VOLTCRAFT DOV series

[Lutz21]

So, what’s the verdict on the DOV70X so far?

My assessment, of course, relative to the price:

PROS

• Very good bandwidth and rise time for the promised 70 MHz
• Very stable and detailed signal display (12-bit).
• Good memory depth and waveform update rate.
• Digital phosphor resembles that of a CRO, depending on the memory depth.
• Color grading is very good.
• Good measurement accuracy, at least on my model (Martin?)
• Comprehensive Android system with USB, LAN, Bluetooth, HDMI, web control, web browser, etc.
• Decent touchscreen. Smooth and easy to use, though better suited for slender fingers. Mouse option via USB or Bluetooth.
• Hackable! (Thanks to Olex).

CONS

• Essentially unacceptable ripple at 200mV/div and 10V/div.
• Questionable FFT function (on my unit).
• Loud fan.
• Stupid layout of the V/div and s/div buttons. Switching between channels is particularly annoying on the 2-channel version (no separate controls per channel).
• Many minor shortcomings and missing features, but for me, these aren’t absolute deal-breakers.

Overall, the device feels half-baked and rushed to market. That seems to be the case quite often these days.

So:
Can the DOV70X be recommended without reservation? No.
Is there a better oscilloscope for signal analysis on the market for (starting at) €149? I don’t know of one.

Your opinions are welcome. I’m sure I’ve forgotten a few things.

There are also a few things that are most likely class leading, if you call entry level 12bit DSO's a class:

• Battery-backed real-time clock
• Lowest USB-C power consumption, the DVO704 uses about 10 W less than the Rigol DHO804
• Highest multi-channel sampling rate

[Aldo22]

The SG-004A has a dedicated mV mode, but unfortunately I didn't realize that until it was too late.

Yes, that's even better.
But it costs significantly more than the SG-003A—almost as much as a DOV702. 

As a hobbyist, I simply use a voltage divider for certain tasks to generate lower voltages and/or minimize drift effects.
EDIT: Attached 100μV with voltage divider.

[Martin72]

My conclusion about the scope:

+ Compact design
+ Can be powered by a battery pack
+ Good display resolution
+ HDMI output
+ Web server
+ Bandwidth significantly higher than the model name suggests
+ 2 12-bit ADCs from TI
+ Price (€199)
+ Good SPO display
+ Memory up to 50Mpts
+ Good waveform update rate

- Original fan is annoyingly loud
- Generally sparse software features
- Menu structures are sometimes confusing.
- No dedicated screenshot button
- ADCs are being operated outside their specifications.
- Basic FFT without axis labels, RBW display, memory display, no averaging, no peak function
- Thermal design (offset drift)
- Models with board rev. V1.0 exhibit unacceptable noise behavior at 200 mV/div and 10 V/div across all channels.
- Noise on the calibration signal due to a design flaw
- No automatic memory management
- Only one USB-A port
- No vertical fine-tuning possible

Overall, the Scope is an attractive alternative to the one-hit wonders and no-name brands in the very low-end price range, if only because it offers “true” 12-bit resolution, has a good display, and features a relatively high waveform update rate.
However, all of this only applies at a price of around €199—and even then, only if the critical ripple issue at 200 mV and 10 V/div is fixed.
It wouldn’t be worth the original price, because for that amount you can get a Rigol DHO800, which is superior to the DOV700 and ADS800.
By now, we've figured out that newer models no longer have this defect because Owon “secretly” revised the circuit board.
In other words, they don’t care that countless people have bought a defective device—anyone who doesn’t speak up (like Microcheap) is simply out of luck.
Or, as in my case, they don’t lift a finger.
This is a very poor example of customer support and quality assurance, and it reinforces my opinion of Owon and Co. once again.
And by now, I also believe that’s why the Voltcraft units are being sold off so cheaply.
Basically, you should steer clear of this Owon series.
If instructions on how to fix the problem are provided later, then this scope will be worth the money.
I mean, this scope at this price.
Otherwise, it’s too expensive even for €199 or less.

Martin

[Aldo22]

- No dedicated screenshot button

There isn't a physical button, but you can either press “Copy” on the screen (next to “Run”) which repeats the last "Save" menu action, or use the Android method by swiping down and tapping the phone icon. See screenshot.

[Aldo22]

Otherwise, it’s too expensive even for €199 or less.

I'm having a bit of a hard time with this assessment.
It's too abstract for me.
My question would be: Which oscilloscope is better than this one for €149?
And I can't think of any (except maybe a used one, but even then it would be difficult to find a decent one for that price).

With a budget of 150 €, I’d buy this one, despite all its shortcomings.

[Martin72]

Hi Aldo,

This statement refers to the Ripple issue.
If that's resolved, I think it's definitely worth 199—maybe even a little more—as long as it stays below DHO800.
As things stand now, we—everyone—has bought a scope that’s effectively defective; that’s the truth.
And it’s not even worth the special price.
On top of that, the FFT isn't working properly on your scope.

[Martin72]

There isn't a physical button, but you can either press “Copy” on the screen (next to “Run”) which repeats the last "Save" menu action, or use the Android method by swiping down and tapping the phone icon. See screenshot.

I know how to take screenshots with this thing, but that doesn't change the fact. 

[the Chris]

I just got mine, the reasons to give in and buy it where mainly:

• due to its form factor, I was sure to be able to integrate it into my lab shelve via 3D printed parts
• the recent Conrad offer pulled it to a new all-time low so the bargain trigger beat my good intentions

I first had to finished the adaption to the shelve so no deeper looks into it performance yet. But I did a quick comparison to my 34411A and the results looked fine in DMM mode. No noise evaluation yet. I'll contribute the results of my sample hopefully within the next few days.

[Aldo22]

This statement refers to the Ripple issue.
If that's resolved, I think it's definitely worth 199—maybe even a little more—as long as it stays below DHO800.
As things stand now, we—everyone—has bought a scope that’s effectively defective; that’s the truth.
And it’s not even worth the special price.

Yes, the ripple issue is a nasty one—a major bug—and I’d be glad if it could be fixed.
But how bad is it really in practice, and does it actually render the scope useless when comparing models in this price range?

I think there are many ways to deal with it:

1. Be aware of the problem and ignore it. Often, it’s just a cosmetic issue and doesn’t prevent you from taking measurements.

2. Light averaging (4) smooths out the ripples in many frequency ranges.
It doesn’t look any better on other scopes in this price range.
Here’s an example at 170 kHz.



3. View the signal at 500 mV/div, stop, and zoom in. Thanks to 12-bit resolution, the vertically “zoomed” signal still looks better than on other oscilloscopes.
The same example as above, paused at 500 mV/div:



4. At frequencies below about 1 MHz, you can switch the probes between 10x and 1x to overcome the ripple resolution.

So yes, it’s ugly, but I can live with it. You have to compare it to the competition in this price range.

[Martin72]

You're the kind of customer that sloppy manufacturers can only dream of...
But hey, always looking on the bright side of life and finding a way—that's a good attitude otherwise.

[Aldo22]

You're the kind of customer that sloppy manufacturers can only dream of...

I don't know—if I'd paid 300, I wouldn't accept the defect.
But I really do judge almost everything in the context of its price.

But hey, always looking on the bright side of life and finding a way—that's a good attitude otherwise.

I think it's fun finding ways to achieve a goal using simple/cheap stuff.
Just as a hobby, of course. 

[Martin72]

Nah, that would be a waste of my time.
Or rather, in this case, I did it for several reasons.
First, to show that there are limits to how cheap you can go, and that you shouldn’t go below a certain level.
Second, because this forum has helped me a lot, and I want to give back from time to time by doing reviews.
Sounds corny, but it’s true.
And I’m always curious to see how low the bar can be set in the “cheap limbo” before you can’t get through anymore.

[Aldo22]

First, to show that there are limits to how cheap you can go, and that you shouldn’t go below a certain level.

Oh, come on, they almost made it. 

