Rigol HDO1000 and HDO4000 12bit oscilloscopes launched in China

[skander36]

I would not be surprised if XILINX's FPGA above some size are in embargo (if aimed to CHINA) due to current international situation.

PLA (the army in China) for sure is affected. Although we can't find this line of Xilinx products on the list, there is a kind of shortage.

Rigolshop.eu told me about the early december for HDO1K ...

[trinacria]

Just out of curiosity, how much are purchase fee in USA for this kind of products ?

Here in Italy we have a stunning 22% 

For me, it's 0%, but it varies by state.

Does anyone other than Dave have a review yet of the Rigol HDO1000?

[markone]

I would not be surprised if XILINX's FPGA above some size are in embargo (if aimed to CHINA) due to current international situation.

PLA (the army in China) for sure is affected. Although we can't find this line of Xilinx products on the list, there is a kind of shortage.

Rigolshop.eu told me about the early december for HDO1K ...

Dunno what to think, i asked them twice over two different channels, AMAZON and Rigolshop.eu, the answer was always "not before mid december".

I cannot understand why so long when those models seem plenty available in US right now.

[coppice]

I would not be surprised if XILINX's FPGA above some size are in embargo (if aimed to CHINA) due to current international situation.

The biggest fastest FPGAs and the highest performance ADCs are the two key areas which are ALWAYS under embargo.

[Kleinstein]

The ADC seems to be a RIGOL branded ASCI. The ADC is impressive, but not super fast. Not sure though if it is really made in main land China or a Taiwanese fab.
The FPGA is not really cutting edge (more like mid range) and may not be subject so many restrictions (maybe still a nogo for north Korea, Iran, Afganistan and now Russia).

There can are general supply problems with even more simple parts like voltage regulators that can halt / slow down production. The parts sold in the US may just a first relatively large batch and the 2nd one for europe might have got hit by the chip crisis. The transport got a bit slower and less predictable. AFAIK the distance to the US is a little longer than to Europr and may need to ship them first to get simultaneous availabilty. Another point could be higher than expected orders for the 4000 series. Another reason to slow down the delivery could be a delay with the version that uses the smaller cheaper FPGA.

[Fungus]

I cannot understand why so long when those models seem plenty available in US right now.

Simple: They haven't got enough production capability yet so they're only selling in one market.

(outside of China)

[Martin72]

Or, or, or...

[Fungus]

Or, or, or...

Or maybe American consumers have lower standards and aren't so critical of early firmware/hardware as the Europeans are.   

[Martin72]

A "new" teardown of the 1000 could be interesting, when they´re everywhere avaible since a couple of time..

[markone]

A "new" teardown of the 1000 could be interesting, when they´re everywhere avaible since a couple of time..

More expensive SKUs (HDO 1104, 1202, 1204) are already available @ Rigolshop EU (also at Amazon), Conrad has something in catalog but with higher prices (than official listing), here Rigol is aiming to make some decent margin NOT selling models like HDO1072/74 in first place.

Being not afraid about alpha stage FW status i would kindly pay the current listing price for bargain SKUs, what i saw in Dave video is enough to bet on the new toy, that seems more responsive than MSO5000 scopes, even if i'm convinced that there is a lot of space for further optimizations considering the application SOC used.

I also appreciated a lot the Rigol choice for optical encoders, i had some bad experience in the past with mechanical ones failing after few years from the purchase ...

But I need badly a new DSO with 9-10" touch screen at home lab, 12 bits would be a nice addition but not mandatory, so if HDO1072/1074 don't appear soon in EU market I'll go for  something else like an MSO5074 or  a SDS2104X-PLUS.

The latter has a "quiter" analog acquisition and bigger screen while the Rigol has a much higher sample rate and it's cheaper, a real struggle to decide if HDO1000s do not come in

[markone]

A "new" teardown of the 1000 could be interesting, when they´re everywhere avaible since a couple of time..

Do you smell a complete HW redesign for HDO1000 series ?

