not sure the answer to that if it really bothers the customer try either changing the dimmer or changing the circuit breaker whichever is easier or cheaper
Assuming the dimmer is compatible with the load type, the problem is caused by a combination of the brutal way that the switching is achieved in triacs.
More expensive phase-control dimmers have large inductors in series with the triac (or SCR), often in conjunction with a capacitor across the load. The following illustrations show the difference - although bear in mind these are very basic, to illustrate the point, and may not include all components required for a real design):
For a dimmer built into a unit intended to be fitted in place of a light switch, it's not really possible to calculate what you can get away with for the L-C filter (because that's what it really is), and the impact on the dimmer switch if the load goes open-circuit, because they are often potted and you can't see what components are used in the product.
Realistically, you might want to try:
Swapping out the dimmer, or trying another manufacturer's product
I had a similar problem with a client who chose mismatched lighting with the control gear.
All to do with the difference between leading edge and trailling edge mains current switching in the dimmer circuitry. affecting the LED control circuits or drivers.
Leading edge causes unreliable results and lots of noise (hum) althought it may work with some drivers. Trailing edge control behaves more consistent with the use of LED lighting.
Note. some dimmers will require a minimum load to work. that means more LEDs but then you'll have large start up currents.
I suggest you seek advice from the appropriate manufacturers.
As an aside has anyone else tried to dim the Osram (LEDvance) 'bulbs' that are supposed to dim and found they turn themselves off after about an hour, but then recover if you let them cool down ?
Of course one needs to understand that trailing edge dimmers don't use triacs. Triacs are rather old technology, and not as cheap as IGBTs which are used for trailing edge dimmers. The leading edge ones tend to cause noises because they switch on (large current) at high voltages. Trailing edge switch off ( to zero current) at similarly high voltages. Less noise from magnetics, but potentially high spike voltages instead.
Agreed - IGBT trailing edge, or even PWM, are often far better than leading-edge phase control in terms of not causing "buzzing" or "humming" noises.
I guess very soon, "smart" LED drivers (i.e. those with in-built power control capability) will be the order of the day, and we will forget we ever had these kinds of issues. Now, who remembers rheostat and leakage-reactance dimmers ?
