Home
JAQForum Ver 20.06
Log In or Join  
Active Topics
Local Time 04:51 29 Mar 2024 Privacy Policy
Jump to

Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me.

Forum Index : Electronics : Siglent ISFE Isolated Front End module

     Page 1 of 2    
Author Message
nickskethisniks
Guru

Joined: 17/10/2017
Location: Belgium
Posts: 404
Posted: 09:53am 15 May 2021
Copy link to clipboard 
Print this post

Hi, I was  looking for another scope when I found this module:
https://www.siglent.eu/product/1142088/siglent-isfe-isolated-front-end-module

I want to look in to the switching signals in a half bridge in the near future, so measuring the 2 gates at the same time, and looking at the two D-S transitions at the same time would be my main goal. This looked to be a good option. Would this be suitable? 25khz would be the minimum frequency I like to test at, is 1Mhz bandwidth enough to see the oscillations for example?

I know you can't expect the world for this price.

I came across some discussion here:
https://www.eevblog.com/forum/testgear/siglent-isfe-scope-front-end-isolator-any-opinions/
Edited 2021-05-15 19:53 by nickskethisniks
 
Warpspeed
Guru

Joined: 09/08/2007
Location: Australia
Posts: 4406
Posted: 12:13pm 15 May 2021
Copy link to clipboard 
Print this post

One Mhz bandwidth is pretty limited and I doubt if you could really see much which would be useful. I cannot see this as a bargain at the price.

How about an old Tektronix A6902B ?
Two input channels fully isoloated to 500v above ground, dc to 20 Mhz bandwidth, 20mmV to 500v per division displayed on your oscilloscope.

These are now about twenty years old and obsolete, but they do come up quite regularly on e-bay and from secondhand equipment brokers. Prices vary a lot depending on physical condition and the probes/accessories that come with it.

Here is the service manual for it:
https://archive.org/details/tektronix_A6902B/page/n1/mode/2up
Cheers,  Tony.
 
Haxby

Guru

Joined: 07/07/2008
Location: Australia
Posts: 417
Posted: 10:12pm 15 May 2021
Copy link to clipboard 
Print this post

I recently had the same question here

Seems like a common problem for us inverter builders.
 
nickskethisniks
Guru

Joined: 17/10/2017
Location: Belgium
Posts: 404
Posted: 01:08pm 18 May 2021
Copy link to clipboard 
Print this post

Hi Haxby,

Would the micsig dp10007 be a better choice voor gate measurements? It should have a lower noise level, 15mV instead of 40mV.

Here is a comparison between the pintek dp25 and de micsiq dp10013:
https://www.youtube.com/watch?v=892RcsIUAtU

Because I can't find the dp-25 in EU, I'm looking at the pintek dp-65 now, should fit my needs and beyond at this moment but 369eur...
https://www.eleshop.nl/pintek-dp-65-differential-probe.html

For the same price I could buy 2 micsig's dp10007 or higher voltage rated...
Something to sleep on.
 
poida

Guru

Joined: 02/02/2017
Location: Australia
Posts: 1387
Posted: 10:56pm 18 May 2021
Copy link to clipboard 
Print this post

I have both the Tek A6902B and the DP-25
The DP-25 has much less noise, is accurate and stable immediately upon power up
and is very compact.

The A6902B is none of the above. I could improve calibration with a couple of
hours fun with internal adjustments and the service manual.
Zero drift was about 3 divisions over a 30 minute warm up.

One thing I noticed was how nice it is to have two channels.
The DP-25 is only single channel.
2 channels means you can put both low and high side signals through
identical signal paths and then have the result effected by the same delay, freq. response and gain.

The much larger noise in the output of the A6902B can be dealt with nicely
with the DSO "average" capture modes since we are looking at repetitive signals
that can be triggered off of reliably and with good time precision.

