EEVblog #842 - Rohde & Schwarz HMO1202 Oscilloscope Teardown

[zapta]

Hameg/R&S gear is very polished + updates long after release date. My HMO1524 is a treat to use and this seems very similar. Wish the CAN/LIN decode was a bit cheaper tho

The CAN/LIN/UART/SPI/I2C decoders come free with this older model $1K oscilloscope.

http://www.tequipment.net/Rohde-&-Schwarz/HMO-Max/Digital-Oscilloscopes/?v=0

I placed an order this morning for one. This video was very helpful.

[lem_ix]

That's a great deal, you'll enjoy using it for sure

[Gunb]

Hameg/R&S gear is very polished + updates long after release date. My HMO1524 is a treat to use and this seems very similar. Wish the CAN/LIN decode was a bit cheaper tho

?

Compare it to Agilent / Keysight carefully and it will show HOW MUCH you get from Hameg/R&S in comparison to Agilent/Keysight.
THEN you know what's cheap and WHAT's expensive.

I'm using Agilent / Keysight, Tek, Rigol and Hameg/R&S for years now, and without preferring any of these brands I can say that Hameg/R&S is the winner concerning FFT/bus decoding especially
in contrast to Agilent/Keysight.

I've used the InfiniiVision MSOX4054 (20k €!!) for many months now using it for micros including the expensive bus decoding options. Finally I've replaced it with an Hameg/R&S 3054 since it has many options more than the crap of Agilent/Keysight.

The HMO 1002 / 1202 R&S series has much better bus decoding functionality than Keysights and Rigols, and the FFT is a dream for any communication electronics engineer.

So, if the review would be better it wouldn't always deal with open the housing and have a look on a stupid PCB but would reveal more the details I've mentioned above.

Concludingly, if using the scope mainly for micros the Hameg/R&S are offering plenty of features where the Rigols fail. For most purposes you don't need millions of wfm/s. It's just a feature easy to read from
the datasheets, and unfortunately for many users the only criteria.

One might have the most beautiful sports car, but not very helpful for outdoor purposes.

Sorry for being a bit inpolite, but it's simply the truth.


Rgds
Gunb


BTW: Rigol's FFT is crap.

[Gunb]

I prefer Keysight scopes. Hameg / Rohde & Schwarz is good only if you need a very compact scope with a silent fan.

Bullsh**t, just that simple.

P.S.: Please don't post videos, we know them all.

[rf-loop]

I prefer Keysight scopes. Hameg / Rohde & Schwarz is good only if you need a very compact scope with a silent fan.

You can prefer what ever scope just as also who ever even if do not have any knowledge or experience about anything.

Comment about R&S is total bullshit and tell only that you are talking things what are outside of your knowledge. Do you even know what is Rohde & Schwarz. Have you never seen any kind of R&S equipments. (If I win enough money from lottery I know what I buy and name plate is on nearly all equipment "true" R&S without any doupt. Do not even think.)

(I do not talk about these R&S adopted "semi R&S" scopes what is here in this teardown what are based perhaps to quite old design by Hameg or whoever and then named by R&S what is quite sad but even these have poor things and nice things, and very good building quality, not state of art but good.)

[lem_ix]

Hameg/R&S gear is very polished + updates long after release date. My HMO1524 is a treat to use and this seems very similar. Wish the CAN/LIN decode was a bit cheaper tho

?

Compare it to Agilent / Keysight carefully and it will show HOW MUCH you get from Hameg/R&S in compariso to Agilent/Keysight.
THEN you know what's cheap and WHAT's expensive.

I'm using Agilent / Keysight, Tek, Rigol and Hameg/R&S for years now, and without preferring any of these brands I can say that Hameg/R&S is the winner especially
in contrast to Agilent/Keysight.

I've used the InfiniiVision MSOX4054 (20k €!!) for many months now using it for micros including the expensive bus decoding options. Finally I've replaced it with an Hameg/R&S 3054 since it has many options more than the crap of Agilent/Keysight.

The HMO 1002 / 1202 R&S series has much better bus decoding functionality than Keysights and Rigols, and the FFT is a dream for any communication electronics engineer which is included in todays scopes.

So, if the review would be better it wouldn't always deal with open the housing and have a look on a stupid PCB but would reveal more the details I've mentioned above.

Sorry for being a bit inpolite, but it's simply the truth.


Rgds
Gunb

Don't see what you're getting worked up about? I didn't compare it to anything and I'm very well aware of equivalent keysight pricing although it's a promo deal for some scopes atm right? Working on a competition robot  with some mates, decided on canbus for inter-board communication. Can't justify the price for that, a bit more money and I could get a rigol ds2072, hack it and voila canbus decode plus another scope... Not gonna do any of that, decode is nice but not crucial in any way. Bottom line, a very personal wish, can you blame me ?

