Rigol's New Function Generator and DMM: DG800Pro / 900Pro / DM858

[tszaboo]

I'm surprised the DG800Pro series contains the hardware for the higher sample rates of the DG900Pro series.

The teardown ( https://www.eevblog.com/forum/testgear/rigol-new-dg822pro-teardown ) shows a TI DAC38J82, which is a 2.5GSPS dual channel DAC (1.25GSPS per channel).

Strangely, TI has a pin-compatible 1.6GSPS part in the same lineup at almost half the price: DAC37J82. This lower spec part would make more sense in the DG800 series, which tops out a 625 MSPS/channel, or 1.25GSPS combined.

I wouldn't be surprised if later DG800Pro builds started coming with the cheaper, lower sample rate variant DAC. Seems like Rigol is leaving money on the table by putting the faster DAC in the slower model when a cheaper variant is available.

Maybe it does. These lower specced parts are sometimes the same die, that doesn't meet the specifications, but you can unlock the extra speed or bits if you use the same code.

[thm_w]

Maybe it does. These lower specced parts are sometimes the same die, that doesn't meet the specifications, but you can unlock the extra speed or bits if you use the same code.

I get what you are saying but how is that a "maybe", the teardown photo clearly shows the part number as being DAC38J82I. If it were a DAC37 it would be marked DAC37.
https://www.ti.com/product/DAC38J82 $68/ea

[DaneLaw]

Any way to validate if the DG821PRO are able to obtain the full sample rate of the DG900PRO series.
Which on paper can muster 1.25GSa/s - 16bits - 200MHz -32Mpts.

A DG922PRO (32M) is here in Europe a tad under 1900 USD (incl. 25% EUVAT) if you can land those specs with the entry [DG821PRO] at a fraction of the price, and if the performance is linear, that would elevate the demand quite a bit.
Not least in the US where the prices seem a tad lower. https://www.saelig.com/product/dg821-pro.htm
At the end of the day, a modern sig-gen with "1.25GSa/s - 16bits - 200MHz -32Mpts" at fx 399$ certainly doesn't hang on the trees 

The big 7" touchscreen (1024x600) on these modern signal generators that nowadays are packed with settings & features, is a very welcome addition, as it can get quite cumbersome on a small display like fx Rigol's current DG1000 series, which Rigol is still selling' that out of the gate are higher priced, but also quite telling about the extensive jump that has occurred on the fundamentals.
DG1022A (700€) - 100Msa/s - 14bits - 25Mhz - 4K.  https://rigolshop.eu/function-generator-dg1022.html

DG900PRO datasheet
https://static.eleshop.nl/mage/media/downloads/DG900Pro_DataSheet_EN.pdf
Manual
https://static.eleshop.nl/mage/media/downloads/DG900Pro_UserGuide_EN.pdf

Miscellaneous DG821PRO screendumps from December/January https://tinyurl.com/4ddye75e *old fw it came with.

Of the DG821PRO 3x USB sockets. (1x USB-A _ 1x USB-B  1X USB-C) it's the USB-B in the back that supports TMC (Test & Measurement Class]
while the 1xUSB-A in the front can act as a host for fx dongles etc... at least the impression I get from the manual, but I haven't tried it, only a wireless mouse so far, and that part worked great.
I haven't checked if the USB-C (main power) also can tolerate data, alongside power with a Power Delivery/PD passthrough dongle.

[Construct]

Any way to validate if the DG821PRO are able to obtain the full sample rate of the DG900PRO series.
Which on paper can muster 1.25GSa/s - 16bits - 200MHz -32Mpts.

Both DG800Pro and DG900Pro specify the maximum sequence data rate as 312.5 MSa/s.

The TI DAC part has a maximum input data rate of only 1.23 GSa/s, so technically it can't (shouldn't) be used with a full 1.25 GSa/s input anyway.

I suspect they're using the interpolation feature of the DAC to arrive at the advertised sample rate.

