Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronic Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

Determine Transformer Loss

Swend

Full Member level 4
Joined
May 14, 2019
Messages
237
Helped
10
Reputation
20
Reaction score
9
Trophy points
18
Location
Roskilde, Denmark
Activity points
2,304
Hi friends.

I would like to know the loss (power/energy) in a air-core pulse transformer of my own fabrication.

What we are looking at is this:
Screenshot from 2019-08-20 16-25-14.png

I'm measuring voltage with two identical 100:1 voltage dividers 100MΩ∥22p/1MΩ∥2.2nF.

I'm measuring current with two uncalibrated but identical Rogowski coils, I suspect the error in current magnitude could be ±10% but in any case the error will be the same on both coils.

Here are my measurements from both primary and secondary side of the transformer.
Figure_8.png
X-axis = nanoseconds

The voltage is being stepped down, but the current is almost identical on both side, which is puzzling to me. I would have expected the current to increase on the secondary side as voltage is stepped down.

So how do I go about determining the transformer loss?
 
Last edited:

Swend

Full Member level 4
Joined
May 14, 2019
Messages
237
Helped
10
Reputation
20
Reaction score
9
Trophy points
18
Location
Roskilde, Denmark
Activity points
2,304
OK, I will try to rephrase my question.

Accorcing to electronics-tutorials.ws in a ordinary power-transformer with sine AC, the efficiency is
Screenshot from 2019-08-21 12-25-02.png

And the power is calculated like this
Screenshot from 2019-08-21 12-25-32.png

Which is all fine, but since I have a arbitrary pulse, can I calculate average power by summing the individual product of e.g. (pri-voltage * pri-current) samples, and the divide the sum by the total number of samples? Then obviously do the same thing for secondary and calculate efficiency?

And my second question is: As it can be seen in the traces, the voltage is being stepped down, but the current is almost identical on both side, which is puzzling to me. I would have expected the current to increase on the secondary side as voltage diminish, what is the reason for this unexpected behaviour?
 

aryajur

Advanced Member level 3
Joined
Oct 23, 2004
Messages
793
Helped
124
Reputation
248
Reaction score
37
Trophy points
1,308
Location
San Jose, USA
Activity points
7,788
I did't understand your statement about measuring with 2 100:1 dividers. The transformer will step the voltage down and then depending on the load on the secondary the current should be set. If the primary current is almost the same as the secondary with the voltage being stepped down theoretically it indicates a large power loss in the transformer. Can you recheck the measurements?
 

Swend

Full Member level 4
Joined
May 14, 2019
Messages
237
Helped
10
Reputation
20
Reaction score
9
Trophy points
18
Location
Roskilde, Denmark
Activity points
2,304
Hi arajur

I did't understand your statement about measuring with 2 100:1 dividers.
Here is a picture of one. I use one to measure primary voltage and a second one to measure secondary voltage.
IMG_20190822_175105.jpg

The transformer will step the voltage down and then depending on the load on the secondary the current should be set. If the primary current is almost the same as the secondary with the voltage being stepped down theoretically it indicates a large power loss in the transformer.
This is exactly my point, it must be very inefficient, but how do I quantify it? can I calculate average power by summing the individual product of e.g. (pri-voltage * pri-current) samples, and the divide the sum by the total number of samples? Then obviously do the same thing for secondary and calculate efficiency?

Can you recheck the measurements?
Sure, but why? Not that I haven't done it many times.
 

aryajur

Advanced Member level 3
Joined
Oct 23, 2004
Messages
793
Helped
124
Reputation
248
Reaction score
37
Trophy points
1,308
Location
San Jose, USA
Activity points
7,788
Thanks for explaining it better. I requested to remeasure because it seemed the losses in the transformer may be excessive. But yes to quantify it I would just subtract the power delivered to the output from the intput power. That would give the total loss. To go a step further we could quantify the DC losses by measuring the primary and secondary coil resistances. If you don't plan to operate the transformer for high frequencies then you can probably neglect skin effect. Once you know the winding loss the rest of the loss would be due to core losses.
 

Swend

Full Member level 4
Joined
May 14, 2019
Messages
237
Helped
10
Reputation
20
Reaction score
9
Trophy points
18
Location
Roskilde, Denmark
Activity points
2,304
I requested to remeasure because it seemed the losses in the transformer may be excessive.
I think I just have come up with a extremely bad design that is extremely lossy.

But yes to quantify it I would just subtract the power delivered to the output from the intput power. That would give the total loss.
Thank you, And how about the way I calculate power i.e. can I calculate average power by summing the individual product of e.g. (pri-voltage * pri-current) samples, and the divide the sum by the total number of samples? Then obviously do the same thing for secondary and calculate efficiency?

To go a step further we could quantify the DC losses by measuring the primary and secondary coil resistances.
Yeah, but there isn't really any DC , or is there? After 45uS the current becomes DC, how to do that? Should I make a hybrid calculation i.e. AC up to 45uS and thereafter for DC?

If you don't plan to operate the transformer for high frequencies then you can probably neglect skin effect. Once you know the winding loss the rest of the loss would be due to core losses.
The operating frequency is at the moment around 33KHz, does that induce the skin effect?
 

aryajur

Advanced Member level 3
Joined
Oct 23, 2004
Messages
793
Helped
124
Reputation
248
Reaction score
37
Trophy points
1,308
Location
San Jose, USA
Activity points
7,788
Thank you, And how about the way I calculate power i.e. can I calculate average power by summing the individual product of e.g. (pri-voltage * pri-current) samples, and the divide the sum by the total number of samples? Then obviously do the same thing for secondary and calculate efficiency?
Yes I think you need to do the product of the samples and then sum and divide them as you mention. Basically integrate the VI. Same thing for secondary.

Yeah, but there isn't really any DC , or is there? After 45uS the current becomes DC, how to do that? Should I make a hybrid calculation i.e. AC up to 45uS and thereafter for DC?
I shouldn't have said DC. I meant ohmic losses i.e. I^2 R losses in the transformer wire. You already have the currents in primary and secondary if you know the resistances then you can calculate what losses happened in the coils

The operating frequency is at the moment around 33KHz, does that induce the skin effect?
At 33KHz the skin depth for copper is 0.36mm. If your wire radius is close to it or smaller than it then you can neglect it.
 
Last edited:
  • Like
Reactions: Swend

    Swend

    points: 2
    Helpful Answer Positive Rating
Toggle Sidebar

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Top