Caps only tend to fail with too much ripple current or too much voltage (or reverse polarity, which isn't the problem).
In the car, does the rail voltage ever get near the rated voltage of the caps?
In the car, does the rail voltage ever get near the rated voltage of the caps?
The SLD5k amps that I have, all railcaps are ok. In some SPL car's they use lithium 16V, so railvoltage get's higher then normal.
Output filtercaps in class d section are blown, Samyong 100uF/200Vand red filmcaps blown..:.
Amp 1. flitercaps 10uF/250V film blown, replaced them with 10uF / 250V and changed 100uF / 200V Samyong to 100uF / 250V Nichicon. Working ok now.
Amp 2. filtercaps 100uF/200V blown. Replaced both 10uF/250v and 100uF / 200V with same as amp 1. After starting up in protect, power supply and all the mosfets are ok.
Output filtercaps in class d section are blown, Samyong 100uF/200Vand red filmcaps blown..:.
Amp 1. flitercaps 10uF/250V film blown, replaced them with 10uF / 250V and changed 100uF / 200V Samyong to 100uF / 250V Nichicon. Working ok now.
Amp 2. filtercaps 100uF/200V blown. Replaced both 10uF/250v and 100uF / 200V with same as amp 1. After starting up in protect, power supply and all the mosfets are ok.
No perry, I measured the rail both negative and positive, and even at 15v of battery, this never reaches (not even minimally) the maximum tolerable threshold of the capacitors (which is 200v).
The maximum I got (On the dyno) with 15v power was +/- 167v.
Unless there is a phenomenon of "BUS PUMPING" which would tend to increase one or two rails, the only reason for the failure of the caps remains too much RIPPLE to dispose of (which would explain why it happens with the car in function and at zero volume, in fact, I imagine that if the amplifier is used and stressed even minimally, the ripple is consumed and disappears from the caps, saving them).
But Bus pumping is something that occurs and is accentuated during operation, and the harder the operation (loud or very loud) the more accentuated, this is not the case.
The weird thing is that last time I also monitored the rails with my scope, and apart from some (normal) wobble, I didn't seem to see any ripple. I set a very large time base (to make sure I catch the ripple on a large scale) but I didn't see anything.
I was thinking, if it were possible, interrupt the connection between the diodes and the caps and interpose an LC filter of adequate size (as is done in some more refined amplifiers) and see what happens.
Last edited:
The failure seemed a bit random. When the amp is driven hard in the vehicle, do some of the rail caps heat up and others stay cool?
Is the supply in this amp regulated or not?
Is the supply in this amp regulated or not?