[TurboTom]

It's a pity that Rigol didn't use the opportunity when designing the DG XXX Pro series to eliminate the shortcoming of CH2 to be disabled whenever the frequency counter function is active (as the manual states) -- this "feature" was also present on the "non-Pro" models. Except for the higher frequency capabilities, the bigger screen and the "funny" fonts, this new model appears to provide little (technical) improvements over the older specimen...

Whatsoever, if the "uphack" is and stays possible, the entry level "Pro" model probably isn't a bad choice, provided Rigol won't "disimprove" their firmware too much... 

[DaneLaw]

It's a pity that Rigol didn't use the opportunity when designing the DG XXX Pro series to eliminate the shortcoming of CH2 to be disabled whenever the frequency counter function is active (as the manual states) -- this "feature" was also present on the "non-Pro" models. Except for the higher frequency capabilities, the bigger screen and the "funny" fonts, this new model appears to provide little (technical) improvements over the older specimen...

Whatsoever, if the "uphack" is and stays possible, the entry level "Pro" model probably isn't a bad choice, provided Rigol won't "disimprove" their firmware too much... 

What exactly are you looking for, that you feel are lacking on this 16-bit entry PRO series?

DG811, wasn't it 10Mhz sine, 5MHz square, 200kHz Ramp, 5MHz pulse, 5MHz harm and amplitude 1-2.5 vpp, seq 2K to 30MSa' out of the box (125Msa).
That's all quite a bit higher with 25MHz sine and 20MHz square, 1MHz Ramp(5x), 10MHz Pulse, 10MHz harm and 1mV- 10Vpp amplitude (20vpp HighZ) seq 1μSa to 312.5 MSa...
it's pretty much all around the basic specs got quite a hump up on DG821PRO vs DG811, at least as far as I can tell.

Alongside a modern 7" touchscreen, num-dial, and also "mobility" so your not locked down to an AC socket, which is a feature that I value a lot.
I don't recall how many harmonics orders and sequences you could stack on DG811 out of the box.? but on DG821 I recall its 20 harm and 64 seq.
Also looks to be quite a difference in square rise-time that claims around 9ns on DG811 & co and 3ns on DG821.
I'm curious what else one could expect on an entry signal gen in this price bracket, that you feel are missing, - as all the fundamentals got elevated quite a bit.? 

// Seems its 8 harmonics on DG811.

[TurboTom]

I'ld just expect that when buying a state-of-the-art 2-channel AWG with an additional frequency counter function, to be able to use the two AWG channels while the counter is activated. What I don't expect is that one of the AWG channels drops out of operation when I enable the frequency counter.

See here, "Tip" on pg. 79

[gbix]

I'ld just expect that when buying a state-of-the-art 2-channel AWG with an additional frequency counter function, to be able to use the two AWG channels while the counter is activated. What I don't expect is that one of the AWG channels drops out of operation when I enable the frequency counter.

after activation counter the 2nd channel is go off

[DaneLaw]

I'ld just expect that when buying a state-of-the-art 2-channel AWG with an additional frequency counter function, to be able to use the two AWG channels while the counter is activated. What I don't expect is that one of the AWG channels drops out of operation when I enable the frequency counter.

See here, "Tip" on pg. 79

I get that, but ain't that the norm on Rigols UI, also recall that was the case on the more expensive & older models, that it limited ch2 when activating the counter?.. if it's critical for you, to have a counter and also 2 channels.
I would look at a dedicated freq-counter..the counter seems more like an afterthought in these entry signal gens, and far from the best or you can hope that Rigol optimize their UI. (but don't hold your breath, its Rigol)

But as you can compare, there are upgrades across the plate on the fundamentals in DG821PRO (1ch) vs DG811 (1ch).. the jump is quite significant out of the box but you need to dig down into the datasheet and compare.

[DaneLaw]

A compare DHO814 & DG822PRO cirquit board.

[csuhi17]

I'ld just expect that when buying a state-of-the-art 2-channel AWG with an additional frequency counter function, to be able to use the two AWG channels while the counter is activated. What I don't expect is that one of the AWG channels drops out of operation when I enable the frequency counter.

See here, "Tip" on pg. 79

The simplest solution would be to buy 2 DG811Pro and upgrade them to the DG922Pro level.

