What's missing from the market?

[wraper]

Regarding GW-Instek GDS-2000E series: Can the 70 MHz version be "upgraded" to the 200 MHz version?

So far no. Thought price difference between the models is is not that high as with most of the other scopes. 100 MHz version is pointless IMO as costs almost like 200MHz one.

[C]

I vote for "Reasonably priced oscilloscope Current probes"   Also differential probes for high voltage.     I look at the prices and just say I will not own them.

Ah ha, but with a Isolated channel DSO there's no need for differential probe for other than HV.
The need for current probes are reduced too as simple current shunt measurements and be made without worry of Gnd loops.

Isolated inputs are suitable for single ended signals which have a low common mode impedance to ground.  That covers almost all off-line applications but not differential probe applications. 

You imply isolated inputs are not differential inputs. 
Considering a 1M scope impedance input and the further impedance (10M, 100M, 1000M) with the correct isolated scope probe (well insulated reference lead and clip), where's the problem?
Educate me please.

tautech
  How much of electronics today is built in a way to make limited test equipment usable?
  Why is the last stage before a high speed ADC a differential stage?
What all changes when you go to fully differential circuits?
Use a simple thought process. Go back to point to point wire even for power. Now change all the wires to have some resistance.
Look at what you have.

Fully Differential Output:
First you should note an output is no longer using one of the power wires to change it's output and the signal is not using one of the power wires for a return path. When power is needed to change the outputs, equal power is needed from each power wire. Your power bypass cap becomes part of a low pass filter and no longer a coupler of noise on one power wire to the other power wire.

Differential input:
Look at the inputs. The load is between the two output signals. If you use two resistors as a load then the center tap is the center of the signal. No this center tap should not be directly connected to center of power. When the differential input becomes more single ended a signal starts appearing here. You really want the average of this point to be the center of power, a high resistance low pass. Here a single ended output with two resistors to power allow slow correction of power center if power is floating.

There is a cost to do Fully Differential and a lot of gains.
At the very low level the difference between P & N types is compensated for with difference between outputs correct while the center of the outputs change with respect to power center.

You need to look at Fully Differential as the difference between outputs and not as two single ended outputs.

The sad thing is Fully Differential existed back in the tube and transistor days.     
 

[tautech]

I vote for "Reasonably priced oscilloscope Current probes"   Also differential probes for high voltage.     I look at the prices and just say I will not own them.

Ah ha, but with a Isolated channel DSO there's no need for differential probe for other than HV.
The need for current probes are reduced too as simple current shunt measurements and be made without worry of Gnd loops.

Isolated inputs are suitable for single ended signals which have a low common mode impedance to ground.  That covers almost all off-line applications but not differential probe applications. 

You imply isolated inputs are not differential inputs. 
Considering a 1M scope impedance input and the further impedance (10M, 100M, 1000M) with the correct isolated scope probe (well insulated reference lead and clip), where's the problem?
Educate me please.

tautech
  How much of electronics today is built in a way to make limited test equipment usable?
  Why is the last stage before a high speed ADC a differential stage?
What all changes when you go to fully differential circuits?
Use a simple thought process. Go back to point to point wire even for power. Now change all the wires to have some resistance.
Look at what you have.

Fully Differential Output:
First you should note an output is no longer using one of the power wires to change it's output and the signal is not using one of the power wires for a return path. When power is needed to change the outputs, equal power is needed from each power wire. Your power bypass cap becomes part of a low pass filter and no longer a coupler of noise on one power wire to the other power wire.

Differential input:
Look at the inputs. The load is between the two output signals. If you use two resistors as a load then the center tap is the center of the signal. No this center tap should not be directly connected to center of power. When the differential input becomes more single ended a signal starts appearing here. You really want the average of this point to be the center of power, a high resistance low pass. Here a single ended output with two resistors to power allow slow correction of power center if power is floating.

There is a cost to do Fully Differential and a lot of gains.
At the very low level the difference between P & N types is compensated for with difference between outputs correct while the center of the outputs change with respect to power center.

You need to look at Fully Differential as the difference between outputs and not as two single ended outputs.

The sad thing is Fully Differential existed back in the tube and transistor days.   

Thanks.

While there are particular difficulties in isolated channel front end design it can be done and is.
A recent vid Dave did examined the isolated frontend in some detail.
There are several HH options available currently, but there are zero modern benchtop DSO's with isolated channels.

Once upon a time they were commonly called Power scopes and those I've seen were in uni labs for the instruction of 3 phase power electronics. Tell me that's not taught anymore. 

