Which scope with isolated inputs?

[hibone]

Buy the time the technician graduates, all scopes would be isolated anyhow, if this new paradigm shift becomes a reality, giving everbody the freedom to measure between 2 random measure points. Exactly the way it should have been from day one, as all you want to measure is the potential difference between 2 measure points. Nobody should have to care about which measure point is and has to be ground. Why do you have the tip and the clip in the first place otherwise (2 measure points and not 1!)?. In the current constellation they could even consider to drop the clip as it always has to be ground anyhow

That's beyond the point, I think.
You are simply suggesting that all the telescopes should have smoked lenses so that everyone can freely look at the sun with it.
A tool is simply a tool. It shouldn't be idiot proof so that every one can be safe around it. On the other hand, the user should be given the education to use the tool, rather than the freedom to fiddle with it.

What if a mechanic starts loosening the nuts on your car just to try out its new wrench?
 

[hibone]

The Micsig can, supposedly: http://www.micsig.com/en/product_detail.asp?id=19
"Serial bus decode   UART?LIN?CAN?SPI?I2C"

It doesn't seem there is any indication on the maximum voltage rating between channels. I may be wrong but the channels may have a shared ground in the MS3000 series of micsigs scopes. That could be nasty.

[richpike]

I looked at the R&S and this seems like a nice option but for some reason it does not have CAN decoding.....

I'd be interested to know if any isolated channel DSO has decoding. 

The R&S scope offers I2C, SPI and UART decoding as an option.

Yes, but not as one option, however wouldn't it be nice if it was a bench DSO?

It appears there are some promotions on it that make the decode/trigger options free (along with the MSO and advanced trigger) - but it's only on the 4ch models:
https://www.scope-of-the-art.com/en/lab-promotion?rs_cid=sota_lab-promotion_2015_landingpage_rsint_scope-promotions_2014_landingpage

The 2ch offer is different:
https://www.scope-of-the-art.com/en/field-promotion?rs_cid=sota_field-promotion_2015_landingpage_rsint_scope-promotions_2014_landingpage

-Rich

[tautech]

HH isolation is a lot easier to design, there are less I/O's to consider and seemingly manufacturers think so too.

It would seem R&S get around this ^^^ for their HH by offering wireless LAN connectivity as an option, but is their std RJ-45 LAN connection isolated ?  Possibly not.     

[nctnico]

LAN is isolated by definition so no problems there.

[robert_]

I looked at the R&S and this seems like a nice option but for some reason it does not have CAN decoding. I'll skip the BK precission/Siglent due to the usual Siglent firmware problems according to several threads on this forum.

CAN decoders (and a few others) are under development, as i heard from R+S.

Oh, and as for isolation:
i measured around 30pf from channel (common/BNC shell) to channel, about the same from channel to the remaining parts (ie, logic port etc). Pretty good, floating measurements with some tens of V/ns common mode voltage work fine (didnt really work on THS720).

[nctnico]

Still the decoding options from R&S are prohibitively expensive. I see no way to sell it to my customer that having CAN decoding on a 3000 euro scope costs 1000 euro extra. I'll already have a hard time selling the isolated scope for that price when they can buy a scope with a similar bandwidth starting from 500 euro.

[tautech]

I need to select an oscilloscope for one of my customers. I think it is best if they get one which has isolated inputs so the can measure between any random point without needing to care about common grounds.........

This can be addressed to some degree by use of any HH DSO for one channel but no more unless there is isolation between subsequent channels and if isolation is only occasionally required this may be sufficient, but:
HH specs often specify the max voltage elevation above mains GND that is permitted while operating on battery power and this need be observed.

However
After 3 pages of this thread the OP has not yet stated a required spec. 

Be it:

BUDGET ?
# of channels ?
BW ?
Memory depth ?
Sampling rate?
I/O's ?
HH or bench top ? (Is high portability required)
Other features?

It would seem there is as requirement for Decoding.....which types?



I'm truly not wanting to offend the OP in asking this and while there is some choice of HH isolated channel DSO's available, for Bench top isolated channel DSO's the selection is very limited. 

