I'm repairing a Luxman L410 integrated amplifier and I've traced the problem to the output transistors, both shorted. The complementary pair, 2SC3182N and 2SA1265N, are no longer available, at least not from a source I'm confident in.
My question is: Is my proposed sub, FJA4310, close enough to the original 2SC3182N? (My main concern is the difference between the maximum collector-base voltages.)
2SC182N
Material of Transistor: Si
Polarity: NPN
Maximum Collector Power Dissipation (Pc): 100 W
Maximum Collector-Base Voltage |Vcb|: 140 V
Maximum Collector-Emitter Voltage |Vce|: 140 V
Maximum Emitter-Base Voltage |Veb|: 5 V
Maximum Collector Current |Ic max|: 10 A
Max. Operating Junction Temperature (Tj): 125 °C
Transition Frequency (ft): 30 MHz
Forward Current Transfer Ratio (hFE), MIN: 55
FJA4310
Material of Transistor: Si
Polarity: NPN
Maximum Collector Power Dissipation (Pc): 100 W
Maximum Collector-Base Voltage |Vcb|: 200 V
Maximum Collector-Emitter Voltage |Vce|: 140 V
Maximum Emitter-Base Voltage |Veb|: 6 V
Maximum Collector Current |Ic max|: 10 A
Transition Frequency (ft): 30 MHz
Forward Current Transfer Ratio (hFE), MIN: 50
So, are these close enough? Also, how close is close enough wrt to the transition frequency? I've seen other transistors that are very close to the OEM ones, but with ft values ranging from 20 to 70 MHz, instead of 30 Mhz.
For reference, I've appended the schematic for the amplification stages. The output transistors are shown on the right.
My question is: Is my proposed sub, FJA4310, close enough to the original 2SC3182N? (My main concern is the difference between the maximum collector-base voltages.)
2SC182N
Material of Transistor: Si
Polarity: NPN
Maximum Collector Power Dissipation (Pc): 100 W
Maximum Collector-Base Voltage |Vcb|: 140 V
Maximum Collector-Emitter Voltage |Vce|: 140 V
Maximum Emitter-Base Voltage |Veb|: 5 V
Maximum Collector Current |Ic max|: 10 A
Max. Operating Junction Temperature (Tj): 125 °C
Transition Frequency (ft): 30 MHz
Forward Current Transfer Ratio (hFE), MIN: 55
FJA4310
Material of Transistor: Si
Polarity: NPN
Maximum Collector Power Dissipation (Pc): 100 W
Maximum Collector-Base Voltage |Vcb|: 200 V
Maximum Collector-Emitter Voltage |Vce|: 140 V
Maximum Emitter-Base Voltage |Veb|: 6 V
Maximum Collector Current |Ic max|: 10 A
Transition Frequency (ft): 30 MHz
Forward Current Transfer Ratio (hFE), MIN: 50
So, are these close enough? Also, how close is close enough wrt to the transition frequency? I've seen other transistors that are very close to the OEM ones, but with ft values ranging from 20 to 70 MHz, instead of 30 Mhz.
For reference, I've appended the schematic for the amplification stages. The output transistors are shown on the right.
Last edited:
The maximum ratings are the parameters that the circuit must not exceed lest the transistor fail, so higher values for Veb, Vcb, Vce, Pc, Ic are all advantages if anything. hFE is usually not critical as any bunch of transistors will have a range of values, but generally higher is better and in some circuits pairs of transistors are ideally matched by test and selection.
Best quality performance is usually achieved with a close complementary match of hFE or beta over the full working range of current but that's only likely if the datasheets show similar curves and you have adequate stocks of parts to test and choose from. Otherwise, it looks like ONsemi FJA4310-O + FJA4210-O will be fine. Are the drivers OK though?
Thanks johnmath and Ian Finch. I don’t have the stock or test gear for matching the complementary transistors, so I will have to trust to luck in that regard.
Ian, with my DMM in diode mode, the driver transistors measured fine.
A bonus question, if I may. I understand that exceeding the specified ft value can lead to instability. But what about going in the other direction: is it okay to use a transistor with a ft value below the spec’ed value?
Ian, with my DMM in diode mode, the driver transistors measured fine.
A bonus question, if I may. I understand that exceeding the specified ft value can lead to instability. But what about going in the other direction: is it okay to use a transistor with a ft value below the spec’ed value?
That's an interesting question because solid state audio began with very low Ft transistors, much worse than even 2N3055 that we still find in beginner and DIY circles. However, sound quality improved by quantum leaps in the 1980s, when Toshiba introduced their multi-emitter, high beta linearity, 30MHz transistors and these soon became the popular standard amid stiff competition from other Japanese manufacturers. Together, those improvements to BJTs remain essential to high quality, linear class AB linear audio amplification technology.
High Ft isn't everything but it does play a significant role in reducing audible distortion with switching designs (i.e. typical class AB). The simple answer is basically that the amplifier was designed to use those 30MHz transistors and you can only do worse by substituting the wrong (say, old standard 1-4MHz Ft) type power transistors there. Use the right type parts and you'll be fine.
High Ft isn't everything but it does play a significant role in reducing audible distortion with switching designs (i.e. typical class AB). The simple answer is basically that the amplifier was designed to use those 30MHz transistors and you can only do worse by substituting the wrong (say, old standard 1-4MHz Ft) type power transistors there. Use the right type parts and you'll be fine.
The main difference seems to be that the FJA4310 is designed for switching use as its
Vsat is 0.5V at 5A, not 2V as for the 3182. Also the leakage currents are much higher,
(maybe for same reason?). The output capacitance and other values are close though.
Vsat is 0.5V at 5A, not 2V as for the 3182. Also the leakage currents are much higher,
(maybe for same reason?). The output capacitance and other values are close though.
Let me throw another option out there. It has a a collector to emitter saturation voltage (IC = 5 A, IB = 0.5 A) of 2V but works with a transition frequency of 50MHz. Is that tf value too high? Any preference for the FJA4310 (Vsat of 0.5V) or the 2SC3519A (ft of 50MHz)?
Edit: here's the full specs on the alternative transistor:
2SC3519A
Material of Transistor: Si
Polarity: NPN
Maximum Collector Power Dissipation (Pc): 130 W
Maximum Collector-Base Voltage |Vcb|: 180 V
Maximum Collector Current |Ic max|: 15 A
Transition Frequency (ft): 50 MHz
Forward Current Transfer Ratio (hFE), MIN: 50
Collector to Emitter Saturation Voltage (@IC = 5 A, IB = 0.5 A): 2V
Last edited: