Isolated power - IEC61010

[bodewonkinobe]

Hi everyone

I am designing a product which needs an isolated measurement input which will connect to mains (as well as various CATII circuits up to 1kVrms).

I am designing to IEC61010 and need to get communication across an isolation barrier as well as power across the isolation barrier to the analogue measurement circuits (as well as some digital circuitry).

IEC61010 describes the overall product testing requirements and creepage/clearance etc but it does not detail the requirements of the isolators themselves.

There is are a number of standards that apply to the opto isolators (I will use opto isolators to transfer data across the barrier), such as the IEC Standard is IEC60747-5-2.

The Broadcomm ACNV4506 opto is good for an operating voltage of 2262V, I need 2kV isolation so this is perfect.

However, I am not sure which standard to refer to for the isolated power.

I think I will need to design my own isolated PSU as i can’t find an off the shelf model offering a 2kV continuous working voltage.

My requirements are;

Max Vpk continuous : 2kV
600Vrms CAT III Minimum (higher if possible)
1kVrms CAT II

I need to check the creepage and clearance in IEC61010 of the ACNV4506, at 13mm, I imagine this will be ok but I will make no assumptions.

So my real question is: What standards relate to the isolation transformers in switching PSUs that are used to power isolated measurement channels?

Do you know of any off the shelf solutions to meet the requirements of IEC61010 for a continuous working voltage of 2kVpk?

I have read in this article - (https://www.edn.com/Pdf/ViewPdf?contentItemId=4441292) It describes how IEC61010 gives guidance for safety tests of a product and how it details the creepage and clearances required from a product design perspective. But IEC61010 does not detail component specific requirements.

So as far as the isolation transformer for the power to the isolated board is concerned, other than the primary to secondary creepage and clearances from input to output leads, what standards are applicable to the overall isolation offered by a transformer?

I am not a transformer designer, is the only option to contact a transformer company to design a 2kV insulated transformer? Which standard should it meet?

Thanks

bode

 

[KlausST]

Hi,

Go to a DCDC converter manufacturer and use the selection guide to find devices with the expected isolation voltage.

Klaus

 

[Easy peasy]

where do you get 2kV continuous isolation stress from? we have designed psu's up to 120kV peak operating isolation

2kV continuous requires a 1min opto rating of > 7500Vac ...

 

[FvM]

I don't understand the statements about IEC 61010 in post #1. It does specify required test voltages, they can be also applied for DC/DC converters and isolators. There are specific standards e.g. for isolators, but the test voltage requirements are almost similar.

I presume, you are designing reinforced insulation

1 kV CAT II demands 5400 VAC 5s (IEC 61010-2010 Table K.2) and 4400 VAC 60s (IEC 61010-2010 Table K.8)

In addition, you should refer to the working voltage specifications of DC/DC and isolator vendors.

- - - Updated - - -

The said test voltages are also good for higher secondary circuit voltages as long as the nominal mains voltage doesn't exceed 1000Vrms, see table K.10.

- - - Updated - - -

You can get DC/DC up to 10 kV test voltage, 5kV partial discharge free from specialized manufacturers like

 

[bodewonkinobe]

I don't understand the statements about IEC 61010 in post #1. It does specify required test voltages, they can be also applied for DC/DC converters and isolators. There are specific standards e.g. for isolators, but the test voltage requirements are almost similar.

I presume, you are designing reinforced insulation

1 kV CAT II demands 5400 VAC 5s (IEC 61010-2010 Table K.2) and 4400 VAC 60s (IEC 61010-2010 Table K.8)

In addition, you should refer to the working voltage specifications of DC/DC and isolator vendors.

- - - Updated - - -

The said test voltages are also good for higher secondary circuit voltages as long as the nominal mains voltage doesn't exceed 1000Vrms, see table K.10.

- - - Updated - - -

You can get DC/DC up to 10 kV test voltage, 5kV partial discharge free from specialized manufacturers like

Hi FvM

Your presumption is correct about reinforced insulation.

I understand the test requirements of IEC61010.

Am I correct in saying that if the design “survives” these elevated test voltages of 1 min (eg 4400VAC) then the product will survive 1kV CATII environments continuously?

