Very confusing Timegrapher results

[VWatchie]

3 hours ago, Endeavor said:

As if it is not complicated enough, another variable in the mix ! How good / sensitive are you ears?

Actually, 25 cm was from memory but now that I listen it's more like 10 to 15 cm from one of my ears. Pretty loud anyway...

[JohnR725]

6 hours ago, VWatchie said:

The manual (Chinese to English via Google translate) didn't make much sense.

As your manuals in Chinese I'm attaching a PDF of the English version. It may not be any better for meaning because they may have used Google in the first place to translate it.

To understand the problem timing machine faces I have a link to a video. Its purpose is to demonstrate another escapement. So ignore everything for about the first minute and then you can see the lever escapement in action. You will note in the video that there are five sounds but two of them are so close to the other one that normally on our oscilloscope you only see three unless you're having a problem. For the timing machine the key sound is the unlocking where the roller jewel hits the fork. This is a very quiet sound but key to everything.

 

https://youtu.be/g5c5RK4WFV8

WeiShi Timegrapher 1900.PDF

[JohnR725]

Depending upon where you're located in the world there are companies that will calibrate witschi timing machines. Unfortunately you have to physically send them the machine dug up to the GPS calibrator. In addition to witschi making a GPS calibrator they make this the micro signal generator. I'm sure it's astronomically expensive. But it be nice if we could borrow one and try it out for little while. Because this way you can verify your timing machine is indeed working correctly. Perhaps we should get clever and make one.

 

https://www.witschi.com/en/group-of-devices/measurement-of-mechanical-watches/microsignalgenerator/microsignalgenerator.html

[Delgetti]

I had a similar problem. In my case the point was simply that when the movement was attached directly to the timegrapher the cannon pinion was slightly rubbing against the spring of the machine. Of course I dont know if this is the case here.

[JohnR725]

21 minutes ago, Delgetti said:

I had a similar problem. In my case the point was simply that when the movement was attached directly to the timegrapher the cannon pinion was slightly rubbing against the spring of the machine. Of course I dont know if this is the case here.

If you look at the very first picture the discussion you'll notice how you're supposed to hold the movement in the holder. In other words there some notches in the metal parts and in the spring part to hold it up off the base and you should never touch the spring. If you can opinion was touching the spring because it was a really big watch it after you really big you could always flip it over the other way around. In any case it works better if the watch is up in the notches of the holder.

[praezis]

12 hours ago, JohnR725 said:

In addition to witschi making a GPS calibrator they make this the micro signal generator. I'm sure it's astronomically expensive. But it be nice if we could borrow one and try it out for little while. Because this way you can verify your timing machine is indeed working correctly. Perhaps we should get clever and make one.

Already happened! I built such a device years ago, though not quite as chic as Witschi's. It not only is able to test every aspect of TMs, but can test a single microphone, too. Essential for repair and development of timing machines

Frank

Edited March 25, 2021 by praezis

[VWatchie]

Today I made a small black mark on the balance wheel rim exactly where the impulse pin is located and recorded a slow-motion video of it. Unfortunately, it no longer seems possible to upload iPhone slow-motion videos to YouTube. So, I paused the video at the turning point where the balance wheel stops and took a screen shot of it (image above).

In the image the balance wheel has just rotated clockwise and has come to a full stop before beginning its anti-clockwise rotation. As evident, at this point the balance has travelled well over 270 degrees from its position dead centre between the banking pins. The amplitude may not be 300 degrees but not far from it. So, it appears that I made a good job servicing this beautiful movement after all. If you feel like it, you may now congratulate me! 

The big question though is how we should relate to our cheap Chinese timing machines? If we can’t trust them, or find a reliable way to use them, they aren’t all that useful, are they? I guess slow-motion videos is an option, but I would assume only for slow-beat movements.

If anyone happens to have better used Witschi for sale, drop me a PM! (Just checked the prices on Cousins for these professional TMs and O.M.G!)
 

EDIT: Just came to think of it, the movement was in the movement holder when I recorded the video. Hmm... 

Edited March 25, 2021 by VWatchie

[Endeavor]

Let me be the (first?) one to congratulate you ! Anything over 270 degrees is pretty good and now you can wear the Omega with proud !!

As for the Weishi 1000, I've no problems with it, that's to say that I'm happy what it tells me. Whether that is the truth I don't know and never checked, but the fact is that all my serviced watches do run fine.

