Odd timegrapher plot - could it be caused by pallet ?

[mikepilk]

I've been working on an FHF 150 for a while. It needed quite a lot of work - new balance staff, escape wheel, roller.

The amplitude was a little low, and I noticed the entry pallet jewel had a lot of lock - at least 1/2 jewel width. So I reduced it to about 1/3 (see pic).

The timegrapher plot was looking good for a while, but then it suddenly runs fast. It's not re-banking, and I've re-cleaned and demagnetised. I cannot see any coils sticking - but even with slo-mo it's hard to tell.

I was pondering .... trying to think if there could be a fault caused by the pallet causing the plot.
e.g. If there was insufficient lock on the entry jewel and it slipped a tooth?  I'm struggling to figure out what would happen.

One thing I noticed is that the banking pin limiting the entry pallet is bent outwards (see pic).

I'm guessing that such a sudden increase must be sticking coils?

Any ideas?

 

 

 

Edited August 14, 2023 by mikepilk

[Waggy]

Just shooting from the hip here, but could the fast trace be the 'real' value, and the slow line be the problem phase? If this was the case then something would be bogging the power train down, and then temporarily freeing itself?

The amplitude looks reasonable on the fast bit, what is the amplitude on the slow bit?

If its low then this could add weight to my crazy idea 

[mikepilk]

Just now, Waggy said:

Just shooting from the hip here, but could the fast trace be the 'real' value, and the slow line be the problem phase? If this was the case then something would be bogging the power train down, and then temporarily freeing itself?

I don't see how it could be. The 'slow' trace is so steady, sometimes for a minute or so, and it responds as expected to small movements of the timing adjuster.  

[Neverenoughwatches]

3 minutes ago, Waggy said:

Just shooting from the hip here, but could the fast trace be the 'real' value, and the slow line be the problem phase? If this was the case then something would be bogging the power train down, and then temporarily freeing itself?

There are always possibilities that we miss something, but he second half of the display does look erratic. 

4 minutes ago, mikepilk said:

I don't see how it could be. The 'slow' trace is so steady, sometimes for a minute or so, and it responds as expected to small movements of the timing adjuster.  

And you cant see anything visually happening at that point of change Mike 

[mikepilk]

39 minutes ago, Waggy said:

The amplitude looks reasonable on the fast bit, what is the amplitude on the slow bit?

The odd thing is, the amplitude doesn't seem to change. I wondered if it was a timegrapher error, but then you would expect the amplitude reading to change.

33 minutes ago, Neverenoughwatches said:

but he second half of the display does look erratic. 

No sh1t Sherlock 

33 minutes ago, Neverenoughwatches said:

And you cant see anything visually happening at that point of change Mike 

It's hard to look through the microscope and at the timegrapher at the same time. I'll give it a go.

I just tried it under the microscope for several minutes, and now it doesn't want to misbehave !

Edited August 14, 2023 by mikepilk

[Neverenoughwatches]

35 minutes ago, mikepilk said:

No sh1t Sherlock 

My deduction was elementary good sir

36 minutes ago, mikepilk said:

The odd thing is, the amplitude doesn't seem to change. I wondered if it was a timegrapher error, but then you would expect the amplitude reading to change.

No sh1t Sherlock 

It's hard to look through the microscope and at the timegrapher at the same time. I'll give it a go.

I just tried it under the microscope for several minutes, and now it doesn't want to misbehave !

What else can increase the frequency rate of a hairspring's oscillations.  How about a slipping collet ? Occasionally.  The weight of the balance would be removed or partially removed temporarily until it catches again. In a vertical position the collet might slip  less ?

[mikepilk]

16 minutes ago, Neverenoughwatches said:

What else can increase the frequency rate of a hairspring's oscillations.  How about a slipping collet ? Occasionally. 

Good thinking. I'll check that and also check the roller. They seemed OK when I fitted a new staff, but maybe the ultrasonic has loosened something.

[Neverenoughwatches]

4 minutes ago, mikepilk said:

Good thinking. I'll check that and also check the roller. They seemed OK when I fitted a new staff, but maybe the ultrasonic has loosened something.

