The concern is thermal run away. As bipolar transistors heat, up they tend to want to conduct harder which makes them hotter, which makes them conduct harder...
The emitter resistors force the outputs to share the load equally. That's why P/L manual calls out a load sharing test. The bias transistor has 2 functions. The first function is to set the bias. The second function takes advantage of the bipolar bias transistors' desire to conduct harder when it heats up. When the outputs heat up and the heatsinks therefore heats up, the bias transistor does it's thing (conducts harder) and reduces the overall bias.
On the other hand, if the bias transistor opens, due to a broken transistor lead, or broken connection to the driver board, the outputs turn full on and the effect is they go to war with the power supply. The power supply wins, the outputs fail. The fuses are of no help as they are too slow and the fuses only prevent a bigger fire.
I had this happen on a 700B about 35 years ago. I had to eat 12 new 2SD555's. You don't forget that lesson soon.
Joe's technique of soldering the bias transistor directly to the backplane board eliminates the stranded lead wire in the factory configuration. As he points out in the instructions, a good thermal contact to the chassis is critical. I have seen several photos where the case of the bias transistor rotated when the mounting nut was tightened down, and added some stress to the soldered leads. This opens up the possibility of a broken transistor lead, if not today, down the road.
The emitter resistors force the outputs to share the load equally. That's why P/L manual calls out a load sharing test. The bias transistor has 2 functions. The first function is to set the bias. The second function takes advantage of the bipolar bias transistors' desire to conduct harder when it heats up. When the outputs heat up and the heatsinks therefore heats up, the bias transistor does it's thing (conducts harder) and reduces the overall bias.
On the other hand, if the bias transistor opens, due to a broken transistor lead, or broken connection to the driver board, the outputs turn full on and the effect is they go to war with the power supply. The power supply wins, the outputs fail. The fuses are of no help as they are too slow and the fuses only prevent a bigger fire.
I had this happen on a 700B about 35 years ago. I had to eat 12 new 2SD555's. You don't forget that lesson soon.
Joe's technique of soldering the bias transistor directly to the backplane board eliminates the stranded lead wire in the factory configuration. As he points out in the instructions, a good thermal contact to the chassis is critical. I have seen several photos where the case of the bias transistor rotated when the mounting nut was tightened down, and added some stress to the soldered leads. This opens up the possibility of a broken transistor lead, if not today, down the road.
