EEVBlog 121GW Discussion thread

[CDaniel]

This issues are known for a long time if you read the posts .
Autorange is slow .
50ohm range may be redundant because they did it wrong . The protection thermistor PTC3 little drift with temperature is responsible for getting different results all the time . Maybe there is some self heating from the measurement current itself , not just the internal temperature variation .  2 wire low ohm measurement is not that simple to do it right .

Later edit

I played a little with the board out and ... it's not the thermistor heating up ... The problem is that the measurement current is low ( maybe too low )  and it's using the lowest range possible with 10 times internal amplification ( in HY3131 configuration is like that ) . Every solder joint is a thermocouple at this  low level . That's why it drifts all the time and can't be stable .

With  the thermistors shunted , any solder joint I heat up ( in input zone ) results in massive 1000 counts drift . Thats 1 ohm . In other ranges the drift is lower , as expected . What's impossible to say at this point is if they use HY3131 correctly in 50ohm range as in configuration pdf document .

[EEVblog]

The question is are the new batch of meters any better than the kickstarter meter?
Is the new meter still using the spacer?
Has the new meter any new components and specification?
Does the new meter still have the existing firmware bugs?
Dave is there any difference with the new over the old?

The latest meters shipping have two hardware changes. A copper shield, and a two transistor daughter board to replace the TVS.
They still use the spacer shims until existing blank PCB stock is used up.
Firmware is the current release 1.57
Specs have not changed.

[IanB]

The latest meters shipping have two hardware changes. A copper shield, and a two transistor daughter board to replace the TVS.

Can we see pictures, especially of the shield?

These seem like things that might be relevant retro-fits.

[EEVblog]

The latest meters shipping have two hardware changes. A copper shield, and a two transistor daughter board to replace the TVS.

Can we see pictures, especially of the shield?
These seem like things that might be relevant retro-fits.

Attached.
The shield is a multi-layer plastic sandwich construction. You could DIY, but watch for shorts and clearances etc. Connected to the negative battery clip on the other side.
Again, this will only be for existing stock of blank PCBs.

I don't believe the DIY transistor mod is recommended because it likely requires recalibration.

[joeqsmith]

Sure, there's some new shield and clamping...   but what's this I see!!!? 

Is that a TI CD4053B with an absolute maximum Vsupply of 20V??  The same part I have said several times that I went to in an attempt to get the meter to survive some basic transients???!!!!   Looks like someone is trying to spoil my fun.    If they get the clamping right, this meter should do just fine. 

Looking forward to seeing the meter finished up.   

[EEVblog]

Is that a TI CD4053B with an absolute maximum Vsupply of 20V??

Yes, sorry, forgot about that change.

[1anX]

Sure, there's some new shield and clamping...   but what's this I see!!!? 

Is that a TI CD4053B with an absolute maximum Vsupply of 20V??  The same part I have said several times that I went to in an attempt to get the meter to survive some basic transients???!!!!   Looks like someone is trying to spoil my fun.    If they get the clamping right, this meter should do just fine. 

Looking forward to seeing the meter finished up.   

Joe, does changing to this spec chip alter calibration?
What "basic transients" or voltage levels did you apply during your testing which destroyed the 4053?

[EEVblog]

Sure, there's some new shield and clamping...   but what's this I see!!!? 

Is that a TI CD4053B with an absolute maximum Vsupply of 20V??  The same part I have said several times that I went to in an attempt to get the meter to survive some basic transients???!!!!   Looks like someone is trying to spoil my fun.    If they get the clamping right, this meter should do just fine. 

Looking forward to seeing the meter finished up.   

Joe, does changing to this spec chip alter calibration?

No, it doesn't. It's a DIY mod if you are really that keen, I personally would not bother.

What "basic transients" or voltage levels did you apply during your testing which destroyed the 4053?

I can't speak for Joe's tests, but FYI the previous chip passed all UL CAT III standard tests just fine.

[1anX]

Thanks for the info Dave!

I have bought the BM235 and the 121GW from you and have given the BM235 to my son for his automotive work. I'm looking to buy another 121GW type spec meter and have been considering the Brymen 869. How long before the next gen 121GW becomes available with the new PCB and updates?

I'm quite happy to hardware mod my current 121GW and would love to see a video on EEVBlog on the modding and re-calibrating this meter. I'm sure plenty of others with the kickstarter meter would also like to mod and update this meter with a instructional video from you!

Still need to buy another meter, but making the current kickstarter meter a project would be fun and educational.

