Rigol DHO804 Test and Compare Thread

[TurboTom]

Here's a very peculiar finding: https://youtu.be/3TAeSheBUl4
I guess this is due to the AWG using the highest output level configuration via the THS3095 which is most probably supplied via rails that aren't regulated on the AWG piggyback board and that may slightly get affected by the power consumption (balance) of the Rigol's frontend chip which apparently draws some substantial current that may be affected by the input signal / trace position input. What a sentence... When trying to talk in a foreign language while recording a video clip, I'm less eloquent 

[Martin72]

However when working with shunts one does need custom attenuation which it seems Rigol has not allowed for.

When is the magic word..
I don't think that's a flaw that would be game-changing.
For one or the other it might be, then he has to buy another scope, but seen on the majority this is rather void.

[tautech]

However when working with shunts one does need custom attenuation which it seems Rigol has not allowed for.

When is the magic word..
I don't think that's a flaw that would be game-changing.
For one or the other it might be, then he has to buy another scope, but seen on the majority this is rather void.

Yet when you consider this DSO is targeted at the price conscious buyer that's very unlikely to have the additional budget for a $ current probe the lack of custom attenuation might be just that.
These are features the discerning buyer needs to be aware of.

[tautech]

Uh, oh!

Is somebody compiling a new Rigol vs. Siglent comparison chart?

Best you stay away then as it's more complex than getting a USB stick to work. 

[2N3055]

Custom ?
How many ?

There is no such thing, and with 28 options, you don't need it.
For example, we have many differential probes with 200x, 500x, 1000x and even 1500x.
And we have mostly scopes that "offer" only 3 fixed ratios, I find manual input annoying, such a selection as the DHO offers it more practical.

For the record, I have been using custom probe ratios for shunts, homemade probes and for adjusting for probe gain to make them more accurate. Or you can map some signal range to some more round value for the sake of easier reading: you map 0-33mV to 0-100mV as a proxy for 0-100 %...
None of it is "must have" but very useful. It is one of those things that you start inventing new ways of doing things when you know you have that freedom. For instance, a common mode choke becomes impromptu AC current probe in few minutes...

I got used to it with Picoscope....

For standard use with standard probes not needed... This simply shows it is simple, not analytical scope.

[2N3055]

Here's a very peculiar finding: https://youtu.be/3TAeSheBUl4
I guess this is due to the AWG using the highest output level configuration via the THS3095 which is most probably supplied via rails that aren't regulated on the AWG piggyback board and that may slightly get affected by the power consumption (balance) of the Rigol's frontend chip which apparently draws some substantial current that may be affected by the input signal / trace position input. What a sentence... When trying to talk in a foreign language while recording a video clip, I'm less eloquent 

It almost looks like channel offset DAC loses resolution when AWG is enabled. So you get same offset steps as if you are on 2V/div. This might even be code bug, where programer did some mess with numerous variables named offset_xxxx

Does it happen with all channels?

[TurboTom]

It almost looks like channel offset DAC loses resolution when AWG is enabled. So you get same offset steps as if you are on 2V/div. This might even be code bug, where programer did some mess with numerous variables named offset_xxxx

Does it happen with all channels?

Yes it "works" on all channels identically.

I'm certain it's a hardware thing since it requires some particular configuration settings in order to behave like this: The signal must originate from the internal AWG and the AWG has to be configured to high output amplitude so the signal gets routed through the THS3095 output amplifier. I tried feeding in an identical signal from an external AWG which made the effect disappear. Same situation if the input to the low pass network is just terminated with 50 Ohms.

I'm sure shifting the trace vertically is accomplished by feeding an analog current/voltage into Rigol's analog front end chip to modify its internal DC balance. Obviously, this may have an effect on its presumably considerable current consumption (just look at the power consumption/dissipation in general) and change its balance. If this just slightly changes the pre-regulator rail voltages that (once again presumably) also supply the AWG's high-level output amplifier chip, this observation is easily understandable. One would have to measure the internal rail voltages when moving around the traces.

I'm almost certain that the low-level output amplifier of the AWG (AD8009) is supplied by a pre-regulator so the front-end-fed supply imbalance wouldn't affect its performance.

Edit: Just verified - changing any trace position affects the DC output component of the AWG in its higher output amplitude ranges by some ten millivolts -- well done, Rigol 

[iMo]

Custom ?
How many ?

There is no such thing, and with 28 options, you don't need it.
For example, we have many differential probes with 200x, 500x, 1000x and even 1500x.
And we have mostly scopes that "offer" only 3 fixed ratios, I find manual input annoying, such a selection as the DHO offers it more practical.

The famous 1:21 is missing.. 

[ebastler]

However when working with shunts one does need custom attenuation which it seems Rigol has not allowed for. 

