EEVblog Electronics Community Forum

EEVblog => EEVblog Specific => Topic started by: EEVblog on November 16, 2017, 06:57:26 am

Title: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: EEVblog on November 16, 2017, 06:57:26 am
What's inside the 4CH 100200MHz entry level Uni-T UPO2104CS oscilloscope?

https://www.youtube.com/watch?v=ywewix7DlpI (https://www.youtube.com/watch?v=ywewix7DlpI)
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: ataradov on November 16, 2017, 07:27:09 am
Video went down?
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: technix on November 16, 2017, 07:33:12 am
Why is there 20 seconds of dead air between 2:35 and 2:55?
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: chriswebb on November 16, 2017, 07:35:04 am
The display artifacts were because the gradient was using black on the low range. When you made it one color, you saw that it was the signal and not a bug in the display.
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: OzOnE on November 16, 2017, 07:42:35 am
@technix

Haha - I KNEW Dave couldn't let that go.

(may have been some other missing sections maybe?)

Looks like he's re-uploading the vid, and just when I was about to post an "amusing" comment. lol

Interestingly, I could still watch a fair bit of the vid as it was cached, but I made the mistake of refreshing, thinking that the error while posting the comment was a networking issue.

Oh well. I'll await the re-upload.

OzOnE
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: ataradov on November 16, 2017, 07:43:59 am
It was not a mistake, I was just watching the video and it just abruptly stopped anyway.
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: OzOnE on November 16, 2017, 07:51:33 am
Do you mean it stopped before the end anyway?

Ahh, OK. Just a rendering glitch then. No biggie.

OzOnE.
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: ataradov on November 16, 2017, 07:52:29 am
Do you mean it stopped before the end anyway?
Yep, stopped mid sentence, and static, and "Video is no longer available" or something like that.
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: OzOnE on November 16, 2017, 07:54:31 am
Yep, that's where Mr Dave is likely re-uploading.

OzOnE.
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: nidlaX on November 16, 2017, 08:26:37 am
Good to finally see a teardown of this scope, I recall there was a lot of speculation over its capabilities and OEM when it was released.
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: iceisfun on November 16, 2017, 08:34:55 am
Are the function lockout messages at the end because independent time bases are enabled?
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: EEVblog on November 16, 2017, 08:57:59 am
Are the function lockout messages at the end because independent time bases are enabled?

Yes, you can't turn off any of the channels with the ind timebase enabled.
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: snoopy on November 16, 2017, 10:56:17 am
You can see it here ;)

https://www.bitchute.com/video/GIKrPcdRxYM/ (https://www.bitchute.com/video/GIKrPcdRxYM/)
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: EEVblog on November 16, 2017, 12:15:48 pm
Fixed video is now back up.
Took a while, had to come back into the office after dinner and re-render.
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: ChrisLX200 on November 16, 2017, 12:20:48 pm
Thanks for the great review! I think at that price point I would wait to see the Siglent offering though. Did I see a date of April 2017 on your example? Maybe Uni-T already fixed some of the display issues you noted, but even so, too expensive to compete I think.
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: tzok on November 16, 2017, 02:32:47 pm
Are the function lockout messages at the end because independent time bases are enabled?

Yes, you can't turn off any of the channels with the ind timebase enabled.
I think he meant the Function Disabled for Decode due to Independent Mode enabled.
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: NANDBlog on November 16, 2017, 02:44:59 pm
Dave, you call a bunch of resitors around 10:50 termination resistors.  I would've guessed they are for configuration. First, only 1 of each pair is on the board. Also, all of them are 01C = 10KOhm. Maybe this scope can be hacked for higher speed/more memory very easily??
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: Ondre on November 16, 2017, 03:14:30 pm
I saw that UNI-T is using thermal reliefs on the vias. Is there a good reason to do so? I usually connect them directly to the plane, unless I plan to solder bodge wires to them :-)
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: don.r on November 16, 2017, 04:54:49 pm
If you do a shootout, please include the Micsig in there.
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: NivagSwerdna on November 16, 2017, 05:25:28 pm
Too many bodges for me.  UniT won't be on my Christmas list.  (Nice sqiggles though)
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: Hydron on November 16, 2017, 05:44:14 pm
Hi Dave,

The Analog Devices "DAC" part you found is actually an ADCMP562B (i.e. the bit after ADCMP looks similar to 5628). This is a "Dual High Speed PECL Comparator", which makes sense for a trigger section.

Thanks for the teardown, always interesting to see how they've spent their money in low cost stuff like this.

Edit: regarding the channel bias, my RTB2004 has a "Zero Offset" control in the vertical menu, which seems to do the same thing. I've found it handy a couple of times dealing with active probes with (small) output offsets - no idea of the range of the control.
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: smashIt on November 16, 2017, 08:15:55 pm
I'd guess that the "bodge" at minute 20 is just a flipped resistor.
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: Someone on November 16, 2017, 10:13:45 pm
Same as the GW Instek GDS-1000B the power supply earthing connection (and other aspects of the mains wiring) don't meet Australian basic safety standards from AS3100, and would fail other regions standards too. No wonder you don't find many sellers for it.
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: nctnico on November 16, 2017, 11:53:43 pm
Edit: regarding the channel bias, my RTB2004 has a "Zero Offset" control in the vertical menu, which seems to do the same thing. I've found it handy a couple of times dealing with active probes with (small) output offsets - no idea of the range of the control.
I've also found a bias control in a Yokogawa DSO but in the end it it controlled the same DAC as the trace offset.
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: kcozens on November 17, 2017, 04:06:08 am
We know that Bob's your uncle but who is this Flynn person?  ;)  The two DDR memory chips were upside down. I was waiting for you to point out the electrons are going to fall out.

The UNI-T has better colour choices for the four traces vs the colours used by the Rigol 1054Z. I wonder if the 50 second boot time is due to the UNI-T possibly running a version of Windows based on it having a PrtScr button. For proper price comparison you have to look at the Rigol 1104Z. It has a price of $619 for the Rigol 1104Z so the UNI-T is only $20 less. The Rigol 1054Z wins on price if a user hacks the scope to enable its full capabilities. Have you hit the Display button to see if it has the XY display mode?
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: cpuerror on November 17, 2017, 04:27:10 am
What a hot mess of a scope. The front panel looks like a cheap imitation of the DS1054. Also, whats up with using Arial in the UI? The last time I have seen something that ugly it was a Visual Basic app running in Windows 3.1.

This scope is a loser, it's too expensive, it's underdeveloped and poorly imitates its competitors. At $299 it might have a shot at the low end market, at $599 its just setting good money on fire.
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: Mr. Scram on November 17, 2017, 05:50:25 am
It's interesting to note that both Uni-T and Siglent are taking design cues from Rigol. I guess that means they're becoming a real contender.
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: Fungus on November 17, 2017, 08:20:02 am
For proper price comparison you have to look at the Rigol 1104Z. It has a price of $619 for the Rigol 1104Z so the UNI-T is only $20 less. The Rigol 1054Z wins on price if a user hacks the scope to enable its full capabilities.

Does any hobbyist NOT hack their Rigol?  :popcorn:

This scope is a loser, it's too expensive, it's underdeveloped and poorly imitates its competitors. At $299 it might have a shot at the low end market, at $599 its just setting good money on fire.

Yep. At that price and with no hacks it's dead in the water. If I were Uni-T I'd be thinking about "leaking" some hacks.

If the bandwidth hacks are hardware mods (move resistors) then it's still dead in the water once the Siglent arrives. You'd have to be a total Uni-T fanboy to buy one over the Siglent.

So much R&D money down the drain...
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: NivagSwerdna on November 17, 2017, 08:53:29 am
We know that Bob's your uncle but who is this Flynn person?  ;)
Dave used to run a dating web site so it should really be "In like Dave" but the reference is to Errol.

According to IMDB "Errol Flynn was born to parents Theodore Flynn, a respected biologist, and Marrelle Young, an adventurous young woman."  Marrelle was born Mary but changed it to Marrelle and was born of 'seafaring folk'.. I wonder if she was a pirate?
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: Hydrawerk on November 17, 2017, 05:43:56 pm
Is this an old-style 50Hz iron-core transformer? You hardly see a switched PSU and a 50Hz transformer in a scope.  :o
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: kx5 on November 17, 2017, 06:26:39 pm
Is this an old-style 50Hz iron-core transformer? You hardly see a switched PSU and a 50Hz transformer in a scope.  :o
That has to be the standby power supply. What a mess of a design  :-\
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: Hydrawerk on November 17, 2017, 09:58:08 pm
I think that there is no back-up battery for real time clock in this scope.
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: thm_w on November 17, 2017, 10:32:43 pm
I saw that UNI-T is using thermal reliefs on the vias. Is there a good reason to do so? I usually connect them directly to the plane, unless I plan to solder bodge wires to them :-)

Zero reason to do this. When you add thermal relief rule to Altium, the default is to apply to all pad types. So you have to specifically add another rule to exclude vias from reliefs, which people either don't realize or don't know how to do (its a commonly asked question).
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: Ardutronic on November 18, 2017, 05:46:57 pm
What's the difference between DC bias and the vertical offset? When I want to see a signal outside the range of the ADC I just use the vertical offset and it works :-//.
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: nctnico on November 18, 2017, 07:29:58 pm
If the bandwidth hacks are hardware mods (move resistors) then it's still dead in the water once the Siglent arrives. You'd have to be a total Uni-T fanboy to buy one over the Siglent.
Everybody seems to be totally excited about the new Siglent 4 channel scope and forget that every piece of Siglent gear is riddled with bugs when released and chances are some bugs are never fixed. If the Uni-T has firmware which works then it might be a better buy. Especially if you want to use your new scope for measuring instead of help filling a thread about bugs in a scope on this forum.
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: Fungus on November 18, 2017, 08:37:20 pm
Everybody seems to be totally excited about the new Siglent 4 channel scope and forget that every piece of Siglent gear is riddled with bugs when released and chances are some bugs are never fixed.