[joeqsmith]

Changing to the CD4053B in of itself would do nothing to improve the meters robustness.   We are talking about a couple of volts vs thousands of volts we are applying.    The problem is coming up with a clamp that is not set so low that it interferes with the high voltage supply used for the diode check.  Yet it can't be so high that it allows us to exceed the absolute Vs rating of the mux.   You have a 15-16V supply and an IC rated for 18V (I think) absolute max.   Increasing this margin a couple of volts eased the effort of sorting out the clamp.   

I am hoping that what led them to use the TI part was to try to improve the meter's robustness.  I won't know how it behaves until I run one. 

The video for the prototype is still on-line.  It was first damaged with a 2KV peak, 100us FWHH with a 2ohm source.   There were a lot of low cost meters that were damaged at these low levels.    For those wanting to point out the 87V as a gold standard, make sure you check all the facts first.   Again, the data is here:

https://docs.google.com/spreadsheets/d/1cXzYpIoyVm9QJUju4KXqM22CEQZP3_xwWvDyeVwxTy4/edit#gid=400910915

Video of the 121 proto can be found here:
https://youtu.be/X28bwdTBW8g?list=PLZSS2ajxhiQDBDdtQNjVnGxShaVQ3nUMY&t=1444

[ChunkyPastaSauce]

The latest meters shipping have two hardware changes. A copper shield, and a two transistor daughter board to replace the TVS.

Can we see pictures, especially of the shield?
These seem like things that might be relevant retro-fits.

Attached.
The shield is a multi-layer plastic sandwich construction. You could DIY, but watch for shorts and clearances etc. Connected to the negative battery clip on the other side.
Again, this will only be for existing stock of blank PCBs.

The shield - a fix for the jumping values when interacting with the membrane buttons?

[EEVblog]

How long before the next gen 121GW becomes available with the new PCB and updates?

Well past this batch of 700, they have a lot of blank boards yet. Can't give a timeline.

I'm quite happy to hardware mod my current 121GW and would love to see a video on EEVBlog on the modding and re-calibrating this meter.

Most people do not have the capability to calibrate the meter, it requires precise generated values from a proper meter calibrator.

[EEVblog]

The shield - a fix for the jumping values when interacting with the membrane buttons?

Correct.

[GeoffreyF]

Is there a suggested process to make (or purchase from EEVBlog) a shield in earlier models?

[EEVblog]

Is there a suggested process to make (or purchase from EEVBlog) a shield in earlier models?

I could ask Kane to send me a whole bunch I guess for those that want them.

[1anX]

Is there a suggested process to make (or purchase from EEVBlog) a shield in earlier models?

I could ask Kane to send me a whole bunch I guess for those that want them.

Dave, start a list of those wanting the button shield and add me to it!
I would appreciate the added stability it would bring to the meter.

[darik]

Is there a suggested process to make (or purchase from EEVBlog) a shield in earlier models?

I could ask Kane to send me a whole bunch I guess for those that want them.

Dave, start a list of those wanting the button shield and add me to it!
I would appreciate the added stability it would bring to the meter.

I'm interested.

[TikhonovC]

Is there a suggested process to make (or purchase from EEVBlog) a shield in earlier models?

I could ask Kane to send me a whole bunch I guess for those that want them.

And a two transistor daughter board to replace the TVS.

[lowimpedance]

Is there a suggested process to make (or purchase from EEVBlog) a shield in earlier models?

I could ask Kane to send me a whole bunch I guess for those that want them.

And a two transistor daughter board to replace the TVS.

As Dave stated above this would mean a re-calibration which uses equipment that most users would not have ready access to, so is best to leave as is. Of course having the PCB available to those that do have the required gear would be nice too  .

[TikhonovC]

I think that the cost of this board is 2-5 USD.
The cost of the shield is 5-10 USD.
Shipping costs will be at least 10-15 USD.
This PCB can be added to the package. For future calibration. Anyway it can be useful.

[CDaniel]

That's the output protection for resistence / capacitance ranges , so it's not that difficult to recalibrate . Only high resistance ranges and maybe low capacitance range will be affected anyway .

[Brumby]

The comment was not on how easy/difficult recalibration might be - but about the equipment required to do so properly.

[CDaniel]

equipment = some high precision resistors and maybe capacitors

[Brumby]

What?  No high precision voltage standards?

Therein lies the question: what equipment and standards are required?  Do you know?

[CDaniel]

Anybody familiar with the schematic will see from the picture that the replacement 2 transistor board is for output protection in resistace/capacitance mode ... D7-D8 1N4007 . Output = the current source or charge / discharge circuit used for measuring resistors or caps .