The key words are "emitter ballast"
see Inside the RF Power Transistor
and http://www.google.com.my/url?sa=t&s...sg=AFQjCNE4yy7VfsjNCWlZsOI0_PqTMSuKrg&cad=rja
see Inside the RF Power Transistor
and http://www.google.com.my/url?sa=t&s...sg=AFQjCNE4yy7VfsjNCWlZsOI0_PqTMSuKrg&cad=rja
Fascinating! I also looked up 'ring-emitter transistors', the principle (i.e. using multiple, ballasted emitters) seems similar as in those RF power types.
jan didden
Jan,
there´s a inherent trade-off between power handling and bandwidth as power is inversely proportional to the square of frequency. Electron mobility in materials that can withstand large electric fields is slower, but the time limits are decisively set by the necessarily larger capacitances (larger charge storage required for operation).
By the way, why the sudden interest in the subject? I happened to read, by chance really, some of your previous statements de-emphasizing the importance of a wide operating frequency rage. Have you reassessed your stance?
there´s a inherent trade-off between power handling and bandwidth as power is inversely proportional to the square of frequency. Electron mobility in materials that can withstand large electric fields is slower, but the time limits are decisively set by the necessarily larger capacitances (larger charge storage required for operation).
By the way, why the sudden interest in the subject? I happened to read, by chance really, some of your previous statements de-emphasizing the importance of a wide operating frequency rage. Have you reassessed your stance?
Sorry I do not understand this
Simple: I had an event last weekend (ElektorLife!) where I was going to give a presentation about Doug Self's power amp consisting of 64 parallel opamps, as published in Elektor magazine. In the preparation for that I wondered why output devices seem to be necessarily slow.
But as far as I know, I never said operating freq range should not be wide. What makes you think so?
jan didden
Jan,
I think that power transistors (BJT) are not necessarily slow, but manufacturing processes for fast transistors are expensive, because of necessity of parallelling many small transistors structures, which are faster than big ones, because of smaller capacities and interconnection resistance. This is well known in CPU architecture, where very small transistors can permit to reach very high frequencies (if it is possible to sink large amount of heat in small devices... that's the big problem)
I think that power transistors (BJT) are not necessarily slow, but manufacturing processes for fast transistors are expensive, because of necessity of parallelling many small transistors structures, which are faster than big ones, because of smaller capacities and interconnection resistance. This is well known in CPU architecture, where very small transistors can permit to reach very high frequencies (if it is possible to sink large amount of heat in small devices... that's the big problem)
Great! I hate very long and winding posts; these are better saved for an article.
There's some info on my private website Jan Didden audio diy and other human frailties place . The "Arch Nemesis", Nelson Pass' variation/improvement on the original Nemesis (by Jean Hiraga, described in L'Audiophile), is described in my bookzine Linear Audio Volume 0 available from linearaudio.net.
jan didden
I made a new thread which is related, some of you might be interested.
http://www.diyaudio.com/forums/solid-state/177906-speed-bjts-fairchild-vs-nxp-bc550.html
- keantoken
http://www.diyaudio.com/forums/solid-state/177906-speed-bjts-fairchild-vs-nxp-bc550.html
- keantoken
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.