed on a disassembled mainplate, unless it's a very serious problem or on an inner coil you can't get to i would recommend trying this in part because the spring will mostly ALWAYS look flat when the balance wheel is not installed on the cock. Install the balance wheel back in the cock and the cock with the cap jewels in the mainplate, with it installed even the most imperceptible twist will be pretty obvious. Just make sure the balance jewels are also installed.

In these movements, if 99% of the time any twist or problems gonna be around the terminal curve, twists often near the stud, especially if you've been removing their awfully designed studs from the balance cock. If there's a reliable way to install those things without introducing a minor twist to them i haven't found it. they're an awful design.

INstall the balance on you mainplate WITH your capstones/chatones and look sideways across the balance under magnification. Look for the highest or lowest spot of the now not flat hairspring. 180 degrees from that high or low spot is where your twist will be. You have to be VERY careful with twists as the tiniest over or under twist brings it out of flat. If the area i have to untwist is near the stud i just use a pair of #5 tweezers or some very sharp fine curved tweezers and gently see which way i twist it corrects the twist, then i hold the very fine point of the tweezers on the hairspring and just gently squeeze and check it and see if it started to correct it, then i rinse and repeat. Tiny adjustments is what you want, don't close them all the way, and dont' introduce a flat spot in them. it's just a little nudge by closing the tweezers on it at an angle. . 

If the twist is not near the stud, i grasp the hairspring with some very fine curved tweezers as gently as possible while still holding it then i do that same sorta tweezer squeeze with my #5's. 

[BobG]

What you’re saying is that if the entire coil is lifted or tilted either up or down it is due to a slight twist in the spring? Usually next to the stud? 
I’ve chased this before and in one instance the spring came out of the stud and I had to reinstall it and secure it in place. 

Do you find that the angle (not twist) the spring leaves the stud not only affects how it centers in the regulator pins but also the coil spacing? This could be caused by a distortion in the spring as it grabs and release the pins. I find turning the stud counter clockwise opens the coils 180 degrees away from it. So do I adjust the stud in relation to the regulator pins or to the coil spacing? 

I also find that unless I remove the balance from the cock I can’t see the slight distortions in the spring. In my picture the distorted spring looked fine when installed. Do you use Alex’s technique of sliding the regulator arm up and down the curve to find distortions or can you visually see them? 

 

Gettin back to positional variation for a moment. The angle of the regulator block is set by the factory and I’ve noticed it’s always the same. I have seen that sometimes the pins are too closed and grab or can grab the spring. If you open the pins to do the curve check what guide do you use to close them back down to the correct angle?