I've been reading with interest the various posts on the output stage of the 3020 and its relatives, especially those concerning using emitter resistors with new output devices to prevent thermal runaway.
I own a 3020B and bought a 3225PE for my daughter. Earlier this year after recapping the 3225PE, before giving to my daughter, I shorted the outputs and of course destroyed one channels' output devices and related driver transistors.
I bought the replacement parts and had no problems getting it working again, but during Thanksgiving she brought it back to me with THE SAME channel blown. The other channel still has the original output devices dated 1989.
The only thing I can figure is I didn't do my research here and learn that I needed the emitter resistors with new output devices. So having learned my lesson I've ordered everything I need and will add the parts to the underside of the board.
But a couple of questions: After adding the resistors to the emitters (and probably adding the 4.7 ohm base resistors too), do I bias the amp the same? That is, remove the solder jumper and adjust for 28 to 30 mV across the 1 ohm resistor in the collector leg? How will the emitter resistors affect the bias, crossover points, etc.?
And what about that .33 ohm 3-watt resistor's ability to dissipate heat on the underside of the board? A larger wattage is probably too thick to fit. Is 3 watt enough (considering mounting position) or do you guys have any other suggestions?
Thanks!
I own a 3020B and bought a 3225PE for my daughter. Earlier this year after recapping the 3225PE, before giving to my daughter, I shorted the outputs and of course destroyed one channels' output devices and related driver transistors.
I bought the replacement parts and had no problems getting it working again, but during Thanksgiving she brought it back to me with THE SAME channel blown. The other channel still has the original output devices dated 1989.
The only thing I can figure is I didn't do my research here and learn that I needed the emitter resistors with new output devices. So having learned my lesson I've ordered everything I need and will add the parts to the underside of the board.
But a couple of questions: After adding the resistors to the emitters (and probably adding the 4.7 ohm base resistors too), do I bias the amp the same? That is, remove the solder jumper and adjust for 28 to 30 mV across the 1 ohm resistor in the collector leg? How will the emitter resistors affect the bias, crossover points, etc.?
And what about that .33 ohm 3-watt resistor's ability to dissipate heat on the underside of the board? A larger wattage is probably too thick to fit. Is 3 watt enough (considering mounting position) or do you guys have any other suggestions?
Thanks!
It should be fine for the power output from this amplifier. There will not be much heat dissipated in the transistor.
You can bias the amp by the same method, but a more effective way to check the bias current is simply to measure across one of the emitter resistors. 28-30mV across 1 ohms translates to 28-30mA of current using V=IR. You can do the same sort of calculation for 0.33 ohms.
The story for why you need the emitter resistors is this - when the 3020 was first made, the output transistors used a construction known as hometaxial. With this method, the resistors weren't necessary for thermal stability. However, with modern devices, they are produced using a construction called epitaxial. This is more efficient, but requires some local feedback (the emitter resistors) to be thermally stable.
You can bias the amp by the same method, but a more effective way to check the bias current is simply to measure across one of the emitter resistors. 28-30mV across 1 ohms translates to 28-30mA of current using V=IR. You can do the same sort of calculation for 0.33 ohms.
The story for why you need the emitter resistors is this - when the 3020 was first made, the output transistors used a construction known as hometaxial. With this method, the resistors weren't necessary for thermal stability. However, with modern devices, they are produced using a construction called epitaxial. This is more efficient, but requires some local feedback (the emitter resistors) to be thermally stable.
Most texts that discuss hometaxial types are either >30 years old or draw on same and probably even older material. I guess there will be a few obscure devices still produced for special purposes, such as high-reliability or research applications but as far as the interweb knows, the hometaxial manufacturing process for discrete devices died out in the early 1980s. RCA had already changed much of their power range by 1976.
NAD seemed to source their 2N2055s from a number of unlikely manufacturers - even Matsushita! Perhaps this arose because of the need to continue with hometaxial types but I can't say.
This thread also makes interesting general reading and gives a few leads: http://www.diyaudio.com/forums/soli...dels-quasi-complementary-power-output-17.html
NAD seemed to source their 2N2055s from a number of unlikely manufacturers - even Matsushita! Perhaps this arose because of the need to continue with hometaxial types but I can't say.
This thread also makes interesting general reading and gives a few leads: http://www.diyaudio.com/forums/soli...dels-quasi-complementary-power-output-17.html
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.