Ok, Im tired, but this thought occured to me...
Amplifier with mosfet outputs, if it takes less then 14V to fully turn on a mosfet, why does the input and driver stages need to operate from the same +/- voltages as the output devices.
Am i so tired i am missing something, why couldnt the driver stage run from say +/- 15-18V or something????
Zeeerooo Sleep now....
Amplifier with mosfet outputs, if it takes less then 14V to fully turn on a mosfet, why does the input and driver stages need to operate from the same +/- voltages as the output devices.
Am i so tired i am missing something, why couldnt the driver stage run from say +/- 15-18V or something????
Zeeerooo Sleep now....
Hi
the gate to source (GS) voltage sets the current in the drain to source (DS).
The DS current will be about zero at 3v GS an rises to a maximum at say GS of 20v. If the product of DS current and voltage exceeds the power capacity of the mosfet then bang.
The simple way to limit the DS current is to limit the GS voltage with a zener. Often you see circuits with 6v upto 15v zeners across GS. These are current limit protection and need a gate resistor to be effective. The zeners also prtotect the mosfet from exceeding the maximum GS voltage stated in the data sheet.
Now imagine the source voltage swinging up & down following the music signal. The gate voltage must also swing up in time with the source. In fact the process is the reverse. The gate swings up and the DS current forces the source voltage to follow in step. Now if the source has reached the the peak output voltage (Vpk) and is sending Ipk through the load then the gate must be sitting at Vpk plus Vgs above steady state voltage.
So it seems the drive rails need to be about 5 to 8 volts higher than the output rails to get best efficiency out of the power supply. That is why some designers choose to use two different voltages for their supply rails. A regulated low current supply for the voltage and driver side and a cheap & cheerfull high currrent supply for the output transistors is occasionally the chosen solution.
Hope this helps
regards Andrew T.
the gate to source (GS) voltage sets the current in the drain to source (DS).
The DS current will be about zero at 3v GS an rises to a maximum at say GS of 20v. If the product of DS current and voltage exceeds the power capacity of the mosfet then bang.
The simple way to limit the DS current is to limit the GS voltage with a zener. Often you see circuits with 6v upto 15v zeners across GS. These are current limit protection and need a gate resistor to be effective. The zeners also prtotect the mosfet from exceeding the maximum GS voltage stated in the data sheet.
Now imagine the source voltage swinging up & down following the music signal. The gate voltage must also swing up in time with the source. In fact the process is the reverse. The gate swings up and the DS current forces the source voltage to follow in step. Now if the source has reached the the peak output voltage (Vpk) and is sending Ipk through the load then the gate must be sitting at Vpk plus Vgs above steady state voltage.
So it seems the drive rails need to be about 5 to 8 volts higher than the output rails to get best efficiency out of the power supply. That is why some designers choose to use two different voltages for their supply rails. A regulated low current supply for the voltage and driver side and a cheap & cheerfull high currrent supply for the output transistors is occasionally the chosen solution.
Hope this helps
regards Andrew T.
"Amplifier with mosfet outputs, if it takes less then 14V to fully turn on a mosfet, why does the input and driver stages need to operate from the same +/- voltages as the output devices.
Am i so tired i am missing something, why couldnt the driver stage run from say +/- 15-18V or something????"
The driver stage must provide the voltage you want at the speaker PLUS the extra 10V or so needed to turn the MOSFET on.
A 100W Bryston 3B ran on ±52V with bipolar outputs, where as the 100W Hafler DH200 ran on ±63V with MOSFETs.
Hafler TransNova, QSC, many Altec, EV, and Crown power amps use a grounded output stage with the speaker hooked to the transformer center tap. With this design all you need is ±15V or so on your front end. You need one transformer per channel.
http://k-amps.8m.com/cgi-bin/i/PowerAmps/Semicond/hafler_915c.jpg
Am i so tired i am missing something, why couldnt the driver stage run from say +/- 15-18V or something????"
The driver stage must provide the voltage you want at the speaker PLUS the extra 10V or so needed to turn the MOSFET on.