Because then it has a 2x7" display and 4CH 200Mhz AWG or 3CH AWG + 1CH counter or 2CH AWG + 2CH counter.

I don't think anyone will be able to buy such a set for ~800 USD...

If I might be wrong and there is such a 4-channel 200MHz version from another manufacturer, I would be happy to look at it.
Of course, for a price of ~800 USD.

[TurboTom]

I've been just putting this issue up since I wanted to point out a missed opportunity by Rigol (among many others...). I.e. first deliver a "finished" firmware for the "non-Pro" versions  . If I need a decent frequency counter, I use one of my hp 53310A MDAs. None of the AWG-integrated counters stands a chance against these "dinosaurs"!

[csuhi17]

I understand, but for such little money, don't expect more, I think there is no match for the price-value ratio, or at least I haven't found one.
If you really need the amount of channels mentioned above, there is a very good method for it.

It is possible that this is not a software limitation.
If it only costs that much because of the limitation, then I'm happy with it. Personally, I rarely use the counter.
If I look at its specification and compare it with other AWG
A 1000 USD one with a similar base, but the AWG part is max 120MHz, and the counter only measures up to 200MHz. And less memory.

This includes up to 200MHz AWG and up to 1GHz Counter.
I can't say which AWG has a clearer signal.
And Rigol certainly has bugs.

I've been just putting this issue up since I wanted to point out a missed opportunity by Rigol (among many others...). I.e. first deliver a "finished" firmware for the "non-Pro" versions  . If I need a decent frequency counter, I use one of my hp 53310A MDAs. None of the AWG-integrated counters stands a chance against these "dinosaurs"!

Ah, it was a shame to read the user manual hp 53310A MDA. My head hurt.
I don't think that the simple counter function is worth comparing with a full-fledged "Modulation Domain Analyzer".
I didn't understand half of what I read in the manual.
It has many more features.

Unfortunately, if it is no longer produced, you cannot get it.
If someone has one, I don't think they will sell it if they can't find a better one, or if they're not forced to. Or he doesn't need it anymore.

[csuhi17]

I have already found an annoying bug.
I turn on the MOD/SUM function on channel 1.
It can be selected at the bottom in the Waveform drop-down tab.
Sine, Square, Ramp, Noise, Arb, CH2.
Arb in the manual is "Arb - standard arbitrary waveform other than DC"
If I understand correctly, I should be able to choose the one I want to see on the basic wave from the 148 built-in waves.
There is no menu item to set it up.

If I select the other channel "CH2", I can see the set wave.
If I want to change CH2 to Arb, it says:
"The waveform disallows overlay to another channel, and the other channel overlay waveform has been changed".
It sets CH2 to what I selected, but changes CH1's mod/waveform to something else. often to Sine.

However, if I set CH2 first and then change CH1's mod/Waveform to CH2, it accepts it.
But after that, no matter how many times I want to set the CH2 to another Arb, it keeps dropping it. Leave only when selecting basic waves.

[tszaboo]

A compare DHO814 & DG822PRO cirquit board.

They seem to be using the same screws.

[Njk]

It seems arbitrary waveforms are no longer supported. Only those built-in Arbs. Also, the number for jitter looks copy-pasted from the previous series. DG8/9/2k has 250 Mhz DDS clock and 200 ps rms jitter (whatever this means). The Pro uses 1.25 GHz clock but the jitter figure remains the same. Anyway the copy-past method is safe because the number can't be worse with higher clock speed. For the first time, the accuracy of the frequency counter is documented.

[DaneLaw]

A compare DHO814 & DG822PRO cirquit board.

They seem to be using the same screws.

Your sharp 
its the same board template.. so a lot of rehash of the secondary components/sockets DHO800/900, DG800/900PRO, DM858/E.
Quite populated these signal gens and not much seems to be unpopulated (if any) but on paper, the DG800PRO allows only up to 8M (optional) and 2M native while DG900PRO its 32M optional and 16M native..
A board picture of the DG900 would be helpful to see if any of the chip differs.

[csuhi17]

A compare DHO814 & DG822PRO cirquit board.

