English PW Timegrapher question

toshik

Registered User
Sep 6, 2013
37
0
0
PVB
Country
Region
I was always wondering how are my old watches performing and borrowed the 1000 timegrapher from a friend. I'm puzzled by results of testing of 2 watches. Could someone explain what this means? Is there some kind of reference for decoing those results? :)
 

Attachments

Last edited:

gmorse

NAWCC Member
Jan 7, 2011
12,623
2,216
113
Breamore, Hampshire, UK
Country
Region
Hi toshik,

I was always wondering how are my old watches performing and borrowed the 1000 timegrapher from a friend. I'm puzzled by results of testing of 2 watches. Could someone explain what this means? Is there some kind of reference for decoing those results? :)
View attachment Test and Witschi measuring technology mechanical watches.pdf

Although this document, (from the Witschi website), is not surprisingly about their own products, there's also some good generic information about interpreting timing machine traces.

Which features in particular are puzzling you?

Regards,

Graham
 

toshik

Registered User
Sep 6, 2013
37
0
0
PVB
Country
Region
I never said I'm puzzled by fatures, but the RESULTS. This is why I posted the picures of testing of 2 watches. Is there a reference material explaining different displayed results?
In another words whatd these random dots an dotted lines mean?
Your document is in PHP format, do you have anything in PDF?
 
Last edited:

gmorse

NAWCC Member
Jan 7, 2011
12,623
2,216
113
Breamore, Hampshire, UK
Country
Region
Hi toshik,

That document is a PDF, (or at least it was when I posted it!). An alternative source of information is De Carle's "Practical Watch Repairing", which gives several examples in Chapter 24.

Regards,

Graham
 

watchwldr940

Registered User
Nov 26, 2008
399
20
18
Hello. Maybe I can add to the confusion, or maybe tell you what you already know. The first trace shows a watch that is pretty out of beat and running 4-1/2 minutes slow per day, give or take. The amplitude is shown at 258degrees, which might or might not be correct, depending on the angle of lift that was entered into the machine. The figure of 7.6 ms is how out of beat this watch is. Compare this with the other trace which shows a much better beat error of .1 ms which is great. Notice the rows of dots on the second trace. In this trace, showing a series of lines, there are actually two rows of dots in each line, and the closer these dotted lines are to each other the better the beat is, and a trace showing a "perfect" beat will show up as a single line. Keep in mind, also, that the resolution of the machine will also have a bearing on the appearance of the dotted lines. The second watch is in very good beat, and is also running somewhat slow. It is shown as having an amplitude of 238 deg. which is subject to the same explanation as the other trace-- one must specify the lift angle to let the machine give you an accurate amplitude reading, and there are other threads about how to determine the lift angle. On modern movements this data is usually published somewhere in the manufacturer's literature. Lastly, the 18,000 is the nominal beats per hour value, which is automatically determined by the machine, or this value can also be manually entered. In closing, remember that this represents the performance of this watch only at this particular time of examination; the state of wind, position, if the watch has a calendar or some other complication, and if so, if that complication is engaged, are some other factors that affect the trace given by a watch at any particular time. One more thing-- on a freshly serviced American pocket watch in good condition with a new mainspring, an amplitude of 270 degrees in the dial up or down position, full wind, is considered about right. Hope this answers some of your questions. George
 

toshik

Registered User
Sep 6, 2013
37
0
0
PVB
Country
Region
Thank you! So with the low beat error one (right picture) all I need is to speed it up a bit with the regulator and two lines should become parallel to the long side of the sceen? All I found is that I should test the pocket watch in at least 2 positions: face up and crown up...
Is there any reference material on the subject?
 
Last edited:

watchwldr940

Registered User
Nov 26, 2008
399
20
18
You will want to check it in several positions. Also, the watch should be in good running order and capable of the accuracy you're trying for. There are any number of references on interpreting timing machine traces. If you can get a copy of a manufacturer's instructions that would be helpful
 

toshik

Registered User
Sep 6, 2013
37
0
0
PVB
Country
Region
You will want to check it in several positions. Also, the watch should be in good running order and capable of the accuracy you're trying for. There are any number of references on interpreting timing machine traces. If you can get a copy of a manufacturer's instructions that would be helpful
When I test watch in horizontal position it performs fine as soon as I go vertical crown up hell brakes lose. I mostly work on Hampden 16s.
I think it's timegraher 1000 -there are abosolutey no explanations in its manual...
 

Dr. Jon

Moderator
NAWCC Member
Dec 14, 2001
6,537
931
113
New Hampshire
Country
Region
Your timeogragher looks fine to me. These do not suffer poor performance gracefully. Once a watch is more than few minutes fast, its trace looks crazy. It does look like you got the beat right.

One thing I do is to lower the gain as much as can. I push the down arrow button until just before the trace goes away. That reduces a lot of stray noise pick up.
 

Electro

Registered User
Nov 22, 2011
222
0
0
Alberta Canada
Country
Region
You also need to set the lift angle to get an accurate result. When the lines are diagonal like that, it's a good indication of magnetism.
 
Last edited:

gmorse

NAWCC Member
Jan 7, 2011
12,623
2,216
113
Breamore, Hampshire, UK
Country
Region
Hi toshik,

If you want to get a better understanding of timing machines in general, I suggest that you download the free version of the eTimer software from Delph Electronics. It does everything that the full version, (which comes with a microphone; costs vary with the type of mic), does except working in real time from a microphone, and it has good help screens. You can use it to replay traces which have been recorded using other means, such as Audacity, and it comes with several example traces showing various faults; you can learn a great deal by playing with it. I use it with a home-made piezo microphone recording onto Audacity, and it's far more comprehensive than most of the hardware timers on the market.

Electro is right in so far that the diagonal lines may be caused by magnetism, but the gaining or losing rates that they indicate can be caused by many other factors as well.

Regards,

Graham
 

Al J

Registered User
Jul 21, 2009
679
101
43
Canada
Country
Region
That reduces a lot of stray noise pick up.
These cheaper machines really do pick up a lot of stray noise, and you need to have the room you are in pretty quiet, and turn the gain down as suggested.

Cheers, Al
 

Forum statistics

Threads
167,024
Messages
1,455,483
Members
87,253
Latest member
javalink
Encyclopedia Pages
1,057
Total wiki contributions
2,914
Last edit
E. Howard & Co. by Clint Geller