Hi !
This circuit might produce smoke, you use a bc548c in the vas,
it has a max Vce of 30volt, you feed it with up to 80volts, and
Q6 will dissipate about 0.4Watts, quite much for a TO92 !
This gives ~105°C, hot enough to boil water !
You might consider a bd139 here... (With small heatsink maybe)
And i think c3 is way too big, 220pF here might cause oscillation.
A typical value here is 22pF.
Mike
This circuit might produce smoke, you use a bc548c in the vas,
it has a max Vce of 30volt, you feed it with up to 80volts, and
Q6 will dissipate about 0.4Watts, quite much for a TO92 !
This gives ~105°C, hot enough to boil water !
You might consider a bd139 here... (With small heatsink maybe)
And i think c3 is way too big, 220pF here might cause oscillation.
A typical value here is 22pF.
Mike
you need to fix bias, 1/1.5 ma in diff pair at 0 V out is very wrong
R17 could be adjusted to balance the input Q currents
better would be to current mirror load the diff pair
also integrated darlingtons usually suck, discretes let you control driver bias better, but then you're getting to another amp altogether
Linear.com has the SwCad free spice simulator, no component/net limits, fast, solid - somewhat less sophisticated analysis tools but I haven't touched my OrCad Demo since trying LtSpice (even downloads faster and takes lots less room than OrCad)
R17 could be adjusted to balance the input Q currents
better would be to current mirror load the diff pair
also integrated darlingtons usually suck, discretes let you control driver bias better, but then you're getting to another amp altogether
Linear.com has the SwCad free spice simulator, no component/net limits, fast, solid - somewhat less sophisticated analysis tools but I haven't touched my OrCad Demo since trying LtSpice (even downloads faster and takes lots less room than OrCad)
...or the bias problem could be the other end, R18 could also be adjusted for diff pair bias symmety, but don't go too far and unbias the zener, the zener won't be a good regulator without several mA flowing through it
transient response is the only way to go for "debugging" - after you have the biasing correct
but it should be done at different scales with varying loads (esp. cable cap load)
really helpful for compensation is a small pulse riding on a slow lage amplitude sine to check operating point gain/comensation interaction
transient response is the only way to go for "debugging" - after you have the biasing correct
but it should be done at different scales with varying loads (esp. cable cap load)
really helpful for compensation is a small pulse riding on a slow lage amplitude sine to check operating point gain/comensation interaction
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.