If you are thinking of getting a A6902B, there are a couple of probe options.
The option with 500V probes is probably the best. These probes are quite like
normal 1:10 Tek probes. They are much more compact than the 3000V type that is
an option.

If I was starting again, I would be happy with the A6902B only.
But I would want one that has both 500V probes, looks in perfect condition
and was tested for correct operation of all controls and 2 channels.
The attenuation knob is a flimsy plastic linkage and could be prone to breaking.
And likely an irreplaceable part.
Prices range from $90 US for a one probe (3000V) and zero accessory
to about $180 for both probes, manual, good condition and tests OK.
It's power supply is universal and only needs the voltage selector to be correctly
positioned for your mains voltage.

A subtle thing with the 2 channels is that the ground clips for both channels
are isolated from signal ground/supply AS WELL AS EACH OTHER.
So you can put one ground clip on phase 1 of 240V 3 phase, the other clip on phase 2 and measure phase 3, all without exceeding the 500 V probe rating, and not blowing things up. But there is a fair bit of common mode interference.
wronger than a phone book full of wrong phone numbers
 
Warpspeed
Guru

Joined: 09/08/2007
Location: Australia
Posts: 4406
Posted: 12:05am 19 May 2021
Copy link to clipboard 
Print this post

That is all extremely helpful Peter.

In spite of its massive size an "clunkiness" of the A6902B, I agree that having two matched and totally isolated channels is a big advantage, and its definitely the lowest cost way to get TWO totally isolated inputs.

For what we inverter guys will mostly be using it for, ultra low dc drift and noise are not really a major concern, mainly just sufficient speed to see ringing and overshoot and measure dead times and so on, with higher amplitude digital waveforms.

The Tektronx should also be repairable if it ever does take a dump.
Not sure about the others, as these days schematics, service information, and spare parts are never available from anything that ever comes out of China.
Multilayer surface mount boards are also just about impossible to trace out, and many of the components will have no identifiable markings on them. If you burn a resistor for example, no way possible of knowing what it was.

The Tek on the other hand uses all ancient through hole readily identifiable and available parts.
Cheers,  Tony.
 
poida

Guru

Joined: 02/02/2017
Location: Australia
Posts: 1387
Posted: 01:47am 19 May 2021
Copy link to clipboard 
Print this post

Warp, I used AC coupling on the DSO to cure the DC drift completely

Nicks, if you want to start experimenting now, and have a 4 channel DSO
how about you set it up like so:

All DC coupled inputs
channel 1 = trigger/sync
ch2 low side Gate, referenced to DC ground 5V/div
ch3 high side Source, referenced to DC ground 5V/div
ch4 high side Gate, referenced to DC ground 5V/div
DC offset ch3 and ch 4 to fit the 50V peak in the middle of the display.
It may not be possible to get enough DC offset from DSO control so use
different V/div..

capture the event, maybe use average or high res. mode

Then save the capture as a CSV file. (my Rigol can do this, not sure
if your DSO can)

Import into Excel, add an extra column, which will be (ch4 - ch3)
Look at results.

This approach will cancel all signal delays between high and low Gate voltages.
The DSO should have fine adjustment for channel to channel delay matching.

The reason I did not use built in math functions of the DSO
is because they are very poor quality results.
MAybe your DSO has usable math functions, i.e. display (ch4 - ch3) on screen
But the Rigol's math function is not good enough.
wronger than a phone book full of wrong phone numbers
 
nickskethisniks
Guru

Joined: 17/10/2017
Location: Belgium
Posts: 404
Posted: 06:53pm 19 May 2021
Copy link to clipboard 
Print this post

Thanks for the replies, difficult choice.

Nope, I don't have a 4 ch scope, it's on my long "to buy" list, I own an owon SDS7072, nothing special.

But since the pintek comes out better, I think I can expect a similar result from the micsig probes and the Tektronix. I think the newer ones have more benefits than the tektronix, Tektronix is serviceable and so a big benefit, but I don't expect the new ones to fail in a short period.