You can always use a math channel to convert to your custom scale.
Strike it off the sales pitch list, it's a bit lame.

[tautech]

However when working with shunts one does need custom attenuation which it seems Rigol has not allowed for. 

You can always use a math channel to convert to your custom scale.

Good idea and overlooked.

So now we need investigate Math capabilities.

[2N3055]

However when working with shunts one does need custom attenuation which it seems Rigol has not allowed for. 

You can always use a math channel to convert to your custom scale.
Strike it off the sales pitch list, it's a bit lame.

Yes you can use math channel for that.. I checked datasheet and it has Ax+B function. But you use up math channels fast that way. And they are very limited in a frist place so you will sometimes need several just to achieve some math..

[ebastler]

But you use up math channels fast that way. And they are very limited in a frist place so you will sometimes need several just to achieve some math..

You are obvioously a more advanced user than me -- I don't see myself using up 4 math channels anytime soon. No doubt that one can construe use cases where one could use more, but certainly not a deal breaker for me...

[2N3055]

But you use up math channels fast that way. And they are very limited in a frist place so you will sometimes need several just to achieve some math..

You are obvioously a more advanced user than me -- I don't see myself using up 4 math channels anytime soon. No doubt that one can construe use cases where one could use more, but certainly not a deal breaker for me...

Well filters is in math channel, FFT is in math channel, you have 4 arithmetic operations and few functions. But you can use only 1 at the same time. So if you want to filter out PWM on 2 channels that's 2 math gone and then you multiply them there goes 3rd...
That is my point. Whether people use math channels (how or at all) is up to them..

I'm not even criticizing here, just point out realities.. 
I already pointed out this is very basic scope.
But people say "it also has 4 math channels"  without realizing not all math is created equal. For instance Micsig has filter on every channel. It has only 1 math channel but has advanced math where you can use complex formula. With that you can create many things you cannot with DHO800/900/1000/4000. And that is inexpensive scope. Touchscreen Siglents have 4 math channels that are even more advanced than Micsigs one. But those are advanced analytical scopes and price is quite different....

That being said, DHO has basic math that will do basic stuff we expect.

[Martin72]

So now I have actually "moved" with the Rigol....
And the USB port extender also arrived today, very nice.
Then I can continue with the testing...

[Martin72]

I had tried a little bit with the demo board earlier.
That has two outputs for the test signals, one directly after the DAC, unfiltered, the other with a filter.
The Rigol seems to run a filter, the staircase signal does not look so pronounced, compared to another 12-bit scope.
The last two pictures show both signals from the demo board.

[tautech]

Also, does the "DC1M" on the Siglent meant that it has a 1 MHz lowpass activated?

No.
It's the channel tab showing DC coupled with 1M Ohm input vs 50 Ohm.

[ebastler]

Also, does the "DC1M" on the Siglent meant that it has a 1 MHz lowpass activated?

No.
It's the channel tab showing DC coupled with 1M Ohm input vs 50 Ohm.

Ah, right, that scope can switch in internal 50 Ohm termination. Luxury! 

[Martin72]

And finally, in the second pair of screenshots (including CH2) it looks like multiple overlaid captures are shown in the Rigol screenshot, vs. a single capture on the Siglent, which makes them hard to compare.

in both cases I had simply pressed the stop button and made a screenshot.
Evtl I have a "wrong moment" caught at the rigol, I make another picture.
Oh well, earlier I had again had a crash at the rigol.
For the second time a phenomenon where the scope after plugging in the usb connector from the demoboard just "turns off the light" and nothing happens until you turn it off and on again.

[Martin72]

Picture made, compare it to the siglent pic 19.
And I´ve found out what made the scope crashing, it´s the USB splitter.
When you connect it while the scope is running/booting, the scope will crash.

[Martin72]

The signal is a 50Hz sine with 3.3V amplitude, filtered to one output, to the other directly from the DAC.
If one then resolves accordingly, one can well recognize the stair steps of the non-filtered sine.
And that surprises me, that it looks with the Rigol rather like the filtered signal.
Edit: Ignore the arrow in the picture...

[Martin72]

I´ve tried it a dozen of times* but yes, for you I´ll try it one more time...

*)I had tried to set the same values at the rigol, but that would not work in the sense of no signal on the screen then.

Here is a picture of the siglent sds1104x-e, but with 50µs, I will add 100µs later...

[tautech]

Here is a picture of the siglent sds1104x-e, but with 50µs, I will add 100µs later...

Better comparison would be to be using channels 1&3 to maintain sampling rates, but you already know such as that's how you use your HD. 

[Martin72]

Therefore new screenshots, Rob..

[Fungus]

Picture made

Did you know you can turn off the "DHO804" overlay in your screenshots?

[Martin72]

Now yes, thank you !