Rating of the chopper circuit: Output Voltage: 27V Output Current: 650A Ripple Voltage (max): 0.5% Oscillating frequency: 9.678 Khz The inductor used in the circuit implementation has an iron core. The core temperature of the inductor ~60 degree centigrade above ambient. Will this have a negative effect on the effectiveness of the circuit in the future?
This link states that a class-H transformer is allowed to have a winding temperature rise of 150 deg in a 40 deg C environment. This is 190 deg C and the core should be close to the winding temperature.
I dont quite understand what youve stated Dave your saying this is descibed as class H but when you read through these notes this has to do with the laminations used in the transformer "Correct me if I've misunderstood " , plus any eddy currents present, frequency of operation they are quoting here max 120hz
How do you classify this as Class H when we dont have the type of matarial that the core is manufactured from i.e the laminations If this were for example normal ion i would think this would have significant detremental effects on the circuit seeing as this operates at 9.678 Khz HF
He also hasnt mentioned what the frequency of the core is rated at or if this is a torroidal core and they too have ther own colour codes which relate to the frequency of operation particularly if you are talking about switch mode power supplies which is what this is
Quote from other sites Impossible to know without full information about the performance of the core. Chances are the "Iron Losses" will go through the roof at that frequency, and it will rapidly overheat.
Further information Quote from: mikeselectricstuff on February 07, 2011, 11:00:38 AM
Eddy current losses increase with frequency, which is why ferrite is used instead of laminated iron at higher frequencies. If we know the full information about the toroidal core, is it possible to calculate all the parameters in Model A and Model B? Thanks. o figure out if you're going to saturate it, you need to use the calculation:
This is only valid for sinusoidal signals, but it's close for others. If it's an iron core, you need to keep Bmax under about 1T, and if it's ferrite then 200uT is a good rule of thumb. You need to know N (the number of turns) and A (core area) which might be tough if the datasheet isn't really good.
The core is of CRGO http://www.transformercore.com/tech_articles.htm (I cannot use ferrite cores as the huge frame ferrite coores is not readily available)This is the link for it's chemical composition.Thee cores generally operate at 50Hz.I would like to know the effects this has in the circuit.Thank you
Did you read any of the notes i made reference too They all suggest that it does make a big difference especially at 10khz at 50hz yes I might agree with you but at 10khz They all say it makes a massive difference and has extremely negative effects of the operation
They can’t all be wrong can they or are you suggesting they are From my own experience in building pwm supplies at different frequencies of operation I know dam well they are right and if you don’t match the cores operating frequency to the operation of the supply It does just that It overheats in seconds and gets baking hot so much so that you land up with shorted turns on the primary and secondary coils So they cant all be wrong can they !!
Quote fomr the questionairs specs very limited as well hes saying quote ,"Thee cores generally operate at 50Hz.I would like to know the effects this has in the circuit.Thank you " unquote
Then is his specs he say what !! Quote , " Output Voltage: 27V ,
Like the bloody electrician who fits a shielded cable to run loudspeaker sound on a 100 volt line through a building and uses shielded pair phono lead about 0.5 mmm if that or cuts a bloody hole in the ceiling right above a bloody staircase so that when you have to service this or replace this You have to erect five stories of scaffolding because he / she didn’t think about it at all ( when designing it and when you tell them how you want the cable run which is so simple They cant even do that properly or to spec also a bit like your builders with plasterbourd with supposidly 400cm based battoning or cross battoning and they say Oh yes thats all be done you know untill you have to mount a 30 to 40 kg braket on there only to find Oh where are the battons they tell us are behind this , "Missing altogether !!"
I do see lots of this you know Top class designer Yeah ok !! Like the mains wiring underneath a leaking pipe no conduit Super Jobs reflective of super engineers and super duper designers, super duper managers , Super duper directors who manage to waste 27 000 pounds with no schematics and set fire to cars Oh Cool babes !! isn’t that just wow
i think you forget i see this every day of my life and have done for years with major rows with corperate companies who dont listen at all
I think Il stick with the blokes in the armed forces I might stand less chance of getting myself killed or my building set alight because of these blithering idiots
In fact come to think of it at least they listen dam sight more than what I can say about some of these people ive met or had dealings with
An iron core, or more correctly silicon steel, such as 3% grain oriented silicon steel 1 mil thick (see link) is suitable for use at 10 kHz. It appears that with a Curie temperature of 730 deg C the core temp is not the limiting factor and that the insulating materials are. The Curie temperature of this material is 730 deg C.
