Author Topic: Rigol DHO800 memory depth / bandwidth upgrades and noise level!  (Read 18042 times)

0 Members and 1 Guest are viewing this topic.

Offline Bob773Topic starter

  • Newbie
  • Posts: 9
  • Country: us
Rigol DHO800 memory depth / bandwidth upgrades and noise level!
« on: November 06, 2023, 03:04:55 am »
It seems like nobody is talking about how the memory depth cuts back when adding channels… goes from 25Mpts single channel to 10Mpts with 2 channels, why not 12.5?… then drops to 1Mpts with 3/4 channels?!!  This alone is a major deal breaker for me.  I use single triggers with multiple channels very often.  Anybody have insight to why this would be?
« Last Edit: November 07, 2023, 03:27:36 am by Bob773 »
 
The following users thanked this post: egonotto

Offline PELL

  • Regular Contributor
  • *
  • Posts: 56
  • Country: cn
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #1 on: November 06, 2023, 04:25:38 am »
https://www.eevblog.com/forum/testgear/hacking-the-rigol-dho800900-scope/msg5148330/#msg5148330

follow this guide hack the mem-depth, then you can get 10M at 4 channels

I tested it under firmware 1.01 and it works to me  ;)
 
The following users thanked this post: egonotto, Bob773

Online Fungus

  • Super Contributor
  • ***
  • Posts: 17519
  • Country: 00
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #2 on: November 06, 2023, 09:14:18 am »
It seems like nobody is talking about how the memory depth cuts back when adding channels… This alone is a major deal breaker for me.

We've all hacked ours to 50Mpts so we get 10Mpts per channel with all 4 channels on.

That's why it doesn't get much discussion - it's a non-issue!  :)

All you need is an Ethernet cable...

PS: While you're doing that "upgrade" you can also change the bandwidth of a DHO804 to 200MHz (measured).
 
The following users thanked this post: egonotto, Bob773

Online csuhi17

  • Frequent Contributor
  • **
  • Posts: 351
  • Country: hu
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #3 on: November 06, 2023, 09:53:38 am »
Yes, but why do we lose 21M of memory on the 800 and 10M on the 900?
Fnirsi oscilloscope = waste&regret
 

Online Fungus

  • Super Contributor
  • ***
  • Posts: 17519
  • Country: 00
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #4 on: November 06, 2023, 10:40:47 am »
Yes, but why do we lose 21M of memory on the 800 and 10M on the 900?

All we know is that it's a marketing decision not a hardware thing.
 
The following users thanked this post: ebastler

Offline magic_smoke

  • Newbie
  • Posts: 7
  • Country: 00
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #5 on: November 06, 2023, 11:09:25 am »

We've all hacked ours to 50Mpts so we get 10Mpts per channel with all 4 channels on.

That's why it doesn't get much discussion - it's a non-issue!  :)

All you need is an Ethernet cable...

PS: While you're doing that "upgrade" you can also change the bandwidth of a DHO804 to 200MHz (measured).


Is there a downside to these upgrades? That goes against the official maximum bandwidth (100Mhz) afaik.
 

Online ebastler

  • Super Contributor
  • ***
  • Posts: 7375
  • Country: de
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #6 on: November 06, 2023, 12:05:59 pm »
Is there a downside to these upgrades? That goes against the official maximum bandwidth (100Mhz) afaik.

It is the "nominal 100 MHz" bandwidth. Turns out the actual bandwidth of the scope's front end is larger than stated.

(Which is good and bad -- you need to be a bit more careful about potential aliasing when using the scope in 4-channel mode, i.e. with only 312 MSa/s.) The early DHO800 firmware even had a provision for an "official" bandwidth upgrade, but it has apparently disappeared with the more recent firmware release. I don't think anybody has reported side effects beyond the potential aliasing -- which should apply to the unmodified DHO814 as well.

There is also an alternative hacking approach which lets the DHO800 think it is a DHO900, which would potentially enable even higher bandwidth. But this comes with side effects, namely offset error (at least in some units) which don't go away upon auto-calibration. It also displays the digital channel control in the bottom row of the screen, which remains non-functional of course because some hardware is missing in the DHO800. Hence, this is not a recommended way to "upgrade" your scope at this time.
 
The following users thanked this post: magic_smoke

Online Fungus

  • Super Contributor
  • ***
  • Posts: 17519
  • Country: 00
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #7 on: November 06, 2023, 01:31:12 pm »
There is also an alternative hacking approach which lets the DHO800 think it is a DHO900, which would potentially enable even higher bandwidth.

Yep. I've had my DHO804 up to 280Mhz measured bandwidth.

But this comes with side effects, namely offset error (at least in some units) which don't go away upon auto-calibration.

I'm sure that's fixable with a bit more hacking, but... 200MHz is enough for me.  :)
 

Offline Warhawk

  • Frequent Contributor
  • **
  • Posts: 836
  • Country: 00
    • Personal resume
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #8 on: November 06, 2023, 01:42:51 pm »
Do I get it right that to get the extra memory you need to make the scope thinking that it is the 900 series?

