Swapping IRFP9240/240's for SC5200/SA1943's??

[Zero Cool]

Any thoughts on swapping some IRFP9240/240 mosfets in place of some 5200/1943 bipolars???

[DarkHorse]

Originally posted by Zero Cool
Any thoughts on swapping some IRFP9240/240 mosfets in place of some 5200/1943 bipolars???

hi,that need to check with your driver stage,maybe need more adjusting the value of device to fit with the MOSFET's Ugs(3~4V),but the BJT Ube just about 0.6~0.7V...

regards!
WINCO

[jacco vermeulen]

Hi Zero,

you will need to calculate for the number of devices you plan to use what the slewrate will be.
Without any other changes except the output devices the drivers will need to feed the outputs capacitance with enough current.
Slewrate = driver current / total gate capacitance (Ciss)

Unless you alter the voltage for the output stage you will loose power because of the Mosfet threshold voltage.

IRF's have relatively high Ciss, 1300 pF.
To obtain a slewrate of 50 V/uS with four IRF devices the driver needs to deliver 260 mA.
The Toshiba's have a high current amplification, you will have to check if the drivers can deliver as much

[richie00boy]

Don't worry about any calculations Unless you are modding a hyper fast amp you are wasting your time. It's even possible that the rest of the stages might not slew that fast anyway.

As the BJTs have relatively low current gain the driver stage should be up to the task. In practice it's a straight swap as long as you mod the Vbe mulitplier (bias generator) to generate the required increased bias voltage. If you are lucky the existing Vbe multiplier may have enough adjustment to achieve this without changing anything.

As mentioned the downside will be slightly less available max power output/voltage swing.

[Allexx]

The best way to do it is to make Sziklai combination of BJT and MOSFETs with drains connected to load via small current sensing resistors. Average values of resistors are 33 Ohm in emitter of BJT driver with other end connected to drain of MOSFET, about 330 ohms in collector and collector connected to gate via 150 Ohm resistor to prevent oscillation. It is up to you to use any protection Zener diode connected between source and gate or parallel to 330 Ohms. Such a circuit enables you to get higher voltage swing at the output compared to source follower arrangement. Next you need only 2Vbe bias from Vbe multiplier. The best way to compensate temperature drift of output transistors is to put Vbe multiplier transistor at the same small heatsink of any BJT driver transistor.

[richie00boy]

Forgot to add on my post that you will need gate stopper resistors as close to the gate lead as possible. I achieved this by snipping the gate lead with about 4mm sticking out and soldering in a 0.125 watt resistor which then is connected to the board where the gate lead would normally be soldered. I used 470 ohms with IRFP240 and 510 ohms with IRFP9240 in my particular application. These values should suit most designs.

[richie00boy]

Originally posted by Allexx
The best way to do it is to make Sziklai combination of BJT and MOSFETs with drains connected to load via small current sensing resistors. Average values of resistors are 33 Ohm in emitter of BJT driver with other end connected to drain of MOSFET, about 330 ohms in collector and collector connected to gate via 150 Ohm resistor to prevent oscillation. It is up to you to use any protection Zener diode connected between source and gate or parallel to 330 Ohms. Such a circuit enables you to get higher voltage swing at the output compared to source follower arrangement. Next you need only 2Vbe bias from Vbe multiplier. The best way to compensate temperature drift of output transistors is to put Vbe multiplier transistor at the same small heatsink of any BJT driver transistor.

All true, but we don't know if the design in question already has CFP output or EF output It would take some major modding to implement a CFP output if it's an EF output.

[jacco vermeulen]

I would love to see the schematics if it is available and read the outcome after the swap.

[Zero Cool]

Thanks for all the info. The driver stage uses A968 and C2238 TO-220 devices so they should have enough current available.

I am aware that I would need to add gate resistors. I usually use 220 ohms.

The schematic provided shows D2155/B1429 devices, but later models use C5200/A1943 devices. And even later models changed again.....

I have 2 of these modules and both have suffered a similar fate. I believe that are breaking into oscillation and blowing up. This is a Bass guitar amp so they get pushed pretty hard. the amp is rated for 350 Watts into 4 ohms and some players use a lot of slap and pop playing style and so this amp could see a lot of transients etc.

The output stage is very similar to the output stage of the Soundcraftsmen amps i work on so I am thinking a swap should be pretty easy. I’m looking to add stability and strength over audiophile type sound quality.

NOTE: Q311, Q312 are there for protection.

Also Not TH302, this Thermister mounts on the heat sink between the output devices.

The BIAS transistor is a C3200 and seems to be a pretty small device. I am thinking it’s not big enough to handle the Mosfets added requirements.

I would appreciate any suggestions for added stability. The only reason I want to switch to Mosfets is because they seem to be more robust.

R321 is a 150 ohm 2 watt resistor and always smokes when the amp blows. It runs VERY warm anyway. So this might need to be changed to a larger rating?? But my Soundcraftsmen’s use a 180 ohm resistor here and I don’t think it’s more that 2 watts so?


Zc

[Zero Cool]

Forgot the attachment.


ALSO....The schematic is for reference only and is the respective property of its maker. Im sure its copyrighted.


ZC