True. I hope Dave is thorough.

If the Uni-T has firmware which works then it might be a better buy. Especially if you want to use your new scope for measuring instead of help filling a thread about bugs in a scope on this forum.

Or get a Rigol DS1054Z for $200 less.
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: tautech on November 18, 2017, 09:00:03 pm
If the bandwidth hacks are hardware mods (move resistors) then it's still dead in the water once the Siglent arrives. You'd have to be a total Uni-T fanboy to buy one over the Siglent.
Everybody seems to be totally excited about the new Siglent 4 channel scope and forget that every piece of Siglent gear is riddled with bugs when released and chances are some bugs are never fixed. If the Uni-T has firmware which works then it might be a better buy. Especially if you want to use your new scope for measuring instead of help filling a thread about bugs in a scope on this forum.
The usual anti Siglent BS from you.
You should know better, this from a thread you've participated in:
https://www.eevblog.com/forum/testgear/testing-dso-auto-measurements-accuracy-across-timebases/msg1204964/#msg1204964 (https://www.eevblog.com/forum/testgear/testing-dso-auto-measurements-accuracy-across-timebases/msg1204964/#msg1204964)

And need we remind you of the bug reports you made of other brands you own.  :P
https://www.eevblog.com/forum/testgear/siglent-sds1204x-e-released-for-domestic-markets-in-china/msg1291644/#msg1291644 (https://www.eevblog.com/forum/testgear/siglent-sds1204x-e-released-for-domestic-markets-in-china/msg1291644/#msg1291644)

Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: nctnico on November 18, 2017, 09:38:33 pm
And need we remind you of the bug reports you made of other brands you own.  :P
Bending the truth again through your pink glasses!? Here is my text you are referring to from that thread to put things into the right context:

The GW Instek GDS-2204E had two or three minor bugs which where solved within weeks. The dissaster with the Siglent SDS2000 made me want to verify all functions before committing to buying it. Since then a handfull of minor bugs got fixed (exotic stuff like things going wrong if there is a space in a free math expression send to the scope using an SCPI command). I got the MicSig TO1104 specifically for a review where they installed a beta version of the decoding, so yes that deserved some remarks. Both these scopes don't have a bug list as long as a typical Siglent unit after it has been released and no trivial bugs which should discovered by the factory's testers.

You can write all you want but that doesn't change the cold hard fact that it takes Siglent months to years to fix all bugs in a scope if they even care to do so. It is up to the buyers if that is acceptable for them or not so leave it at that. After all you get what you pay for.

Reading through the threads about the bugs in the low end scopes (in general) makes me a little sad. I see a lot of relatively inexperienced people struggling with whether the effect they see on their scope is a problem in their circuit, user error or whether it is a bug in their scope. You can't do much about the first two but the last one can be mitigated by spending a bit more (or settle for something with lesser specs and/or second hand) especially if your time is worth something. Good tools make working on a project way more pleasant.
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: Mr. Scram on November 18, 2017, 10:23:15 pm
Can you please take your lovers' quarrels elsewhere?
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: DutchGert on November 21, 2017, 03:39:23 pm
I niticed the pcb designer used thermal relieves on the via's, bit wierd
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: Bashstreet on November 21, 2017, 03:49:59 pm
I think David was rather complimentary of the Uni-T maybe he is hoping for some UNI-T swag  :-DD

Cannot say i was impressed... for the price... you would hope something bit more... refined
Uni-T keeps trying to offer lot of functionality in nice case hiding the absolute rubbish workmanship horrible components and cost cutting done to extreme.

Now i do not mind cost cutting when it is done sensibly.. but when one cuts fraction of cents to build a product with absolutely inferior rubbish components risking despair and disappointment of their customers is not good logic.

Uni-T must evolve or it will die slowly.

I have made the mistake purchasing Uni-T product in the past .. never again once company ruins their reputation with me that's it.
Only if they can grow in future good reputation and change their ways no thx for me.
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: Hydrawerk on November 21, 2017, 08:47:12 pm
absolute rubbish workmanship horrible components and cost cutting done to extreme.
What exactly? Is it worse than Siglent and Rigol??
Title: Re: EEVblog #1038 - Uni T UPO2104CS Oscilloscope Teardown
Post by: Mr. Scram on November 23, 2017, 01:47:22 am
I think David was rather complimentary of the Uni-T maybe he is hoping for some UNI-T swag  :-DD

Cannot say i was impressed... for the price... you would hope something bit more... refined
Uni-T keeps trying to offer lot of functionality in nice case hiding the absolute rubbish workmanship horrible components and cost cutting done to extreme.

Now i do not mind cost cutting when it is done sensibly.. but when one cuts fraction of cents to build a product with absolutely inferior rubbish components risking despair and disappointment of their customers is not good logic.

Uni-T must evolve or it will die slowly.

I have made the mistake purchasing Uni-T product in the past .. never again once company ruins their reputation with me that's it.
Only if they can grow in future good reputation and change their ways no thx for me.
Can you point out how they cut costs to the extreme. Dave doesn't quite seem to share your opinion.