Online Fungus

  • Super Contributor
  • ***
  • Posts: 17519
  • Country: 00
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #9 on: November 06, 2023, 02:02:37 pm »
Do I get it right that to get the extra memory you need to make the scope thinking that it is the 900 series?

No.
 

Online Fungus

  • Super Contributor
  • ***
  • Posts: 17519
  • Country: 00
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #10 on: November 06, 2023, 02:05:43 pm »
I'm sure this question is going to be asked a million times over the next few years so I'll put this here as a reference.

My 'scope says "DHO804" on the front. I've modified the bandwidth and memory options.

Here's my new "About" screen:


Here's my new "Options" screen:


Here's my new memory depth with 4 channels enabled:


Here's my measured bandwidth:


nb. I'm using the bandwidth math from here and Martin72 confirmed the "100Mhz" bandwidth with a 200Mhz sine wave here.

(I don't have a SIGGEN that goes that high so I can only do rise times  :( )

All changes are harmless and 100% reversible. There are NO side effects.

Notice the difference in rise time and the shape of the yellow waveform between the last two images. That's because 321.5MSa/s isn't enough for a 197Mhz 'scope, ie. I'm seeing some aliasing.

(Solution: (a) Be aware, and (b) Turn off some channels when viewing fast signals and see if anything changes. If it does? Aliasing!)
« Last Edit: November 06, 2023, 05:18:04 pm by Fungus »
 
The following users thanked this post: egonotto, ebastler

Online Fungus

  • Super Contributor
  • ***
  • Posts: 17519
  • Country: 00
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #11 on: November 06, 2023, 02:23:09 pm »
I just removed my bandwidth upgrade (proof that it's reversible!):


Now my measured bandwidth is 128MHz:


I have no idea where Rigol get their "70Mhz" from... but don't complain!  :D
« Last Edit: November 06, 2023, 05:17:22 pm by Fungus »
 
The following users thanked this post: egonotto

Offline Warhawk

  • Frequent Contributor
  • **
  • Posts: 836
  • Country: 00
    • Personal resume
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #12 on: November 06, 2023, 03:24:49 pm »
Thanks, I lost interest in the other thread when it went off-topic with the gibbs' ears. I did not know that there is generally an option for extended memory. I found it now. I am happy for you and others. This was really annoying limitation.

Online ebastler

  • Super Contributor
  • ***
  • Posts: 7375
  • Country: de
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #13 on: November 06, 2023, 03:27:59 pm »
I am wondering whether the 0.45 factor to convert risetime to bandwidth is a bit too aggressive here. Historically a factor of 0.35 has been used, which I understand is correct for simple 1- or 2-pole bandwidth limiting filters. The recent transition to 0.45 is justified when steeper, higher-order filters are used for bandwidth limiting.

But looking at the nice step response of the DHO800, without overshoot and ringing, I would assume that it uses a relatively "soft" lowpass filter. So maybe the classical factor of 0.35 is closer to the truth here? It would put the bandwidth closer to Rigol's specs.

Anybody here with access to a tunable sine wave generator which goes up to 200 MHz to measure the bandwidth in the frequency domain? And ideally an even higher bandwidth scope to check the generator output amplitude?  8)
 

Offline Warhawk

  • Frequent Contributor
  • **
  • Posts: 836
  • Country: 00
    • Personal resume
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #14 on: November 06, 2023, 04:17:35 pm »
Anybody here with access to a tunable sine wave generator which goes up to 200 MHz to measure the bandwidth in the frequency domain? And ideally an even higher bandwidth scope to check the generator output amplitude?  8)
I do have the access but don't have the scope  :-DD

Offline TomKatt

  • Frequent Contributor
  • **
  • Posts: 529
  • Country: us
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #15 on: November 06, 2023, 04:29:44 pm »
I'm a bit surprised the shared DAC and 312smps at 4 channels isn't more of a concern.
Several Species of Small Furry Animals Gathered Together in a Cave and Grooving with a PICt
 

Offline Serg65536

  • Regular Contributor
  • *
  • Posts: 134
  • Country: ua
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #16 on: November 06, 2023, 04:53:17 pm »
I'm a bit surprised the shared DAC and 312smps at 4 channels isn't more of a concern.
It's NOT a DAC, but ADC!
And you are right: 250 Msampl/s, as other brands have at a slightly higher prices is much better! And don't forget about noise and quantization step: it's also much higher in competition. So yee.. it's not a good scope...  :-DD  :popcorn:
 

Offline TomKatt

  • Frequent Contributor
  • **
  • Posts: 529
  • Country: us
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #17 on: November 06, 2023, 04:55:53 pm »
Yes ADC - duh.  The DST time shift always puts me off.