If not the collet then maybe the new staff slipping the balance wheel ? I considered the roller but couldn't come up with a reason why it would increase the rate. I was thinking the roller would only slip during it engagement with the fork.

10 minutes ago, mikepilk said:

Good thinking. I'll check that and also check the roller. They seemed OK when I fitted a new staff, but maybe the ultrasonic has loosened something.

Thinking about it again Mike, if the staff has some movement within the wheel, the most torque on the wheel is at full extension and contraction of the hairspring, it might be at that point it slips. Thats going to seriously affect the beat error though ?

[nickelsilver]

Did you do all the checks on the escapement before and after moving the entry stone?

 

Edited August 14, 2023 by nickelsilver

[Knebo]

Very interesting one! 

If I may ask, how do you exclude the possibility of rebanking? I would have guessed that from the timegrapher (and assuming that the amplitude reading isn't accurate). Related to that, do you know the lift angle for sure - especially after adjusting the pallet stones? Maybe you need calibrate it (with the 180 degrees / visual method). 

I could imagine that the amplitude slowly increases until it hits the rebanking point. Then it misbehaves. But as soon as the rebanking has slowed it down, it goes back to normal. Increases again, rebanks, normal,... 

Your observation that amplitude doesn't seem to change may be a faulty reading from the Weishi 1000. I observed in previous movements that the Weishi will just keep the last clear/actual reading for quite a while until "admitting" that I has lost the trace. 

 

PS: what kind of smartphone do you have? Mine allows me to take slow-mo videos that are good enough to visually observe the true amplitude. I find this very useful and avoids going through the lengthy process of calibrating via the 180-degrees-sharpie-method.

Edited August 14, 2023 by Knebo
Added PS

[mikepilk]

6 minutes ago, Knebo said:

If I may ask, how do you exclude the possibility of rebanking?

I can hear from the ticking - it's not galloping.

Yes my phone will do slow motion - I'll try it

[Knebo]

3 minutes ago, mikepilk said:

Yes my phone will do slow motion - I'll try it

Cool. Check for settings that may allow you to change (=max out) the frame rate. If necessary, you can also use a loupe (with low-medium magnification) and just hold it onto the phone camera. 

[mikepilk]

1 hour ago, nickelsilver said:

Did you do all the checks on the escapement before and after moving the entry stone?

I printed out your words of wisdom when you first posted them 

It's the first time I've really followed them. I can't imagine managing without a microscope.

From what I can determine, the 'clearances' are OK, and I cannot unlock the escapement (fiddly, I used an old sharpened oiler) when it shouldn't.

I'm not so clear on checking "run to banking". Looking at the pallets, I notice the point of drop, and then see if there is any more movement of the pallet fork ? Is this correct. If so, it's OK

[nickelsilver]

5 minutes ago, mikepilk said:

I'm not so clear on checking "run to banking". Looking at the pallets, I notice the point of drop, and then see if there is any more movement of the pallet fork ? Is this correct. If so, it's OK

That's about all there is too it. The fork should continue to move a bit until it hits the banking, which is the total lock (drop lock + run to banking [also called slide]= total lock). Check again that you have guard pin clearance with the fork banked, rotating the balance fully until the roller jewel "rebanks" against the fork. It's not uncommon for a roller table to be eccentric for whatever reason, and you can end up with weird things happening in specific positions as gravity pulls things down.

 

Still looks like you have a fair bit of total lock on the entry; I would try bending the banking closer to straight, which will reduce that, checking of course that there is still sufficient fork horn and guard pin clearance.

 

Doesn't sound like this is an escapement issue though. Sometimes the way the hairspring is situated and centered can cause it to touch the center wheel or stud when the amplitude hits a certain point. That can cause a drastic increase in rate and odd looking timing machine results.

[JohnR725]

which you need here is a different timing machine and/or timing machine software. If you're having a power fluctuation from the mainspring to the escapement you can see interesting effects on the timing. to basically get a different way to look at rates and amplitude and you can see the effect of geartrain issues for instance. Bent pivots and other things that cause powertrain fluctuations. Although your fluctuation is rather severe. You might try changing the averaging time of the machine to being really short to see if you catch the amplitude fluctuation if that's what it is.

