New toy(?) scope, DSO154pro, 1ch, claimed 40MS/s

[Fungus]

Please someone tell me which oscilloscope can show lossless 3mhz square wave?  thanks

No oscilloscope is "lossless" but a 50Mhz oscilloscope will show a 3MHz square wave very well.

[techcornertv]

Hi everyone,

My humble contribution for this thread, I just released yesterday a video about it.


https://youtu.be/ugzLNDy6DEs



Cheers!
Hugo

[Matamipru]

there will be no worldwide interest in this product.it is open dangerous for alternating voltage. it only works for that.  so dso150 is safer and satisfies. they should have concentrated all efforts on the main parts. No in battery, probes and other things the customer of this type of product likes to do. if the software was open it would have already sold a zillion units.they think they keep and have the amazing secret.  profit policy on resale of parts is ignored.  will get rich selling the software 

[hallo1]

Got this from the vendor:

Hi, dso154pro there was an update, it mainly added zeeweii icons, there is not much difference in function. Due to hardware limitations, it does not support firmware updates.

The scope is VERY likely to support firmware update considering there is an "unused" USB to serial converter chip on board.

[Fungus]

The scope is VERY likely to support firmware update considering there is an "unused" USB to serial converter chip on board.

But it may be very difficult to hack if we don't have example firmware to work with or any clues about the hardware.

[Dave_g8]

For general information, I received a DSO154PPro recently and noticed that the PCB was marked “V1.4”.



This has a few notable differences compared to the previous versions. The changes may be due to component obsolescence, functional updates and/or cost reduction.

- The generation of the negative power supply rail has been modified.
This now uses a similar arrangement to other Zeeweii models, with a switched mode step-up regulator followed by a step-down regulator. This may be to ensure correct behaviour at low battery levels. The large electrolytic decoupling capacitors associated with the previous switching regulator have been removed and since the updated regulators have a higher switching frequency, smaller ceramic capacitors have been used.

- The device used for the data acquisition and waveform generation has been changed.
The device is marked “AG32VF303KCU6”, which appears to be an MCU with a RISC-V core with RV32IMAFC support and should be suitable for signal processing applications.

- The ADC has been changed.
The device is marked “MS9280”, which is a 10-bit ADC. The least significant 2-bits are not connected. There is a minor change in the analogue circuit around the 74HC4051 multiplexer which may be related to the ADC update.

- The waveform generator output circuit has been modified slightly.
The resistor ladder network has been changed to discrete resistors and there are some modifications around the “8091” output amplifier.

I also own a Rigol oscilloscope which is more capable for higher frequency, complex triggering and serial decoding, but the DSO154Pro is still a handy pocket oscilloscope with built in waveform generation and has a quick and simple user interface.

There are freely available 3D printable case models, which I used to provide some degree of protection to the unit.



Overall, for the cost, the DSO154Pro provides excellent functionality and performance compared to other devices in the price bracket.

[mojoe]

I'm supposed to get mine on Wednesday. I have four better scopes, but I ordered this to take with me in my carry-on luggage. For use with microcontroller hobby stuff while away from home, this should be sufficient. I could take my HDS272S, but it's much larger, and I pack very light.

[mojoe]

My itty bitty scope arrived a day early. I'll do some testing later, after it's charged.

Mine is version 1.3. I don't know if that is good or bad. As we all know, later versions could either be due to improvements or bug fixes, or due to cost cutting where they removed components to save half a cent.

Test results to follow.

Edit: as usual, the charging is not real USB-C.

[mojoe]

Test results.

Up to 8 MHz, a 1 Vpp sine wave looks good and measures at 1 Vpp. Beyond that, the voltage starts dropping slowly. I didn't bother looking for the 3 dB point, as from 10 MHz to 18 MHz, everything starts dancing, and the measured voltage isn't stable. At 18 MHz, it's still somewhere around 890 mV, but fluctuating wildly. Beyond 19 MHz, it just won't sync.

As for square waves, at 1 Mhz, the waveform starts to look like Bart Simpson, but is still quite usable. At 5 MHz, you've lost all but the 3rd harmonic, but can still tell it's a square wave.

So, it looks like the bandwidth is actually about 18 MHz, and any harmonics beyond are lost.

Yes, a toy scope, but still quite useful for up to 5-10 MHz, depending on the waveform. For my travel use, it will do nicely. If I drop it or run over it, I won't feel too bad, considering the cost. This would make a nice, inexpensive gift for a budding hobbyist.

Those of us who have been around electronics for quite a while will remember when we wished we could have afforded a boat anchor scope that only did 5 MHz (and you could heat the house with it). So don't sneer too much at this little scope.

[Dave_g8]

Hi,
The traces look similar to what I have seen on the V1.4 hardware (see attached), so I expect they are functionally the same.




The ripple on the square wave trace is due to the sin(x)/x interpolation that is used.
A simple example of reconstructing a sampled wave is shown attached.
- Equivalent to 40Msa/s and a 5MHz square wave.
- The red trace is the reconstructed signal



It would be useful to be able to turn off the interpolation in some cases to see the actual sampled waveform, but provided the limitations of the oscilloscope are understood, all is as expected.