[Gunb]

Don't see what you're getting worked up about? I didn't compare it to anything and I'm very well aware of equivalent keysight pricing although it's a promo deal for some scopes atm right? Working on a competition robot  with some mates, decided on canbus for inter-board communication. Can't justify the price for that, a bit more money and I could get a rigol ds2072, hack it and voila canbus decode plus another scope... Not gonna do any of that, decode is nice but not crucial in any way. Bottom line, a very personal wish, can you blame me ?

Well, you've mentioned CAN/LIN. The R&S decode is not as expensive as Agilent/Keysight, and faster with more features than Rigol provides. Got Rigol and Hameg/R&S at home, Agilent/Keysight at work. So not only a theoretical sight of view. Can't understand to pay approx. 600€ for serial decode for the Agilent without the option to display characters in ASCII !!! The HMOs have got all options included, and that's only one example.

Comparing a hacked scope with a non hacked one rather prefers the hacked one.

Decode is more crucial today than ever before. Dealing with embedded every day it's permanently in use in parallel to analog / mixed signal. And then one might need proper implementation of all these features than having a simple scope only. Rigol lacks in a few things concerning this. Neither 140 Mpts memory nor high wfm/s will change these facts.

Besides bus decode the HMOs are proper devices with much better rotary encoders than these of the Rigols. Had to change one after short usage. Even the Agilent 4054X I've mentioned above has got a broken rotary encoder of less quality.

Don't want to make the Rigols bad, commonly good scopes. But there are reasons for the HMOs and to spend a bit more money, too. I'm missing these details in the review.

[Hydrawerk]

Do you even know what is Rohde & Schwarz. Have you never seen any kind of R&S equipments.

Yes.
https://www.scope-of-the-art.com/en/

But this handheld R&S scope is new for me. http://www.2-minutes.com/

[Fungus]

this handheld R&S scope is new for me. http://www.2-minutes.com/

Nice.

If I had $6000 to spare I might get one....

[Neganur]

Don't know about the memory debate. It's a bit too much comparing numbers rather than what is enough for the task it is designed for.
It's an oscilloscope, not a digitizer. I think it'll do it's job well enough.

Imagine it had 1 Gigabyte of the fastest memory you can imagine. What would it be good for and how often would you need that?

(Edit: 640KB ought to be enough, no? I'm probably failing to see the need.)

[nctnico]

For the price of the HMO1202 (especially if you want the decoding options) you can certainly do better without needing to resort to Rigol or Siglent!

[zapta]

Don't know about the memory debate. It's a bit too much comparing numbers rather than what is enough for the task it is designed for.
It's an oscilloscope, not a digitizer. I think it'll do it's job well enough.

My previous oscilloscope was a TDS210 with 2500 points/ch and it worked just fine, so the X200 memory size of the R&S HMO1002 should be more than enough for me.

I ordered it after the DS1054Z arrived. No offense but the Rigol felt noisy, non responsive, cramped, and even crude (e.g. some of the fonts it uses). Not something I would enjoy using in the next 10 years, even if it has larger memory and two more channels. This is subjective of course, YMMV.

Comparing numbers is not everything.

[rf-loop]

Don't know about the memory debate. It's a bit too much comparing numbers rather than what is enough for the task it is designed for.
It's an oscilloscope, not a digitizer. I think it'll do it's job well enough.

One aspect is: For maintaining fast sampling.

Why oscilloscopes need deep acquisition memory by Joel Woodward, Senior Product Manager, Oscilloscopes,  Agilent Technologies.
There he tell also two other reasons.

But this whole question is perhaps so importaant that somewhere is good to ppen thread for: Why oscilloscopes need deep acquisition memory but in some cases 10 sample memory lenght is enough.

[hjups]

Does anyone have any insight into the ESD topology of the logic probe pod? I have been racking my brain over it since this teardown was posted. They must be doing something special, since I would imagine it should be able to measure TTL signals and the ADCMP562s are only rated for up to a 3V input.

[EEVblog]

My previous oscilloscope was a TDS210 with 2500 points/ch and it worked just fine, so the X200 memory size of the R&S HMO1002 should be more than enough for me.

A few megs is enough for most things, especially general purpose scope usage.
In this case other factors can drive the buying decision.
For example, the FFT mode on the R&S is very nice (128Kpoint), was playing with it today, and might do a comparison video with some other scopes for that.

I ordered it after the DS1054Z arrived. No offense but the Rigol felt noisy, non responsive, cramped, and even crude (e.g. some of the fonts it uses). Not something I would enjoy using in the next 10 years, even if it has larger memory and two more channels. This is subjective of course, YMMV.

Yes, quite noisy. In fact I'm starting to get a bit annoyed by it now. The R&S is much more polished, like the Keysight X series, although you pay a hefty price premium for them.

[EEVblog]

So, if the review would be better it wouldn't always deal with open the housing and have a look on a stupid PCB but would reveal more the details I've mentioned above.

It's not a review video, it's a teardown video. Usually I don't even turn the scope on in a teardown video, but in this case I had a little play.

