I am testing this amp with 10R in parrallel with 1uF. I am getting ringing which is considered to be a sign of instabilitly from what i understand. I have used a miller cap but the value needed to be about 220p before im just left with the initial overshoot and no ringing, but then the bandwidth is small. What can i try instead of miller caps. I also noticed the ringing amplitude varys depending on where i attach the scope probes 0V lead...
Attachments
1uF would be considered an extreme capacitive load. In real life, a speaker with xover network will never be anywhere near that capacitive. The ringing seems well damped, and the amplifier is not getting stuck in oscillation as a result, so it's safe to say it's stable.
You should still have a Miller cap but 100p Max should be plenty.
You should also have a Zobel network, 10 ohms + 100nF is the usual combination.
You should still have a Miller cap but 100p Max should be plenty.
You should also have a Zobel network, 10 ohms + 100nF is the usual combination.
overlapping post...
1 uF without any series R is just silly - was popular but doesn't really model any realistic load - even the ESL loudspeaker it is supposed to model
output series L, Zobel make amp output much more load insensitive without costing audio frequency loop gain
but ringing at that expected low frequency (you really should tell us) changing with 'scope gnd location is not a good indication for bypassing, gnd, ps layout/implementation quality
1 uF without any series R is just silly - was popular but doesn't really model any realistic load - even the ESL loudspeaker it is supposed to model
output series L, Zobel make amp output much more load insensitive without costing audio frequency loop gain
but ringing at that expected low frequency (you really should tell us) changing with 'scope gnd location is not a good indication for bypassing, gnd, ps layout/implementation quality
Member
Joined 2009
Paid Member
I believe even John Curl has conceded that 1 uH or less output series L is probably inaudible if not using ferromagnetic core, kept away from magnetic hardware
I don't understand exotic Japanese low Ccb VAS Q paired with MJE350 - I think both Q should be of similar "quality" - when 3 different data sheets don't have ft or C values I get suspicious...
and the bias is way too low for good ft, or ability to drive the MOSFET parasitic C at higher frequency
I don't understand exotic Japanese low Ccb VAS Q paired with MJE350 - I think both Q should be of similar "quality" - when 3 different data sheets don't have ft or C values I get suspicious...
and the bias is way too low for good ft, or ability to drive the MOSFET parasitic C at higher frequency
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.