EEVBlog 121GW Discussion thread

[amspire]

It has been found that the switch contact issues were ultimately caused by a slight PCB thickness difference between the pre-production and production boards.
Pre-production boards were on the high side of the tolerance allowance, and production boards on the low side. Actually, the production board thickness was actually slightly outside of UEi's board spec.
A 0.95mm shim (as has been found by others) is the fix for existing units and current stock boards (i.e. all Kickstarter ones and for sometime afterwards). Future volume production boards shouldn't need a shim however.
Shims are on the way to me now for cycle testing.

Does "the low side" mean that the board is too thin? Is the 0.95mm shim meant to be attached to the circuit board? Just confused a bit.

[EEVblog]

It has been found that the switch contact issues were ultimately caused by a slight PCB thickness difference between the pre-production and production boards.
Pre-production boards were on the high side of the tolerance allowance, and production boards on the low side. Actually, the production board thickness was actually slightly outside of UEi's board spec.
A 0.95mm shim (as has been found by others) is the fix for existing units and current stock boards (i.e. all Kickstarter ones and for sometime afterwards). Future volume production boards shouldn't need a shim however.
Shims are on the way to me now for cycle testing.

Does "the low side" mean that the board is too thin? Is the 0.95mm shim meant to be attached to the circuit board? Just confused a bit.

Yes, on the thin side.
The shim goes between the plastic range switch and the case, as per how others have done their own DIY solution.
For those that want to do it themselves, the shim thickness is 0.95mm.

[amspire]

Does "the low side" mean that the board is too thin? Is the 0.95mm shim meant to be attached to the circuit board? Just confused a bit.

Yes, on the thin side.
The shim goes between the plastic range switch and the case, as per how others have done their own DIY solution.
For those that want to do it themselves, the shim thickness is 0.95mm.

So that means that the fact the board is too thin is not being addressed unless the shim is causing the shaft clip to press down on the switch rotor. It just seems to me that if the board it too thin, there should be a shim under the rotor clip.

[1anX]

Does "the low side" mean that the board is too thin? Is the 0.95mm shim meant to be attached to the circuit board? Just confused a bit.

Yes, on the thin side.
The shim goes between the plastic range switch and the case, as per how others have done their own DIY solution.
For those that want to do it themselves, the shim thickness is 0.95mm.

So that means that the fact the board is too thin is not being addressed unless the shim is causing the shaft clip to press down on the switch rotor. It just seems to me that if the board it too thin, there should be a shim under the rotor clip.

Like this picture from forum member lowimpedance previous post?

[amspire]

Like this picture from forum member lowimpedance previous post?

I think that is the fix Dave is talking about. I am just asking if there should be a shim on the plastic clip on the actual switch rotor on the PCB and not just one on the shaft.

That shim in the photo is certain a way to reduce the wobble in the switch knob which is a second problem.

[ChrisG]

Like this picture from forum member lowimpedance previous post?

I think that is the fix Dave is talking about. I am just asking if there should be a shim on the plastic clip on the actual switch rotor on the PCB and not just one on the shaft.

That shim in the photo is certain a way to reduce the wobble in the switch knob which is a second problem.

Agree with you. I think the range switch, the plastic knob, could be redesgined a bit so that a wobble is less likely to occur. Let's wait for final testing by Dave and team for a prolonged period 50.000 cycles or more if this is the solution. Nonetheless happy to read that a difference and potential root cause has been identified.

[EEVblog]

Like this picture from forum member lowimpedance previous post?

I think that is the fix Dave is talking about. I am just asking if there should be a shim on the plastic clip on the actual switch rotor on the PCB and not just one on the shaft.

That shim in the photo is certain a way to reduce the wobble in the switch knob which is a second problem.

Agree with you. I think the range switch, the plastic knob, could be redesgined a bit so that a wobble is less likely to occur. Let's wait for final testing by Dave and team for a prolonged period 50.000 cycles or more if this is the solution. Nonetheless happy to read that a difference and potential root cause has been identified.

The switch wobble is separate problem that is fixed with a slightly redesigned white plastic indent ring. Backers with existing meters will be shipped a new knob, plastic indent ring, and a shim.

[amspire]

Can someone please measure the input impedance when the mVDC is selected.

It is about 9.9M.

[1anX]

Dave commented much earlier on the probability of Murphy playing a role in this meter's fate. The comment that comes to mind ran along the lines of delivery dates could be impacted by such things as the USA coming into conflict with North Korea.

Hopefully UEi can deliver before this comes into operation and hopefully diplomacy will find a way. https://www.washingtonpost.com/opinions/trumps-team-is-floating-an-attack-on-north-korea-americans-would-die/2018/02/27/8e6cdf66-1826-11e8-92c9-376b4fe57ff7_story.html?utm_term=.5808f5ad42ac

[Bud]

Stay away from politics.