A 100W Bryston 3B ran on ±52V with bipolar outputs, where as the 100W Hafler DH200 ran on ±63V with MOSFETs.
Hafler TransNova, QSC, many Altec, EV, and Crown power amps use a grounded output stage with the speaker hooked to the transformer center tap. With this design all you need is ±15V or so on your front end. You need one transformer per channel.
http://k-amps.8m.com/cgi-bin/i/PowerAmps/Semicond/hafler_915c.jpg
Is the glass half empty, or is it half full?
There is no difference to the speaker if it is connected to the outputs (hot) on one end and ground on the other end (along with the center tap) vs ground on one end (along with the outputs) and the other end hooked to the center tap (hot)
To further confuse the issue, consider suspended supply operation.
It looks like a 'normal' amplifier with the outputs (hot) connected to the speaker, and the other end of the speaker returned to ground (along with the center tap). What makes this different is the front end runs off of a ±15V supply with its ground reference returned to the speaker hot.
There is no difference to the speaker if it is connected to the outputs (hot) on one end and ground on the other end (along with the center tap) vs ground on one end (along with the outputs) and the other end hooked to the center tap (hot)
To further confuse the issue, consider suspended supply operation.
It looks like a 'normal' amplifier with the outputs (hot) connected to the speaker, and the other end of the speaker returned to ground (along with the center tap). What makes this different is the front end runs off of a ±15V supply with its ground reference returned to the speaker hot.
For grounded emitter amps like the QSC, forget the concept of "ground" for a moment.
If you took a battery and a light bulb and a pot all wired in series, you could turn the pot and vary the bulb current and thus its brightness. You can fo this without grounding anything, or you could connect any one point in it to a ground. Point is that there is a circuit for the current. it is a loop with all three elements in series.
The output circuitry of the amp is equivalent to a big pot controlling current, the power supplies are equivalent to batteries. The speaker is the bulb.
We usually ground the point where the speaker and power supply meet. But the grounded emitter amp just moves the ground point to the other end of the speaker, where it meets the emitters.
I find it is more intuitive if you visualize the speaker, power supply, and transistors as one big series loop. Connect any point in it to ground. If there were no ground, you would have the speaker connected between the emitters and the center tap. In the conventional arrangement, the speaker connects to ground, and if you think of it, the chassis is just a big wire over to the center tap.
Crown and others call it a "flying rail" design, and if you scope the supply rails with respect to ground, they are indeed moving with the signal. If you have +/-80VDC, there will be a constant 160V between them, but the center point will be moving with the signal. SO in that sense, yes it modulates the power rails.
If you took a battery and a light bulb and a pot all wired in series, you could turn the pot and vary the bulb current and thus its brightness. You can fo this without grounding anything, or you could connect any one point in it to a ground. Point is that there is a circuit for the current. it is a loop with all three elements in series.
The output circuitry of the amp is equivalent to a big pot controlling current, the power supplies are equivalent to batteries. The speaker is the bulb.
We usually ground the point where the speaker and power supply meet. But the grounded emitter amp just moves the ground point to the other end of the speaker, where it meets the emitters.
I find it is more intuitive if you visualize the speaker, power supply, and transistors as one big series loop. Connect any point in it to ground. If there were no ground, you would have the speaker connected between the emitters and the center tap. In the conventional arrangement, the speaker connects to ground, and if you think of it, the chassis is just a big wire over to the center tap.
Crown and others call it a "flying rail" design, and if you scope the supply rails with respect to ground, they are indeed moving with the signal. If you have +/-80VDC, there will be a constant 160V between them, but the center point will be moving with the signal. SO in that sense, yes it modulates the power rails.
Some MOSFET amps have two power rails. I used one with +/- 50v for the output stage, and +/- 60v for the driver and input stages. In this way, the gate voltage can exceed the drain voltage, and turn the fet on harder, yet dissipation in the output stage is reduced compared to running the output stage at +/- 60v. It increases the efficiency and reliability of the amp. The mains transformer had five taps on the secondary to supply the necessary voltages.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.