EEVblog #842 - Rohde & Schwarz HMO1202 Oscilloscope Teardown

[blueskull]

A piece of beautiful sh*t.
Small memory, deep in thickness, non touch, high price, outdated design and gold plated connectors (yes, they suck if you plug them often).

2M is fine only with segmented memory. With continuous memory, 2M=1ms at 2Gsps. Besides, at 2Gsps throughput, the capability of Cyclone 4 can easily dump it into larger DRAM, instead of SRAM.
It is really thick in size, compared to ultra thin designs which are becoming popular, like the owon ones (I own an original 2010 SDS7102, and I really appreciate its size).
The design uses expensive and outdated NS ADC, while new designs use Hittite converters that consume a fraction of power and cost a fraction of the cost, plus come with native 12bit mode.
The fact is, no one cares about performance of the ADC, since in such high bandwidth applications, most noise comes from front end, not ADC.
The application system is old too. A new design using Zynq can be cheaper while providing more on die SRAM and 10x CPU-FPGA bandwidth with DDR3 hard IP, and the effort is acceptable -- all HDL can be reused.
Compared to mid range Rigol scopes at the same price, I do not know how can it compete, besides better building quality. In addition, Rigol scopes can be hacked.
Finally, I really hate the gold plated connectors. When mated and unmated often, the gold layer will wear out, revealing half gold half nickel finishing. I'd rather get a scope with high quality thick nickel plating instead of soft gold plating.

For 1/3 of the price, DS2072 can be hacked to have 300MHz BW, 140M (?, not sure) memory and AWG.
For the same price, MSO4000 series offers tons of new features and much better hardware specs.
At 2x the price, MSOX2022A offers 100x capture rate with segmented memory and a bunch of other options that could be free if you are lucky.

[zapta]

I am in the market for a new oscilloscope (replacing a TDS210 almost 15 years ago) and the R&S seems interesting. Is the HMO1202 a new model?  I cannot seem to find it on equipment.

What I could find is a HMO1102, 100Mhz, $1220

http://www.tequipment.net/Rohde-&-Schwarz/HMO1102/?v=0

and a HMO Max 100Mhz, $1095
http://www.tequipment.net/Rohde-&-Schwarz/HMO-Max/Digital-Oscilloscopes/?v=7401

Anybody knows how they compare to each other and to the HMO1202.

My needs are modest. I was happy with the TDS210 until it died on me and for long captures and decoding I use a Saleae logic analyzer which works well for me. Also, I am aware of the lower cost Chinese products but want to explore a better quality product.

[Muxr]

I would like to see this thing run against the Rigol DS2000 to compare its capture performance. Price wise for 300Mhz it is virtually the same as a Rigol DS2302A which is $2450US on Tequipment. Compared to other big names you can also go tektronix and spend over $8000 to get a 300Mhz scope that was obsolete 10 years ago.

Personally I think the Rigol is much more attractive but beauty is highly subjective.

I have the DS2072A (hacked to DS2302A), was a bit over $800. It's a great scope when you look at the performance per dollar. But I never came to grips with its UI. It's as annoying to use today as it was when I first got it. If I was doing it all over again I would seriously consider the HMO1202 instead.

[nowlan]

The 200mhz is odd. The original budget model comes 50mhz std, with license unlocks to 70 or 100mhz.

The 1202 seems to be a new model from November 2015.
https://www.rohde-schwarz.com/au/news-press/press-room/press-releases-detailpages/new-r-s-hmo1202-mixed-signal-oscilloscopes-offer-top-performance-at-a-great-price-press_releases_detailpage_229356-152704.html

[Paul Moir]

Finally, I really hate the gold plated connectors. When mated and unmated often, the gold layer will wear out, revealing half gold half nickel finishing. I'd rather get a scope with high quality thick nickel plating instead of soft gold plating.

I'm wondering why I'm supposed to be cheering for gold BNCs solely supported by solder rather than a nut.

[blueskull]

I'm wondering why I'm supposed to be cheering for gold BNCs solely supported by solder rather than a nut.

Probably they have done the math, after all it is a piece of over engineered gear.
With no severe temperature cycling, mechanical strength of solder should be fine.
But with gold contact, I don't know. IMC of Au and Sn is notoriously unreliable with thermal cycles.
That's another reason why we stopped using Au to bond power semiconductors other than cost factor.
It will probably be fine since the connector may not experience high temp or thermal cycle at all.
But still, I do not consider this properly. Nut and washer is always better.

