I have been reading up on SSRs and optoisolated gate drives and one thing that have struck me is that most SSRs are slow like only turning off in about 0.1ms and on in about 1ms compared to Opto. gate drive + MOSFET which seams to a factor 10-100 faster. Why is this? The schematic seam to be the same in theory…
I’m definitely interested in the theory behind this as well, since some SSRs use optoisolators internally.
It depends a bit on how the SSR is implemented. It can be that the optocoupler output feeds a thyristor (for AC types) or to back-to-back coupled MOSFETs (for DC capable types).
From the days of studying semiconductor device physics 15 years ago I seem to recall that the thyristors are inherently slow because they need to fill up with injected carriers before they start to conduct, then the current is maintained through an internal generation of free carriers…and they will conduct for ever as long as the minimum current is maintained, even if you remove the gate bias/carrier injection. This is why they only turn off when the AC voltage crosses zero and the free carriers recombine or is carried away to the load/source. But well, nowadays they are fancier, and can turn off through reverse biasing, etc.
I have never actually used a DC capable SSR, so I won’t claim any knowledge on those, but in principle they should be much faster!
Ah I see. Intresting to know how the zero crossing was implemented. That would explain the zero crossing types but the one I have seen have been DC.
I have started to think that it might have with the capacitance on the gate. That would still make it slow even if the gate drive is fast.
Right now I’m leaning to use SSR ICs due to space constraints. That darn space would be better with 4D