which one design has more efficient with regards to power dissipation.
a class AB with mosfet as step drive?
or class AB with BJT as step drive?
a class AB with mosfet as step drive?
or class AB with BJT as step drive?
The loss through a BJT can be lower than the loss through a mosFET.
The mosFET has a hard set Rds-on that cannot be lowered.
On that basis I see BJT as being more efficient than mosFET.
The mosFET has a hard set Rds-on that cannot be lowered.
On that basis I see BJT as being more efficient than mosFET.
If You want to use switching rail class H, use MOSFET, as it's easy to use, and simple to drive.
Iy You want to use modulated rail class H, use BJT, as it gives better linearity.
Sajti
Iy You want to use modulated rail class H, use BJT, as it gives better linearity.
Sajti
With BJT, power dissipated = I * Vce(sat) = I * 0.2
With MOSFET, power dissipated = I^2 * Rds
So MOSFET has higher losses when Rds*I > 0.2
Let me pick a random PSU... say the first PSU currently on Jonnyguru's PSU review list - Corsair SF450 450W, using FCP104N60F for switching, with an Rds of 91mohm.
With these numbers, the MOSFET will be less efficient than BJT when I > 2A
And MOSFET can be paralleled to reduce that Rds without requiring additional 0.1ohm or 0.22ohm resistors in series, which oops is higher than the typical Rds.
Typical BJT have Vce sat of 0.05V to 0.1V
If one takes your 2A as a starter, then your Rds on for a saturated mosFET has to be 25mohm to 50mohm to match that BJT loss.
There are many very low Rds on mosFETs, are any suitable for the duty that Clovi requires?
If one takes your 2A as a starter, then your Rds on for a saturated mosFET has to be 25mohm to 50mohm to match that BJT loss.
There are many very low Rds on mosFETs, are any suitable for the duty that Clovi requires?
Too many for me to choose... problem of being a DIYer - the day job isn't about squeezing pennies while the DIY parts count is too low to justify squeezing pennies
Higher end PSUs use parts in the 20mohm ~ 5mohm range (Corsair Vengeance 650M 650W - 0906NS - 4.5mohm) so I'll pick from that range,
For TO-220, maybe STP60NF06 for USD$1.50?
Or IRLB8743PBF at the same price with 2.5mohm Rds, but lower Vds(max) of 30V.
One more thing to consider tho, and I'll have to borrow Elliott Sound Product's schematic to explain:
Unless there is a high voltage low current tap dedicated for gate driver (or base driver), the highest voltage the gate or base can go will only be the same as the incoming Vcc. And often worse, because of control circuits and stuff.
In the best case scenario, Vds = Vgs, or Vce = Vbe, essentially turning them into diode-connected transistors. BJT's forward voltage is around 0.7V, but MOSFET's Vth is usually 1.5V or more, and it doesn't get lowest Rds until the transistor is fully turned on at maybe Vgs = 5V or 10V.
And as can be seen in the above schematic, the more voltage drops in the preceding circuit, the less voltage goes to gate or base and the higher the Vds or Vce becomes. Ultimately there is going to be inevitable voltage drop across the transistor regardless of BJT or MOSFET, the efficiency difference might be minor, but BJT would win.
Higher end PSUs use parts in the 20mohm ~ 5mohm range (Corsair Vengeance 650M 650W - 0906NS - 4.5mohm) so I'll pick from that range,
For TO-220, maybe STP60NF06 for USD$1.50?
Or IRLB8743PBF at the same price with 2.5mohm Rds, but lower Vds(max) of 30V.
One more thing to consider tho, and I'll have to borrow Elliott Sound Product's schematic to explain:
An externally hosted image should be here but it was not working when we last tested it.
Unless there is a high voltage low current tap dedicated for gate driver (or base driver), the highest voltage the gate or base can go will only be the same as the incoming Vcc. And often worse, because of control circuits and stuff.
In the best case scenario, Vds = Vgs, or Vce = Vbe, essentially turning them into diode-connected transistors. BJT's forward voltage is around 0.7V, but MOSFET's Vth is usually 1.5V or more, and it doesn't get lowest Rds until the transistor is fully turned on at maybe Vgs = 5V or 10V.
And as can be seen in the above schematic, the more voltage drops in the preceding circuit, the less voltage goes to gate or base and the higher the Vds or Vce becomes. Ultimately there is going to be inevitable voltage drop across the transistor regardless of BJT or MOSFET, the efficiency difference might be minor, but BJT would win.
Last edited:
sorry if someone is misleading..what i meant is for switch rails only,not a smps or psu.
so when a signal reaches close to a first Vrail,the next rail pull up so it won't saturate. the thing is if i use mosfet or bjt as switching, which would dissipate less power?..
i am concern with this since it will be use for mobile application in a remote area where no electricity is present. so power source is much limited
this exclude a class D amp ofcourse since its hard to DIY compare to class AB
so when a signal reaches close to a first Vrail,the next rail pull up so it won't saturate. the thing is if i use mosfet or bjt as switching, which would dissipate less power?..
i am concern with this since it will be use for mobile application in a remote area where no electricity is present. so power source is much limited
this exclude a class D amp ofcourse since its hard to DIY compare to class AB
Attachments
Context is pretty clear when you mentioned class G and H
Which one being more efficient depends on your design, how much voltage drop would that transistor see. BJT would be easier to make more efficient in most practical scenarios, although for the sake of discussion MOSFET is ultimately more efficient if you do it right. If you have multiple rails being switched, use BJT for the highest, MOSFET for the rest. But ultimately the difference in efficiency may be negligible.
And since nobody has mentioned the elephant in the room, you won't be able to make a high efficiency class H amplifier unless the tracking rail itself is SMPS - the tracking rail itself is in essence a class D amplifier.
Which one being more efficient depends on your design, how much voltage drop would that transistor see. BJT would be easier to make more efficient in most practical scenarios, although for the sake of discussion MOSFET is ultimately more efficient if you do it right. If you have multiple rails being switched, use BJT for the highest, MOSFET for the rest. But ultimately the difference in efficiency may be negligible.
And since nobody has mentioned the elephant in the room, you won't be able to make a high efficiency class H amplifier unless the tracking rail itself is SMPS - the tracking rail itself is in essence a class D amplifier.
suppose i have an additional 2 rails on my class AB and let's say max rail is +/-120vdc, you mean the first rail is MOSfet and last is BJT driven?..
many commercial design uses 3rails driven by mosfets.but if BJT is used, only 2rails are commonly available.
does it increase its efficiency if i made 3rails amp and both are BJT driven?..
- Status
- Not open for further replies.