A suitably sized C-core made with 1 mil thick material would do the job on this 18 kW switching power supply.
I’m not arguing with you Dave but I cant help but notice that the question isnt right to start with neither are the facts but I do know enough to be able to tell you that it does have significant impact both practically and theoretically at 10khz Try it yourself build a pwm supply and see ( do it !! ) That’s the only way you going to prove this to yourself
I came across the same problem on a pump out of a washing machine at 50 hz ceasing due to eddy current laminations over heating causing the pump to fail then the timer with control circuitry It blew the Triac up in the end on a Miele washing machine top of the range That was the reason and it didn’t show for at least 20 minutes until you felt the pumps laminations would work A ok for at least 20 to 30 minutes then fail for that reason
I had to probve this to the customer in the end who also didn’t believe me to start with untill he saw it for himself And ive had to do exactly the same and change a manufactures specification on car amplifier also due to incorrect tolerance on windings and incorrect gauge used on swicth mode transformers and IT DOES MAKE A MASSIVE DIFFERENCE TO OPERATING TEMPERATURE THE LOT
As I said before I received better training from someone in the SADF than I gained from other companies In fact I received better training from Panasonic than any other manufacture out there in the Uk most of them ex service men
They know what they are on about very seldom they dont
The power supply, the specs, the circuit, and the transformer core temperature question are quite satisfactory. I strongly suspect the power supply is an electroplating power supply. A power supply having similar specs - low voltage, high current, and low ripple is shown in the link below.
INever suspect I get the facts first I learnt that lesson a long time ago and never just assume anymore because 9/10 assumptions prove to be incorrect especially when some says ," It is and actually isn’t " I hope you learn this quickly as well for your own sake not that I’, trying to upset you I just know from shear hard experience and having to learn the hard way Sometimes the best way to learn you need to fail now and again to get that wake up call
When you have got the answer to that question perhaps you might like to also ask for all the details complete re that schematic inclusive a decent schematic showing exactly what that supply is intended for and a complete explanation on how its supposed to work Id be very interested to see that answer if the questionnaire actually gives you the full information that’s another story all on its own
This power supply is going to be used in the army for something that has not been divulged to me as yet.However the spec sheet and the circuit diagram are both correct.I know that there will be huge overheating and my question is that will it affect the properties of the core of the choke(inductor) in any way?If so in what way?(choke formation is from E-I stampings) [At this moment it is working I just want to know of the long term effects]
And for similar reasons I cannot divulge the entire schematic however if you have any further enquiries I would be happy to help you with the specifics
I do not consider a temperature rise of 60 degrees to be excessive and it is a perfectly reasonable design parameter. Operate the insulating materials within their safe operating temperature limits and the temperature should not be a problem.
✨ The discussion addresses the impact of a 60°C temperature rise above ambient on an iron core inductor used in a high-current (650A), low-voltage (27V) IGBT-based chopper circuit operating at approximately 9.678 kHz. The core material is CRGO (Cold Rolled Grain Oriented) silicon steel, typically designed for 50 Hz operation. Several contributors debate the suitability of this core at high switching frequencies, noting that while the core's Curie temperature (~730°C) is not a limiting factor, eddy current losses and core heating at kHz frequencies can cause significant performance degradation and potential failure due to overheating and insulation breakdown. It is emphasized that operating the core and insulating materials within their thermal limits is critical. Practical experience with PWM supplies indicates that using a core not rated for high frequency can lead to rapid overheating, shorted turns, and circuit failure. The consensus suggests that although a 60°C rise is not inherently damaging if within design limits, the mismatch between core material frequency rating and operating frequency can negatively affect long-term reliability. The power supply is intended for military use, and full schematics are undisclosed. References to transformer insulation classes (Class H) and core construction (E-I stampings) are made, highlighting the importance of core material properties and frequency compatibility in high-frequency chopper circuits.
TL;DR: At 9.678 kHz and ~17.6 kW, a +60°C inductor core rise is acceptable if insulation and flux are within limits; “A 60 deg C rise above ambient will not harm the iron core.” [Elektroda, DAVID CUTHBERT, post #21669782]
Why it matters: Designers of IGBT choppers for high‑current DC loads need quick guidance on heat, core choice, and ripple control—without guessing.
