I’ve recently purchased two mono’s amplifier of an italian manufacturer (North Star Design Monoblock).
The power (measured) is about 110W/8R, and increases up to 340W/2R.
I’ve not the schematic, but from a “visive” inspection I know that:
- the input stage is a double-differential, fet amplifier, peraphs with cascode load
- the VAS...I’m not sure, but I’ve read something about common-base configuration, it sounds unusual, and a little strange
- the output stage is a simple darlington with mje 15030/31 drivers and 8 MJL1302A/MJL3281A (4 per rail) mounted on two separate heatsink (P-N); no resistor between the emitter of the drivers and output TR base.
- the VBE multiplier seems to be made by two TO-126 BJT, one mounted directly on one output transistor (P), and the other on the heatsink of the opposite bank (N)
- the value of emitter resistors of finals is 0.22R
I’ve measured the voltage across the output resistors, and it was originally set at about 4 mv: it means a bias current less than 20 mA per final, so I tried to increase it, but the thermal stability became very critical. The differences between the 4 finals also increase with high bias.
I’ve noticed that the bias strongly depends also from the power supply, with variations (according with the main) of 2-3 volts per rail the current rise up to double, and the amplifier get into thermal runaway. I’ve a class-B amplifier during the day, and class-AB in the night, when the voltage from the main is the fully 220V!
Is this behaviour normal? What can I do to solve this problem? My old Adcom 585 seems to be more stable, even with its triple-darlington output stages and +/- 80V supply...much more refined design! And better sound, too!
Thanks
Cicero
The power (measured) is about 110W/8R, and increases up to 340W/2R.
I’ve not the schematic, but from a “visive” inspection I know that:
- the input stage is a double-differential, fet amplifier, peraphs with cascode load
- the VAS...I’m not sure, but I’ve read something about common-base configuration, it sounds unusual, and a little strange
- the output stage is a simple darlington with mje 15030/31 drivers and 8 MJL1302A/MJL3281A (4 per rail) mounted on two separate heatsink (P-N); no resistor between the emitter of the drivers and output TR base.
- the VBE multiplier seems to be made by two TO-126 BJT, one mounted directly on one output transistor (P), and the other on the heatsink of the opposite bank (N)
- the value of emitter resistors of finals is 0.22R
I’ve measured the voltage across the output resistors, and it was originally set at about 4 mv: it means a bias current less than 20 mA per final, so I tried to increase it, but the thermal stability became very critical. The differences between the 4 finals also increase with high bias.
I’ve noticed that the bias strongly depends also from the power supply, with variations (according with the main) of 2-3 volts per rail the current rise up to double, and the amplifier get into thermal runaway. I’ve a class-B amplifier during the day, and class-AB in the night, when the voltage from the main is the fully 220V!
Is this behaviour normal? What can I do to solve this problem? My old Adcom 585 seems to be more stable, even with its triple-darlington output stages and +/- 80V supply...much more refined design! And better sound, too!
Thanks
Cicero
Hi !
No, this behaviour is NOT normal, it indicates bad designer...
What to do to solve the problem ? I would say, return these
monoblocks and return to your adcom 585 ?
If it's that sensitive to voltagesupply-variations, it's likely it selfdestructs
in a few years, when parameters of devices slightly change...
Mike
No, this behaviour is NOT normal, it indicates bad designer...
What to do to solve the problem ? I would say, return these
monoblocks and return to your adcom 585 ?
If it's that sensitive to voltagesupply-variations, it's likely it selfdestructs
in a few years, when parameters of devices slightly change...
Mike
Changes in the bias should NOT result in thermal runaway...
There are obvious some failure in the design or in the units you got....
If your amps get unstable with reference to the common main supply, there are for sure something wrong!!! An amp should have no problems what ever the main is 190 or 240 VAC
There are obvious some failure in the design or in the units you got....
If your amps get unstable with reference to the common main supply, there are for sure something wrong!!! An amp should have no problems what ever the main is 190 or 240 VAC
This is obviously bad design
The most usual problem is caused by the VAS bias current increasing as the mains AC voltage increases and the Vbe of the output devices decreasing as their Vce increases [given the same Ic, the higher the Vce the lower the Vbe drop]
The increase of current through the Vbe multiplier increases its Vbe and reduces its negative feedback effect, and thermal runaway will happen in case the circuit has already low thermal negative feedback
Also, independent measurements from me and other people have shown the concept of placing the Vbe sense transistor over the plastic package of one of the output devices to be not advantageous at all, so relocating this sense transistor to the heatsink, just near the body of one of the output devices, could be an easy way to provide better bias tracking without further modifications
The most usual problem is caused by the VAS bias current increasing as the mains AC voltage increases and the Vbe of the output devices decreasing as their Vce increases [given the same Ic, the higher the Vce the lower the Vbe drop]
The increase of current through the Vbe multiplier increases its Vbe and reduces its negative feedback effect, and thermal runaway will happen in case the circuit has already low thermal negative feedback
Also, independent measurements from me and other people have shown the concept of placing the Vbe sense transistor over the plastic package of one of the output devices to be not advantageous at all, so relocating this sense transistor to the heatsink, just near the body of one of the output devices, could be an easy way to provide better bias tracking without further modifications
Thanks to you all for the replies.
I also think that my new amps are poorly designed, so I'll call the manufacturer and hope to know something more of the schematic; obviously, I'll tell them about the strange bias behaviour.
After, I'll try to disassemble
the amps and fix the problem; in the worst case, I can recicle the BIG (500VA) toroids, the 16 expensive capacitors, the 16 output devices and, of course, the cabinet and I'm going to build a DIY amps! I've just an idea of the schematic...
I also think that my new amps are poorly designed, so I'll call the manufacturer and hope to know something more of the schematic; obviously, I'll tell them about the strange bias behaviour.
After, I'll try to disassemble
the amps and fix the problem; in the worst case, I can recicle the BIG (500VA) toroids, the 16 expensive capacitors, the 16 output devices and, of course, the cabinet and I'm going to build a DIY amps! I've just an idea of the schematic...- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.