Member
Joined 2009
Paid Member
It may be sufficient that they are both at the same ambient temperature if the amp is being used for reasonable home listening. I prefer to have them in contact.
The issue is that the bias current through the drivers causes the drivers to warm up - simple heat dissipation. And when bipolar transistors heat up the bias current increases because of the physics of how they operate. The drivers control the bias current through the output power devices and they draw a lot of current. This increase in current causes more heating - and so it's a death spiral that ends in smoke!
The bias generator, Q9, is in a circuit that is designed to control the bias current through the drivers and it is designed so that an increase in temperature of Q9 produces a fall in bias current through the output. So by placing Q9 in contact with the drivers you create a thermal control over the bias current.
Last edited:
Simply, the output stage of P3a is a CFP (complementary feedback pair) design. With this type of power amplifier design, the temperature variation of the drivers provides the correct thermal reference for controlling bias, not that of the output transistors which usually results in thermal runaway or just an unstable bias level as power levels change.
The output transistors work fine as a temperature reference with more common and typical Emitter Follower output stages. That's why people confuse the reqirements for stable performance of CFP types. There are plenty of texts and web references to search on the 2 output stage design types.
The output transistors work fine as a temperature reference with more common and typical Emitter Follower output stages. That's why people confuse the reqirements for stable performance of CFP types. There are plenty of texts and web references to search on the 2 output stage design types.
Member
Joined 2009
Paid Member
you're in the wrong thread for DX my friend ! - for EF output you want around 25mV across the emitter resistor of each output device usually better to be slightly higher - aiming for 55mV across the 2 emitter resistors usually is good. Best wait for the amp to reach normal working temperature too and trim the dc-offset first.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.