This FAQ is for power‑electronics engineers optimizing inductors for low‑voltage, high‑current choppers.
Not by itself. If winding insulation stays within its temperature class and the core does not saturate, +60°C is fine. Verify flux density and copper loss, then confirm thermal steady state in your enclosure. “A 60 deg C rise above ambient will not harm the iron core.” [Elektroda, DAVID CUTHBERT, post #21669782]
Will CRGO laminations struggle at 9.678 kHz?
They can. Eddy-current and hysteresis losses increase with frequency, especially with thicker laminations. Use very thin laminations and validate flux density and loss. If losses are high, reduce ripple current, increase turns, or change material. “Iron losses will go through the roof at that frequency” reflects common field experience. [Elektroda, Mark Harrington, post #21669783]
What long-term effects can overheating cause in the choke?
Excess heat accelerates insulation aging, increases copper resistance, shifts inductance, and can lead to shorted turns. A known edge case: a mismatched laminated core at high frequency can overheat within minutes, damaging windings and semiconductors. Plan for continuous worst‑case duty, not bench bursts. [Elektroda, Mark Harrington, post #21669786]
How do I check flux density (to avoid saturation) quickly?
Use Bmax = Vrms/(4.44·f·N·A). 1) Estimate Vrms across the choke at ripple. 2) Insert switching frequency, turns, and core area. 3) Keep Bmax below your core’s safe region and iterate N or A if needed. [Elektroda, Mark Harrington, post #21669783]
What is CRGO steel?
CRGO is cold‑rolled grain‑oriented silicon steel used for low‑frequency transformers and chokes. It offers low core loss near 50–60 Hz due to grain orientation and lamination. At higher kHz, verify lamination thickness and loss before committing. [Elektroda, Tridib Baidya, post #21669784]
What is an IGBT chopper in this context?
An IGBT chopper rapidly switches the DC link to control average output voltage and ripple. Here it regulates 27 V at 650 A with ≤0.5% ripple at 9.678 kHz, demanding a low‑loss, thermally managed inductor. [Elektroda, Tridib Baidya, post #21669781]
Does Curie temperature limit this inductor?
Not here. For silicon‑steel, Curie temperature is far above operating temperatures. Insulation systems and core/winding losses typically limit you first. Focus on keeping windings within their class and validating loss at frequency. [Elektroda, DAVID CUTHBERT, post #21669789]
What ripple target makes sense for plating or defense loads?
Low‑voltage, high‑current supplies often target ≤1% ripple. Your stated 0.5% aligns with such sensitive loads, including suspected plating applications. Lower ripple reduces heating and improves process stability. [Elektroda, DAVID CUTHBERT, post #21669791]
Could a C‑core with thin laminations work at ~10 kHz?
Yes, with very thin (e.g., ~1 mil) silicon‑steel laminations and correct sizing, a C‑core can be viable at ~10 kHz. Validate temperature rise and loss under full load. [Elektroda, DAVID CUTHBERT, post #21669789]
What quick tests should I run before finalizing the design?
Do a full‑load thermal soak to steady state, measure ripple and inductance versus temperature, and confirm insulation margins. If within limits, temperature is not the problem; loss is. Document airflow and enclosure conditions. [Elektroda, DAVID CUTHBERT, post #21669795]
What does ≤0.5% ripple mean in practice?
Ripple voltage is the AC variation superimposed on the DC output. At 27 V, 0.5% means ≤135 mV peak‑to‑peak at the load. Ensure the choke and output capacitors meet this across duty and temperature. [Elektroda, Tridib Baidya, post #21669781]
How much power is this supply handling and why does it matter?
It delivers about 17.6 kW (27 V × 650 A). That power level stresses the inductor’s copper, core loss, and cooling. Design cross‑section, turns, and airflow for continuous operation at this load. [Elektroda, Tridib Baidya, post #21669781]
Does core geometry (E‑I vs toroid) change the outcome?
Yes. E‑I stacks from 50 Hz service at kHz frequencies can incur higher losses due to joints and lamination thickness. Validate with your E‑I choke build and adjust turns or material if losses are excessive. [Elektroda, Tridib Baidya, post #21669793]
Are ferrite cores an option here?
Ferrite is common at kHz, but large frames may be unavailable in your context. The project uses CRGO due to supply constraints; design accordingly and verify losses at frequency. [Elektroda, Tridib Baidya, post #21669784]