If I was able to buy a Tektronix locally for around 200 eur ( 250US) I think I would go for it. But I only see units from US and that would cost me at least 150eur for a unit and 90 eur for the transport without import and declaration costs, another 31%, roughly 320eur total. I contacted the sellers with ultra-cheap price quotation, but heard nothing back about transport costs to europe.

So I'm tempted to go for 2 pieces DP10007 since it's "only" 169eur/piece, I could buy a higher rated one later if needed, or keep an eye on a A6902B when it pops up.

When the scope disconected from earth, when testing D-S signals in a half bridge I could also maybe test with the ground clip connected to the ac output and 1 probe to + rail and 1 probe to the negative rail and just invert the signal measuring the ground. Gate signals would be difficult that way.
 
poida

Guru

Joined: 02/02/2017
Location: Australia
Posts: 1387
Posted: 10:41pm 20 May 2021
Copy link to clipboard 
Print this post

or make one.
this will perform well
Check out the testing he did with it.

Then a more experienced E.E. had a go and improved it
and it works really well

It looks like this:



All SMD of course.
here are the files for the project including the BOM.





Edited 2021-05-21 08:50 by poida
wronger than a phone book full of wrong phone numbers
 
gaspo
Regular Member

Joined: 25/06/2018
Location: Australia
Posts: 61
Posted: 09:48am 04 Dec 2023
Copy link to clipboard 
Print this post

  poida said  or make one.
this will perform well

Looking at the schematic of this probe it does not seem to have galvanic isolation.
Can this DIY probe be safely used with the grounded oscilloscope to measure signals on the  the inverter (both primary and AC side)?

I can source components for two probes for less than $120-140, just not sure of safety of it.
 
phil99

Guru

Joined: 11/02/2018
Location: Australia
Posts: 1721
Posted: 11:40am 04 Dec 2023
Copy link to clipboard 
Print this post

  Quote  Looking at the schematic of this probe it does not seem to have galvanic isolation.
It doesn't but that is not always necessary. As long as the two attenuator divider chains have an adequate voltage rating, as this does, it should be safe.
Provided the attenuator end is  enclosed in plastic to avoid accidental contact I would use one.
 
gaspo
Regular Member

Joined: 25/06/2018
Location: Australia
Posts: 61
Posted: 11:04am 29 Dec 2023
Copy link to clipboard 
Print this post

I've finally built the two probes shown in the previous post, and they appear to be working. However, I'm a bit confused about their measuring capabilities.

According to the specifications:

- Common Mode Range > ±340V (240VAC produces a sine wave with a peak-peak value of 679V).
- Differential Voltage Range > ±25V for 240VAC common-mode, ±25V for 0V common-mode.

I have a question about the "Common Mode Range > ±340V." Although the probe has a 1/250  attenuation for working with mains, the signal between the inputs should still be within ±25V; otherwise, it may get clipped.

My main question is how and where to connect the probe inputs to achieve a 240VAC common-mode and then measure it accurately. Additionally, I wonder if I still need to use a voltage divider in front of the probe to ensure the signal remains within the ±25V range.








 
phil99

Guru

Joined: 11/02/2018
Location: Australia
Posts: 1721
Posted: 11:53am 29 Dec 2023
Copy link to clipboard 
Print this post

  Quote   I wonder if I still need to use a voltage divider in front of the probe to ensure the signal remains within the ±25V range.

That is my understanding of how it works.
To check connect the inputs to a 30V transformer, if it clips you have the answer.
  Quote  My main question is how and where to connect the probe inputs to achieve a 240VAC common-mode and then measure it accurately.

If it does clip in the above test a 10:1 attenuator will be needed to measure a differential of 240V. 1M ohm across the probe tips with 4.5M from each tip to the 240V. For a 50Hz fundamental that may be enough but to accurately read harmonics and spikes small adjustable compensation capacitors will be needed across the 4.5M resistors.
Use a square wave to adjust them to remove any rounding or peaking of the edges.

