Ive just aquired a nice Sansui AU-222 integrated w/ a blown amp that I would like to Cloneize. I took a voltage check from the three PS wires and got 37volts on each side. This does mean Ive got a 75VCT transfo right? I checked the application sheets for the LM3875 and it states that it can operate on 20-84 volts. It seems that 35volts is about the most anyone recommends for a GC so whats up?
I would really like to use this transformer but is 37.5 X 1.4 = 52.5volts going to fry the chip? And why does NSC claim a 84volt maximum?
amt
I would really like to use this transformer but is 37.5 X 1.4 = 52.5volts going to fry the chip? And why does NSC claim a 84volt maximum?
amt
"And why does NSC claim a 84volt maximum?"
Death and destruction point.
"I would really like to use this transformer but is 37.5 X 1.4 = 52.5volts going to fry the chip? "
75VCT is ±52.5V, or 105V total. See above.
"It seems that 35volts is about the most anyone recommends for a GC so whats up? "
The LM chips have protection circuits that are sensitive to both current and heat. If you try and run them at too low of a load impedance with too high a supply voltage they make funny sounds (current protection).
If your heatsink is not big enough it will activate the protection circuit and make funny noises too (it activates the current protection to keep cool).
A 24-0-24V transformer will give about ±35V no load, realistically, that's about it for these chips.
While complex, you can parallel chips for more current at low impedance, and bridge them for high voltage swings at high impedance.
One company parallel bridges SIX chips for a high voltage/high current amplifier (250W/4R).
Death and destruction point.
"I would really like to use this transformer but is 37.5 X 1.4 = 52.5volts going to fry the chip? "
75VCT is ±52.5V, or 105V total. See above.
"It seems that 35volts is about the most anyone recommends for a GC so whats up? "
The LM chips have protection circuits that are sensitive to both current and heat. If you try and run them at too low of a load impedance with too high a supply voltage they make funny sounds (current protection).
If your heatsink is not big enough it will activate the protection circuit and make funny noises too (it activates the current protection to keep cool).
A 24-0-24V transformer will give about ±35V no load, realistically, that's about it for these chips.
While complex, you can parallel chips for more current at low impedance, and bridge them for high voltage swings at high impedance.
One company parallel bridges SIX chips for a high voltage/high current amplifier (250W/4R).
You can run it on 52 volts (as well as 84) but it's all down to ohms law which speakerload you can drive. With 35 volts a single 3875 can drive 8 ohm loads without problems.
The maximum power for a 3875 is around 60 watts. The current needed to drive a load is I=U/R. The power that a current and resistance makes is P=R*I^2. Go from there and get the load for your voltages and see if it is enough.
Anders
The maximum power for a 3875 is around 60 watts. The current needed to drive a load is I=U/R. The power that a current and resistance makes is P=R*I^2. Go from there and get the load for your voltages and see if it is enough.
Anders
NSC is quite correct in their spec sheet. The LM3875 can handle 80-84V single-ended power supply, where the neg PS pin is connected to 0V, the output to speaker needs DC bocking cap. But split rails the max has to be divided by two, consequently +35V/-35V is really 70V. Certainly you can go as high as +42V/-42V (=84V), but I don't recommend it.
Joe R.
Joe R.
thanks for the help
Looks like this integrated will have to wait for a rebirth. I'll look for an easy design that can use the voltage it has.
thanks amt
Looks like this integrated will have to wait for a rebirth. I'll look for an easy design that can use the voltage it has.
thanks amt
- Status
- Not open for further replies.