I just wonder what the Lilliput guys were smoking to release this device into production like this.
The ripple issue is really impossible to miss.
I actually think that Owon was trying to build a competitive scope here. That’s why the ripple hardware flaw is truly baffling to me.

[Filippo52]

Good morning, and sorry.
I've carefully read all the developments in this interesting oscilloscope thread. I hoped someone would find the source and cure for the 40 kHz hum, but it didn't work out.
However, I remember that Martin initially found another flaw in the front end. I think they were talking about intermodulation, and this would be impossible to fix without heavily modifying the analog input layer.
How did it end? Did I miss something?
Do those who accept the 40 kHz hum flaw also experience intermodulation distortion?
And there's also the suboptimal FFT, but this seems more common, and I've even seen users complain about Rigols like the MHO954.

Thanks to everyone who helped.

Filippo

[Aldo22]

I noticed something strange today.

Could it be that “self calibration” is only performed for the selected V/div setting?
I wanted to measure low voltages (1 mV DC), and the result was very inaccurate (offset). Performing a “Self calibration” at 5 mV/div did not improve the situation.
However, a calibration at 500 µV/div yielded excellent results.
I didn’t expect that.
Can anyone confirm this?

The screenshot shows 0V and 1mV DC.

[Olex]

Hallelujah!
It was all because of that buffer!

[Martin72]

Hi,

After taking these measurements I wanted to wait a bit longer to see if the missing capacitor value in the front end would be identified (it wasn't added for no reason).
But I'm glad that was the main cause.

[Martin72]

Could it be that “self calibration” is only performed for the selected V/div setting?

At least that would explain why the self-calibration routine is so fast.
When I used it for the first time, I was surprised at how fast it was—especially for a 4-channel scope...

[Aldo22]

Hallelujah!
It was all because of that buffer!

Congratulations! 
So you removed the ST3SK, moved one 0R up, and replaced 01Y and 01A (if I'm reading this correctly) with 0R?
Anything else?
Did you use hot air for this?

[Olex]

Hi,

After taking these measurements I wanted to wait a bit longer to see if the missing capacitor value in the front end would be identified (it wasn't added for no reason).
But I'm glad that was the main cause.

I will definitely measure the voltages and test the oscilloscope, haven't assembled it yet.
Thought since I'm already digging inside, I might as well fix the calibrator at the same time.
Though in any case, some extra capacitance at the input won't hurt.

[Olex]

Hallelujah!
It was all because of that buffer!

Congratulations! 
So you removed the ST3SK, moved one 0R up, and replaced A10 and 01A (if I'm reading this correctly) with 0R?
Anything else?
Did you use hot air for this?

Spot on, exactly like in this photo - https://www.eevblog.com/forum/testgear/voltcraft-dov-series/?action=dlattach;attach=2844874;image

I was too lazy to look for where I stuffed my hot air gun, so I just did everything with a soldering iron.
I've got plenty of experience soldering tiny components by hand.

[Martin72]

To document this properly later on, I decided to “modify” my good old Andonstar microscope camera a bit so that I could fit a few more components into the frame.
And it didn't disappoint me...

[Aldo22]

To document this properly later on, I decided to “modify” my good old Andonstar microscope camera a bit so that I could fit a few more components into the frame.
And it didn't disappoint me...

Thanks, now I can read it.
So the removed resistors are 01Y (1Ω ±1%) and 01A (100Ω ±1%).

What SMD size is that? 0603?

[Martin72]

Yes, unfortunately 0603—I have the 1206 and 0805 assortments here for 0-ohm resistors.
But 0805 might still work.

So the removed resistors are 01Y (1Ω ±1%) and 01A (100Ω ±1%).

That makes sense, too—if the buffer fails, all the current from the 2.5V reference will be used up.

[Olex]

To document this properly later on, I decided to “modify” my good old Andonstar microscope camera a bit so that I could fit a few more components into the frame.
And it didn't disappoint me...

Wish I could also figure out whether that missing capacitor under the relay - the one going from the op-amp input to ground - is actually needed.