Design a New Precision LCR Tweezers with Much Lower Cost 【Shannon Tweezers ST42】

[SilverSolder]

The MS8911 is an amazing product for its price point - it strikes an impressively good balance between quality, features, and price.  They did a nice job.   

They needed to as their previous product was a POS even compared to my now ancient ST3 from 2007.
I've seen a few rave about the later 8911 but maybe they have never used an A brand SMD tweezer. 

Shannon knows what's required and will likely present us with a top class product.

I have never used anything better than the 8911, but I have used worse!

come on, you are wealthy SilverSolder, and we are SnPbSolder

try to change your taste once in a while, the steak is delicious, even the bread could make you full

Back in college, one of the instructors (who I knew was a millionaire, retired, and teaching just because he liked to do that) drove to work in an incredibly beat up old Fiat.  I asked him, "Sir, I know you are a millionaire, why are you driving such a terrible car?" - his answer stuck with me for life:  "Priorities.  I have a crappy car, a big sailboat, a beautiful woman, and I take 6 months vacation a year.  What are your priorities, son?"  -  What can you really say against that?  I tried driving crappy cars for years, but none of the other compensating events ever happened! 

[tautech]

Shannon
Have you decided on a warranty scheme ?

Are you ready to show a simple video of your tweezers in use ?

[Shannon]

Shannon
Have you decided on a warranty scheme ?

Are you ready to show a simple video of your tweezers in use ?

tautech
we don't make the decision at present, we may support 1-year limited warranty.

Due to the menu and UI is still modifying these days, we are plan to show the video early next week.

[toli]

I'm very glad to see this thread progress, and hold my fingers that indeed this will turn out to be a good instrument in both accuracy and user experience. Personally, I think an instrument with good accuracy and UI that has repeatable and reliable results over time is much better than one with excellent accuracy that sacrifices UI or long term stability (like the UT61E which I've measured over time and seen it drift out of spec quite quickly ).

Good job Shannon for the work so far, including being open to so much feedback from the people here on the forum. It can be difficult to come up with something that will make everyone happy (probably impossible ) with everyone pulling to a different direction. So far it seems to be on the right track of finding the common ground without pushing size/price too far.

I will definitely keep watching this thread to see what you come up with and how/when it will be available.

[Shannon]

I'm very glad to see this thread progress, and hold my fingers that indeed this will turn out to be a good instrument in both accuracy and user experience. Personally, I think an instrument with good accuracy and UI that has repeatable and reliable results over time is much better than one with excellent accuracy that sacrifices UI or long term stability (like the UT61E which I've measured over time and seen it drift out of spec quite quickly ).

Good job Shannon for the work so far, including being open to so much feedback from the people here on the forum. It can be difficult to come up with something that will make everyone happy (probably impossible ) with everyone pulling to a different direction. So far it seems to be on the right track of finding the common ground without pushing size/price too far.

I will definitely keep watching this thread to see what you come up with and how/when it will be available.

toli
Thanks for your support

Long-term stability is really a key value for a metrological instrument, such as high precision DMM.
We build the self-calibration and manual zero and short calibration functions for these LCR tweezers, which is very helpful for long-term stability.

UI is under continuous optimization, and a simple video is coming soon, hope you like it

[Shannon]

Below link is the first video presentation, and it's still under continuous optimization. Any comments are welcome.
We use 3D printing material for the shell for this prototype, so it's white. After mass production, it will be changed to black or other colors ABS.

[tautech]

Looking good ! 

Can you make it sense a diode in Auto mode so to not have to select Diode mode ?
There also seems room on the display to place a Vf diode measurement, can we do that with the current levels of excitation ?

[SilverSolder]

Looking good, you are the artists! 

I only noticed one bug:  "Cancel" only has one L  (not CANCELL, but CANCEL)

Too early to demo the Beep on Capture in the Hold function?

[tautech]

BTW Shannon your efforts to develop a new SMD tweezer for the marketplace has also been noticed here:
https://forum.allaboutcircuits.com/threads/lcr-pro1-plus.160725/post-1603452

[kripton2035]

nice !
just the esr values when you measure capacitors seems odd to me ?

[indman]

Del

[Shannon]

Looking good ! 

Can you make it sense a diode in Auto mode so to not have to select Diode mode ?

There also seems room on the display to place a Vf diode measurement, can we do that with the current levels of excitation ?

Thanks tautech

we really want to implement this feature, but to be honest, it's another feature need balanced.
sometimes, we would use these LCR tweezers to measure PCBA components, such as the cap on the power rail, which would git rid of the influence of the internal equivalent diode of IC. So if auto mode select the diode mode, the result would be the diode forward voltage

the test method for the diode is a DC current excitation, in principle, we could use DC current to test the DUT and justify it's a diode or not, and if not, we use AC excitation signal to test the L/C/R.
but there are some bad case need to process
1. A diode parallels with a large-cap, it would cost much time to charge this cap firstly, and then to justify this is a diode or not.
2. if the DUT is an MLCC, the DC voltage would have an influence on the cap value
3. High forward voltage diode such as LED would lead to more uncertainty for the test mode.

at least, we will try to optimize the hotkey sequence to make diode mode is easy to be selected


Sure, we will add the forward current information for the diode test

[Shannon]

Looking good, you are the artists! 

I only noticed one bug:  "Cancel" only has one L  (not CANCELL, but CANCEL)

Too early to demo the Beep on Capture in the Hold function?

Thanks SilverSolder

we will modify the CANCELL definitlly.

For the beep on capture in the hold function is too early, I'm a little confused about your comments.

We have implemented a simple hold function at present, the value would hold when the tips leaving DUT as the video shows, and it shows the stable value before leaving DUT. Test speed of these tweezers is much faster than the other, so the user experience is not bad.
more smart processing method as you said, we have recorded this function and would consider about that

[Shannon]

nice !
just the esr values when you measure capacitors seems odd to me ?

Thanks for your thumbs up. 

You mean the Rs or the Rp value is not stable or not correct, when measuring the caps? could you give me more details, I will try to analyze the result.

One screenshot attached, the test mode is 10nF cap serial 10k res, which shows the high accurate test result.

[Shannon]

Thanks you for the video!
I have some questions and suggestions:
1. By measuring the resistances, everything is fine, only you did not show the closure of the test contacts to each other and the stability of the zero readings on the display? I would also like to see the calibration process?
2. I showed on the photo 1 place where the unit of measurement of reactance is indicated not completely "m", "G". This is clear to me, but may not be very clear to another user. I understand that there is not enough space on the display to accommodate all the information, so I suggest slightly reducing the size of the beep icon. You can simply display the speaker icon when the sound is on and the strikethrough icon when the sound is disabled.
3. I showed in photo 2 the moment that when measuring a capacitance of less than 1 μF, it is desirable to output not Rs, but an important secondary parameter D. The Rs parameter is usually shown when measuring a capacitance of more than 1 μF.
4. I showed in photo 3 that when measuring the inductance, they show the secondary parameter Q if the Q-factor of the coil is more than 1 and show the parameter Rs if the Q-factor of the coil is <1. I am also sure that the test frequency of 10 kHz will not be enough for stable measurement of small values ​​of inductance. This will require a frequency of 50-100 kHz.
This is just my opinion, I do not insist on 100% certainty.
[/quote]
Thanks a lot for your suggestion!

1. Cal process support selfcal/open/short calibration, as the first attached
2. Thanks a lot for your suggestion, we will try to modify that
3/4. D/Q mode is supported, which need go to measure menu and select R mode or Q/D mode, as 2 and 3 attached, 50k-100kHz is our next goal, which needs more support, we focus on the 10kHz on this version

[indman]

Del

[SilverSolder]

Looking good, you are the artists! 

I only noticed one bug:  "Cancel" only has one L  (not CANCELL, but CANCEL)

Too early to demo the Beep on Capture in the Hold function?

Thanks SilverSolder

we will modify the CANCELL definitlly.

For the beep on capture in the hold function is too early, I'm a little confused about your comments.

We have implemented a simple hold function at present, the value would hold when the tips leaving DUT as the video shows, and it shows the stable value before leaving DUT. Test speed of these tweezers is much faster than the other, so the user experience is not bad.
more smart processing method as you said, we have recorded this function and would consider about that

I am not one to make the "perfect" the enemy of the "good",  the hold function is already much better than anything else I've seen.

Given how the readings can wander up and down while probing, it may be better the way you have done it (i.e. leave it to the user to determine when a reading is valid, by "letting go" of the component)

Keep up the nice work, this is looking good, including the work you have done on making the tweezer stable and accurate mechanically -  I agree, the cheaper testers can be a little difficult to use for manipulating small parts (even if they work fine on component that are soldered in a circuit).

[Shannon]

Looking good, you are the artists! 

I only noticed one bug:  "Cancel" only has one L  (not CANCELL, but CANCEL)

Too early to demo the Beep on Capture in the Hold function?

Thanks SilverSolder

we will modify the CANCELL definitlly.

For the beep on capture in the hold function is too early, I'm a little confused about your comments.

We have implemented a simple hold function at present, the value would hold when the tips leaving DUT as the video shows, and it shows the stable value before leaving DUT. Test speed of these tweezers is much faster than the other, so the user experience is not bad.
more smart processing method as you said, we have recorded this function and would consider about that

I am not one to make the "perfect" the enemy of the "good",  the hold function is already much better than anything else I've seen.

Given how the readings can wander up and down while probing, it may be better the way you have done it (i.e. leave it to the user to determine when a reading is valid, by "letting go" of the component)

Keep up the nice work, this is looking good, including the work you have done on making the tweezer stable and accurate mechanically -  I agree, the cheaper testers can be a little difficult to use for manipulating small parts (even if they work fine on component that are soldered in a circuit).

Thank you very much for your encouragement

[Shannon]

3/4. D/Q mode is supported, which need go to measure menu and select R mode or Q/D mode, as 2 and 3 attached

@Shannon, I would like to see these parameters not through menu selection, but in automatic mode, which you demonstrated.

THANKS A LOT FOR YOUR REMIND.

[indman]

Del

[Shannon]

Shannon [/ b], I'm sure you will show all these modes in the next video?

HaHa, not yet, sorry about that, a new video is ready

Cap Measurement Noise performance video

The screen measurement data update rate is 2 samples per second, which is much faster than the other high precision SMD testers, and there are about 100 samples record for 5pF cap measurement, 6fF peak to peak noise performance

[indman]

Del

[SilverSolder]

Can we get the measuring voltage down to 100mV?   That is a Yuuuuuge feature of the Mastech, that makes it possible to measure things in-circuit without triggering any silicon junctions.

[tautech]

Looking good ! 

Can you make it sense a diode in Auto mode so to not have to select Diode mode ?

There also seems room on the display to place a Vf diode measurement, can we do that with the current levels of excitation ?

Thanks tautech

we really want to implement this feature, but to be honest, it's another feature need balanced.
sometimes, we would use these LCR tweezers to measure PCBA components, such as the cap on the power rail, which would git rid of the influence of the internal equivalent diode of IC. So if auto mode select the diode mode, the result would be the diode forward voltage

the test method for the diode is a DC current excitation, in principle, we could use DC current to test the DUT and justify it's a diode or not, and if not, we use AC excitation signal to test the L/C/R.
but there are some bad case need to process
1. A diode parallels with a large-cap, it would cost much time to charge this cap firstly, and then to justify this is a diode or not.
2. if the DUT is an MLCC, the DC voltage would have an influence on the cap value
3. High forward voltage diode such as LED would lead to more uncertainty for the test mode.

at least, we will try to optimize the hotkey sequence to make diode mode is easy to be selected


Sure, we will add the forward current information for the diode test

Yes some things are difficult in Auto mode......however if we need more detail we can use a Diode mode.
I look forward to your solutions. 

[Shannon]

Can we get the measuring voltage down to 100mV?   That is a Yuuuuuge feature of the Mastech, that makes it possible to measure things in-circuit without triggering any silicon junctions.

We think so, and try to add this option
the spec announcement is a problem for this option in the datasheet

BTW, I found the number of visits to the video is quite small, do you have any idea about how to make more guys know about this design