Perhaps me and my watches, as far as the amplitude is concerned, are living in blissful ignorance ?

Somehow that seems to work

Any measuring device needs calibration, the expensive Witschi's are not exempt of that fact either.

Edited March 25, 2021 by Endeavor

[VWatchie]

Just recorded another video, this time with the movement mounted directly on the microphone stand and I get the very same result. That is, near 300 degrees.

Edited March 25, 2021 by VWatchie

[Endeavor]

3 minutes ago, VWatchie said:

Just recorded another video, this time with the movement mounted directly on the microphone stand and I get very same result. That is near 300 degrees.

Did this time the Weishi 1000 confirm the video results?

[VWatchie]

6 minutes ago, Endeavor said:

Let me be the (first?) one to congratulate you !

Well, you happen to be the first, so thank you very much! 

[VWatchie]

3 minutes ago, Endeavor said:

Did this time the Weishi 1000 confirm the video results?

Silly me just assumed it would as I tried it on and off the movement holder repeatedly before I wrote my first post. I'll test it again. Back in a few minutes...

EDIT: Will hopefully be back within an hour. Wife calling! 

Edited March 25, 2021 by VWatchie

[VWatchie]

43 minutes ago, Endeavor said:

Did this time the Weishi 1000 confirm the video results?

Indeed not! I get the very same readings as described in my first post in this thread. About 240 degrees when mounted directly on the microphone stand and around 300 degrees when the movement holder is in between. It shall be interesting to see what I get once it has been cased. Perhaps these cheap Weishi machines were basically made for collectors and configured to do well with cased movements!? BTW, that brand name is sort of similar to Witschi. A pure coincidence, I'm sure! 

Edited March 25, 2021 by VWatchie

[Endeavor]

I remember that JohnR725 did send me a comparison he made between the Weishi 1000 and a professional Witshi; the results were very impressive, with the Weishi 1000 in very close range. For the slight difference but the huge difference in purchase price it was a no-brianer for me.

Of course, he may have had a good Weishi 1000? Hard to tell unless you compare a whole bunch of the Weishi 1000's with a proper, calibrated Witshi.

Perhaps some of those test were done?

Edited March 25, 2021 by Endeavor

[VWatchie]

17 minutes ago, Endeavor said:

I remember that JohnR725 did send me a comparison he made between the Weishi 1000 and a professional Witshi; the results were very impressive, with the Weishi 1000 in very close range.

I have a Weishi 1900 with an adjustable gain. I just tried it with the movement mounted directly on the microphone and when I decrease the gain to a minimum the amplitude reads around 300 degrees (more like the truth), but when I increase the gain to the maximum the amplitude drops to around 220 degrees. So, a massive difference!

Perhaps a reasonable strategy would be to increase and decrease the gain until you get the most favorable amplitude and consider it close enough?

Edited March 25, 2021 by VWatchie

[VWatchie]

Just placed one of my Vostok Amphibian divers (massive stainless steel case) on my Weishi 1900 and this time around changing the gain setting doesn't seem to have much affect at all. Perhaps TMs are like the universe, impossible to comprehend 

[Endeavor]

Before I bought the Weishi 1000, I had the self-build Watch-o-scope. The gain was (not easy) adjustable as well. With the gain set too sensitive it picked up nearly every sound in the room and when my wife, the freezer or the refrigerator did cut in, you could forget about any decent readings. The W.O.S. did display raw data and any room-noise was clear to see.

The W.O.S. could not make any sense of my freshly serviced Rolex and I was afraid that I had scr*wed up. With pure the movement clamped in the microphone all seems fine, but with the movement in the housing the reading were all over the place. I didn't know which reading to trust. Luckily the Rolex ran constant and on time.

The very first thing, when the W-1000 arrived, was checking my cased Rolex and the W-1000 had no problems detecting the proper signal; to my big relieve all the readings were 100%.

As I mentioned on the top, vibrations or resonance (in my case the thick Rolex submariner housing) did / can cause (huge) problems.