[chickenHeadKnob]

Dave: I would like to request a head to head comparison video versus the Rigol 2072 not the 1074 you have as I feel they are more direct competitors. Partial idea of some things to test,

• the quasi- averaged higher resolution modes of the Hameg on under <20 Mhz signals versus Rigol 0.500 mV setting and their respective noise floors. 
• quality of the intensity gradient display versus the Rigol's
• Wfrm/s
• Navigation of the capture memory, ease of use

 
I am pretty sure other forum members  would be similarly interested.
Thanks

[EEVblog]

The 1202 seems to be a new model from November 2015.
https://www.rohde-schwarz.com/au/news-press/press-room/press-releases-detailpages/new-r-s-hmo1202-mixed-signal-oscilloscopes-offer-top-performance-at-a-great-price-press_releases_detailpage_229356-152704.html

Yes, I forgot to mention that, it's a new model just released.

[coppice]

The 1202 seems to be a new model from November 2015.
https://www.rohde-schwarz.com/au/news-press/press-room/press-releases-detailpages/new-r-s-hmo1202-mixed-signal-oscilloscopes-offer-top-performance-at-a-great-price-press_releases_detailpage_229356-152704.html

Yes, I forgot to mention that, it's a new model just released.

It has an 11 year old ADC that has already reached the "not for new designs stage" (although a very nice quiet 7.4 ENOB ADC), limited storage and less than stellar waveforms per second. This looks like a design that must have been in gestation for a long long time.

[Fungus]

Yes, I forgot to mention that, it's a new model just released.

It has an 11 year old ADC that has already reached the "not for new designs stage" (although a very nice quiet 7.4 ENOB ADC), limited storage and less than stellar waveforms per second. This looks like a design that must have been in gestation for a long long time.

Or maybe they got a sweet deal on the last batch of chips.

[TinkerFan]

I noticed a little Y on the PCB ground plane (which is not connected) next to the sample memory (18:45) and I was wondering, if that is some sort of a sign for the Pick & Place machine? Is that common? I thought, they would do it with silkscreen markings or so.

You never do fiducial markings with silkscreen, there can be alignment errors.

So I assume it is a common method of aligning the chip with a pick and place machine?

[zapta]

The 1202 seems to be a new model from November 2015.
https://www.rohde-schwarz.com/au/news-press/press-room/press-releases-detailpages/new-r-s-hmo1202-mixed-signal-oscilloscopes-offer-top-performance-at-a-great-price-press_releases_detailpage_229356-152704.html

Yes, I forgot to mention that, it's a new model just released.

Do you know how it compares to the previous R&S models ?

HMO1102, 100Mhz, $1220
http://www.tequipment.net/Rohde-&-Schwarz/HMO1102/?v=0

HMO Max 100Mhz, $1095
http://www.tequipment.net/Rohde-&-Schwarz/HMO-Max/Digital-Oscilloscopes/?v=7401

[rf-loop]

I have been long long  time big fan of R&S equipments. (specially older high grade - state of art equipments. Just so amazing building quality - and prices)

But this do not look at all like these real R&S equipments.

After reading specifications, this video and so on I'm very confused.
Specially if this is really new design and launched 2015.


After reading Trigger specifications I need really ask do this scope have true digital side trigger system or do it have conventional old analog trigger system? If it is analog trigger and launched 2015 and brand is R&S I have very mixed feeling.



Then some other things..

10kwfm/s speed. Really?
32 level intensity grading?
Small 6.5"  display.
I can not find gated measurements.
I can not see segmented memory acquisition.

I can see small 2M memory today, really in new 2015 launched scope from R&S.
I can see ADC what is classified "Not for new design".

Of course I can see this is not crap and there is many nice things but... 

Is this at all R&S design?

[coppice]

After reading specifications, this video and so on I'm very confused.
Specially if this is really new design and launched 2015.

Maybe it spent 6 or 7 years waiting for a supply of lemon soaked paper towels.

[nfmax]

More likely it got delayed while the negotiation & due diligence for R&S's acquisition of Hameg took place - this looks like a Hameg product.

[Groucho2005]

this looks like a Hameg product.

Indeed, entry level Hameg scope.
The Hameg HMO3000 series scopes are already much better - 4GS / 500 MHz / 200,000 waveform update rate.
The "real" R&S scopes ("Scope Of The Art") make me drool - up to 20GS / 4 GHz / 1,000,000 waveform update rate.

[nowlan]

I assumed it was a minor change from the older models. But they have made a few changes between series.

HMO Compact Series
70 to 200 MHz 2/4 Channel Mixed Signal Oscilloscopes
http://www.hameg.com/0.616.0.html

This model.
https://www.rohde-schwarz.com/us/product/hmo1202-productstartpage_63493-142720.html
http://value.rohde-schwarz.com/vi/value/oscilloscopes/r-srhmo1202-digital-oscilloscope.html



HMO compact rear


HMO1002 RS rear


DVI replaced by ethernet. Different USB socket.
External trigger moved to front.

So did they just update the FPGA and ports?

[bitseeker]

+1 on Dave's comment about the front panel layout. The spatial relationships between the control and input sections don't make sense. I'd like to know what rationale went into that. Looks like a case of trying to be different.

[rf-loop]

At 2x the price, MSOX2022A offers 100x capture rate with segmented memory and a bunch of other options that could be free if you are lucky.