Currently, benchtop DSO isolation when needed can be provided in 2 accepted (safe) common ways:
A-B, uses 2 channels for 1 measurement and suffers from poor CMRR. 
Differential probes, for which a set of 2 or 4 will cost more than many DSO's. 


So do we take a moment to explore the commonly required/wanted specifications of an isolated channel bench top DSO?

WHY?

There will be manufacturers that will see this thread and the feature set members can wisely suggest. 

[nctnico]

Actually I wanted to compile some sort of a list with potentially viable choices. First step of selection is to get an overview of what is out there in order to set realistic requirements. If money needs to be spend then it needs to be spend but there has to be good value for money.

[tautech]

Actually I wanted to compile some sort of a list with potentially viable choices. First step of selection is to get an overview of what is out there in order to set realistic requirements. If money needs to be spend then it needs to be spend but there has to be good value for money.

To share with us?

[nctnico]

Actually I wanted to compile some sort of a list with potentially viable choices. First step of selection is to get an overview of what is out there in order to set realistic requirements. If money needs to be spend then it needs to be spend but there has to be good value for money.

To share with us?

Ofcourse... Sharing is caring! working on it...

[nctnico]

The customer in question does a lot with batteries for vehicles and many of those batteries have a switched minus which gives all kind of potentials between the communication ports and the rest of the vehicle. They want to use an oscilloscope to look at currents & voltages coming in & out of vehicles and also look at what happens during a short circuit. Currents can exceed 1kA for several micro-seconds. However the people over there are not the kind that use an oscilloscope everyday. Combine that with batteries which can push enough power into a probe to vaporise it in seconds it is clear the measurement solution needs to have some safety build into it. Also some tests can be destructive; the easiest way to capture an event without needing to rely on a trigger is to put the oscilloscope in roll mode and press stop when the event has occured. For this purpose the oscilloscope needs at least 1Mpts of memory. So far I have been using current shunts, a differential probe and a normal oscilloscope (Yokogawa DL1740) to do measurements for them. Those measurements have also showed that a rugged, portable scope could be beneficial; my DL1740 had a rough landing on a concrete floor which didn't make it look better.

I also mentioned vehicles and with that comes the CAN bus. They do have a CAN analyser so decoding on a scope isn't necessary but they will want to look at the signal waveforms. But even then they could create a ground loop between an oscilloscope with a current shunt connected and the CAN bus.

So far I have compiled this list:

Handhelds
Keysight
U1620A  $3500
isolated inputs, 2Mpts (1ch) / 1Mpts (2ch), CAN triggering, VGA color, 200MHz BW, DVM

R&S
RTH1002 $3400
isolated inputs, 500kpts/250kpts, color, 60MHz, DVM
+ 4 channel version available (no DVM).
+ upgradeable to 500MHz

Bench models
Lecroy
WaveSurfer 3022 $3400
10Mpts, 200MHz
- no isolated inputs so potential short when regular probes used or other grounded equipment
- diff probe needed

Keysight
DSOX 3012T $3500
1Mpts effective, 60MHz
- no isolated inputs so potential short when regular probes used or other grounded equipment
- diff probe needed

I also looked at offerings from Rigol (desktop) and Siglent (handheld) but the bugs and limits (Siglent's handhelds have a normal and deep mode indicating something hasn't been properly integrated) are preventing me from considering these. Fluke, Digimess, Metrix and Tektronix handhelds all have too short memories. B&K Precision seems to be a rebranded Siglent indeed! Tektronix' DPO2000 series could be a desktop contender but unfortunately the display resolution got stuck in the 90's.

All in all it seem the Keysight U1620A comes out on top because it has CAN triggering so at least it will produce a stable picture of a CAN bus.

[tautech]

R&S
RTH1002 $3400
isolated inputs, 500kpts/250kpts, color, 60MHz, DVM
+ 4 channel version available (no DVM).
+ upgradeable to 500MHz

Minor error, 4 Ch version has DVM and uses channel the BNC inputs but the fully featured DMM is only in the 2 ch versions that have the Banana sockets.



With all the proposed candidates you've listed, for the money involved wouldn't you consider a much cheaper benchtop DSO with Differential probes could offer considerable savings? 
Is that not worth exploring?