Edited March 25, 2021 by Endeavor

[jdm]

20 minutes ago, VWatchie said:

Just placed one of my Vostok Amphibian divers (massive stainless steel case) on my Weishi 1900 and this time around changing the gain setting doesn't seem to have much affect at all. Perhaps TMs are like the universe, impossible to comprehend 

On this class of machines there is no indication of the strength and quality of the input signal. There is no input level meter, no waveform display, not even an audible repetition of the sound through the speaker. Which is bad, but also a case of "get what you pay for".

However as you have found different distances from the mike, as well other elements like presence and type of a case can affect the readings because of the introduced attenuation, distortion and resonances. All one one can do is to adjust gain to get the cleanest and most consistent reading. However In my experience that is most often not necessary.

[VWatchie]

1 minute ago, jdm said:

a case of "get what you pay for".

 

1 minute ago, jdm said:

All one one can do is to adjust gain to get the cleanest and most consistent reading. In my experience that is most often necessary.

I think you're dead on! Thanks!

[VWatchie]

5 hours ago, Endeavor said:

I remember that JohnR725 did send me a comparison he made between the Weishi 1000 and a professional Witshi; the results were very impressive, with the Weishi 1000 in very close range.

@JohnR725 If you still have that comparison, I'm sure many of us, including myself, would be interested in seeing it.

[JohnR725]

5 hours ago, VWatchie said:

The big question though is how we should relate to our cheap Chinese timing machines? If we can’t trust them, or find a reliable way to use them, they aren’t all that useful, are they? I guess slow-motion videos is an option, but I would assume only for slow-beat movements.

If anyone happens to have better used Witschi for sale, drop me a PM! (Just checked the prices on Cousins for these professional TMs and O.M.G!)

As mentioned by Endeavor Up above I've done comparison of timing machines. I own a witschi watch expert two which is now considered obsolete but still a nice machine.  Because people know of my interest in timing machines locally as able to borrow a Chinese 1000 machine first then later a Chinese 1900. The 1000 machine came with a bonus the owner was concerned that it wasn't timing his watches correctly.

Testing procedure is more or less the same. The peculiar wording of more or less the same means that ideally I would try to photograph both machines at the exact same time because if you don't do that you will end up with minor variations. Both machines were timing the same watch the same time because are clipped on the same watch at the same time. That means they should show things identically but they'll be a little variation. The other thing I did with the Chinese machines was always try to remember to change the averaging time to 20 seconds which is what I keep the witschi machine.

Anyone paying attention in the group will notice that I recommend the Chinese 1000 or 1900 machine because for the most part their outstanding. Side-by-side comparison which the witschi they do a really really good job. There missing a few features of the witschi machine has but that's probably not a bother most people.

Where you get interesting problems is what happens on extremes? Or as what's happening here if you're getting bad pickup. At least that's my guess what the problem is. In the case of the watch expert there is no on oscilloscope but I do have a nice knob I can change the gain adjustment or volume adjustment or whatever it is typically it defaults at a certain level and works fine there.

The other extreme is our if you go too low of amplitude the Chinese machine has an issue the witschi does not have an issue. There are graphical display issues that affect both machines are more common with the Chinese because it was smaller display at least for the 1000 machine. Like for instance I just the beat of your watch deliberately and go out of beat on the 1000 machine ill roll over at I don't remember like maybe 4 ms. The island up with the graphical display showing one thing the numeric display showing another but there are ways to fool the witschi machine also. Unfortunately these machines are not perfect but they still do a really really good job on the Chinese machines at their cost arch truly outstanding machines.

If you are upgrading to the witschi machine typically the wood by the watch expert now in its fourth generation found at this link

https://www.cousinsuk.com/product/witschi-watch-expert-4th-generation

Also a review from someone else

https://omegaforums.net/threads/witschi-watch-expert-iv-timegrapher-review-initial-thoughts.48957/

I had to do a British pound dollar conversion and the pound price looks nice the dollar price does not either gets $1000 more than what I paid for my machine. I wonder if ordering from another country other than UK would get you better price?

That we get the something different? All of the current machines as far as I can tell are now basically PCs. Analyzing the update for the machine at work it's running Windows CE. We even make a reference somewhere that their machine is basically a PC. So what if you toss the PC and keep the microphone and throw in some software? Then you get something else like the ChronoMaster And the ChronoMaster  pro There's also a couple more found on the witschi website that cousins doesn't have so I can't compare prices. I know there's one version with a full automatic microphone but that's can have incredible sticker shock. Disappointing with witschi is that instruction manual for the microphone covers the microphone does not cover the software it refers you to the software manual and they don't conveniently have the software manual?