[Fungus]

I ordered it after the DS1054Z arrived. No offense but the Rigol felt noisy, non responsive, cramped, and even crude (e.g. some of the fonts it uses). Not something I would enjoy using in the next 10 years, even if it has larger memory and two more channels. This is subjective of course, YMMV.

Yes, quite noisy. In fact I'm starting to get a bit annoyed by it now.

Yep, Rigol fan noise is annoying.

How about a video where you set all your fancy thermal imagers and auido measuring gear on a Rigol and see if the fan is really even necessary. If so, what's a good way to silence the thing?

[EEVblog]

How about a video where you set all your fancy thermal imagers and auido measuring gear on a Rigol and see if the fan is really even necessary. If so, what's a good way to silence the thing?

Install a better quality quieter fan.

[Fungus]

How about a video where you set all your fancy thermal imagers and auido measuring gear on a Rigol and see if the fan is really even necessary. If so, what's a good way to silence the thing?

Install a better quality quieter fan.

Recommendations? What's the lowest speed it could run at? How much measured difference would it actually make? etc.

You said it annoys you. Don't you want to 'engineer' it?

[EEVblog]

Recommendations? What's the lowest speed it could run at? How much measured difference would it actually make? etc.
You said it annoys you. Don't you want to 'engineer' it?

I think some other have done it on the forum.
I was thinking that Papst fan in the R&S is damn nice...

[Groucho2005]

For example, the FFT mode on the R&S is very nice (128Kpoint)

Wow, I didn't realize that they upgraded the FFT length. My HMO722 has "only" 64K points and gets a bit sluggish when using that resolution. I wonder if they use a faster processor to make 128K usable.

[HighVoltage]

I think some other have done it on the forum.
I was thinking that Papst fan in the R&S is damn nice...

Although Papst fans are really nice, they can also be extremely loud.
But there is one type of Papst fans that I like the best for instruments, it is the Type 612 MI
They are extremely quiet and have a little thermistor to regulate the speed internally.

[Wuerstchenhund]

One aspect is: For maintaining fast sampling.

Why oscilloscopes need deep acquisition memory by Joel Woodward, Senior Product Manager, Oscilloscopes,  Agilent Technologies.
There he tell also two other reasons.

While that is correct in principle, you have to see this in the context of situation, which here is about low-end scopes with low bandwidth and low sample rates which all come with somewhat reasonable amounts of sample memory. The HMO1202 has 2GSa/s and 2Mpts, which should be plenty for what a scope of that category is used for, and because of the low sample rate there's little benefit having excessive amounts of sample memory available, especially when (like with Rigol scopes) they don't come with a proper search function and can't even be used for FFT (which on all Rigol scopes uses just a few thousand points which is a bad joke).

Putting 50Mpts or more of sample memory in a measly 2GSa/s scope is silly and only serves marketing purposes. For an low-end entry-level scope, as long as the sample memory is in Mpts and not just kpts then it should be fine. It's a different story in the mid-range sector though, where bandwidths increase to 1GHz and sample rates to 4GSa/s, plus scopes start to come with more advanced functionality.

But this whole question is perhaps so importaant that somewhere is good to ppen thread for: Why oscilloscopes need deep acquisition memory but in some cases 10 sample memory lenght is enough.

I don't think this really needs further discussion. As far as I remember even Dave has explained this long ago, and there are also lots of other sources which explain the relation of sample rate and sample memory. It's not really complicated.

[zapta]

One aspect is: For maintaining fast sampling.

Why oscilloscopes need deep acquisition memory by Joel Woodward, Senior Product Manager, Oscilloscopes,  Agilent Technologies.
There he tell also two other reasons.

My TDS210 had 2500 pts/channel but one feature I found to be very useful and used by default is the min/max ('envelope' in Dave's R&S video) capturing mode. It shows the min/max value that were observed in each of the 2500 time slots. This way I could tell if the signal has short spikes that would not otherwise be captured.

BTW, anybody knows how the per time slot min/max values are computed? Is done the min/max done in the analog or digital domain?

[rf-loop]

Putting 50Mpts or more of sample memory in a measly 2GSa/s scope is silly and only serves marketing purposes.

Equivelent is that perhaps then 25M for 1GSa/s scope?

Silly perhaps but do this in attached image even with this 1GSa/s samplerate using less memory or example with this R&S. Just, no way - end of game. What next,  job is still waiting to do.

Silly, is it. Only for marketing?  Do you really think that if you do not need something it is true for all others who of course are wrong because you know what is silly and what is not.

Is this something what is not normal oscilloscope use. If it is not normal use, define what is normal use I do this kind of things many times.
(it was ultra simply crap cycle-cycle  jitter (or something like it)  check for 10ms period 20ns pulses with also infinite persistence running long time).

Many times I miss lot of more memory just for keep full samplerate with low timebases. Other way is segmented acquisition.