7 hours ago, mikepilk said:

escape wheel, roller.

7 hours ago, mikepilk said:

banking pin limiting the entry pallet is bent outwards

3 hours ago, nickelsilver said:

Did you do all the checks on the escapement before and after moving the entry stone?

you made escapement changes and bent banking pin isn't a good sign. This is where you need to go through and make sure everything is right not just the depth of the pallet fork there's an entire procedure for checking the escapement

2 hours ago, nickelsilver said:

Doesn't sound like this is an escapement issue though. Sometimes the way the hairspring is situated and centered can cause it to touch the center wheel or stud when the amplitude hits a certain point. That can cause a drastic increase in rate and odd looking timing machine results.

if you look at your timing machine results notice you go a long ways basically several revolutions the escape wheel before you of a problem. So big it's unlikely sort of to be in escapement issue I do think your palate stones look like at least the one picture looks like the lock is a little too much. But that is also dependent upon you checking the rest of the escapement the make sure everything else is right.

what happens if you let off half the power basically wind the watch halfway up do you still have the same problem?

[mikepilk]

I give up on this watch, I cannot regulate it !  

The first pic is at half wind, the second at full. 

It's never going to give a nice smooth plot, there's wear on some of the gear train pinion teeth (it's not worth spending any more money).
It's just a cheap Art Deco evening wear watch, but it's bugging me, and I'd like to get it within say 30s/day.

The balance has been through the cleaning process twice, and demagnetised twice.
Then the hairspring cleaned in acetone and IPA.  
I'd fitted a new balance staff, so I've checked the staff, roller, roller jewel, collet are all secure.
The hairspring coils are nicely spaced. There is nothing to interfere with them either at the collet or stud.

I straightened the banking pin as @nickelsilver suggested.

I've stared at the hairspring long and hard through the microscope, but can see no obvious sticking. Though I'm not sure I'd notice as it's going so fast, and I cannot get good enough slow motion vid using my phone.

But touching hairspring coils must be the cause (?)

All I can think of doing is pulling the entry pallet jewel out a touch to lower the amplitude.

 

 

[Galilea]

Awwwww, you sound so dejected and frustrated. It looks like rocket science to me, I couldn't get that far. Of course, I have no advice for you but I did bring  and . Wasn't sure if you wanted a  so I brought one just in case. Chin up!

[JohnR725]

1 hour ago, mikepilk said:

It's just a cheap Art Deco evening wear watch, but it's bugging me, and I'd like to get it within say 30s/day.

did we ever get a picture of the complete movement so we know what you're talking about or maybe a model number so I can look it up and see what it looks like?

1 hour ago, mikepilk said:

I straightened the banking pin as @nickelsilver suggested.

the banking pins are part of a safety feature they're not supposed to be straight EF there where there supposed to be. They're not just arbitrarily in a straight position because it looks pretty back see have a functionality and they need to be where there supposed to be for the particular watch. This is why are American pocket watches are typically movable because the early watches needed to be adjusted considerably more than a modern watch so it's find the straightness but did that have any consequences anywhere else even though this is probably not what your problem is

 

1 hour ago, mikepilk said:

But touching hairspring coils must be the cause (?)

All I can think of doing is pulling the entry pallet jewel out a touch to lower the amplitude.

the two different timing machine results what did you do? I would personally take a little more time to understand what's going on before you start modifying and changing things. Then when it's supposedly going at that higher amplitude does the balance wheel actually look like it's going at a higher amplitude? Sometimes on timing machines if the amplitude is too low the timing machine gets confused or if it can't get a good clean signal it gets confused and it will not read the right part of the waveform and then the amplitude is not right at all. Typically it will read way too high when it's not that I had all.

Out of curiosity how do you lubricate your escapement?

 

 

 

[Knebo]

1 hour ago, Galilea said:

Awwwww, you sound so dejected and frustrated. It looks like rocket science to me, I couldn't get that far. Of course, I have no advice for you but I did bring  and . Wasn't sure if you wanted a  so I brought one just in case. Chin up!