Sorry, I don't have a better camera, I only have a microscope, but I should have take a couple of 100 pictures with it.
DG912Pro and DM858.






The only difference I found.
Under the add-on card.
On the DG822Pro and DM858
SEC 319 BYMA K4B4G0846E
On the DG912Pro
SEC 322 BYMA K4B4G0846E

[DaneLaw]

A compare DHO814 & DG822PRO cirquit board.

Sorry, I don't have a better camera, I only have a microscope, but I should have take a couple of 100 pictures with it.
DG912Pro and DM858.

(Attachment Link)

(Attachment Link)


The only difference I found.
Under the add-on card.
On the DG822Pro and DM858
SEC 319 BYMA K4B4G0846E
On the DG912Pro
SEC 322 BYMA K4B4G0846E

Looks good csushi17  thanks for sharing.. 
You took the plunge on both DG912PRO & DM858..?

Whats your impression sofar? 
Took a web-look at the different chip-markings your posted (as different) but my web-kungfu is weak sauce, so nothing popped up-

[csuhi17]

Looks good csushi17  thanks for sharing.. 
You took the plunge on both DG912PRO & DM858..?

Whats your impression sofar? 
Took a web-look at the different chip-markings your posted (as different) but my web-kungfu is weak sauce, so nothing popped up-

I think it could be.
https://semiconductor.samsung.com/dram/ddr/ddr3/k4b4g0846e-byk0/
I don't know what SEC 322 ls SEC 319 can mean.


Yes.
I haven't tested them much yet.
This is my first serious instrument group in this category.
Before that, I used FY6800 and simple multimeters, Owon's 20000 count DMM.

The display is large and clearly visible and readable. Due to the flat design, it fits more easily on the table, and if necessary, I can take it with me because of the USB-C.
But it should be noted that this is much more vulnerable than a manual instrument.

The DM858 does not contain a fan and is quite quiet.
However, the back panel is the same as the DHO900, so dust can fall through the holes, only a thin plastic covering sheet was added.

I didn't notice a big difference in its accuracy, after being cold and warming up for several hours.
I have only used it to measure voltages so far.
I somehow managed to freeze it once, I don't remember what I did.

The flat design has two disadvantages, one is that you can easily lean forward if you pull the measuring cable, and the other is that it is difficult to plug it in, so it requires two hands.

You can see what it can do in the description, and as far as I can tell, it does.

The main reason for my purchase is its flat design, large display, and the hope that I can make loop recordings with it.
I mean when I record the change in voltage in smaller groups of, for example, 500k points for several days with the highest possible accuracy at 125 Read/s.
Unfortunately, this function is not included.
There is some kind of "Record" for the Histogram in the description, but it doesn't appear for me. It is possible that it is not yet complete and this function has been removed from it.
I hope Rigol will improve on it.

I can tell you after a short search that this is a very basic DMM in its class.
Someone wrote earlier that the DMM chip in it is not the best.
TruRMS is 8kHz and its accuracy in the measurement ranges is poor even at lower frequencies.


The DG912Pro has made a positive impression so far.
You can see his abilities in his manual.

There is no easy way to load your own wave yet, one forum member described the currently usable method, but it is quite fiddly and incomplete, if I remember correctly it cannot be parameterized.

I couldn't measure how accurate the 5,000,000,003 kHz 1,000.01V sine wave is, I don't have the right instrument.

It is easy to use, I managed it even after reading the manual.

However, now that it turned out that the 800pro and 900pro are the same inside, I'm a little saddened, but I'm also happy at the same time.
So maybe if someone buys the cheaper version and hacks it, they will be fine with it.

What bothers me is that the program is not on it like it is on the DHO800/900 memory card, so I can't make a backup of it.

For me, these two will completely cover my interests if I get the two things I'm missing.

If anyone has any questions or requests, I will be happy to answer them and test them.

Unexpectedly, I can't say anything else about the two devices.


+Its fan is horrible.

[tszaboo]

A compare DHO814 & DG822PRO cirquit board.

They seem to be using the same screws.