[JonM]

I was away and missed the live stream. Upon returning last night I saw the pop-up notification which took me to a recording of the live stream. David2's "plot" was there, shortly before Dave has to leave he holds a tablet up to the camera. First was pass/fail data, then a bit later "raw" data. I think that there was some good explanation of why some data are bad. In the reproducible research world (if making a document), it is best to show two plots with one including the "bad" data and the explanation of why it is bad. You don't want the scale set so that "real" variation cannot be seen. Of course, I don't think that they are planning a document on the testing.

[bitwelder]

Any plans regarding sigrok support for this meter?

[Tinkerlad]

Any plans regarding sigrok support for this meter?

Once I get my hands on one I will be looking at adding sigrok support if someone hasnt already. Im also very interested in seeing what sort of data throughput the BT module can get and if I can get a low data rate stream over Bluetooth into sigrok lol.

[EEVblog]

If you go back now and look, it appears the data was removed.  I watched it after it was live and assume it would have had to been cut and reloaded.

No, that's not how Youtube live streams work. It records the whole thing and makes it available afterwards. I did not edit it or even upload it.

Strange as I don't remember it being anything that I considered that out of the ordinary.  They were messing with the jig a fair amount and maybe just want to show a clean data set.  It sounded like they may run a few other meters as well.

No, we only ran the one meter.

[1anX]

That's odd.  I know I saw the plots the first time I watched the videos. Maybe I am just not able to find it.

Hey Joe, are you going to be getting a production run meter from the next batch?

The first release meters even to backers were really prototypes, so the next batch will really be the first production run released having undergone substantial testing.

I'm hoping to see you test the new updated meter and look forward to your unbiased results!

[IanB]

I really have not seen a lot of testing going of of these first meters, not like I had hoped for anyway.   It's premature to think about it.

I think the timing of the Kickstarter announcement made it most accessible to people in the USA (it was in the middle of the night in Europe), so many (even most?) orders came from here. And as you know, the USA batch has not been delivered yet. I think you can expect to see much more discussion and testing once the USA meters get into people's hands.

[EEVblog]

Results from the 50,000 cycles with the shim solution

[TheSteve]

Very cool! The results are pretty darn good for 50K cycles.

Going to have a giveaway for the highest use 121GW in existence?

[EEVblog]

Going to have a giveaway for the highest use 121GW in existence?

Fraid not, we only have two beat up meters in the lab.

[amspire]

If you put some thin Kapton tape on both sides of the spacer and heat it to set the glue, it would probably stop the abrasion.

The normal tape is HN Kapton. There are non-Dupont FEP coated (teflon) polyimide tapes available. I just do not know if the Kapton FN (FEP coated) film is available as a tape.

Two layers of Kapton tape varies in total thickness between 0.13 mm for the 1mil tape to 0.23mm for the 3mil tape. Not sure if 0.13mm is too much or not - may have to thin the spacer a little. If you only add the tape between the spacer and the case, the thickness can drop to 0.065mm.

I have also mentioned before the method I used with polyimide tape around the outside of the knob to reduce clearance between the knob and the case so it cannot wobble much. Seems to work great. You can reverse the camber of the knob edges to match the case.

[EEVblog]

UPDATE: We completely screwed the pooch! We installed the shim upside down!
There was supposed to be no contact friction between the shim and the plastic case (which created all the dust). Apparently we were too dumb to read UEI's instructions precisely. Doh.? 

[amspire]

UPDATE: We completely screwed the pooch! We installed the shim upside down!
There was supposed to be no contact friction between the shim and the plastic case (which created all the dust). Apparently we were too dumb to read UEI's instructions precisely. Doh.? 

Why have they routed the recess then? It isn't meant to fit between the case and the clip, is it? Otherwise it seems they have thinned the part of the spacer taking all the force.

[Brumby]

Why have they routed the recess then?

That's the question that came to my mind as well.

[EEVblog]

UPDATE: We completely screwed the pooch! We installed the shim upside down!
There was supposed to be no contact friction between the shim and the plastic case (which created all the dust). Apparently we were too dumb to read UEI's instructions precisely. Doh.? 

Why have they routed the recess then? It isn't meant to fit between the case and the clip, is it? Otherwise it seems they have thinned the part of the spacer taking all the force.

The recess fits over the protruded tab on the actual top of the range switch, I showed this on the latest live feed.
The circlip actually resides flush with the case as it's in a concave part of the case molding.

[1anX]

The recess fits over the protruded tab on the actual top of the range switch, I showed this on the latest live feed.
The circlip actually resides flush with the case as it's in a concave part of the case molding.

How about a link Dave!
I couldn't find the live feed where you show the correct position of the spacer.