EEVBLOG/UEI 121GW Part 1, Functional Test and Teardown

It's been over two years since I looked at the 121GW prototype. I wanted to allow Dave and UEI plenty of time to address the problems before reviewing it.
In Part 1, I perform a basic functional test of the meter and have a quick look inside.
Feel free to join the thread:
www.eevblog.com...
An old video explaining the transients I apply:
• Programmable Transient...
The data for the meters I have tested may be found here: docs.google.co...
Answers to Frequently Asked Questions may be found here: docs.google.co...

[@joehubler4965]

Hey Joe, the new firmware 2.02 fixes some of the issues you saw

[@Todestelzer]

Joe Hubler yeah. Isn’t he active in the EEV forum I wonder? And there is a video up too on the eev KZbin Chanel addressing it.

[@joesmith-je3tq]

Joe, I pinned your comment. If you follow my posts, you may be aware that I had spent a fair amount of time testing various versions of the firmware. My goal with this first part was only to perform an initial check of the meter to make sure they were functional. I will more than likely devote one full segment to the firmware but for now my focus will remain on the hardware. Like one of the other viewers, I was a bit surprised that the two meters were not shipped with the latest firmware. If you are interested in seeing some of the tests I had performed with the firmware, you can check out the following videos: kzbin.info/www/bejne/iIjUgWOdgsSEnJI kzbin.info/www/bejne/Z2TRp2CEqLaLetU

[@joehubler4965]

@@joesmith-je3tq keep up the great work, I have that meter and had similar problems until I up dated, seems good now.

[@joesmith-je3tq]

@@joehubler4965 I'm guessing you didn't take the time to watch that first video where I demonstrate the increased noise levels with each new release of firmware. In the threads, I had posted some images of what the roll off looks like (my software can automatically test that). The AC line rejection was also changing. I suspect these changes were being made in an attempt to improve the settling time at the expense of noise. Personally, I will take slow and accurate any day over rapid inaccurate measurements. But as I said, I will devote a episode to the firmware. I gave up chasing it at version 1.57 so it will be interesting to see what has changed. One thing is certain, the firmware is not going to correct that poor switch design and it's not going to improve the alignment.

[@TylrVncnt]

joe smith - definitely v2 (“2.02” now) is muchhh much better to use from a functionality/speed standpoint, so it would be greatly insightful for myself (and I’m sure many other viewers!) if you ran similar noise tests etc. on the v2+ firmware! From my anecdotal “user perspective”, it really feels like a completely new meter after I updated (mine shipped with v1.57) Cheers!

[@isoguy.]

You are one in a million Sir, A person who isn't in it for the money. People are remembered for their achievements and what they pass onto the next generation. Thank you for taking your time to make these vids.

[@javierpallalorden]

Joe, thank you for keeping your channel ad-free!

[@cygnuselectrical]

Thank you for taking the time to make this video, it has been very helpful.

[@joesmith-je3tq]

There's plenty more to come. We are just getting started.

[@georgegherghinescu]

Great to see some multimeter action again, cant wait to see the next episodes :)

[@chrisblank9090]

this is indeed the FAIREST UNBIASED and UNFAVORED REVIEW/TEST channel here in youtube. kudos mr smith.

[@joesmith-je3tq]

After running so many handheld meters, I certainly have my favorite brands based on the data I have collected. Using a standard test setups removes the bias. You basically get to see how the brands compare no matter if it's a $3 or a $300 meter. The transient generators don't care.

[@Chris_Grossman]

I too have had issues with sub 1000pF capacitance measurements with my 121GW. I have not tried it latest firmware yet since I have many alternatives for measuring components. I truly appreciate your reviews and careful measurements and characterizations you do. I know how much work you put into them and I have learned quite a bit about input protection circuits from watching them. Thank you!