[nctnico]

Which low cost DSO isn't buggy with the currently available firmware and has a reasonable screen resolution and the required memory depth? The whole low-end range from Keysight and Tektronix are a no-go for that reason. The Rigol 1000Z series still has many issues. But my main concern is that my customer will (not could) create a ground loop when using a probe or a shunt combined with a PC connected to another part of the battery. A differential probe in combination with a current shunt isn't a good idea (some are 0.5mV/A !) because you really don't want the extra attenuation which comes with the differential probe. Another option would be a clamp on current probe with a >1kA DC-1MHz range. I have not seen one so far.

[tautech]

Which low cost DSO isn't buggy with the currently available firmware and has a reasonable screen resolution and the required memory depth? The whole low-end range from Keysight and Tektronix are a no-go for that reason. The Rigol 1000Z series still has many issues. But my main concern is that my customer will (not could) create a ground loop when using a probe or a shunt combined with a PC connected to another part of the battery. A differential probe in combination with a current shunt isn't a good idea (some are 0.5mV/A !) because you really don't want the extra attenuation which comes with the differential probe. Another option would be a clamp on current probe with a >1kA DC-1MHz range. I have not seen one so far.

I know too well the position you find yourself in......customer has some experience but not quite enough so one can let them loose with the appropriate tools and know they'll be OK. 
Do I also correctly suspect you might be driving them remotely?

The other factor that's not been discussed is purchase timeframe and that also relates to your so-called buggy low cost DSO's as they progressively become more mature.

So let's explore Benchtop DSO's further:
BW 60 MHz + ?
Sampling 1 Gsa/s +?
Channels 2 or 4
Memory 1 Mpt +

and has a reasonable screen resolution

Define reasonable?

The above are potentially easy and many DSO's could be suitable until we get to Isolated channels and CAN decoding. So let's address isolation at this time as this requirement can be addressed in other ways.
Needs are:
Voltage and current measurements (shunt type)
Resolution required for each unstated.

We suspect current measurement will need to be DC level capable and therefore the simplest is with a shunt and for isolation a Differential probe is a perfectly acceptable solution.

From a know brand differential probe manufacturer.
Just one from their range:

1mV High Sensitivity Model
 65 Vp-p/30MHz Max.
 3 Range: x1/ x2/ x10
 Input Impedance (Between Input): 2M? // 1.7PF


So that leaves CAN decoding......

And so-called buggy DSO's for a cost effective solution.

I know there soon be further information to assist you by way of a respected members review of the SDS1000X to add to the only apparent other one hastily done by Dave.

[nctnico]

The purchase timeframe is yesterday so I need something which is useable today. If there was more time I might have put another candidate on the list with bench scopes.
Reasonable screen resolution is >=640x480. CAN decoding isn't strictly necessary but CAN triggering is probably very usefull. A differential probe with only 65V peak capability isn't going to cut it. The voltages they work with can be / sometimes are higher than that.

[Ian.M]

It appears to be an exercise in muppet-proofing.
Would it be viable to hack the probes by putting inline fuses in the ground clip leads so mistakes with an ordinary scope become non-catastrophic?

[_Wim_]

At a customer I work for, we have a couple of fluke 225c, and if I remember correctly, it has CAN decoding. It a very basic scope (for $$$$), but it is really robust and can be used by people that have no clue how to operate a scope (which is the mean reason the customer ONLY allows isolated scope that are CAT III rated...)

[_Wim_]

Just looked it up, it is only bus healt monitoring, no actual decoding...

http://www.tequipment.net/Fluke225Cscope.html

[nctnico]

It appears to be an exercise in muppet-proofing.
Would it be viable to hack the probes by putting inline fuses in the ground clip leads so mistakes with an ordinary scope become non-catastrophic?

I wouldn't call it muppet-proofing. Even I could make a mistake and batteries are typically very unforgiving. Fuses in probes don't sound like a good idea either. The battery management boards inside the batteries have some less-well protected interfaces as well which would still be blown up because a fuse is too slow. Secondly over time some of the probes will have blown fuses which nobody cares to replace so every now and then time gets wasted by probes which don't work.

Anyway, I have recommended the U1620A to them and the R&S RTH1002 as an alternative but to me the U1620A is the best fit for that particular customer.