Tell me a well designed scope with 4 isolated channels can't be offered to the market for less that a currently available DSO plus the cost of 4 differential probes. 

[nctnico]

CAN decoding support.  The Rigol DS1054z can already do SPI, I2C, and RS232, why can't it do CAN? 

This is a capability I need, but I can't find it on a 4 channel scope under $2000.

GW Instek GDS2000E series to the resque again! I had the same requirement and the GDS2000E series does CAN decoding very well.

I also second the suggestion for a low cost isolation amplifier. I'm working on one (schematics for prototype ready) but need some spare time to burn on this project to see if my idea works or not.

[David Hess]

There are several HH options available currently, but there are zero modern benchtop DSO's with isolated channels.

Does the Tektronix TPS2000 count.

http://www.tek.com/oscilloscope/tps2000

Tell me a well designed scope with 4 isolated channels can't be offered to the market for less that a currently available DSO plus the cost of 4 differential probes. 

I doubt the high price of high voltage differential probes reflects the cost of design and manufacturing.  I have never seen an hobby level one so the market must be very limited.

I actually outlined what would be needed to hack together a limited performance and bandwidth one (below 20 MHz) using standard x10 probes a couple days ago for someone.

[Someone]

There are several HH options available currently, but there are zero modern benchtop DSO's with isolated channels.

Does the Tektronix TPS2000 count.

http://www.tek.com/oscilloscope/tps2000

For all the hate the TPS2000 is a quality little unit, old fashioned but the lack of competition in that area of isolated channel scopes shows just how small and undemanding the market it.

Tell me a well designed scope with 4 isolated channels can't be offered to the market for less that a currently available DSO plus the cost of 4 differential probes. 

I doubt the high price of high voltage differential probes reflects the cost of design and manufacturing.  I have never seen an hobby level one so the market must be very limited.

I actually outlined what would be needed to hack together a limited performance and bandwidth one (below 20 MHz) using standard x10 probes a couple days ago for someone.

There is an example differential probe here: https://www.eevblog.com/forum/projects/diy-100mhz-differential-probe/
Its configured for ?300V RMS input limited by dissipation in the DC divider, but the design is scalable in both directions to higher or lower voltages with more or less gain, getting lower noise needs a discrete FET input design but the most time consuming part was mating to the active probe interfaces and sourcing a suitable multi conductor cable.

[tautech]

There are several HH options available currently, but there are zero modern benchtop DSO's with isolated channels.

Does the Tektronix TPS2000 count.

http://www.tek.com/oscilloscope/tps2000

It did 20 years ago but today with 2.5Kpts of memory depth it's a joke. 

So I'll highlight my statement above. ^^^

Tell me a well designed scope with 4 isolated channels can't be offered to the market for less that a currently available DSO plus the cost of 4 differential probes. 

I doubt the high price of high voltage differential probes reflects the cost of design and manufacturing. 

Of course not.
Currently we have every conceivable feature added to modern DSO's except arguably the greatest feature that would promote simplicity and ease of use. Channel to channel reference and Ground isolation.
Hell, even a reasonable bench DMM provides for 500V+ of isolation from mains. (Voltage Bias offset)

I have never seen an hobby level one (differential probe) so the market must be very limited.

The absence of cheap differential probe adds additional cost to scope users that need one and that cost should NOT exist for want of a slightly more expensive isolated channel modern DSO.
All of us at some time have needed/wanted isolation for the signal reference (Gnd clip/lead), haven't you?
IMO this is an opportunity for some manufacturer to create a market demand, such are the benefits of simplicity and ease of use.

[C]

Is an isolated device or channel  the same as a differential input device?

Why not test and see if they are.
You may have a bunch of isolated devices. Your DMM for example.

To test, you just need a signal generator or function generator with the higher frequencies showing more. Protect the output with a series resistor of the output impedance.

So you have signal generator on one side of resistor with scope and one lead of  isolated device on other side.

Look at scope while changing frequency from low to high with following.
1. No device connected to resistor.
2. DMM negative lead connected to resistor.
3. DMM positive lead connected to resistor.

Are all three steps the same?
#3 is like one input of something with differential inputs

[tautech]

Is an isolated device or channel  the same as a differential input device?

Why not test and see if they are.
You may have a bunch of isolated devices. Your DMM for example.

To test, you just need a signal generator or function generator with the higher frequencies showing more. Protect the output with a series resistor of the output impedance.

So you have signal generator on one side of resistor with scope and one lead of  isolated device on other side.

Look at scope while changing frequency from low to high with following.
1. No device connected to resistor.
2. DMM negative lead connected to resistor.
3. DMM positive lead connected to resistor.