Mate, 

I truly feel your pain. 

I'd say two things:

1. Since rates are so different between half and full wind, could it be the mainspring? 

2. Take a pragmatic approach! Wear it for a few days, winding it once a day and calculate the average "de facto" deviation per day. Then regulate from there. Practically, you may get a decent overall timekeeping result (@JohnR725 would be so proud of me). 

Edited August 23, 2023 by Knebo

[rehajm]

I’ll defer to the pros but my money is on knocking. The balance builds a head of steam before it begins hitting somewhere. did it ever work? Mainspring too strong?

Visually that balancespring is a heavy breather…perhaps spinning itself together before it has an opportunity to rebank?

Edited August 23, 2023 by rehajm

[Nucejoe]

2 hours ago, Galilea said:

Awwwww, you sound so dejected and frustrated. It looks like rocket science to me, I couldn't get that far. Of course, I have no advice for you but I did bring  and . Wasn't sure if you wanted a  so I brought one just in case. Chin up!

 Very nice of you Galilea , but mike doesn't seem interested in the cake, perhaps the coffee, so if you don't mind I,ll have the cake.

 

[Galilea]

13 minutes ago, Nucejoe said:

 Very nice of you Galilea , but mike doesn't seem interested in the cake, perhaps the coffee, so if you don't mind I,ll have the cake.

 

Here you go, have his and a slice of Persian love cake  (that's what it's called people) also some rose tea.   Enjoy!j (Just in case, here's a nice cognac 

[mikepilk]

3 hours ago, JohnR725 said:

did we ever get a picture of the complete movement so we know what you're talking about or maybe a model number so I can look it up and see what it looks like?

It's an FHF 150. Looks like this

3 hours ago, JohnR725 said:

Then when it's supposedly going at that higher amplitude does the balance wheel actually look like it's going at a higher amplitude? Sometimes on timing machines if the amplitude is too low the timing machine gets confused or if it can't get a good clean signal it gets confused and it will not read the right part of the waveform and then the amplitude is not right at all. Typically it will read way too high when it's not that I had all.

Out of curiosity how do you lubricate your escapement?

Yes, it does look like it's going to a higher amplitude. 

Escapement lube, as I usually do. Fixodrop on escape wheel, pallet jewels, cap jewels.
9010 for balance jewels. Run for 5 mins, then 9415 on pallet jewels - under microscope to get the correct amount.

3 hours ago, Knebo said:

1. Since rates are so different between half and full wind, could it be the mainspring? 

2. Take a pragmatic approach! Wear it for a few days, winding it once a day and calculate the average "de facto" deviation per day. Then regulate from there. Practically, you may get a decent overall timekeeping result (@JohnR725 would be so proud of me). 

1. No. I would be able to hear galloping if it was re-banking, and I can't

2. Sat on the bench, the figures tie up with the timegrapher - it gains 1-2 mins/day.

3 hours ago, rehajm said:

I’ll defer to the pros but my money is on knocking. The balance builds a head of steam before it begins hitting somewhere. did it ever work? Mainspring too strong?

Visually that balancespring is a heavy breather…perhaps spinning itself together before it has an opportunity to rebank?

Not knocking (re-banking), see 1 above. 

The spring might be moving more than it's meant too ? Is this possible ?

[rehajm]

 

1 hour ago, mikepilk said:

The spring might be moving more than it's meant too ? Is this possible ?

The first tg looks similar to the de Carle vibrograf suggesting too much mainspring power. You do have strong amplitude, even with the error...so couldn't it be hairspring contact? Though it looks concentric and you've already checked for cleanliness but couldn't it be rubbing somewhere, or the coils rubbing together?

[mikepilk]

7 hours ago, rehajm said:

 

The first tg looks similar to the de Carle vibrograf suggesting too much mainspring power. You do have strong amplitude, even with the error...so couldn't it be hairspring contact? Though it looks concentric and you've already checked for cleanliness but couldn't it be rubbing somewhere, or the coils rubbing together?

I'm thinking it must be hairspring contact, but surely it would be designed to allow 360° rotation before coils become too close?

This is how it looks