Ferrite Core Transformer question

Thread Starter

bokbunja ju

Joined Sep 12, 2015
Hey guys, simply why inductors constructed using Ferrite cores usually include an air-gap ?

I know air gaps help increase the total reluctance of the core, hence the flux of the core. But the questions seems to be asking something else, i might not be aware about.

Thanks guys


Joined Jun 17, 2014

The air gap helps with the DC current operation of the core because it changes the overall characteristics of the construction. The often present offset DC current causes the operating point to move up on the BH curve which means the core could saturate more easily. By inserting a small air gap the total permeability goes down and thus it takes more peak current to saturate allowing the core to operate without danger of saturation.
The drawback is more turns are required to achieve a given target inductance value, so more turns have to be added. This in turn increases the potential to saturate, but not as much as it increases inductance so a happy medium is attained when the air gap is introduced and a certain number of turns are added to restore the original inductance requirement.
If you look at a typical BH curve, the air gap tends to spread the curve out more horizontally, which means it takes more current to get near the saturation flux density.

Even a very small gap can make a large difference in the permeability because the intrinsic permeability of a typical core material can be high like 1000 while that of air is only 1.
For example, a small core with 10 turns measured 50uH before any gaps where introduced. The core was then cracked into 4 pieces, then reassembled using super glue. The super glue is very thin, so the four resulting gaps were very narrow, yet the reconstructed core with 10 turns measured only 8.75uH. The ability of the inductor to handle DC current (as in a typical switching power supply) would be much better however, but more turns would have to be added to get back up to 50uH. It would still be better for handling DC current however.

The shorter answer is that the air gap reduces the permeability, and a core with lower permeability than another core where both have the same saturation level means the lower permeability core is harder to saturate. The lower permeability core however does require more turns of wire.


Joined Aug 21, 2008
A gap was sometimes chose or even adjusted in order to attain a particular inductance. This was important for filters in the telecommunications industry, they were eventually replaced in the 1960's and 1970's by solid state circuitry that did not rely on inductance for tuning.


Joined Mar 14, 2008
The use of a gap for inductors used for DC current energy storage, such as switch regulators, is related to the fact that the energy stored is directly proportional to the inductance but proportional to the square of the current (1/2 LI^2).
Thus it's a trade-off of more current versus less inductance.
For example if you half the inductance of core using a gap but can then double the current before saturation, the maximum inductor energy storage is increased by a factor of 2.

ian field

Joined Oct 27, 2012
Hey guys, simply why inductors constructed using Ferrite cores usually include an air-gap ?

I know air gaps help increase the total reluctance of the core, hence the flux of the core. But the questions seems to be asking something else, i might not be aware about.

Thanks guys
Ferrites saturate much easier than laminated iron cores - a gap reduces that tendency.

When the core saturates, the inductance vanishes and you're left with the DC resistance of the wire its wound with.


Joined May 23, 2013
Just a small point regarding the practical use of air gaps when using apparently identical cores. I ordered 10 same part number cores for a project from one of the U.K's main component suppliers, 5 were in stock, 5 back ordered. With the first batch, I finalised the design and and determined an air gap of 0.1-0.15mm was correct for the purpose.
Two weeks later, the second batch arrived and I noticed that they were a different colour, slightly more silvery than black. When I tried to use them, the results were totally different, and to avoid saturation, the gap had to be increased to 0.6-0.8mm. Their permeability was much greater despite being exactly the same part number and physically the same dimensions.
I built a simple saturation curve tracer and one "standard" coil to see the difference and subsequently used this to determine the gap for any further cores.
It turned out that they were from different manufacturers who both stated they were the same spec.


Joined Jun 17, 2014

Some good points coming up here, like about adjusting the inductance with the gap.
This was still used in the 80's for adjusting very large inductors for use with high power switching converters. Some of the inductors alone would weigh over 100 pounds each and there could be three or four in one converter. This idea might still be used today.

The inductance goes up with more wire turns, but the efficiency goes down because there are more wire turns assuming the same size wire. Increasing the wire diameter if possible would help that or using more than one insulated strand.

With a switching converter that does not blow out the switch transistor when the inductor saturates a little each cycle, the ripple current goes up dramatically so you can watch the effect as you increase load current (lower the resistive test load). With the introduction of an air gap (and more turns) the ripple at some output current goes down because the inductor does not saturate anymore.

An example was with the cracked core i was talking about previously. In a switching converter application and the non-cracked core with 10 turns of wire (for 50uH) the ripple would shoot up at around 500ma DC output current from about 50mvpp to 200mvpp, and going any higher than 500ma would cause severe core saturation. After cracking the core (and then gluing back together) and adding 20 more turns of wire (for a total of 30 turns to get back to the required 50uH inductance value) the ripple stayed under 50mvpp all the way up to 1.5 amps and that was the limit of the test as i could not go any higher due to the test source power supply current limitation. It may have been able to go even higher but all i needed at the time was 1.5 amps DC output current.