Are all three steps the same?
#3 is like one input of something with differential inputs

Other than my bench DMM all my equipment is Gnd referenced, just as most test equipment is.

Commonly used terminology adds confusion to general understanding, in particular mention of the Gnd leads on scope probes that are indeed common with mains ground on most scopes BUT should be correctly thought of as the signal reference lead.
This is where what would be a simple connection for signal reference with an isolated DSO is entirely a different proposition with a Ground referenced DSO by way of NO risk of introducing ground loops if care is not taken to where the reference (Gnd) lead is connected on ordinary non-isolated scopes.

A fully isolated scope overcomes this.

Edits
Understanding

[JPortici]

I have a tps2014 at work.
Two things i love it for:
- Lowest noise of all the scope we/i have
- sturdy and lovely case
Boss bought it in the 00's to look inside cars because it was the best affordable isolated scope at that time and it still was about 3k€ or something

But it is so god damn slow and i always have to hope to set the trigger+timebase right for single time events otherwise i can only see what's in the screen...
I'd much rather have a 1054z with battery power.

... not really but, you know.. some modern perks like a decent amount of memory. still the fft and the X-Y mode are eons faster 

I had considered going for a handheld scope like the new owons but the specs again are not that better than the TPS in regards of memory (what were those? 16k?) and to obtain battery powered scopes seems to have become impossible even though the technology is there as it has been for decades. what is it, a regulation problem? so the tps2000 is still king (sigh)

[D3f1ant]

Reasonably priced oscilloscope Current probes.

 +1 

[ralphrmartin]

I doubt the high price of high voltage differential probes reflects the cost of design and manufacturing.  I have never seen an hobby level one so the market must be very limited.

I'm going to pick up a couple of these when I am next in China, and see how they work out:
https://world.taobao.com/item/38029768379.htm

[tautech]

I have a tps2014 at work.
Two things i love it for:
- Lowest noise of all the scope we/i have
- sturdy and lovely case
Boss bought it in the 00's to look inside cars because it was the best affordable isolated scope at that time and it still was about 3k€ or something

But it is so god damn slow and i always have to hope to set the trigger+timebase right for single time events otherwise i can only see what's in the screen...
I'd much rather have a 1054z with battery power.

... not really but, you know.. some modern perks like a decent amount of memory. still the fft and the X-Y mode are eons faster 

I had considered going for a handheld scope like the new owons but the specs again are not that better than the TPS in regards of memory (what were those? 16k?) and to obtain battery powered scopes seems to have become impossible even though the technology is there as it has been for decades. what is it, a regulation problem? so the tps2000 is still king (sigh)

Depressing isn't it. 

There are other/better HH choices however, the SHS1000 series and if you have deep pockets the R&S.
http://www.siglentamerica.com/pdxx.aspx?id=93&T=2&tid=2

[David Hess]

There are several HH options available currently, but there are zero modern benchtop DSO's with isolated channels.

Does the Tektronix TPS2000 count.

http://www.tek.com/oscilloscope/tps2000

It did 20 years ago but today with 2.5Kpts of memory depth it's a joke. 

So I'll highlight my statement above. ^^^

The isolation barrier imposes serious throughput limitations between the digitizer and processor so expanding the record length would come at the expense of record acquisition rate.  The old Tektronix real time digitizer design was already low throughput making it a good match for this application.

2.5 ksamples * 8 bits * 200 acquisitions per second = 4 Mbits/second

[tggzzz]

If you could ask the big mfgs and knew they were listening, what would you want from them? This would be both for B2B and B2C.

I'd ask them to produce a device that would tell me what's missing from the market.

Can I have my $1 now, please.

[tautech]

There are several HH options available currently, but there are zero modern benchtop DSO's with isolated channels.

Does the Tektronix TPS2000 count.

http://www.tek.com/oscilloscope/tps2000

It did 20 years ago but today with 2.5Kpts of memory depth it's a joke. 

So I'll highlight my statement above. ^^^

The isolation barrier imposes serious throughput limitations between the digitizer and processor so expanding the record length would come at the expense of record acquisition rate.  The old Tektronix real time digitizer design was already low throughput making it a good match for this application.

2.5 ksamples * 8 bits * 200 acquisitions per second = 4 Mbits/second

There have been significant advances in memory depth across input isolation from Tek TPS2k days as this post from early this year outlines:
https://www.eevblog.com/forum/testgear/which-scope-with-isolated-inputs/msg839059/#msg839059

Even the Siglent SHS1000 that I linked earlier has 2 Mpts memory depth and while these sorts of figures can't be called stunning by todays standards they are enough to be useful.