... Does it need overvoltage protection? For example, due to disconnected load or overdriving? We're talking about a high DF class AB amplifier. A simple pair of back to back zeners on the primary for good measure?
That's frequently done for PA systems, to generate the 70V or 100V distribution line, and the only precaution is to use protection diodes from the output to each supply rail.
They protect the amplifier from kickbacks in case the amplifier is clipping
They protect the amplifier from kickbacks in case the amplifier is clipping
The protection diodes are needed, transformer or no transformer. Current clipping into any inductive load, including speakers, causes flyback.
Yep, that was a dumb question. Then two diodes should do. Another dumb question. Do chip amps such as TDA7293,4 and LM3886 have such protection integrated? I've never seen such diodes being addded at the output.
Fwiw I have made quit a a few Guitar amps with output autoformers, same thing.
I always added output to power rail diodes as Elvee suggests.
Guitar players love 4-8-16 ohm selectable outputs if at all possible, to mix and match any cabinets they have lying around.
That said, currently using Darlingtons (TIP14x) and switching MosFets (IRFP240/250) which already include them.
I always added output to power rail diodes as Elvee suggests.
Guitar players love 4-8-16 ohm selectable outputs if at all possible, to mix and match any cabinets they have lying around.
That said, currently using Darlingtons (TIP14x) and switching MosFets (IRFP240/250) which already include them.
Something else to keep an eye on: DC offset. For a speaker, 50mV is mostly harmless, but with an ungapped transformer having a very low primary resistance, it could be problematic
Quite correct. I'm currently trying to understand if chip amps, assuming there is a decoupling capacitor in the feedback loop, have some kind of residual offset. Or could it develop dynamically? Or a servo network should be preferred over the capacitor?
Yes, it can be problematic.Something else to keep an eye on: DC offset. For a speaker, 50mV is mostly harmless, but with an ungapped transformer having a very low primary resistance, it could be problematic
Best is to momentarily disconnect amp from transformer, zero offset and then reconnect it.
Otherwise offset may appear smaller than it really is.
Servo is a good idea, if you have it, otherwise a balancing trimmer is fine.
With a capacitor in the ground side of the feedback network, the offset will become the root offset of the chip; for a tda7293 for example, it is +/-10mV max.Quite correct. I'm currently trying to understand if chip amps, assuming there is a decoupling capacitor in the feedback loop, have some kind of residual offset. Or could it develop dynamically? Or a servo network should be preferred over the capacitor?
You have to make some calculations based on the transformer's resistance and its tolerance to DC current.
If that's unacceptable, a servo or simply an offset trimmer can be the solution. Using an output capacitor is also a possibility, but you have to be careful not to create unwanted resonances that could degenerate into ferro-magnetic resonance
- Home
- Amplifiers
- Solid State
- If an SS amplifier feeds an output transformer...