EEVblog #810 - Micsig MS310 Handheld Oscilloscope Teardown

[EEVblog]

What's inside the Micsig 100MHz 1GS/s isolated dual channel handheld oscilloscope?
David2 also makes an appearance.
http://www.micsig.com/en/product_detail.asp?id=19

https://www.silabs.com/Support%20Documents/TechnicalDocs/C8051F33x.pdf
http://www.ti.com/lit/ds/symlink/cdcm61002.pdf
http://www.intersil.com/content/dam/Intersil/documents/kad5/kad5510p-50.pdf
http://www.ti.com/lit/ds/symlink/tms320c6748.pdf
http://www.issi.com/WW/pdf/61NLP_NVP12836B_25618A.pdf

[hamdi.tn]

They will freakout with this video. unsoldred pins and that bit a solder flapping around on the chip pins ... that's terrible.
28:00 to 28:30 was really fun to watch  yap everyday is not my day 

[bktemp]

I am no expert for CAT ratings, but the isolations looks inadequate to me. The optocouplers seem to be IL206A made by Infineon. They are standard optocoupler rated for only 2.5kV isolation test voltage. Not the fancy ones with extra high isolation voltage ratings used in other test equipment.
The metal cans around the isolation transforms are connected to the isolated gnd. The gap around the other soldering tab located in the system gnd is rather small. The micro coax cable can easily come off the connector (for example when the unit is beeing dropped) and touch the metal can. If this happens exposed metal parts (like the USB connector) are connected to the input!
I doubt this unit would pass all the safety standards tests.

[Fungus]

I am no expert for CAT ratings, but the isolations looks inadequate to me. The optocouplers seem to be IL206A made by Infineon. They are standard optocoupler rated for only 2.5kV isolation test voltage. Not the fancy ones with extra high isolation voltage ratings used in other test equipment.

Send it to joeqsmith on the TE forum. Let him zap it!   

 

[Rasz]

They will freakout with this video. unsoldred pins and that bit a solder flapping around on the chip pins ... that's terrible.

dont forget broken transformer, its rather obvious $4K price doesnt include QA

I remember micsig post on the forum, looks like to this day they havent hired a single western engineer (translations, proofreading etc).
I guess it is a big government project to establish a strong chinese corporate presence in another sector of the market, like Xiaomi, and they are heavily subsidized. Next step is usually tariffs to kill competition on the domestic market.

8051 is very Shenzhen, at least they didnt put STC(chinese garbage bin special) in there.
Edit: forgot scratched markings, this is pretty FAIL for someone aspiring to play with big boys.

/not impressed

[Lukas]

QFNs with obscure 3-letter markings are pretty much always from Analog Devices:
H1F: ADA4817 FET opamp
HVB: AD8337 Variable gain amp

[Towger]

They will freakout with this video. unsoldred pins and that bit a solder flapping around on the chip pins ... that's terrible.

Hi Dave, What was the story with the missing pin?  It looked as if it was just a blanked out hole on the bottom board. The frontend board looked like a prototype with all the bodges. In 1080p I would see at least another 2 or 3 you did not point out.
At least they did not goto town with the sandpaper :-)

[electr_peter]

They will freakout with this video. unsoldred pins and that bit a solder flapping around on the chip pins ... that's terrible.

Overall quality is not bad, but details are not finished. Connector between inductor, very poor soldering job on all input connectors - these should be built much better for a price.

28:00 to 28:30 was really fun to watch  yap everyday is not my day 

Four in a row! Dave is on a roll!

[Ampere]

What's with the connector between the DC-DC converters? I would think that you would get a lot of switching noise which wouldn't be desirable in what is meant to be a precision instrument.

[Macbeth]

I'm surprised Dave was so lenient on this one. It was laughable all the bits missing, terrible hand soldering, errant pin, chunks of solder shorting stuff out. Total bodges. I think in this case it was definitely a case of "don't turn it on... tear it apart!" because without mopping up after the very poor hand assembly it probably would have fried.

I mean he tore the arse out of the UT-71 multimeter which was way better put together than this thing. 

[Macbeth]

What's with the connector between the DC-DC converters? I would think that you would get a lot of switching noise which wouldn't be desirable in what is meant to be a precision instrument.

I know. Up 'til then I thought good design let down by the hand assembly... 

[c4757p]

QFNs with obscure 3-letter markings are pretty much always from Analog Devices:
H1F: ADA4817 FET opamp
HVB: AD8337 Variable gain amp

That specific style, too. ADI uses a very specific-looking QFN that they like to call "LFCSP" (yeah, sure guys, "chipscale"... it's a bloody QFN ) that has a nice bevel to its edges that you don't see from a lot of other manufs. An analog section with a bunch of small 8-16-pin D/QFN-type devices with those beveled packages - usually ADI.

[mux]

If you really start looking at the details, it really only gets worse. I think Dave had a miss here; this $4k instrument should have been built much, much better. This is unacceptable. This is either a preproduction prototype or a really shitty board/assembly manufacturer messing it up for Micsig.