So your pricing goes like this the watch expert 4  £2,745.00. The  ChronoMaster  £1,845.00 and the pro £2,395.00

https://www.cousinsuk.com/product/witschi-chronomaster

https://www.cousinsuk.com/product/witschi-chronomaster-pro

[JohnR725]

11 hours ago, praezis said:

Already happened! I built such a device years ago, though not quite as chic as Witschi's. It not only is able to test every aspect of TMs, but can test a single microphone, too. Essential for repair and development of timing machines

Nice that you built one it would be even nicer if you would start a new discussion in it tool section on how we could all build something similar or? Start a new discussion on ideas for building a calibration device for timing machines to verify that they indeed are working correctly.

[JohnR725]

8 minutes ago, VWatchie said:

If you still have that comparison, I'm sure many of us, including myself, would be interested in seeing it.

A little background of how and why did the comparison. Seattle's an interesting place for watch repair schools as originally started with one school. Before wostep and Rolex came in they had an evening program open to anybody. Unfortunately wostep and Rolex didn't like that so they were banished it eventually became their own school. For a while I was teaching their teaching hobbyists. This works out nice as I have quite a few lectures to steal pictures from this pictures below were from one such lecture.

The students with his Chinese 1000 machine was confused with what he was seeing. What I get for us is I just went to snipping out the images in the order of the presentation. No notice in the beginning there is a Chinese 6497 movement both microphones clipped on The same time. You can see the timing machines both perform nicely.

Then we go to the Howard watch is in desperate need of servicing and needs to be demagnetized. Normally and watch repair at least are top to be obsessed with magnetic fields and I had never seen a magnetized Watson the timing machine because you're taught to demagnetized the watch before placing on the timing machine. Here I timed them exactly as found. Then the 44° for the Howard watch probably isn't right that's what would've been set for the 6497 but they're both on the same setting they should be the same and it looked pretty good don't they.Then having access to commercial software you can see the effect of the classic era bent pivot or magnetism. It's one of the nice features with paper tape is that it's much longer than a liquid crystal screen. Then once the watch was demagnetized back on the machine and they both look similar.

Next Hamilton 992B that needed to be serviced we can see a before and after Of that. We also see a frustration that I had the non-backlit witschi display is horrible for taking pictures of. There is the other problem that I've pointed out before if the amplitude is too low the Chinese machine will not produce an error message but instead will capture the wrong part of the waveform and give you a happy amplitude which is entirely incorrect. As I said it's easy sometimes the fool the machines if you push things to extremes. It's where your numeric numbers the graphical display and what you visually see all has to look right or you need to figure out why it doesn't. Then after servicing still a minor issue with something the witschi display were circled the nine you can change's basically the magnification of the screen. This is a irritating feature at work as my boss likes to push the magnification the 16 and that makes everything look bad as things tend to look much better at the normal setting of two.

The last watch for the comparison is a sad watch. His wife wanted to purchase and give him a railroad watch so this was her perception of a railroad watch from an antique dealer. It most definitely is not a railroad watch and it is seen better days of its life. The roller jewel had been either replaced or it was in the presentation in a case it looked burnt there were so many things wrong with this watch it would need a major restoration to even run right.Once again we can see when you push things the Chinese machine thinks the amplitude is better when it very definitely was not. More than likely there was a power issue through the gear train that's not magnetism or seeing. The watch did come apart it had a lot wrong a quick servicing did not fix the problems and is in see it looks that will better afterwards but not much.

If you like to see a comparison with the 1900 I can do that much much later today.

 

[Endeavor]

Excellent John, thank you very much for showing and shearing your excellent comparison again !!

I remember after seeing this comparison and struggling with my W.O.S., that the purchase of the Weishi-1000 was a no-brainer. It (still) seems to work very well and so far I had never any regrets, so thanks again !

[jdm]

22 hours ago, JohnR725 said:

So what if you toss the PC and keep the microphone and throw in some software? Then you get something else like the ChronoMaster And the ChronoMaster  pro-

I suppose the maker has embedded a good copy protection scheme in the USB portion of that microphone, otherwise it would be too easy for the Chinese to copy this hardware, ship it with some lousy PC software, but with the knowledge that it would work fine with Witschi software.