Siglent ISFE, scope front end isolator, any opinions?

[Lightages]

I was browsing things on AliExpress and came across this:
http://www.aliexpress.com/item/Isolation-channel-accessories-ISFE-VD-ISFE-Suitable-for-SIGLENT-s-full-range-of-digital-oscilloscope/32239589606.html

It looks like a reasonable solution for isolating scopes for measuring anything in circuit and mains referenced equipment. I was really surprised that rfloop nor Tautech had not mentioned anything about this thing with all the discussions about isolation and differential probes. Maybe I missed it.

It seems to be just what many people have been looking for. Any opinions? An experiences? What say rfloop and Tautech?

[joeqsmith]

I can see people hooking up their standard probes to the input and then touching the two probe cases.  The probes people use may not withstand the 2KV common mode voltage.   That or they themselves are grounded and touch the input probes.   Notice it only comes with one plastic shelled input cable (not that it looks all that safe).     

Seeing they suggest that it is used to measure line voltage, a surge event happens and next thing you know,  BLAM!!!   DSO dead....

Maybe you could explain how you would see it being used in a "safe" way.   Maybe Fluke will make something that would have some safety approvals.     
 

[Lightages]

Yes, of course, if it isn't properly designed and built to prevent surges from getting in it is just as unsafe as any incorrectly rated multimeter. Perhaps it is not safe for mains use. The main benefit of this for many people would be for measuring in circuit signals inside equipment without shorting the circuitry to ground through the scope grounds.

I am thinking of getting one for a review.

[dadler]

Interesting. The bandwidth is really low but could still be a useful tool for certain applications.

I'd love to see a teardown.

http://www.siglent.com/ENs/prodcut-fjxx.aspx?fjid=404&id=25&tid=1&T=2

http://www.tequipment.net/Siglent/ISFE/Oscilloscope-Probe-Accessories/

[pascal_sweden]

You better go for a Pintek DP-30HS, when your main application is to make accurate floating measurements within a limited voltage range.

http://www.pintek.com.tw/product_detail/landersound/index.php?Product_SN=97781&Company_SN=6002&Product_Site_Classify_SN=17072

The Pintek DP-30HS provides a high sensitivity, has a bandwidth up to 30 MHz, and the voltage range goes up to 65V.

You can buy directly from Pintek. I could not find a distributor, and they allowed me to buy directly from them. Bought two DP-30HS differential probes from them for about 300 USD in total.
The build quality is excellent, and it comes in a really nice casing! Professional stuff, you can tell directly looking at how the product is packaged and built.

[nctnico]

It's only has a bandwidth of one megahertz which really isn't enough. If it has the proper ratings I don't see how a surge event could destroy an oscilloscope because oscilloscope inputs are usually pretty rugged. I'm planning/working on creating an alternative product (also 2 channels) with a much higher bandwidth. It will cost significantly more than the Siglent box though but it will also have some nifty features.

The biggest downside of a differential probe is that the CMRR gets lower at higher frequencies. They may have a high bandwidth but the usual banana leads won't allow to use it anywhere near that bandwidth.

[tautech]

It looks like a reasonable solution for isolating scopes for measuring anything in circuit and mains referenced equipment. I was really surprised that rfloop nor Tautech had not mentioned anything about this thing with all the discussions about isolation and differential probes. Maybe I missed it.

You did.

Recently posted in a thread started by nctnico:
https://www.eevblog.com/forum/testgear/isolation-amplifiers-(for-oscilloscope)/msg724098/#msg724098

I have wanted to do a full thread on this Isolated Front End and I still might  but the attenuation of ISFE has not matched the probe attenuation settings available on all Siglent DSO's.
AFAIK this is about to change in future FW upgrades and add 2 to have a full 1, 2 ,5 range of probe attenuation.
When released the new SDS1000X series DSO's will have 1, 2, 5  probe attenuation so OSD measurements WILL be accurate with the ISFE.

For simple measurements ISFE is a neat, cheap safe product that gives users 2 isolated channels for far less cost than 1 differential probe.