A couple of details:
- Silkscreen is very hit&miss. You can see on the HD that they'reusing a very coarse silkscreen, comparable to what I'm getting on my $10 china boards
- Soldering is mediocre. For instance at 13:44 you can see big differences; Q9/Q2 are two SOT23s that are correctly soldered with reflow, but Q3 has way too much solder on the pins. Right next to that is a 'D1' which obviously isn't a diode but has been replaced with a 51k resistor in the WRONG package size and bodged into place (see the crappy solder joint).
- In the same image, they're using a shitty tin/nickel rigid contact connector to connect to a carbonized flex that goes to the keypad. This is just guaranteed to fail, especially in a portable (shaky/crappy environments) application. This should have been at least a spring contact type, but ideally just not a carbonized flex.
- In the close-up shots with larger gold plating areas visible (e.g. 14:15) you can see dirt/tarnishing on the gold plating. This is indicative of a dirty PCB while doing the ENIG process. Either that or there's dust on it (it does look like meh quality plating though)
- Thoroughly insufficient reflow time or unclean contacts on Y1 (crystal), see 14:24. The crystal looks dodgy too, almost like it's either been reclaimed or improperly stored.

Just flipping through a minute of close-up footage I can see many grievous mistakes. This doesn't instill confidence. I've seen 30 year old Flukes that look better inside.

[Rasz]

I'm surprised Dave was so lenient on this one.

Trying not to burn any bridges on the first contact with company having a decent chance of disrupting hiend? or free scope? /joke

[EEVblog]

I think Dave had a miss here; this $4k instrument should have been built much, much better.

It's $1400.
The $4K one might be similar, or it might not, we dont' know.

This is unacceptable.

I agree, they need to fix their production issues.

But heck, even Keysight had huge production soldering issues with U1272A, a lot worse than this Micsig.

[EEVblog]

I mean he tore the arse out of the UT-71 multimeter which was way better put together than this thing. 

IIRC The UT-71 was crap design and crap components with ok soldering. The Micsig is quite good design, quality parts, with some messy soldering.
I think I'd rather take my chances with the later. Production soldering issues aren't that hard to fix, and Micsig is very new. They'll get better.
Uni-T have had a long time to get better and they are still struggling.

[EEVblog]

In 1080p I would see at least another 2 or 3 you did not point out.

Yes, I miss a lot of stuff looking through the camcorder LCD screen. And generally pay little attention while editing (you get that way after a 1000 videos  )

[Rasz]

Micsig is very new

Shenzhen Micsig Instruments Co., Ltd. is ... Established Year 2004

[TheSteve]

The review seemed a little generous to me - you're getting soft Dave!

edit:
Dare I say it almost felt like a very gentle paid advertisement for Micsig.

[mux]

I mean he tore the arse out of the UT-71 multimeter which was way better put together than this thing. 

IIRC The UT-71 was crap design and crap components with ok soldering. The Micsig is quite good design, quality parts, with some messy soldering.
I think I'd rather take my chances with the later. Production soldering issues aren't that hard to fix, and Micsig is very new. They'll get better.
Uni-T have had a long time to get better and they are still struggling.

*some* quality parts, some bargain bin parts. Succeed brand dc/dcs? Noname 8051s? Crappy BNCs? Even the FPGA was arguably not really that impressive - it's 'only' a cyclone III, the cheapest version bar one. Although I have to say the update rate is impressive so I wouldn't knock too much on that.

Quite good design is also something I'd contest on some details, like the oft-mentioned strange internal coax u.FL connector location between dc/dcs.

However if it truly is $1400 and not $4k, this honestly isn't much of an issue. This is about bang on the kind of components I'd expect in a midrange product.

By the way, I do concur that Agilent (and earlier Philips/Fluke) have had plenty of mfg issues with their scopemeters. I used to own a Fluke 99 and it had some serious display issues stemming from either bad solder joints or bad contact alignment. This was a very well-known issue and down to manufacturing. Not at all what you'd expect from what used to be an astronomically priced unit. Worst thing was: you didn't get support from Fluke when the screen went bad after a couple of years, so you would either have to get it repaired at high cost or cut your losses and sell it. This was the reason you could buy $100 Fluke 99Bs in the early 2000s.

[Porto]

I had a BLOODY good laugh, thanks Dave! 

[Phaedrus]

That sticker in the battery compartment says ? (si hui bu bao) and basically means "no warranty if torn" (or something like that)

Edit: Heh, guess the forum doesn't like Chinese characters :p

[Macbeth]

That sticker in the battery compartment says ? (si hui bu bao) and basically means "no warranty if torn" (or something like that)

Edit: Heh, guess the forum doesn't like Chinese characters :p

Worse, it doesn't even allow greek characters which seems absurd to me for an Electronics forum 

[c4757p]

Edit: Heh, guess the forum doesn't like Chinese characters :p

don't need none o' that yuny-code stuff to speak proper strine

[Stonent]

The review seemed a little generous to me - you're getting soft Dave!

edit:
Dare I say it almost felt like a very gentle paid advertisement for Micsig.

I kept wondering when he was going to say "Just kidding! This thing is crap!"