Really. Perhaps I need update my library? It looks more like 5x faster and only with 250 segments.
(20?s is fastest trigger interval what is really extremely slow if we look today oscilloscopes and segmented memory acquisition. It is more like Rigol frame recorder.)

R&S (model in this topic) do not have segmented memory at all.
Max wfm/s speed in R/S is around 10kwfm/s and MSOX2022A  max is around 50kwfm/s  but it is really slow and Limited with segmented memory. (how they even ask extra payment about this slow speed and low memory segmented  acq option). Also MSOX2022A have standard 100k memory. Optional 1M (agen pay more)  and some others give 280M as standard.  Btw, trig recovery time in specifications is 20us.

Keysight Technologies
Segmented Memory Acquisition
for InfiniiVision Series Oscilloscopes

Number of segments
1 to 2000 (5000, 6000, and 7000 Series)
1 to 1000 (3000, 4000, and 6000 X-Series)
1 to 250 (2000 X-Series)

Minimum segment size
500 points (+ Sin(x)/x reconstructed points on faster timebase settings    )

Re-arm time
(minimum time between trigger events)
5000, 6000, 7000: 6 ?s
6000 X-Series: 7.5 ?s
3000 and 4000 X-Series: 1 ?s
2000 X-Series: 20 ?s   

Example in Siglent SDS1000X and SDS2000 and SDS2000X there is 2 ?s  trigger recovery time when standard segmented memory acquisition is in use.
Even Siglent models normal allways bacround working history buffer can better (60kwfm/s or 140kwfm/s depending model) and up to 80000 waveforms in history buffer.  Segment acquisition speed  max 500 ksegment/s (and even bit over).  Depending timebase and memory settings up to 80000 segments.


More and more looked this R&S data sheet and agen and agen this video it looks like it is really old design with analog trigger system. (If I'm wrong, please evidence)

Is it really wise to label this kind of ("who ever can made") boxes with this  Rohde&Schwarz name what is classified as one of most top-level brand when we talk electronic T&M equipments least in my head.   They want to stain this name?

[blueskull]

5000, 6000, 7000: 6 ?s
6000 X-Series: 7.5 ?s
3000 and 4000 X-Series: 1 ?s
2000 X-Series: 20 ?s

My fault. I thought 2000X is also 1Mwfms/s.
I thought they use the same Mega IV as 3000X and 4000X do.

I'm able to get 1Mwfms/s from my 3104A though.

[rf-loop]

I thought they use the same Mega IV as 3000X and 4000X do.

I'm able to get 1Mwfms/s from my 3104A though.

2000X is very different bottom level scope. (starting from around 1200 euro w/o tax)

3000X is very nice and totally very far over 2000X. But DSOX3104A (standard without any options)  price is here over 12000 Euro without tax.

[NANDBlog]

I thought they use the same Mega IV as 3000X and 4000X do.

I'm able to get 1Mwfms/s from my 3104A though.

2000X is very different bottom level scope. (starting from around 1200 euro w/o tax)

3000X is very nice and totally very far over 2000X. But DSOX3104A (standard without any options)  price is here over 12000 Euro without tax.

I've used recently one of their refurbished scope (MSO3000 not T). It came at some ~50% discount, and it was in absolute perfect quality. Nothing missing. Though this might be some gamble, I think, I would recommend buying them again. Still expensive for home, but a good way to convince management to get the tools that I need for work.

[HighVoltage]

Doesnt appear to be any 4 channel versions (no space on board), unlike the older HMO compacts.

They have a 4 Channel version of this available.
With 4 BNC connectors in the front, the  External Trigger is on the back, together with Y-OUT
Here are some pictures of my HMO1024, 100 MHz, 2 GSa scope.
It is a lovely tiny scope.
But you have to get used to the UI, it is a little different from the Keysight UI.

[owiecc]

They have a 4 Channel version of this available.

HMO1002/1202 is a new design. You have an old Hameg design. There is no 4 channel design for the new ones.

[DaWaN]

Dave: I would like to request a head to head comparison video versus the Rigol 2072 not the 1074 you have as I feel they are more direct competitors. Partial idea of some things to test,

• the quasi- averaged higher resolution modes of the Hameg on under <20 Mhz signals versus Rigol 0.500 mV setting and their respective noise floors. 
• quality of the intensity gradient display versus the Rigol's
• Wfrm/s
• Navigation of the capture memory, ease of use

 
I am pretty sure other forum members  would be similarly interested.
Thanks

Yes! At least I am

On my day to day job I have both a Hameg HMO2524 and a Rigol DS1074-Z available.
I prefer to use the DS1074 because I find it much easier to navigate through the memory.
The navigation of the Hameg is terrible: we even had to replace the front panel because the rotary encoders had worn out.
This new series seems to be a step-up in terms of build quality, as HAMEG never really convinced me.
So I am curious how Dave thinks about the usability of this scope!