Valjoux 7750: Rate Behaviour Question

[Hoe]

Dear Watchmakers

I recently serviced a Valjoux 7750 and I noticed that the movement runs too fast in the first let’s say two days until it stabilises to a useful rate. With amplitudes between 300° and 264° and beat errors between 0 ms and 0.2 ms depending on the position on the timegrapher I consider the 30 years old movement as “healthy”. I installed a brand new mainspring and wore the watch during this test. I place it dial up during the night.

Furthermore, I think due to my home office status and the lack of movement, it doesn’t wind enough so it is slowing down since 31st March.

See my Excel diagram in the attachment.

Why does it run so fast at the beginning? Is something wrong or is this a normal behaviour? Somewhere I read that watchmakers regulate watches after they ran for 24 h.

Best wishes

[nickelsilver]

In professional work it's quite normal to let a watch run number of days before final regulation. 7750 is usually quite stable from the start, what would be interesting is to know the true state of wind of yours over this period; it seems you were observing the time not the displayed rate on a timing machine. If the regulator pins are not close enough you can have a quite large difference in rate between relatively normal amplitudes. Overall I would expect a 7750 to be more stable than this.

[Nucejoe]

7 hours ago, Hoe said:

Why does it run so fast at the beginning? Is something wrong or is this a normal behaviour? 

Its quite normal, because lubricants even distribution may take a week or so, only following which it will stablize. 

[Klassiker]

29 minutes ago, Nucejoe said:

lubricants even distribution may take a week or so,

Genuine interest; are you aware of any evidence or references for this?

[Nucejoe]

45 minutes ago, Klassiker said:

Genuine interest; are you aware of any evidence or references for this?

Make your own observation,  add oil to pallete-escape of a stable watch and you see a curve before it re-stablises.

Regs

[Hoe]

Thanks so far for your inputs.

I will stop wearing the watch to measure how long it runs until the movement runs out of power. According to the ranfft website, the VJ7750 has a power reserve of 46 h. This comparison would make it possible to determine the residual power reserve at this moment.

Afterwards, I will restart a test and record the timegrapher values 1-2 times a day.

[JohnR725]

One thing we may be seeing a methodology problem? Then what kind of regulating system does this watch have?

I see several problems with what you're doing when you placed the watch on the timing machine I assume you time dial down only? This would be perfectly acceptable If you can cleverly where your watch always in a dial down position.

Now we get an interesting problem when you're wearing the watch I do wrist it's in multiple of positions. When it's on the timing machine most people only time in one position. They never look at the other positions or don't even think about timing and a lot of positions and doing some math in seeing what the average is supposed to be. Because you may find that that's different than what you see in one position only.

Then your numbers are wasted time unless you give us a complete picture? In other words every time you give us a number of fast or slow we also need the amplitude. This is because if for instance your regulator pins are a little bit  a part or you have the etachrion system not quite adjusted right In other words the regulator pins are too far apart amplitude can have a insane amount of influence on what the watches doing for timekeeping.

You should also try winding the watch up timing it in if you are obsessed of like I would be six positions just because minimum of two positions your minimum of two positions on the dial down and the crown down. Otherwise your obsessive your six positions we dial down and up 4 different crown positions like up downright left. Allow at least 30 seconds between rotating the microphone for the watch to stabilize measure about 30 seconds.  Then let the watch run on your bench for maybe 12 hours do it again 24 hours again I think your watch run a couple of days may be one more time see what the effect of amplitude is on your timekeeping.

Also look at the PDF take the numbers that you got run the math see what it says the average is and at the same time be doing this timing trial cost of physically look at the hands and see how they compare with what you think the timing machine is telling you especially if you do the math to get an average. Oh and I assume your time reference keeps better time than your mechanical watch otherwise that's an added complication to the whole thing.

witschi X-D-DVH-Di-Im-N_EN.pdf

[Endeavor]

Coincidentally, after a tip of JohnR725, I just went through the 7750 regulator-pins exercise. Nickelsilver pointed this out also.

