paskal
New member
using the original component values and 820pf cap i got around 170KHz switching frequency. measured using a scope, the TL074 non-ucd have quite a bad square signal being feeded to IR2110 logic.Basic non ucd schematich... +-50V, IRFP250N,gate resistor 33ohm, 20uH, 1uH
Get hot after 10 minutas heatsink is small an with fan 3x3 cm heatsink is cold.
Output coil is hot (T106-2)
In feedback wih 2.2nF coil is cold but amp distord, probably slow fets, with 4.7n, coil is warm but distorsion is little bit smaller, and 6.8n amp is without distorsion but coil is hot, about 60 deegress C.
Thats why I asked what freq is with 820p, what is with 6.8n and what is with 10n!? Stewin, can you answer on it..
Because its full range sound, probably freq is above 120KHz. Am I right?
on the other hand, i've been testing DETEX IRS900D and it seem to be better than the tl074 non-ucd. i even managed to get to switch at 800khz stable. no excessive heat at any component. well only the 12V regulator have the most heat output.
using the original component value the irs900d have a switching frequency at 270-280khz. further testing are needed but i think it sounds better when switched at 480-500khz at there's less heat from the mosfet at that frequency.
will post more result and signal pictures tomorrow.