I bought a cheap Hantek 6022BE to learn with and I’m just trying to measure some voltage to check if it’s accurate. I’m using openhantek6022 https://github.com/OpenHantek/OpenHantek6022 and also hscope on android. I’m measuring a 19v laptop power supply with the 10x setting on the probe and also a set of leads with a 20x attenuator. When I measure the power supply with my dmm I’m getting 19.22V with the scope I’m getting much higher or lower depending on where I set the voltage per division. Both programs have a calibration function which I did.
20x attenuator showing 19.6v
10x probe showing 20v
10x probe with 2v per division setting only showing 9.81V
Hscope showing 19.9v
I read your question and was wondering how an oscilloscope could be giving such widely-differing values, with the widest being 0.4 volts against itself and nearly 0.8 volts against a separate instrument. Then it dawned upon me that this oscilloscope is a PC-attached scope with some unique operating limits. I say this having come from a background of using only benchtop digital scopes.
The first limitation is that your scope has a very narrow input voltage range, with the manual listing it as +/- 5 volts but damage would only occur at +/- 35 volts. This is voltage measured at the input BNC connector, so it’s before any probe multiplication is accounted for. Whereas if we look at an inexpensive benchtop oscilloscope like the now-fairly-old Rigol DS1052E, it has an input voltage range of +/- 40 volts. The practical result is that to measure something like a laptop power supply, the Hantek must use attenuation probes, whereas the Rigol can measure that voltage directly. Slightly more expensive oscilloscopes have wider ranges, with some being +/- 400 volts.
Attenuation probes are great for measuring wider voltage ranges, but they come at the cost of both precision and accuracy. The loss of precision comes from the fact that the resolution of the oscilloscope is unchanged, but the voltage range is wider. In concrete terms, both the Hantek and Rigol use an 8-bit ADC, meaning that the span of input voltages visible on the display are mapped to 256 discrete values. If the ADC is imprecise by 1 bit, then that will amount to the reading being off by a certain number of millivolts. But something like a 20:1 attenuation probe causes that millivolt error value to be multiplied by 20x. Whereas the Rigol doesn’t need attenuation probes, and thus doesn’t suffer this penalty.
Furthermore, the Rigol has a neat trick: it uses a separate, more-precise internal attenuation circuit for voltages smaller than +/- 2 volts, and then uses its normal-precision input circuit for all other voltages up to +/- 40 volts. The ADC is unchanged in both modes, and the scope switches seamlessly between the two (though usually with an audible click), but this means that a 20:1 probe measuring a laptop charger would actually cause the Rigol to switch into its precision circuit, which means the Rigol might never pay the precision penalty that the Hantek might. Perhaps the Hantek has a similar feature, but it is not listed in the manual.
As for accuracy loss due to attenuation probes, this is not affected by the amount of attenuation, but rather is a function of how accurate the attenuation is. When a probe is marked as 20:1, it could actually be 19:1 of 21:1 or anywhere around there, depending on the manufacturing tolerances. However, accuracy issues can be resolved through calibration, which you’ve done.
Overall, it seems that you are operating at the very limits of what your Hantek scope can deliver, with its 8-bit ADC and limited input range. Yet your test calls for a voltage 4x higher, so some error is to be expected from the 20:1 probe. With the 10:1 probe, the error is a bit smaller, but now you’re outside the affirmative safe voltage range of the scope. Calibration can only fix accuracy issues, but I think your error is now predominantly due to loss of precision, which cannot be resolved after-the-fact.
If your intended use is to measure signals in the range of a laptop charger and require faithful analog voltage measurements, I’m afraid that you may need to find a different instrument.
Thank you for your very detailed response! My goal is to do automotive diagnostics so I’ll be working on the 13v range at the most. It looks like I should have sprung for the 2 channel picoscope which cost around $200. I knew going in that this scope would have limitations, but I didn’t fully understand what they are. Fortunately many sensors used for diagnostics work in the mv range. I’m mostly looking at relative comparisons such as cam and crank sensor correlation, an amp clamp (1mv output per amp), canbus, voltage fluctuations, and pulse sensors.
I wish you the best of luck in your automotive endeavors. But specific to that field, be advised that automobile power can have a lot of voltage spikes, most notably right after the starter motor shuts off after ignition. This could be as high at 15v or around there. So if you’re not probing during this dynamic event, then your scope will likely still be useful.
I will also note that a used benchtop scope can be had for about $200 USD, often with good tactile controls and acceptable bandwidth and voltage capabilities. A cursory search on eBay shows a 2-channel 50 MHz Siglent SDS1052DL with 400 volt inputs. For general technician and hobbyist diagnostics work, that’s a good deal for an instrument that is one step above what a competent DMM can provide.
10x probe with 2v per division setting only showing 9.81V
The scope is clipping (limiting the voltage to protect the internal ADC). Notice CH1 in the bottom left is red. You can fix this by increasing the voltage scale
The OpenHantek6022 manual describes the limits of each voltage scale setting
Edit: I have exactly the same scope. If you need any help, you can reach out to me
This seems like it’s probably within spec (i.e. not a terribly accurate meter, but not far off).
One other thing to keep in mind is that the laptop supply will probably produce a slightly higher voltage without a load – that is, you’d get a value closer to 19v if you were able to measure it while charging the laptop.
I’m just a hobbyist, so take this with a grain of salt, but it matches my understanding.
Yes. First thing I’d do is question the 19.22V as well. Is it really 19.22V or is the power supply maybe outputting an arbitrary voltage up to 20V with no load?
And I don’t really get what the divisions are supposed to be on the graph? Seems it always shows a line at 7.6 div, no matter what you configured? Except in one case.