The movement is a freshly serviced NOS 25-jewels normal grade 7750. Runs in every position with straight lines, lift-angle 49 degrees.

I started off with the regulator pins nearly "full-open" and started slowly turning to reduce the gap between them, reducing the freedom of movement of the hairspring between the regulator-pins.

Here are the results;

1) DD +4 333 0.0, DU +3 320 0.0, CD -13 296 0.1, CU -16 298 0.1, CL -17 292 0.2, CR -18 292 0.1

Started to turn the pins to close the gap and had to adjust the daily-rate downwards;

2) DD +5 327 0.0, DU +5 315 0.0, CD -5 291 0.0, CU -8 289 0.1, CL -3 295 0.1, CR -10 282 0.1

Closed even further till there was nearly any gap left and had to adjusted daily rate further;

3) DD +8 325 0.0, DU +7 318 0.0, CD -2 291 0.1, CU -5 296 0.0, CL -2 290 0.1, CR -7 287 0.2

4) Closed a tiny bit further, but that turned out to be too much and I had weird readings. I had to open the pins up and to start "squeezing" again. These are the readings with the pins as close together as I could get;

5) DD +1 323 0.0, DU 0 311 0.0, CD -7 287 0.2, CU -10 289 0.1, CL -7 287 0.0, CR -13 281 0.1

As can be seen, closing the gap to the point whereby the hair-spring just "breaths" makes a substantial difference in the positional error.

In 1) the biggest difference between the fastest and the slowest daily rate is 4+18 = 24. After adjustment the difference is down to 1+13 = 14

Edited April 2, 2021 by Endeavor

[Klassiker]

On 4/1/2021 at 9:13 PM, Nucejoe said:

Make your own observation

Hi Joe,

I did this. It depends what you mean by stable, but over three not-very-high-quality movements I see a maximum  delta between the rates recorded a few hours after oiling ( escapement only) and those three days later of 3 seconds (dial up). So, on current evidence I think "a week or so" is unnecessarily long. "The length of the power reserve of the watch in question" would be my guess for a maximum necessary running-in time before final regulation following a complete service. I'm sure there must be a lot of data available from other sources, but I don't know where to look.

I don't want to hijack this thread any futher. Is it worth starting a new one?

[Nucejoe]

2 hours ago, Klassiker said:

Hi Joe,

I did this. It depends what you mean by stable, but over three not-very-high-quality movements I see a maximum  delta between the rates recorded a few hours after oiling ( escapement only) and those three days later of 3 seconds (dial up). So, on current evidence I think "a week or so" is unnecessarily long. "The length of the power reserve of the watch in question" would be my guess for a maximum necessary running-in time before final regulation following a complete service. I'm sure there must be a lot of data available from other sources, but I don't know where to look.

I don't want to hijack this thread any futher. Is it worth starting a new one?

I agree,  you just be staying on the safe side by running the watch for a week. On a high precision piece , however,  a week or longer seems a must.

I learned this from whats commonly practiced in my neck of the woods and personal observation and know of no data on the subject. 

Regs

Joe

 

[Hoe]

Update

In the meantime I let the watch run down without wearing it to calculate the residual power reserve on 1st April.

The maximum power reserve measured was approx. 53 h. Remember, I replaced the mainspring (CousinsUK #7750771, 1.50x0.113x600x12).

I am going to fully wind and wear the watch again and observe its behaviour closer with the timegrapher.

Edited April 5, 2021 by Hoe
typo

[WellAdjusted]

For what it is worth, Elgin, inventors of alloy mainsprings with their innovative DuraPower alloy in the late 1940s, argued that these springs had a break-in period. Several mainspring packages came with an insert noting that the amplitude will be higher later on after the settling period.

There's much to be said for letting a watch run for at least a few days before adjusting it.