Your sharp 
its the same board template.. so a lot of rehash of the secondary components/sockets DHO800/900, DG800/900PRO, DM858/E.
Quite populated these signal gens and not much seems to be unpopulated (if any) but on paper, the DG800PRO allows only up to 8M (optional) and 2M native while DG900PRO its 32M optional and 16M native..
A board picture of the DG900 would be helpful to see if any of the chip differs.

I mean there is nothing shared between these two designs. Not the main CPU, not the power supply. It's clearly done by two different teams, one using silkscreen the other doesn't. Somehow the oscilloscope doesn't have any electrolytic capacitors as well.

A compare DHO814 & DG822PRO cirquit board.

Sorry, I don't have a better camera, I only have a microscope, but I should have take a couple of 100 pictures with it.
DG912Pro and DM858.

(Attachment Link)

(Attachment Link)


The only difference I found.
Under the add-on card.
On the DG822Pro and DM858
SEC 319 BYMA K4B4G0846E
On the DG912Pro
SEC 322 BYMA K4B4G0846E

I actually wonder what's below the metal shield. It seems small, Rigol doesn't do DMM chipsets so it will be someone else's chip there.

[gbix]

However, now that it turned out that the 800pro and 900pro are the same inside, I'm a little saddened, but I'm also happy at the same time.
So maybe if someone buys the cheaper version and hacks it, they will be fine with it.

i dont find differences after converting 800pro to 900pro from original 900pro
but i cant check signal accuracy now

[DaneLaw]

A compare DHO814 & DG822PRO cirquit board.

They seem to be using the same screws.

Your sharp 
its the same board template.. so a lot of rehash of the secondary components/sockets DHO800/900, DG800/900PRO, DM858/E.
Quite populated these signal gens and not much seems to be unpopulated (if any) but on paper, the DG800PRO allows only up to 8M (optional) and 2M native while DG900PRO its 32M optional and 16M native..
A board picture of the DG900 would be helpful to see if any of the chip differs.

I mean there is nothing shared between these two designs. Not the main CPU, not the power supply. It's clearly done by two different teams, one using silkscreen the other doesn't. Somehow the oscilloscope doesn't have any electrolytic capacitors as well.

Sure, nobody said otherwise. its pretty clear from the compare-pictures what specific parts differs & which are identical across DHOxxx, DGxxxPRO, and DMxxx, - as you can see there are obviously way more than just the screws that cycle across these models.. the designers are using the same template to fit with the form factor, so secondary components, sockets, etc. will align.
At the end of the day, its widely different products, that are built around the same form factor but it shouldn't really be up for debate what parts differ,, when its highlighted on top of each other, and you can just look.

[DaneLaw]

....
+Its fan is horrible.

Yep, same here.. I think its the same as in the DHOxxx series scope. and it's a hysteric little whiny temp-regulated fan that will cycle up & down with minute differences, that's far from pleasant.
Im supprised to hear that there doesn't look to be any hardware differences between DG800PRO and DG900PRO, if so that should also allow 32M for the DG800PRO series, even though its only optional to 8M.

But its pleasant news that there doesn't look to be any noticeable hardware difference, and what I speculated in' when I took the plunge in November (black Friday) for a DG821PRO in China, in the hope that I down the line would be able to hack it, as we have seen with the latter gen.
as 1.25Gsa - 16bits - 200MHz - 32M (DG922PRO) is nothing to sneeze at, not least for the MSRP prices they are going for in the Far East... hopefully it is linear across the liberated range..

I also got one of these FeelElec signal gens for the last handful of years (FY6900) its been okay (with the price in mind) but its extremely cumbersome to adjust settings on.. that aspect is a joy on these Rigol PRO signal gens.. the big touch screen makes it way more comfortable to dig down into all the features a modern waveform generator can deliver.

I reckon the bootup time on your DM858 are the same as your DG900PRO..looks to be around 63 fo 66 sec on DM858E (from YT vid) so likely no difference on the more premium DM858.

[zrq]

Thanks for posting, now I'm seriously considering getting a DG821Pro instead of the SDG2042X (to be hacked into SDG2122X). But the boot time seems to be long enough to be annoying.