[@joesmith-je3tq]

Sorry to hear that about the lower capacitance readings. I wonder if its a problem specific to a lot or if all of them exhibit this problem. I was actually surprised the uA was so far out on both as well. Normally I don't see these kinds of problems. Glad you are enjoying the videos.

[@KamotsTalksTech]

@@joesmith-je3tq I had similar issues with mine prior to updating, but I intentionally didn't update when I got it just to see the improvements when I did. I expect any other similar issues can also be fixed via firmware. Looking forward to your more destructive tests since any hardware design issues identified obviously couldn't be fixed in firmware. Hopefully it performs well.

[@chronos4244]

​@@joesmith-je3tq Hi Joe, this is the downside of the low burden voltage. The meter shows a high offset with no current, I suppose this is a thermal voltage issue. First null it out and try again.

[@joesmith-je3tq]

@@chronos4244 I assume a joke as most would plainly see that the meter is reading high. Zeroing out the reading is going to bring the actual measurement even lower. It's just not a very tight alignment. Of course, if we have a thermal issue, that Brymen that was placed in series was doing a damn good job compensating for it. lol.

[@pa4tim]

bought the 869s 2 years ago after your tests and use it professional for electronics repair and I love it.

[@joesmith-je3tq]

Hey Tim, it's still my favorite handheld meter of all time.

[@abeleski]

Same although i despise the selector switch. Too small to grip properly

[@pa4tim]

@@abeleski Big hands ? :-) I have no problem with the switch

[@abeleski]

@@pa4tim yeah mate. My big hands make EVERYTHING look small 😉🤣

[@pa4tim]

@@joesmith-je3tq Mine too (handhelds wise, for benchmeters it is my Keithley 7510, I use that every day too. That thing is ultra fast, has almost every function you can think of and some almost insane ranges. Besides that it is so easy to use. I repair for most calibration and measurement gear, besides that often pretty old industrial stuff that needs to be repaired because otherwise they often need to replace very expensive, often custom build, hardware. ( If I repair something interesting I make a webpage about it on my site. www.schneiderelectronicsrepair.nl )

[@cthoadmin7458]

You are sir, the Ralph Nader of multimeters. Very well done.

[@hdmalpas]

Nice to see someone who is doing it not for reward. Nice one Joe.

[@joesmith-je3tq]

I do find looking at the various designs rewarding.

[@hdmalpas]

@@joesmith-je3tq I have just bought a electronic microscope with stand for working on SMD, I was very impressed with it, I got it from Banggoods. cost only £30 about $35. Worth a look. At 68 my eyes as good as they use to be.

[@joesmith-je3tq]

@@hdmalpas It's rare I will solder without a microscope anymore. Mines made from used parts. My stand is partly custom made. It morphs from time to time. This is basically what it looks like today: www.eevblog.com/forum/testgear/(cheap)-lcd-microscope-vs-(cheap)-optical-stereo-microscope/msg2373153/#msg2373153

[@reedreamer9518]

I love the way the EEVBlog meter looks and is laid out, but when I needed to upgrade my meter it was still under development - so I bought the Brymen BM869s instead. No buyer's remorse - and I love how "Joe" always pulls out the Bryman when he needs a reliable reference measurement. ; )

[@JohnAudioTech]

My GW121 refused to measure a 47 uf cap until I updated to firmware 2.02.

[@joesmith-je3tq]

That's pretty bad.

[@tomgeorge3726]

Hi Joe, can you put some identifying tape on the three DMMs please so we can easily follow your proceedure.

[@joesmith-je3tq]

My wife made the same suggestion. They are all now clearly marked.

[@cannesahs]

I like to see this video repeated after updating fw. Im still sceptic about this meter.

[@joesmith-je3tq]

That's no problem. I will more than likely do a lot more tests with the firmware than what was shown in this video. Stay tuned.

