500W PA amplifier with Limiter

bulb is very bright during first power
after the in rush current
bulb start to dimmed
with the power still on and the bulb dimmed, start checking that the circuits are operating correctly.
What is the voltage on the supply rails?
What is the voltage across the output, both DC and AC.
What are the currents flowing in the LTP pair, are they matched?
ETC........
When you are sure all is operating correctly and that nothing is overheating, then it's time to plug straight to mains power.
It will probably blow the mains fuse and maybe the supply rail fuses as well.
Go build a timed relay bypassed soft start circuit to save fuses.
 
with the power still on and the bulb dimmed, start checking that the circuits are operating correctly.
What is the voltage on the supply rails?
What is the voltage across the output, both DC and AC.
What are the currents flowing in the LTP pair, are they matched?
ETC........
When you are sure all is operating correctly and that nothing is overheating, then it's time to plug straight to mains power.
It will probably blow the mains fuse and maybe the supply rail fuses as well.
Go build a timed relay bypassed soft start circuit to save fuses.
I hope that some basic measurement was already done.
 
Last edited:
When you are sure all is operating correctly and that nothing is overheating, then it's time to plug straight to mains power.
It will probably blow the mains fuse and maybe the supply rail fuses as well.
Go build a timed relay bypassed soft start circuit to save fuses.

It takes more than one short plug in on the light bulb before I'm confident enough to throw the switch on it. Initially set the bias so the output stage is barely cracked open and make sure music plays. And that you get graceful dropping of the mains and graceful clipping as the bulb gets brighter. Then set the bias high (maybe 100mA in the output stage) and let it idle for an hour. make sure it doesn't run away. Then go back and re-check everything, re-set the bias to a milliamp or two - and then run it without the limiter.

And when you first fire it up, the first thing you do is check bias stability. Set it to target, and see where it goes after the first hour. If all is well, hook up a dummy load. Parallel the dummy load with a speaker in series with 1k ohms (so you can hear what's going on but not at ear splitting level). Then clip the snot out of it with test tones and music. This is a good time to measure power at 20 Hz, 20KHz if you desire, and hear what it sounds like when you drive it hard. If it takes *this*, then it's ready for serious listening tests.
 
It takes more than one short plug in on the light bulb before I'm confident enough to throw the switch on it. Initially set the bias so the output stage is barely cracked open and make sure music plays. And that you get graceful dropping of the mains and graceful clipping as the bulb gets brighter. Then set the bias high (maybe 100mA in the output stage) and let it idle for an hour. make sure it doesn't run away. Then go back and re-check everything, re-set the bias to a milliamp or two - and then run it without the limiter.

And when you first fire it up, the first thing you do is check bias stability. Set it to target, and see where it goes after the first hour. If all is well, hook up a dummy load. Parallel the dummy load with a speaker in series with 1k ohms (so you can hear what's going on but not at ear splitting level). Then clip the snot out of it with test tones and music. This is a good time to measure power at 20 Hz, 20KHz if you desire, and hear what it sounds like when you drive it hard. If it takes *this*, then it's ready for serious listening tests.
There is no bias adjust on this circuit.
 
It takes more than one short plug in on the light bulb before I'm confident enough to throw the switch on it. Initially set the bias so the output stage is barely cracked open and make sure music plays.
that's where we fundamentally disagree.
I will not plug in speakers to an untested amplifier.
I will not remove the input shorting link until I know that all the operating quiescent currents are as expected.
I will not set the quiescent output stage current (bias) with the output loaded nor with the input shorting link removed nor when the PSU voltage is below the normal operating voltage, i.e no bulb tester in the primary feed.
 
that's where we fundamentally disagree.
I will not plug in speakers to an untested amplifier.
I will not remove the input shorting link until I know that all the operating quiescent currents are as expected.
I will not set the quiescent output stage current (bias) with the output loaded nor with the input shorting link removed nor when the PSU voltage is below the normal operating voltage, i.e no bulb tester in the primary feed.


With a light bulb in place, you can't hurt the speaker - no matter what you did wrong. Even if the amp rails. On the first no-bulb test the speaker is hooked up through a resistor with the brunt of it going to a dummy load. Again, if there is smoke there's no speaker damage.

Making sure I have proper quiescent current control is a normal part of the initial checkout. With the limiter you will not be able to set it properly, but you should check for control. Then I'll back it off to right at class B before switching to a non-limiter test. You may prefer backing it off as far as the pot will go.