For flyback transformers/inductors. I am considering using several layers of pvc or polamide
What are the characteristic tradeoffs if all three ei transformer legs are gapped - instead of just the centre one? For rated inductance, current capability before saturation, efficiency?
There are so many combinations of transformer core size, and material composition. The ability to gap cores for prototypes as needed would be useful if feasible.
Asked another way - why do COTS transformers only gap the centre leg?
For permanent gap I do all three for better rigidity under clamp .Cut slightly small and fix one side with super glue to hold them in position. As most cores have lowest losses around
100 C, the gap material should be rated class B or better. I use kapton for small gaps and photo mount card for 1mm and up. As rel. permeabilty is practically unity, no magnetic effects.
Thanks. Google turns up lots of discussion about grinding. It's good to know that gapping all legs is also viable technique. It certainly looks easier.
For an EI with 3 gaps the total gap is 2 * thickness because of half flux in outers.
That is different from grinding the inner only, which causes the outers to have approx zero gap
Yes, that is fine.
Some things to keep in mind:
- Adhesive tape squashes down over time. Better to use rigid materials. Doesn't have to be fancy, even paper will get you there -- you're probably not going to need above Class A fireproof/temperature ratings for a hand-wound transformer. (Or if you do, well, choose materials accordingly!)
- The effective gap is slightly less than twice the geometric gap, because the cross-section spreads out in the vicinity of the gap (fringing effect). That is, the gap area is effectively wider, which is equivalent to it being thinner. So plan to measure the inductance and try a few sizes of shims to get it on target.
- The same effect also increases apparent saturation current: more leakage loops back around the windings themselves, or within a core half, thus reducing the field strength within the core itself. This isn't significant for small gaps
- Gap size is relative to the minimum aspect of the core. So, compared to the diameter of the center peg (for ER/ETD), for example; but the side legs are narrower, and gap size is relative to the narrower axis. So expect side gaps to count for less, for this reason.
- Fringing field also means external field. Roughly speaking, the gap acts like a wire loop of the same dimensions, carrying whatever amp-turns are going through the core. Nearby metal or wiring (again, nearness is relative to gap height) experiences substantial induction (eddy currents), which can lead to interference or heating. This causes increased copper loss for solid wire near the gap (litz is preferred over solid wire!), and can cause EMI problems as well.
Tim
In early days I did the Hanna, fringing calcs etc, but all that does not eliminate the need to get the gap by prototype test in situ. For ordinary DC or ac inductor I now assume core Ur is infinity and just use the elementary functions to get the turns and gap. Last one I did that method calculated for 2500 uH , est gap 1.5mm tested incremental 2000 uH at zero DC. I haven't tested at full current yet.
Swinging DC chokes need better initial calcs. I don't know if they are used much thses days.
Hello:
After decades of design and mfg, in transformers:
Gapping: Needs diamond tool and precision lap machine. ELNA Ferrite labs and other do this custom. Never hear of a DIY gapping a ferrite.
Best is center gapped with specified Al by ferrite vendor.
We used UL rated "fishpaper" or think thermosetting, never any tape or other material that can change with age, heat of pressure.
Normal gaps can be a few mil>>30 mil.
Most of our designs 70..700W were forward, double forward, , or bridge and did not need a precise gap.
We used EER shapes not square center leg and the mfg bobbins
Kind Regards,
Jon