If I bias a pair of complementary mosfet outputs up to where I get satisfactory xover characteristics (but they run too hot)....Say 100ma. :
Then double the devices (2 pairs) and either bias them down or add source resistors to get the current down to 50ma (per pair) then can I expect the same xover characteristics ?
Then double the devices (2 pairs) and either bias them down or add source resistors to get the current down to 50ma (per pair) then can I expect the same xover characteristics ?
First order (thumb in wind) approximation:
yes.
But truly, Mosfet like a lot more current than that. If you're
intent on low bias, I suggest Bipolar devices.
yes.
But truly, Mosfet like a lot more current than that. If you're
intent on low bias, I suggest Bipolar devices.
"If I bias a pair of complementary mosfet outputs up to where I get satisfactory xover characteristics (but they run too hot)....Say 100ma. :"
If you had a ±50V supply for a 100W amplifier then 100mA bias would be only 5W per power supply rail.
If this is too hot you have no heatsink.
Look at some ready made amplifiers (not receivers, they try and save money in this area) to get an idea of the amount of heatsink you will need.
If you had a ±50V supply for a 100W amplifier then 100mA bias would be only 5W per power supply rail.
If this is too hot you have no heatsink.
Look at some ready made amplifiers (not receivers, they try and save money in this area) to get an idea of the amount of heatsink you will need.
>But truly, Mosfet like a lot more current than that. If you're
>intent on low bias, I suggest Bipolar devices.
These are Hitachi lateral devices, and they "seem" OK with fairly low bias. (observing xover notch w/ 10kHz squarewave). The strange thing (perhaps you can shed some light?) is that unlike bipolar devices (xover notch is worst at lowest levels) the notch seems to manifest as the amplitude of the signals goes up. Worst case right before clipping? Thank You .......... mike
>intent on low bias, I suggest Bipolar devices.
These are Hitachi lateral devices, and they "seem" OK with fairly low bias. (observing xover notch w/ 10kHz squarewave). The strange thing (perhaps you can shed some light?) is that unlike bipolar devices (xover notch is worst at lowest levels) the notch seems to manifest as the amplitude of the signals goes up. Worst case right before clipping? Thank You .......... mike
>If this is too hot you have no heatsink.
I'm trying to utilize "board level" heatsinks. So, yes a little skimpy in that area. Hence my original question to see if an array of smaller heatsinks (and devices) can operate like a single more massive unit...........Thanks............mike
I'm trying to utilize "board level" heatsinks. So, yes a little skimpy in that area. Hence my original question to see if an array of smaller heatsinks (and devices) can operate like a single more massive unit...........Thanks............mike
In the case of Mosfets of a particular process, they tend to
act the same for given die area. Two devices of a given chip
size look a lot like a single die whose area twice as large.
Thus an IRF250 looks a lot like two IRF240's.
act the same for given die area. Two devices of a given chip
size look a lot like a single die whose area twice as large.
Thus an IRF250 looks a lot like two IRF240's.
- Status
- Not open for further replies.