For other DSO brands ISFE users should consider these points:
ISFE is USB powered. (*below email lists PC, this was done so screenshot could be captured to USB stick)
ISFE BNC connector spacing allows for DSO connection with supplied double ended BNC plugs that may not suit the spacings for some DSO's. 
BNC cables would/could then provide the necessary connection to Ch inputs.
For accurate OSD measurement the DSO must have a 200:1 probe attenuation setting
Probes with insulated BNC connectors should only be used as the Reference lead AND BNC connectors will be an DUT potential.
(this is when the GND lead should NEVER be referred to as GND but instead Reference)

Siglent supplies with all their DSO's, insulated BNC probes for safe use with the ISFE.


Email to Siglent re ISFE attenuation:

Hi Jade
When I visited Siglent China I had a discussion with Wells on the ISFE attenuation.
The 200x does not match any input settings for Siglent DSO’s.
This makes any on-screen measurement inaccurate.

The ISFE is a good product for its intended use: isolation.
But to market it aggressively the 200x attenuation will be a problem.

I see 3 solutions:

1.   Add 200x attenuation in all DSO firmware. Without knowledge of the attenuators, I am not sure if this is possible.
Do you know?
2.   Adjust the attenuation in the ISFE to match available attenuation on Siglent DSO’s.
3.   Add a “Siglent ISFE” option (200x)  in the Probe menu with fine adjustment for additional accuracy of measurements

Manual states 2 important things for the intended use of ISFE.

1.   Isolation for floating measurements.
How can  accurate measurements be taken if attenuation of ISFE and DSO input are not the same?

2.   Used to test high voltage.
Floating measurements are also needed for some low voltage work, this must not be overlooked.

I see the ISFE input impedance is 9 M Ohms and BNC to Alligator clip leads are supplied, so this implies a 1:1 probe connection should be used
Tests indicate a 10:1 probe produces an incorrect waveform.
It need be stated in the Manual “for 1:1 connection”


Very Simple Test
Please see attached and/or copied image:
Incorrect measurements are obvious.

Notes
SDS2304 Probe Cal output ~3V p-p @ 1KHz
All Ch 20 M bandwidth limit ON
DC coupling
All Channels set to 1V/div
256 Average ON
ISFE Channels set to GND as in manual
Ch 1 Siglent 300MHz  DSO probe 1:1
Ch 3  ISFE and Siglent DSO Probe 1:1 DSO input set to 100x
Ch 4 ISFE and ISFE alligator lead 1:1 DSO input set to 100x

*ISFE power supplied from PC
ISFE coupled to 2304 with supplied BNC to BNC connectors

 
Why 3 V test?
Small SMPS may use 3 V and isolated connection may be required.

Comments
Solutions to measurement errors  earlier mentioned.
Ch 3 using Siglent probe shows better reproduction of the waveform than ISFE Alligator to BNC cables.

Please consider my comments as very important issues for Siglent to consider re ISFE.

Kind Regards
Tautech

[tautech]

Of course I have one of these units should anybody wish to see specific tests done with it. 

[Lightages]

I don't know how I missed that thread.

It would be nice to see a small tear down. The image you posted shows a bit of a problem with compensation of the output of the ISFE into the scope, or just plain distortion. Is there any way to correct that?

[tautech]

I don't know how I missed that thread.

It would be nice to see a small tear down. The image you posted shows a bit of a problem with compensation of the output of the ISFE into the scope, or just plain distortion. Is there any way to correct that?

No teardown sorry.

Siglent ISFE package contents


Note
As mentioned before, scope BNC input spacings will determine if ISFE can be directly connected to other  DSO's without using BNC cables.

The ISFE output (to scope) BNC spacings are 32mm and match all Siglent DSO's so the provided BNC to BNC connectors (pictured)can be used for simple fast connection.

[tautech]

With the release of the Siglent SDS1000X series DSO's and R2 FW for the SDS2000 series Siglent now has the 200:1 matching attenuation settings for their ISFE.
As a result OSD vertical measurements are now reasonably accurate.

Although the ISFE provides only a 1 MHz BW it does provide a safe method of isolation for scopes and users.