[videobruce]

This "thing" wouldn't even qualify for a "toy" !!

[Fungus]

This "thing" wouldn't even qualify for a "toy" !!

Nonsense, they're awesome for what they are. Nobody's forcing you to read this thread.

This is exactly what you'd expect from a device that can only "see" the first two harmonics of a square wave:


ie. It can see the 5Mhz and 15Mhz sine waves in the signal but not the 25MHz sine wave (or higher)

[Dave_g8]

Hi,
I agree that the DSO154Pro is excellent for the price, but you do need to be aware of its limitations.

The pre-shoot and overshoot on the square wave from the external signal generator is an effect of the sin(x)/x interpolation which is used by the oscilloscope to construct the displayed waveform.
The rise/fall time of the external signal is approximately 10ns.
The images below shows the Rigol (sampling 1Gsa/s) and DSO154Pro (sampling 40Msa/s) displayed waveforms for the 1MHz signal from the external signal generator.



There is no evidence of an anti-aliasing filter on the front end of the DSO154Pro which would limit the bandwidth to the ADC, so the results are as expected.
 
With the in-built signal generator at 1MHz, there is less ringing because the rise/fall time is limited by the waveform generator circuit and therefore the displayed waveform is closer to representing the actual signal.



If the time-base on the DSO154Pro is set to 200us/Div or greater, the sin(x)/x interpolation appears to be turned off. The trace below shows the displayed waveform with the external 1MHz square wave signal. In this case the trace is joining the dots between samples.

[mojoe]

I had the rise time set to the fastest, which is 8.9 ns, if I remember. Thus, the Bart Simpson look.

I'll have to try changing the timebase to get "connect the dots". BTW, "200us/Div or greater" would imply to me more time per divison, not less. It's a matter of perspective. Once I looked at your screen grab, I knew what you meant.

Just a bit ago, I used my "toy" scope for something useful, without having to either take everything to the bench, or find room for a larger scope at the computer (not enough room). I bought some DS3231SN RTC modules, and wanted to test them to make sure they were actually the SN version, and not the M version. The M version only outputs 1 Hz, and cannot do 1024, 4096 and 8292 Hz, like the SN will. Also, the M has slightly worse long term accuracy.

I had previously ordered what were listed as SN, but received M. They were returned.

This type of use is exactly why I bought this tiny scope. I can take it in my carry-on luggage, along with some microcontroller stuff, to have something to do while I'm away for long periods for work.

[Fungus]

I agree that the DSO154Pro is excellent for the price, but you do need to be aware of its limitations.

That's also true of $20,0000 oscilloscopes.

[tatel]

Hi,
I agree that the DSO154Pro is excellent for the price, but you do need to be aware of its limitations.

Of course, however I see more bang for the buck in it than in most other "toyscopes". AFAIK, none of these two-channel low cost models appearing in the last times is able to do a differential measurement. Even if some offer "FFT". If you can't do differential measurements on a low BW 2-channels o'scope, is there much of an advantage in having two channels? At double DSO154Pro's price? Should they have differential measurement, I would think otherwise.

So, as next step in the ladder after DSO154Pro, I would rather get an old Hameg 20 MHz than, say, Zeeweii's DSO2512G. About the same bandwidth, but traces are much more stable after you get over half the device's bandwidth, and they do differential measurements. At first it seemed impossible to get one of these under 100€ in Europe, but now I have a couple of them (I still need to decide which one will remain here). One can get 50 MHz bandwidth for a little bit more (another couple Hameg 604-605 was found). Needed to go on the hunt for about six months however.

Personally I rather prefer these old boat anchors than one of the new, low-bandwith 2-channel o'scopes appearing now. Unless you really need the portability, of course. But when speaking about portability, DSO154Pro is enough for me. With a separated DMM. YMMV.

Don't really know about those Zoyis, etc, but for what they seem to be, I'd stick to my DSO154Pro, I think. Perhaps the Owons, I don't know. They seem to be quite good and portable. I miss differential measurement on them, tho, and have seen complaints about they getting hot, then becoming somewhat unstable. Others will know better I guess. But they look to me as probably the better of the low cost 2 channel oscilloscope-DMM combos, so I do see some value on them. Just not my choice.

Newbies on a really tight budget, looking for a cheap o'scope to begin learning, will find more value on DSO154Pro than in most other low-cost offerings. It will remain useful as a portable thing after they outgrow it, just for peanuts. On the next price step, old CRO's will also remain useful after outgrowing them. To, say, look at PSUs without risking to blow up any brand new 4-channel benchscope. For little more. Just my thoughts.

[BillyO]

In this video @ 15:40 the bandwidth of the 154Pro is tested and does just fine @ 18MHz.  No where near -3dB.

[Dave_g8]

In this video @ 15:40 the bandwidth of the 154Pro is tested and does just fine @ 18MHz.  No where near -3dB.

Hi,
Yes, the analogue bandwidth of the front end is greater than 18MHz.
Because the sampling rate of the ADC is only 40MSa/s, this means that the resultant displayed waveform after sin(x)/x interpolation for a square wave input signal will have ripple before and after the transitions which is not present on the actual signal.
So although you will see reasonable results for high frequency sine wave inputs, for square waves the result is not so good.
If you have sine waves greater than twice the sampling frequency the results will be invalid due to aliasing. For example, a 21MHz sine wave input will be displayed as 19MHz, 22MHz as 18MHz etc..
This behaviour will be the same for all sampling oscilloscopes, but when sampling at say 2GSa/s, the analogue front-end bandwidth will be apparent and interpolation effects will be much less noticeable.

[BillyO]

Hi,
Yes, the analogue bandwidth of the front end is greater than 18MHz.
Because the sampling rate of the ADC is only 40MSa/s, this means that the resultant displayed waveform after sin(x)/x interpolation for a square wave input signal will have ripple before and after the transitions which is not present on the actual signal.
So although you will see reasonable results for high frequency sine wave inputs, for square waves the result is not so good.
If you have sine waves greater than twice the sampling frequency the results will be invalid due to aliasing. For example, a 21MHz sine wave input will be displayed as 19MHz, 22MHz as 18MHz etc..
This behaviour will be the same for all sampling oscilloscopes, but when sampling at say 2GSa/s, the analogue front-end bandwidth will be apparent and interpolation effects will be much less noticeable.

Right, this will be the same for ANY scope, even an analog scope.  Put a 100MHz square wave into a 100MHz analog scope and the result will be far more sinusoidal than square as the front end acts like a low-pass filter removing most of the harmonics above 100MHz.  However, I see everyone rating this scope at 15MHz or even less.  That's just wrong.  It has an honest 18MHz bandwidth.  Bandwidth is always specified for a sine wave.  To display a decent square wave you need to de-rate the BW by a factor of about 10.  So a 18MHz scope will do a fair job of showing a sub 1.8MHz square wave.  This is true of any scope at any price.  Even then you will have considerable "ripple" on both the top and bottom of the waveform due to the harmonics that make it through the effective filtering of the front end.  This is expected behavior.  Of course you can expect a $3000 scope to deal with that better than a $30 scope but to even make the comparison is ridiculous.  So, I stand by my assertion that his is an 18MHz scope and I'll also add that it's insanely good for the $30 it costs.

It completely blows away the 5MHz scope I got in 1978 (which I had to build myself) that cost me the equivalent of $1000 in today's money.  On top of that it is arguably the best $30 scope you can currently get.

That's the way I see it anyway.

[Dave_g8]

Hi Bill,
As I stated in earlier posts, the DSO154Pro provides excellent functionality and performance compared to other devices in the price bracket.
However, I do not agree that an analogue oscilloscope would show the same ripple or wrong frequency, since this is purely an effect of the sampling and interpolation used.

[BillyO]

An analog scope would just show very close to a sine wave, you might have to educe the frequency from the period.

With my 154, using an 18MHz square wave at 5Vrms from a Siglent SDG1062X and properly terminated I get an apparently amplitude modulated sinusoidal waveform.  The amplitude modulation is being caused by aliasing.  I don't get that with a pure sine wave, plus it's not getting the amplitude right either (I have the probe setting at 10X so divide by 10).  But you can see it's showing the correct frequency.  At anything much higher than 18MHz (like 19MHz) it can no longer properly detect the frequency not matter what the waveform.

[Aldo22]

It completely blows away the 5MHz scope I got in 1978 (which I had to build myself) that cost me the equivalent of $1000 in today's money.  On top of that it is arguably the best $30 scope you can currently get.

That's probably true.
Nevertheless, I would pay a few extra bucks and buy a DSO1511G.
In my opinion it is even better value for money.

https://www.aliexpress.com/item/1005001969702720.html

[Dave_g8]

Hi Bill,
If you inject a 22MHz sine wave you will see a stable 18MHz sine wave on the DSO154Pro (as attached).

[BillyO]

There is no evidence of an anti-aliasing filter on the front end of the DSO154Pro which would limit the bandwidth to the ADC, so the results are as expected.
 
With the in-built signal generator at 1MHz, there is less ringing because the rise/fall time is limited by the waveform generator circuit and therefore the displayed waveform is closer to representing the actual signal.

If the time-base on the DSO154Pro is set to 200us/Div or greater, the sin(x)/x interpolation appears to be turned off. The trace below shows the displayed waveform with the external 1MHz square wave signal. In this case the trace is joining the dots between samples.

Yes, and the first case is more accurate.

Below the first image is like yours with the sin(x)/x.  The 2nd image is taken with a 600MHz scope using a 750MHz 1.5pF probe, so it is showing the signal being produced by the Zeeweii pretty accurately.  It matches what the Zeeweii itself shows pretty exactly.

[BillyO]

Hi Bill,
If you inject a 22MHz sine wave you will see a stable 18MHz sine wave on the DSO154Pro (as attached).
(Attachment Link)

Okay, I am not sure what the point is.  Are you saying it's not working right if you use it beyond it's specifications?  Why would you expect it to?

My Siglent SDS2504 will show a 800MHz signal pretty well, but it's a $3000 scope, not $30.

Zeeweii say the 154Pro is an 18MHz scope and it meets this specification.  What more do you need?