Those DC-DC converters are just what I needed! I am not sure why I am finding it so difficult to find converters with high continuous isolation ratings :-(

 

[FvM]

Am I correct in saying that if the design “survives” these elevated test voltages of 1 min (eg 4400VAC) then the product will survive 1kV CATII environments continuously?

That's at least the assumption. There are however mechanisms that may cause long term insulation failure or device functional loss. Thus you should check for additional manufacture specifications. For AC voltage strain, partial discharge is a possible problem. Continuous applied DC voltage may cause something like ion diffusion. Some signal isolators have a fine print in the data sheet about possible function loss under permanent DC voltage.

 

[bodewonkinobe]

Hi,

Go to a DCDC converter manufacturer and use the selection guide to find devices with the expected isolation voltage.

Klaus

Hi KlausT

Thanks for the reply, my issue was partly that I couldn’t find any examples that state a higher isolation voltage of around 1500V.

Regards

bode

where do you get 2kV continuous isolation stress from? we have designed psu's up to 120kV peak operating isolation

2kV continuous requires a 1min opto rating of > 7500Vac ...

Hi Easy Peasy

I do not quite understand your reply, do you mean that you don’t know why I can’t solve the 2kV isolation problem?

How did you go about designing the flyback transformer for a 120kV system??? Or did you use another scheme?

I’m already happy with the opto, the one I mentioned is rated to 7500VAC for one minute.

My issue is getting power across the barrier as I have limited experience designing isolation transformers.

 

[bodewonkinobe]

If I use the following Recom DC-DC converter as an example: https://docs-emea.rs-online.com/webdocs/1373/0900766b813731f2.pdf

The /R8 unit is rated as follows:

4000VAC for 1 minute
8000VDC 1 second

For hypothetical requirements of 1000V CAT II and 600V CAT III the following test requirements apply in IEC61010;

Table K.2
5 seconds - 5400VAC (Reinforced), 3310VAC (Basic/supplementary)

Table K.8
1 minute - 4400VAC (Reinforced), 2200VAC (Basic/Supplementary)

Would I be correct in saying that these Recom units will only meet the requirements of IEC61010 for basic/supplementary as the 1 minute test voltage is only rated to 4000VAC, not 5400VAC or 4400VAC?

 

[bodewonkinobe]

I don't understand the statements about IEC 61010 in post #1. It does specify required test voltages, they can be also applied for DC/DC converters and isolators. There are specific standards e.g. for isolators, but the test voltage requirements are almost similar.

I presume, you are designing reinforced insulation

1 kV CAT II demands 5400 VAC 5s (IEC 61010-2010 Table K.2) and 4400 VAC 60s (IEC 61010-2010 Table K.8)

In addition, you should refer to the working voltage specifications of DC/DC and isolator vendors.

- - - Updated - - -

The said test voltages are also good for higher secondary circuit voltages as long as the nominal mains voltage doesn't exceed 1000Vrms, see table K.10.

- - - Updated - - -

You can get DC/DC up to 10 kV test voltage, 5kV partial discharge free from specialized manufacturers like

Hi FvM, I hope it is ok to come back to your response...

You mention tables in iec61010-1:2010, but shouldn’t I be looking at table k101.2 in 61010-2-030:2010 as I am talking about signals on the measurement inputs?

The 61010-2-030 standard refers to the measurement inputs whereas the 61010-1 standard refers to the voltages at the power input or voltages generated from within the instrument. Am I right?

Eg: For 600V CAT III measurement inputs table k.101 gives 10.5mm clearance (reinforced) and table k.16 gives a test voltage of just over 4950V.

Or am I way off here?

Sorry to be a pain, but it these standards are a little obscure at times!

Thanks

 

[FvM]

I'd say, you are starting an additional topic.

You have been originally asking about test voltages and isolator component specifications. Test voltages are essentially the same in 61010-1 and 61010-2-30 as far as I see. It may be appropriate to consider 61010-2-30 additionally for your device, I'm not sure. I notice that this standard doesn't distinguish between pollution degree and insulator properties, as if it wants to set a worst case requirement that should be e.g. used for terminals of a measurement device.