Max input voltages:
Per channel +/- 600V pk

Max floating voltages:
Channel reference (Gnd) to Earth ground 1000V rms
Between isolated channels 2000V rms


As connected with SDS1202X DSO measuring probe Cal waveforms (3V 1 KHz)
Waveform position and settings adjusted for best viewing. Full BW. No Filters.

Ch 1: Siglent scope probe  1:1
Ch 2: ISFE supplied alligator lead 1:1


USB lead from another device powers ISFE as USB port needed for screen captures


ISFE supplied lead shows least over/undershoot (Ch 2)




Eres mode

[nctnico]

Did you adjust the probe? That could be the source of the overshoot. Overshoot indicates the frequency response will be off (or wrong).

[tautech]

Did you adjust the probe? That could be the source of the overshoot. Overshoot indicates the frequency response will be off (or wrong).

AS the ISFE is optimised for the supplied Alligator leads, AND 1:1 input I suspect scope probes that are for 1M Ohm inputs may not work correctly. 

How would you suggest I check the ISFE front end impedence over a representitive portion of its 1 MHz BW?

[rf-loop]

Did you adjust the probe? That could be the source of the overshoot. Overshoot indicates the frequency response will be off (or wrong).

AS the ISFE is optimised for the supplied Alligator leads, AND 1:1 input I suspect scope probes that are for 1M Ohm inputs may not work correctly. 

How would you suggest I check the ISFE front end impedence over a representitive portion of its 1 MHz BW?

Input Impedance: 10M  ohm (this is enough  for think normal 1Mohm probe comp adjust do not perhaps work)
For input C  can not find specs.

[tautech]

Did you adjust the probe? That could be the source of the overshoot. Overshoot indicates the frequency response will be off (or wrong).

AS the ISFE is optimised for the supplied Alligator leads, AND 1:1 input I suspect scope probes that are for 1M Ohm inputs may not work correctly. 

How would you suggest I check the ISFE front end impedence over a representitive portion of its 1 MHz BW?

Input Impedance: 10M  ohm (this is enough  for think normal 1Mohm probe comp adjust do not perhaps work)
For input C  can not find specs.

Exactly.
Thank you.

[_Wim_]

Does anyone has some actual experience with the Siglent ISFE? I am currently trying to repair the switch mode power supply part of an induction cooking plate (of which I have no schematics, and which has lots of signals riding on mains voltage), and this might be useful if it works reasonably ok. The 1Mhz bandwidth limit should be ok from what I have seen so far (currently I am using 2 unreferenced probes and the A-B math trick for probing floating signals), but because the signals of interest  are reasonably small and riding on 230VAC mains, this approach is too noisy to be useful.
I like the fact that it would allow me to probe 2 signals at one, but the 1/200 attenuation worries me that I will end up having the same noise levels as I have now, so actual user experience or a teardown or a description of the operation principle would be very much appreciated.

[tautech]

From tests that I've done it reproduces more accurate results with the supplied aligator leads than scope probes. Waveforms can be noisey but easy to clean up with the likes of Hi Res mode.

[_Wim_]

Hi Tautech,

Thanks for the quick reply. Did you make a comparison between using A-B and using the ISFE device? Hi-res mode is not an option for me, because the signal is not sufficiently repetitive (switch mode PS tries to restart approx every second and almost immediatly afterwards shut down again because of overvoltage on the secondary side).

[tautech]

Hi Tautech,

Thanks for the quick reply. Did you make a comparison between using A-B and using the ISFE device? Hi-res mode is not an option for me, because the signal is not sufficiently repetitive (switch mode PS tries to restart approx every second and almost immediatly afterwards shut down again because of overvoltage on the secondary side).

@Wim
Sorry never did use Maths on a ISFE and away from base this week so can't post a screenshot ATM.

[tautech]

Some years since I had a play with these but a customer ordered a couple which we need check before delivery.

SDS1104X-E with 2ch ISFE powered from scope USB port.
As can be seen at LV levels such as the 3V probe compensation output waveform noise is a few mV as input sensitivity is set to 10mV/div for a 2V/div for this 200:1 ISFE adapter which is supplied with BNC to Croc clip 1x leads.
Ch4 normal 10x probe as comparison.