[Martin72]

Only "belly-feeling"....For me personally, the here presented HDO1000 is way too near to the much more expensive 4000.

But I need badly a new DSO with 9-10" touch screen at home lab, 12 bits would be a nice addition but not mandatory, so if HDO1072/1074 don't appear soon in EU market I'll go for  something else like an MSO5074 or  a SDS2104X-PLUS.

The latter has a "quiter" analog acquisition and bigger screen while the Rigol has a much higher sample rate and it's cheaper, a real struggle to decide if HDO1000s do not come in

To be honest, if 12 bit are not really important for you there are no reasons to wait for, except the low noise.
And low noise you can also get with a sds2104x+.

[markone]

-snip
To be honest, if 12 bit are not really important for you there are no reasons to wait for, except the low noise.
And low noise you can also get with a sds2104x+.

I mentioned it in my post, it's actually in my wish list even if it's memory management and some other details would not exactly meet my taste, but you cannot get everything at the same time.

But HDO1000 seems great like HW / price ratio, let's see what happens next days.

[Kleinstein]

A "new" teardown of the 1000 could be interesting, when they´re everywhere avaible since a couple of time..

Do you smell a complete HW redesign for HDO1000 series ?

I don't see a major HW redesign: If at all rather minor changes for cost cutting and maybe better availabilty, like different voltage regulators and a one step smaller FPGA, maybe smaller memory chips. This may even happen on the same PCB. The software may changes to make it work with a smaller FPGA and possibly better performance in some areas.
The main parts like front-end, ADC, FPGA, µC for the UI kind of have to stay. So there is not very much to change unless it gets a different product (e.g. version with internal generator or smaller case/LCD).
So far I don't see a major problem that would need a redesign.

[2N3055]

A "new" teardown of the 1000 could be interesting, when they´re everywhere avaible since a couple of time..

Do you smell a complete HW redesign for HDO1000 series ?

I don't see a major HW redesign: If at all rather minor changes for cost cutting and maybe better availabilty, like different voltage regulators and a one step smaller FPGA, maybe smaller memory chips. This may even happen on the same PCB. The software may changes to make it work with a smaller FPGA and possibly better performance in some areas.
The main parts like front-end, ADC, FPGA, µC for the UI kind of have to stay. So there is not very much to change unless it gets a different product (e.g. version with internal generator or smaller case/LCD).
So far I don't see a major problem that would need a redesign.

I think there is misunderstanding here. I agree architecture won't change. But discussion on "redesign" here revolves around the fact that HDO1000 Dave opened wa literally 4000 board (with big FPGA) minus few analog parts and minus one ADC. Even 50 ohm path was there on PCB although 1000 doesn't support it. That sparked discussion by some "enthusiasts" that means HDO1000 could be converted to 4000 if parts were available. It was obvious that Rigol actually didn't make 1000 series boards at the time but just repurposed 4000 BOM.

Fast forward, HDO1000 is very limited availability (might even be only that first, batch zero), and they obviously will "optimize". That points to conclusion that what we will see now is going to be smaller FPGA and whatnot. They might as well use same PCB (it is still standard FR4 so no need to optimize there), but in the end it will be less powerful hardware. FPGA is pretty much hearth of the scope. Not app processor. So smaller FPGA, less memory and only one ADC and that is smaller model. They will keep app processor and glue to "scope parts", because that part of architecture is actually lots of work and lets them standardize development.
They don't have to do "major redesign" to achieve "major difference in end result".

Same thing as with Keysight 2000. Only 1 Megazoom chip instead of 2 in 3000 series. Less memory, limited input BW and crippled software. In theory Keysight could have made unitary PCB for 2000/3000A series.

[Fungus]

Do you smell a complete HW redesign for HDO1000 series ?

No.

[Fungus]

Note that if they've optimized the code to get it into a smaller FPGA, the HDO4000 might get that smaller FPGA, too.

There's very little difference in functionality between the two models.