And I didn't mean to infer the lack of dual ADC's made it unusable, only that sampling rate seems to be an otherwise high priority spec.
« Last Edit: November 06, 2023, 05:01:13 pm by TomKatt »
Several Species of Small Furry Animals Gathered Together in a Cave and Grooving with a PICt
 

Online 2N3055

  • Super Contributor
  • ***
  • Posts: 7466
  • Country: hr
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #18 on: November 06, 2023, 05:08:59 pm »
I'm a bit surprised the shared DAC and 312smps at 4 channels isn't more of a concern.
It's NOT a DAC, but ADC!
And you are right: 250 Msampl/s, as other brands have at a slightly higher prices is much better! And don't forget about noise and quantization step: it's also much higher in competition. So yee.. it's not a good scope...  :-DD  :popcorn:

It is not an absolute value competition, but Sample rate / BW combination.
If you have 130 MHz -3db BW and 312.5 MS/s then it is fine.
If you have 200 MHZ BW and 312 MS/s you are undersampling.


What other scope are you refering to specifically?
"Just hard work is not enough - it must be applied sensibly."
Dr. Richard W. Hamming
 

Online ebastler

  • Super Contributor
  • ***
  • Posts: 7375
  • Country: de
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #19 on: November 06, 2023, 05:12:53 pm »
The ADC sample rate has been discussed at some length -- I don't recall in which of the various DHO800/900 threads.

The unusual thing with those scopes is that it is not the ADC which limits the acquisition rate, but apparently the FPGA or the FPGA/DRAM interface. The same ADC runs at 2 GSa/s in the DHO1000, but only 1.25 GSa/s in the 800/900. And when one activates the digital channels in the DHO900, the ADC even falls back to 625 MSa/s in total -- shared across 1 to 4 channels depending on the setup. The latter fact is unfortunately not disclosed in Rigol's datasheet.

So yes, there are tradeoffs which one needs to be aware of. And the advertised 250 MHz bandwidth in the DHO924 is pushing things a bit far for my taste. But overall, it's still a very nice package, even with the logic anaylzer enabled. Compare the acquisition rates with a Saleae Logic Pro 16, for example. (And compare the price tag too.)
 
The following users thanked this post: Serg65536

Online Fungus

  • Super Contributor
  • ***
  • Posts: 17519
  • Country: 00
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #20 on: November 06, 2023, 05:17:01 pm »
Anybody here with access to a tunable sine wave generator which goes up to 200 MHz to measure the bandwidth in the frequency domain? And ideally an even higher bandwidth scope to check the generator output amplitude?  8)

Yes, and 0.45 works!

https://www.eevblog.com/forum/testgear/hacking-the-rigol-dho800900-scope/msg5132343/#msg5132343
 

Online ebastler

  • Super Contributor
  • ***
  • Posts: 7375
  • Country: de
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #21 on: November 06, 2023, 05:31:37 pm »
Anybody here with access to a tunable sine wave generator which goes up to 200 MHz to measure the bandwidth in the frequency domain? And ideally an even higher bandwidth scope to check the generator output amplitude?  8)

Yes, and 0.45 works!
https://www.eevblog.com/forum/testgear/hacking-the-rigol-dho800900-scope/msg5132343/#msg5132343

Thanks! That is good news indeed -- and on the other hand a call to paying some attention to aliasing.

I just notice that you had already mentioned Martin's measurements a few posts earlier. I had overlooked the further text below the pictures there, sorry.
 

Online Fungus

  • Super Contributor
  • ***
  • Posts: 17519
  • Country: 00
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #22 on: November 06, 2023, 05:37:24 pm »
I'm a bit surprised the shared DAC and 312smps at 4 channels isn't more of a concern.

It is what it is. Just be careful and switch off a channel or two if there's any doubt.

You're also free to leave it at Rigol's "70Mhz" if you want to - 128Mhz is still a decent bandwidth and the 'scope is still a bargain for $399.

nb. You can switch between 70Mhz/100Mhz options in under a minute. The longest part is waiting for it to reboot.
 

Online voltsandjolts

  • Supporter
  • ****
  • Posts: 2549
  • Country: gb
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #23 on: November 06, 2023, 05:42:25 pm »
Perhaps leaving the bandwidth at 70MHz (128MHz actual, apparently) reduces scope noise, compared to 100MHz (200MHz actual, apparently)? I haven't seen anyone posting measurements of ENOB before/after upgrade.
 
The following users thanked this post: egonotto

Online Fungus

  • Super Contributor
  • ***
  • Posts: 17519
  • Country: 00
Re: Rigol DHO800 memory depth is a deal breaker
« Reply #24 on: November 06, 2023, 06:23:25 pm »
Perhaps leaving the bandwidth at 70MHz (128MHz actual, apparently) reduces scope noise, compared to 100MHz (200MHz actual, apparently)? I haven't seen anyone posting measurements of ENOB before/after upgrade.

Leaving it at "70Mhz" is a very sensible option if you often work with four high speed signals or simply want the best possible signal fidelity.

I can easily measure noise before/after. Hang on...
 
The following users thanked this post: Serg65536


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf