Chinese Timegraphers

[Jon]

1 minute ago, Kanie said:

And would that be true on all timegrapher results on all movements read?

Yes, on this timegrapher. In the main menu change the accuracy from 999 sec/day to 99 sec/day and you'll understand better. Only set it to 999 sec/day if the movement is wildly out in respect of the rate; so, gaining or losing more than 100 sec/day otherwise the timegrapher won't be able to give you a reading

[Jon]

Just now, Kanie said:

Like this one?

So, at some point the movement was losing 326 sec/day, but the main display hadn't caught up to it yet to display that

[Jon]

As I said, if you set the display in the main menu to display the reading every 2 seconds then your main display will display the accurate reading every 2 seconds than every 40 seconds and really, not displaying an accurate figure, because the rate has changed before the 40 seconds has elapsed, so the timegrapher reads an average

[Kanie]

Ok,understood…I will do the setting change when I get home from work and retest my movements…only had it one day and really having fun!

Jon,thank you so much for your help!!

[WatchingOracle]

On 8/22/2021 at 12:47 AM, Jon said:

Although your main readout says zero seconds a day, probably because in the main menu you have it set to 999 sec/day instead of a more accurate 99 sec/day, those small numbers are telling you that the movement is losing one second per day. In the main menu you can change how often those smaller numbers are displayed, such as every 2 seconds, 4 sec, 10, 20, and so on. I like to set mine to have a readout every 2 seconds, but that is just my preference.

Once you set the main menu to 99 sec/day, the main display that currently say zero sec/day will then become more accurate and reflect those numbers that say -1 sec/day, because at present it is not giving a reading other than a whole number and you want to see 0.1 sec/day (as an example) rather than rounding it to the nearest whole number which would be zero.

I have the same as model as pictured (Weishi 1000), my menu doesn't have this option discussed to change the display refresh rate as these instructions are inferring. Am confused lol.

 

[WatchingOracle]

On 7/25/2021 at 4:50 PM, oldhippy said:

Going off subject. Have you watched Kalle’s video on Ultrasonic cleaners? Many on here need to change how they clean watch movements.

Which video are you referring as I am having trouble finding this mate.

[JohnR725]

On 8/21/2021 at 7:47 AM, Jon said:

Although your main readout says zero seconds a day, probably because in the main menu you have it set to 999 sec/day instead of a more accurate 99 sec/day,

the problem with this suggestion is it only works for the 1900. It is not a feature of the 1000.

[WatchingOracle]

Thanks @JohnR725 that explains it

[oldhippy]

53 minutes ago, WatchingOracle said:

Which video are you referring as I am having trouble finding this mate.

This is the one.

 

[WatchingOracle]

Oh... I was looking in the wrong place LOL.  Thanks!

[JohnR725]

On 8/21/2021 at 7:34 AM, Kanie said:

Hey,guys…Recently bought a Weishi timegrapher and would like to know what the smaller numbers mean that are under the main readout numbers..have searched the net but no luck…thanks in advance.

just so everyone understands what were talking about I snipped out an image from your question and circle the numbers I assume these are the numbers are asking about?

I have a theory about the numbers we can try an experiment. Go into the menu screen go down to the menu item which says test period which usually by default is four seconds change it to eight seconds and see if the numbers change their spacing on the screen.

 

 

[Jon]

23 hours ago, JohnR725 said:

the problem with this suggestion is it only works for the 1900. It is not a feature of the 1000.

Thanks John, I never realised the 1000 model was that limiting with the timing display. Another feature the 1000 doesn't have is different coloured traces for the tic and toc. I always advise people to spend a few quid more on the 1900

[xyzzy]

I think I could add this feature to tg.  What would be the most useful way to generate the average?

I was thinking a simple moving average of the last X seconds of data points.  For example a 10 Hz movement and 5 seconds period, one would take how many s/d off each of the last 50 beats was and average it.  Or, another way to think of that, is you find the actual time between two beats 50 beats apart and divide that by 5 seconds.  Mathematically, both of those will find the same value.

But one could also use an exponential moving average, which would have the effect of weighting the most recent part of the averaging interval move than the least recent part.

[LittleWatchShop]

I found a teardown report on the 1900 and was happily surprised to find that it has one of my chips in it.  The C8051F360 was developed by a company I started in 1999 and later sold to Silicon Labs.  So my opinion of the Weishi has gone up a notch!!  I even own one!

[JohnR725]

18 minutes ago, LittleWatchShop said:

C8051F360

I'm pretty sure if you look somewhere in the discussion group we have a picture of the inside of the machine. Then thanks for pointing out that this isn't an ordinary 8051 I just thought it was ordinary nothing special. Roffealizing that I didn't grasp the 8051 could be enhanced looking at the 1000 machine it's the same chip. then the reason I know that the pictures exist on the group somewhere I photographed both machines.

What makes both machines interesting is a incredible degree of similarity to the point where they were both made by the same company. Very likely designed by the exact same person with the programming being very very similar.

Then isn't it amazing what you can find out there if you look? Somewhere somebody even did a schematic for the microphone minus all the values for the capacitors because of course are surfacemount.

then for those that don't know what enhanced 8051 is snipped out an image. But don't get too excited is no longer in stock in the US for sale as it's obsolete.

[jdm]

What they were thinking? 

Before I file it down, i'm wondering why the plastic jaw is shaped to project into space or floor a mov.t placed on its top. Maybe there is a reason just can't fathom. 

[LittleWatchShop]

42 minutes ago, jdm said:

What they were thinking? 

Before I file it down, i'm wondering why the plastic jaw is shaped to project into space or floor a mov.t placed on its top. Maybe there is a reason just can't fathom. 

Here is mine

[HectorLooi]

17 hours ago, jdm said:

What they were thinking? 

Before I file it down, i'm wondering why the plastic jaw is shaped to project into space or floor a mov.t placed on its top. Maybe there is a reason just can't fathom. 

Yes. Please file it down. And make the offset even deeper. I had an accident while using using my unmodified timegrapher. The jaws slipped while I was regulating the movement and it knocked off 2 dial indices and mangled the logo.

[TimpanogosSlim]

While i still wonder if there's a debug serial port exposed somewhere that might dump cooked data, I'm strongly considering tapping the audio signal that comes in from the mic and connecting it to a 3.5mm jack hacked into the back of the nr. 1000. So i can use a program on my computer to monitor performance over an extended period. 

[xyzzy]

No need to go to that trouble.  There's already posts about how to connect the 1000/1900 microphone to a computer.  You just need the matching connector, a 3.5mm plug, a usb cable for power, and connect them up.

[Kalanag]

The Chinese microphones to be used with a timegrapher software or app get better and cheaper! Here is what I got today. I used the free tg timegrapher software (tg-timer 0.5.0)

Edited January 9, 2023 by Kalanag

[mikepilk]

Just what I've been waiting for. Can you report back when you have some more experience with it, and the software ?

[ManSkirtBrew]

30 minutes ago, mikepilk said:

Just what I've been waiting for. Can you report back when you have some more experience with it, and the software ?

I use one of those microphones and Watch-o-Scope software, and have been very impressed with both.

The microphone does exactly what it's supposed to. You can put it on its sides to get all 6 positions. I use my casing cushion to isolate it from the wood desk, but if you're less lazy you can put rubber feetsies on it.

Watch-o-Scope is super versatile, and the free version is completely functional. The paid version is just nice-to-haves, IMO.

You can calibrate it to your specific setup using a quartz watch (mine needed a 1.5 second/day adjustment), and every parameter is completely variable, which came in handy when I had a watch that ran at 21,306 BPH.

I also really like the fact that you can see the raw waveform of what the microphone is hearing in real time.

[Kalanag]

My version is not a 3D printed „prototype“ anymore but industrially injection moulded.

The software Tg is well known and works perfectly: download page

There are cell phone apps as well!

 

Edited January 9, 2023 by Kalanag

[LittleWatchShop]

36 minutes ago, Kalanag said:

My version is not a 3D printed „prototype“ anymore but industrially injection moulded.

The software Tg is well known and works perfectly: download page

There are cell phone apps as well!

 

Interesting.  It digitizes the signal and sends it over USB.  This is in contrast to the other solutions talked about around here where the signal goes into the audio port.  Cute