[@SIXSTRING63]

Wow! was surprised to see these production meters were as inaccurate as they were. I was hoping to see a big improvement. Seems like the contrast is better on your prototype version. Brymen 869 smokes it in speed of readings and accuracy. Your prototype performance is much better after your mods and alignments. Can't wait to see the outcome after testing is complete.

[@joesmith-je3tq]

We will get a better idea how the production meters perform as I start to look at them in more detail. I would have liked to have seen that shim addressed with their thicker PCB (according to Dave). After comparing the old and new, where the two switches appear to have the same stackup, I'm concerned that they really don't understand the problem. Oh well. We have seen some meters ground to dust during that 50,000 cycle test. This will be very telling on if they still have a problem or not.

[@SIXSTRING63]

joe smith Just my observation of this meter and it’s continued issues is it went into to production too quick without enough prototype time to clear the glitches. Seems like once the ball started rolling and tooling and fixtures were up and running someone in management didn’t want to revamp no matter the bugs. I know how that can be after being a programmer of CNC machines making complex parts. Some jobs would take days or weeks to fixture up and get running at the most cost effective production. Sometimes management would be on me for the lengthy setup, my response was it takes time to save time. Choosing the right tooling for the job can make or break it. One job I had to hard turn H13 toolsteel at 58 Rockwell C, my predecessor was using ceramic tooling that would need changing after two parts. I switched to a more expensive CBN insert and got 10 parts, the cost saved in less down time in tooling changes and scrap dropped dramatically. In the end it was always more profitable. They still use my programs and setups 6 years since I retired. I still live by the words an old toolmaker in his 60’s taught me when I was an apprentice, the more hurried I get, the more behind I get. He was exactly correct.

[@joesmith-je3tq]

@@SIXSTRING63 Does the name carry more weight than the quality of the product? I don't care about the name on the products I look at but how they perform. That said, with having ran so many products, I can pretty much guess what ones will perform poorly based on the brand. This is really the first UEI product I have looked at. Outside of the prototype, I really have nothing else to go by. I would have guessed they went out of business based on their website. Growing up, my mom had embroidered and framed a sign with "The hurrier I go, the behinder I get" to which she would continually point at. Words I also still live by.

[@SIXSTRING63]

joe smith That is very cool your mom used that same quote. Her wording was exactly the way he said it also, I just wrote it out a little different. I wish the younger generation lived by or even knew of this thought process. Yes, agreed on depending on the name for quality means nothing if it doesn’t put the rubber to the road. I bought all of my Brymen meters on your tests and videos alone. I have no complaints at all so far, they are spot on and very responsive and stable. I showed the 869s to a good friend of mine who is a retired EE. He was very impressed with the features and the responsiveness of it. He said if he buys a new meter that will be the one. He has an old Fluke 87 IV and the old side push button model. He’s a musician like myself and plays keys, rebuilds keyboards for everybody in my area. I send all that repair work to him and sends me most of the old tube amps and analog solid state gear. I leave most of the digital stuff to him. He is teaching me a bit, he’s in in 70’s now and winding down a bit. Not many young people seem to want to learn how to repair anything if it cannot be done with a computer and requires getting dirty hands.

[@joesmith-je3tq]

@@SIXSTRING63 It's not surprising that people would have no interesting in learning repairs. I doubt it has much to getting ones hands dirty. We live in a disposable world where the cost to repair is normally much higher than buying a brand new, better product. There is a reason we no longer have TV repair shops. We don't value that trade. If I were going to go into the repair business, it would have to be on VERY specialized equipment where I could command a decent wage. At 70, I hope they make a microscope that is powerful enough that I can still solder.

[@srvr1007]

I have one of the original batch of kickstarter meters and never liked the shim "solution". Never ended up installing it out of concern it would increase wear. It's not my most used meter by any means, but I haven't seen the issues others reported before the shim shipped out. Maybe I just got lucky or maybe I just haven't used the meter enough. I've been very slow to update firmware so far as most releases seem to make some improvements, but also add new issues. Also, been a bit worried about damaging that sd card slot. Very much looking to following your further tests.

[@Universal_Craftsman]

I don't know and can't believe that they made the shim of fiberglass, why didn't they use PA6, PTFE or anything else that doesn't get ground down like coffee in a coffee mill?

[@Universal_Craftsman]

As an mechanical engineering student the construction of the mode dial shocks me. This plastic shaft located with those prongs snapped straight into the PCB no bushing in the bearing, unbelievable, this thrust washer made of fiberglass, they should have made them from Teflon, or PA6, you don't want to see such construction in an industrial piece of equipment.

[@joesmith-je3tq]

I ran a life cycle test on the selector switch. It starts here: kzbin.info/www/bejne/mqe5Z6OcZt6Ui9U

[@lvxleather]

Really impressed with all these reviews. I'm surprised at all the flaws for a meter at this price point that is supposed to be a higher end meter. The shim seems pretty chintzy, they obviously screwed up their molds.

[@joesmith-je3tq]

I provided them with a fair bit of feedback on the prototype and there was certainly an opportunity to improve the hardware if that was a goal.

[@GadgetReviewVideos]

20:01 the shield is not on the first batch sent out, must be a rev change modification.

[@joesmith-je3tq]

The shim wasn't there as well. We also know they had made other changes.

[@GadgetReviewVideos]

joe smith True, I want to tear mine down now and compare it to yours in the next video. Maybe make a shield for mine and see if I can push “Hold” without it jumping all around. I’m thinking some metal duct strip, kapton tape both sides, cut the wholes and thin double sides adhesive might work.

[@ArnoWelzel]

Well - what is the point of having a multimeter that takes a couple of seconds to get the final measurement? Are the current copies better?

[@joesmith-je3tq]

Many of us who use even higher performance meters are used to waiting for them to settle. But to your point, yes this meter is very slow for what it is, not to mention all of it's other shortcomings. I haven't looked to see if they have done anything more with the firmware or hardware.

[@Newberntrains]

Why would dave put his name on such a slow meter or better yet allow them to ship em with faulty firmware

[@joesmith-je3tq]

What I find even more interesting is after watching so many of his DMM reviews, where he has been critical about so many of them, why this meter? Even if they are doing all the work for free and you get to have some say in the outcome, it seems like your reputation would be worth more. Oh well, it has not effect on you or I and we don't need to concern ourselves with it. One thing you can count on is these videos will once again be data driven.

[@Universal_Craftsman]

This SD-card holder "Looks like an afterthought", hmm, I think I already heard that a few times from someone... 😁

[@worroSfOretsevraH]

That's terrible Muriel. I'll stick with Fluke's thank you.

[@Okurkaarrgghh]

That SD card slot is flapping around in the breeze.

[@joesmith-je3tq]

It's too bad Fluke really doesn't make a meter I would want. I like the old Fluke 189 shown in this video, but I would rather use the Brymen BM869s. Having looked at so many Fluke products now, I will say they are one of the most robust brands on the market.

[@iglooom]

@@Okurkaarrgghh If your interests, Dave explained why SD-card mounted like this kzbin.info/www/bejne/m2rGoX6Ea550rM0 Case already been finished when he though about SD-card :)

[@cthoadmin7458]

joe smith what’s wrong with the fluke 87 Joe? Mine just screams quality though I haven’t subjected it to all the check you have.

[@joesmith-je3tq]

@@cthoadmin7458 I'm fairly certain I have covered that question more than once. I think there was a whole thread dedicated to it.

[@dukeeod1650]

How can I make one of those boxes?

[@KamotsTalksTech]

If you update the firmware the auto ranging is significantly faster.

[@Okurkaarrgghh]

Not sure why they shipped it with an old firmware version.

[@KamotsTalksTech]

Same reason cell phones and computers need updates when you first turn them on. They sat in a box for a while so they are behind. It is not that old, just the last few revisions addressed the slowness. It is extremely easy to update.

[@Okurkaarrgghh]

@@KamotsTalksTech Are you saying those Duracell AA batteries were installed in the factory months ago? That would be bad practice. With a flaw this big that got fixed in the firmware 4 months ago, one would think that all the new meters would be updated before being shipped out.

[@Todestelzer]

Okurka that would cost extra money and something could go wrong. But to be honest a meter shouldn’t need to be updated anyway.

[@Okurkaarrgghh]

@@Todestelzer So they ship a broken DMM instead? It's not a sub $50 DMM that you expect to have flaws.

[@DrFrank-xj9bc]

These 1µA and 10µA measurements would be too far off, i.e. 10µA range defect, or totally out of calibration, if I wouldn't suspect your current source to be very noisy.. too noisy for the 121GW. This noise is already visible by the unsteady display of the 121GW! Try upranging the 121GW one range, so that it resolves 10nA only, like the other DMMs... I bet it is then much better or even spot on. Also use zero before measuring very small currents, voltages, or resistance please. These 43nA offset you see are caused by the probe detection circuit. Superimposed AC noise on the 1µA range will let the internal amplifier clip or go into saturation, but the 121GW won't display that as an overflow. Remember, that the 121GW as well as Daves µCurrent use an internal 100x amplifier to achieve 1nA resolution, which the Brymen and the Fluke don't do, and do not resolve. For the SD card, simply use tweezers, my fingers are also too big to poke it out w/o force.

[@joesmith-je3tq]

Hey Frank. This video was really just a sanity check to make sure the meters were functional. Consider it an unboxing. We have a long way to go. With the various versions of firmware effecting the filter, I wouldn't be at all surprised if I reloaded 1.0 that it would not settle down. I normally run comparison video and will more than likely do the same with these two meters. From what I remember, my modified UT61E has a slightly lower burden than the 121GW in the uA range. The 121GW will out perform it in all the other ranges. I may toss it in the mix for a comparable. All things for future videos. My only real concern with the one brand new 121GW was when it was unable to resolve th1 1M resistor at 8:27. I was also very surprised when it read 0 with the 150pF cap. Still, my biggest concern is that switch and seeing them increase the preload to the spring contacts. I plan to run the 50,000 full cycle test with out cleaning the contacts. Hopefully it holds up as Dave showed with his toggle bot video. I still suspect the switch was the cause of it hunting with the 1M. Hard to say. A lot of testing to go.

[@stevenbliss989]

I was considering this meter, but the specs on accuracy are a little wanting. And now, after seeing your results on quality, NO WAY would I trust it! NOT BUYING!

[@TheTruthSentMe]

The SD card should be spring loaded.

[@Okurkaarrgghh]

No, spring loaded sockets break when you try to pull the SD card out. Just ask the owners of an RPi2B.

[@Drforbin941]

I stick with my Agilent.

[@ExPsy]

A great design which is held back by a flawed implementation. A pity for Dave. If only he had access to some decent SOCs and manpower for f/w development. It's sad to see the cheaper and less fully-featured bm235 being more practical in real life than this.

[@Todestelzer]

He’s not developing the firmware and not getting the source code. It’s done by the manufacturer.

[@ExPsy]

@@Todestelzer Yeah and they don't let others work on it.

[@chrisblank9090]

fluke 87 4, 187, 189, 287, 289, keysight u124x, u125x, u127x,,, my pref meters.

[@joesmith-je3tq]

I've yet to find anything I like more than the Brymen BM869s for my electronics hobby. It's just a good all around basic meter. I do like the Fluke 189 as well and have a bit of a collection of them now. Again, it's just a really nice all around meter. It's too bad Fluke doesn't offer a meter like this today.

[@chrisblank9090]

joe smith I think fluke is transitioning to biomedical. like beckman before.

[@zxrjimmy1]

Just buy a fluke or brymem , can't go wrong