I'm not sure if it applies to all internal circuits of a measurement device, sounds rather arbitrary in my eyes. I have the standard but I didn't yet apply it.

 

[Easy peasy]

The OP is not making a clear distinction between safety standards which are designed to prevent breakdown for occasional over-voltage events - OR - designs where there is a continual high voltage between input and output.

An opto that can handle 7500Vac for 1 min, will not survive 2kVac for 10 years. Similar for many isolation systems. This is a common newbie mistake.

Dear OP, (original poster) please clarify which you are trying to design for...?

 

[asdf44]

I think the OP understand that. The broadcom opto referenced is in-fact specified for 2262V continuous peak working voltage.

I see the OP trying to sort out what makes the regulatory people happy and how to find a DC-DC with the same type of specified working voltage.

I work in 61010 also but not at these types of voltages (And I sure don't claim to fully understand the spec). Above roughly 500 working volts starts getting quite hard to find.

 

[bodewonkinobe]

The OP is not making a clear distinction between safety standards which are designed to prevent breakdown for occasional over-voltage events - OR - designs where there is a continual high voltage between input and output.

An opto that can handle 7500Vac for 1 min, will not survive 2kVac for 10 years. Similar for many isolation systems. This is a common newbie mistake.

Dear OP, (original poster) please clarify which you are trying to design for...?

Hi Easy Peasy

Thanks for the response, as per my original email I am desgning an isolated input to theoretically survive a continuous 600V CAT III and 1000V CAT II “environment”.

I know of other products using the Broadcomm opto in a 600V CAT II rated products, but not CAT III.

I accept that just because the opto will handle 7500AC for 1 min won’t mean it handles 2kV for 10 years. But the only tests indicated in the standards which set out the test voltages for a 1000V CAT II and 600V CAT III rated input are the “1 min” tests.

Looking at IEC61010-2-030, I believe the Broadcomm would pass these tests and allow me to rate the product using the Broadcomm to 600V CAT III, if I can find an isolation transformer for the power to the isolated board.

To answer your question, I am designing for a “continual 600V CATIII or 1000V CATIII” on the input terminals.

- - - Updated - - -

I think the OP understand that. The broadcom opto referenced is in-fact specified for 2262V continuous peak working voltage.

I see the OP trying to sort out what makes the regulatory people happy and how to find a DC-DC with the same type of specified working voltage.

I work in 61010 also but not at these types of voltages (And I sure don't claim to fully understand the spec). Above roughly 500 working volts starts getting quite hard to find.

Hi asdf

That’s exactly what I’m trying to do, I can see that the Broadcomm “should” pass the tests specified in iec61010, but the proof will be in the “pudding”.

DC-DC converters with a continuous isolation rating of 2kV+ are hard to find.

I am beginning to conclude that if the isolation devices pass the voltage withstand tests in iec61010, then this will deem the isolation barrier satisfactory. At the end of the day, the wording in the standard is clear and states the tests needed to verify the isolation barrier.

There is also a test in iec61010 (table k.105) staying the “1 minute test voltages” required to test the measurement inputs for long term stress of the isolation.

I am continuously studying the standard and will not finalise the design until I fully understand the safety isolation requirements.

P.S. Would you classify an opto and a transformer as “solid isolation”?

 

[Easy peasy]

P.S. Would you classify an opto and a transformer as “solid isolation”?

yes they are - or should be

Broadcom should have data on the long term DC and AC volts withstand of their opto's - you will need to email them...

 

[bodewonkinobe]

yes they are - or should be

Broadcom should have data on the long term DC and AC volts withstand of their opto's - you will need to email them...

Thanks for that.

For anyone interested, I am going to go for a push pull custom built transformer using a shielded E20 bobbin (plastic guard to separate the winding so they donÂ’t oberlap and have high insulating plastic between the primary and secondary.

I will use an LT3439 driver.

I will be needing basic insulation from the opto and PSU as I will have additional bonding of all “accessible parts”. This reduces my clearance requirements and test voltages.

I am in communication with Broadcomm to check that I can perform the isolation/withstand tests at the voltages I need.