Edit
Or use a standard probe on the output of a 24V transformer. 20VA or larger for best accuracy. Waveform distortion can become significant under 5VA, though a resistor in series with the primary reduces it at the expense of phase accuracy.
Edited 2023-12-30 06:51 by phil99
 
gaspo
Regular Member

Joined: 25/06/2018
Location: Australia
Posts: 61
Posted: 12:26am 30 Dec 2023
Copy link to clipboard 
Print this post

  phil99 said  If it does clip in the above test a 10:1 attenuator will be needed to measure a differential of 240V.

Yes, it does clip the signal when the measured Vpp > 50V. If I still need to use the external attenuator what is then purpose of probe's built-in 250:1 attenuator.

Is there a special way of connecting the probe or setting up the circuit in "Common Mode" to be able measure up to ±340V peak?
 
phil99

Guru

Joined: 11/02/2018
Location: Australia
Posts: 1721
Posted: 01:01am 30 Dec 2023
Copy link to clipboard 
Print this post

  Quote   what is then purpose of probe's built-in 250:1 attenuator.

Without it the op-amps would have 340V on the pins. They don't like that.
The probe allows you to measure small differences in voltage on a live circuit.
Eg. Many domestic pump controllers use 5V and 24VDC supplies derived from 240V via a series capacitor and bridge rectifier. The DC "ground" is live so this probe would be ideal there.

If the voltage difference between your mains neutral and earth is small you could use a regular probe with the earth clip removed. The scope ground will be via its mains earth. Not perfect but often good enough.
 
gaspo
Regular Member

Joined: 25/06/2018
Location: Australia
Posts: 61
Posted: 02:00am 30 Dec 2023
Copy link to clipboard 
Print this post

  phil99 said  1M ohm across the probe tips with 4.5M from each tip to the 240V.

Thanks Phil, I'll use the voltage divider you suggested.
 
phil99

Guru

Joined: 11/02/2018
Location: Australia
Posts: 1721
Posted: 03:31am 30 Dec 2023
Copy link to clipboard 
Print this post

I forgot to say 1he 1M is the total resistance across the probe tips. The probe's attenuator looks to be about 20M so the added resistor will be about 1.05M

Edit
A simpler option may be to add a 8.8k resistor across the op-amp end of the existing attenuator, multiplying it by ten.
Top of R1 to bottom of R20.

Put a switch in series with it for two ranges.
Edited 2023-12-30 13:46 by phil99
 
gaspo
Regular Member

Joined: 25/06/2018
Location: Australia
Posts: 61
Posted: 05:24am 30 Dec 2023
Copy link to clipboard 
Print this post

  phil99 said  A simpler option may be to add a 8.8k resistor across the op-amp end of the existing attenuator, multiplying it by ten.
Top of R1 to bottom of R20.


That is a good idea; it can be integrated into the enclosure. Just need to apply some 5-minute epoxy to keep the switch in place and prevent it from leaning on the tiny SMD resistors. I used 2K on one end and 6K8 on the other.

 
phil99

Guru

Joined: 11/02/2018
Location: Australia
Posts: 1721
Posted: 05:58am 30 Dec 2023
Copy link to clipboard 
Print this post

Looks very neat.
I hope you checked the accuracy before gluing it, in case the resistor value needs tweaking to allow for tolerances.
If it is already stuck measure any error to calculate a correction factor to apply to readings.
 
gaspo
Regular Member

Joined: 25/06/2018
Location: Australia
Posts: 61
Posted: 06:22am 30 Dec 2023
Copy link to clipboard 
Print this post

I'll glue just the switch and leave resistors accessible for adjustment if needed.
With the current value of 8.8K the scope measures about 1% more than what the DMM is showing (187Vrms vs 185)
 
     Page 1 of 2    
Print this page
© JAQ Software 2024