[2N3055]

Note that if they've optimized the code to get it into a smaller FPGA, the HDO4000 might get that smaller FPGA, too.

There's very little difference in functionality between the two models.

It is not like that. They won't "optimize" in technical sense but in economical sense. That is confusion here.
As I said, when you are prototyping you start with "bigger" resources because many times you develop software/hardware at the same time. You don't really know how much you are going to use, especially when marketing department keeps moving target. Also mid development, competition releases something, and now you must add that too...
So once the product gets finished enough to take final shape, you "optimize" and settle for hardware version that is "good enough" + some reserve for latter development (if you plan to keep the platform for few years.).
That is why I said that Rigol released these two scopes just a bit too early... Scopes we are seeing now for sale, other manufacturers would beta test for some time until production ramps up in it's final shape...

And with software developers being as they are (unfortunately), don't expect code size to go down, only up... As they finish existing features and add all the promised stuff it will go up...

[markone]

Must also be taken into account that heavily forking the HW design involves a subsequent effort increase for separate FW upgrade service, this is a thing that companies tend to avoid especially for complex products.

Personally i believe that we will see a simple shrink for FPGA and DDR size, keeping the same PCB.
 

[asmi]

I would not be surprised if XILINX's FPGA above some size are in embargo (if aimed to CHINA) due to current international situation.

Both of these are low-end parts by modern standards, so I doubt there are any restrictions on them. I suspect the reasons they have both is that's what they were able to procure. I monitor aliexpress for these parts (as I can't get them elsewhere without 24+ months of a lead time), and found that price for these parts tend to be all over the place, and often "higher-end" A100T can be cheaper than A75T or even A50T.

[Someone]

Note that if they've optimized the code to get it into a smaller FPGA, the HDO4000 might get that smaller FPGA, too.

To be pedantic:
https://www.eevblog.com/forum/microcontrollers/artix-7-only-has-three-different-die-sizes/
It would likely be possible to not optimise the code and just ship off identically placed designs into both those parts.

[asmi]

Note that if they've optimized the code to get it into a smaller FPGA, the HDO4000 might get that smaller FPGA, too.

To be pedantic:
https://www.eevblog.com/forum/microcontrollers/artix-7-only-has-three-different-die-sizes/
It would likely be possible to not optimise the code and just ship off identically placed designs into both those parts.

AFAIK that post is no longer correct as there is another die now for 12T/25T devices.

[Fungus]

It would likely be possible to not optimise the code and just ship off identically placed designs into both those parts.

That would be a very risky thing to try at large scale. There's definitely going to be a binning process at the chip FAB. Even if 99% of them are able to run the code, those 1% returns will cost you a lot of money.

You also run the risk of pissing off the chip maker and they can cut off your supply or start charging you more for the smaller chip.

At Rigol's level it's much better to go and negotiate a deal for legit FPGAs. A good, guaranteed price for the next few years is far more valuable to a company than messing around trying to beat the system.

[thm_w]

Note that if they've optimized the code to get it into a smaller FPGA, the HDO4000 might get that smaller FPGA, too.

To be pedantic:
https://www.eevblog.com/forum/microcontrollers/artix-7-only-has-three-different-die-sizes/
It would likely be possible to not optimise the code and just ship off identically placed designs into both those parts.

AFAIK that post is no longer correct as there is another die now for 12T/25T devices.

The post doesn't even mention 12T/25T,  so that doesn't mean that its incorrect, maybe 12T/25T is a fourth die and the rest remain unchanged.

Anyway, if 75T and 100T are the same JTAG ID, Rigol must set some firmware or board level ID? Or is there another register in the FPGA that can be read out.
Either way, the low chance concern that a 100T built device would have some greater performance, should not be a concern, if its another thing that can be hacked.

[skander36]

For those interested, Rigol EU now has a 10% promotion for the 1k series. https://rigolshop.eu/products/oscilloscope/hdo1000.html
But still 1 month delivery. Preorder?