Rigol DHO800/900 new firmware v00.01.04.00.02 2024/11/1

[TheoB]

I measure the input capacitance of the scope as 14.7pF (specification says 15pF). With 50 Ohm that alone gives a bandwidth corner of 215MHz. I don't see how it makes sense to look for any higher bandwidth with a 15pF input loading. Yes it can be done by terminating with 25 Ohm. But it's not very practical. I've measured with 50 Ohm cable and external termination at input of the scope.
What you propose is to add a Schmitt trigger buffer in front of the input and put multiple in parallel. Yes that will lower the drive impedance below 50 Ohm to measure the voltage bandwidth of the scope input. But you still need a non-standard load at the input of the scope to benefit from it.
Conclusion, anyone interested in over 200MHz bandwidth (rise/fall time <1.75ns) should look for a scope with a 50 Ohm input, not these cheap Rigols 

[norbert.kiszka]

I measure the input capacitance of the scope as 14.7pF (specification says 15pF). With 50 Ohm that alone gives a bandwidth corner of 215MHz.

After removing LC filters in two channels I have 0.8 ns rise time.

With 400 MHz square wave I have same amplitude as power supply of this generator (avg acquisition, so noise wasn't included in measurement).

With passive probes (350 MHz), I get around 1 ns.

Without compensation, You will always have R(L)C filter, which lowers bandwidth.

[TheoB]

The LC you removed, does that define the 15pF input load? Have you measured the new input capacitance?
You normally would want to have nice 20dB/decade rolloff. With an LC present at the input that could cause ringing in the transient signal. The classical BW=350/(Trise) is only true when the input is 20dB/decade. The first 90% transition point is reached after -log(0.9)=0.1 time constant whereas the second crossing point at 10% is reached after -log(0.1)=2.3 time constants. The rise time measured is thus 2.2 time constants apart. w=2.2/Tr or in Hz, f=2.2/(Tr*2*pi)=0.35/Tr. But the scope is not a nice first order transfer function with it's anti-alias filters and sin(x)/x interpolation. I just checked the BW using a sine waive and it looks much higher than 216MHz. Perhaps there is digital compensation for the limited analog bandwidth 

[norbert.kiszka]

With an LC present at the input that could cause ringing in the transient signal.

There are two LC filters for each channel - not in frontend, but after it. Less than one centimeter from ADC. First stage is 250 MHz (at least in 924) and second one is 800 MHz. After removing it (and correcting compensation) I have higher bandwidth, lower rise/fall time and signal is not distorted (but noise is higher, which can be cancelled with avg acquisition). They didn't use first better inductors.

Perhaps there is digital compensation for the limited analog bandwidth

Im pretty sure there is no such thing in those scopes.

does that define the 15pF input load

Do You know what is compensation in probes?

Looking at the documentation of similar series DHO4000 with same front end and 800 MHz bandwidth, it's even more. 19 pF to be exact.

[norbert.kiszka]

In attached print-screen You can see I recreated situation with 15 pF input and mentioned compensation. Since this is a simulation with ideal elements with 0% tolerance and generator rise time is 0 ps, at the (right) capacitor, You also have exact same rise time, which is also 0 ps.

Edit: I know, this is an incredible old software running under Wine (which allows me to run software under Linux but written for a crappy Windows).

[TheoB]

Yes I know how that works. For passive probes I don't expect anyone to use probes due to their limited bandwidth and issue with ground lead inductance. I assume all is measured with 50 Ohm above 200MHz (or with Fet probe, also terminated with 50 Ohm).
But you could in principle make a 1:10 attenuator at the input of the scope BNC connector and place a 50 Ohm cable termination in front of it (termination the coax). Hadn't thought about that. That would then be 10M||1.5pF.
1.5pF with 50 Ohm gives a bandwidth of 2GHz. But have you tried this actually to make any practical use of your mod? It looks to me that you show more "synthetic" pictures of rise/fall times. Note something that is 50 Ohm suitable.

[norbert.kiszka]

Best passive probes reaches 1.5 GHz bandwidth and active probes much much more. It all depends how You use them.

In case of direct connection, I don't need to have any termination or compensation, because at 400 MHz wavelength is 0.75 m (75 cm). Between IC in generator and IC in frontend is much much shorter distance.

With 50 ohm termination I was using one meter BNC cables without any significant loss of bandwidth. With those, I need to have at least 5 GHz to see any problems, but this is way outside of 924S capabilities.

[iMo]

Build yourself the Z0 probe - 1k (or similar) resistor is the tip, then 50ohm coax and 50ohm termination at BNC of your scope. No compensation.. Works beyond 1GHz afaik.

[TheoB]

The Z0 probe is still limited by the 15pF input scope capacitance unless you use a scope with builtin termination. As Norbert correctly mentions there are passive probes that go far beyond the scopes bandwidth. I was not aware. I work either with RF or rather low frequency signals (<100 MHz). But using a passive probe still puts a significant load (15 pF again) on your measurement point thereby lowering the measurement bandwidth again. Measuring strong digital drivers should work though up to 200MHz.

[norbert.kiszka]

But using a passive probe still puts a significant load (15 pF again) on your measurement point thereby lowering the measurement bandwidth again. Measuring strong digital drivers should work though up to 200MHz.

Depends on what exactly You try to measure. Mentioned 74AC14 and mentioned 74LVC14AD can drive 15 pF - with 400 MHz as I said.

If You really need very low capacitance input - best way will be with making Your own passive probe - probably differential. Three fast opamps. In a economic scenario, with single opamp You will have fast differential probe with impedance 0.5-5 pF and around 10 kΩ.

[norbert.kiszka]

OPA694ID will do the job. 690 MHz with 2V/V gain and with 1/1 1.5 GHz. With one, You will have 1.5-3 pF and with three it will be around 1 pF.

Edit: OPA847 is also nice and cheap enough.

[gf]

Build yourself the Z0 probe - 1k (or similar) resistor is the tip, then 50ohm coax and 50ohm termination at BNC of your scope. No compensation.. Works beyond 1GHz afaik.

Any feed via coax cable and feed-through terminator to the Hi-Z input is limited to roughly 424 Mhz (-> 25Ohms/15pf lowpass). To prevent the unavoidable reflections (due to the input capacitance) from bouncing back and forth multiple times, the coax should also be source-terminated with 50 Ohms. The same basically applies to Z0 probes as well. So IMO better use a 500R/56R divider at the Z0 probe tip, and not just a 500R series resistor, if the scope end of the cable is connected via feed-through terminator to a Hi-Z input.

[psykix]

I'm green behind the ears with this scope..

I have DHO804 and a while back modded it to DH814 and added wifi dongle etc. It's been a few months and I just went to update firmware to 1.04 but the scope says no firmware found.

Is there a way to update over wifi?

Thanks!

[norbert.kiszka]

You can use adb and update files manually.

apk files (apps) can be installed via pm install xyz.apk

[Fungus]

I have DHO804 and a while back modded it to DH814 and added wifi dongle etc. It's been a few months and I just went to update firmware to 1.04 but the scope says no firmware found.

When you select the file in the update dialog you have to make sure the little circle next to the name is marked.

[psykix]

I have DHO804 and a while back modded it to DH814 and added wifi dongle etc. It's been a few months and I just went to update firmware to 1.04 but the scope says no firmware found.

When you select the file in the update dialog you have to make sure the little circle next to the name is marked.

I think I am missing something here.. I was pressing the Upgrade button on the scope itself and it just says No Firmware Available

I don't know where the update dialog that you refer to is located?

Thanks!

[norbert.kiszka]

As far I remember in my case, this never worked: main menu (bottom left corner) -> upgrade.

But somewhere in the file manager, there was a option to upgrade.

To test it, I have to switch SD card to another, but currently I will not, because Im working on something completely else.

[Fungus]

I don't know where the update dialog that you refer to is located?

Menu->Storage->Upgrade.

[psykix]

I don't know where the update dialog that you refer to is located?

Menu->Storage->Upgrade.

So how do I get the firmware file then? Do I need to download manually and put it somewhere? I followed a link to download and it takes me to Rigol passport thing and it wants login details.

Sorry if I'm being dense!

[norbert.kiszka]

USB stick. Put unpacked .GEL file into.

[psykix]

USB stick. Put unpacked .GEL file into.

Thanks! For some reason I expected it to be able to update over the air!

[norbert.kiszka]

USB stick. Put unpacked .GEL file into.

Thanks! For some reason I expected it to be able to update over the air!

Magic... That scope "by default" is not connected to the internet. You can do it via Ethernet or via usb wifi dongle.

[psykix]

USB stick. Put unpacked .GEL file into.

Thanks! For some reason I expected it to be able to update over the air!

Magic... That scope "by default" is not connected to the internet. You can do it via Ethernet or via usb wifi dongle.

Got there in the end - and my wifi dongle etc still works.

Thanks both!

[Malmen]

How long will it take to update via wifi dongle? Its been "updating" for 30 minutes now. Will I mess things up if I disturb it?

[norbert.kiszka]

How long will it take to update via wifi dongle?

I think this scope is not designed to update firmware via wifi.