Is there a circuit diagram available?
There are many possibilities for overheating, from an incorrect bias or poor bias stabiliser, to poor thermal mounting, inadequate heatsink size, oscillation or being in an enclosure which does not let air circulate on the heatsink.
Some help on what the circumstances are would be useful.
There are many possibilities for overheating, from an incorrect bias or poor bias stabiliser, to poor thermal mounting, inadequate heatsink size, oscillation or being in an enclosure which does not let air circulate on the heatsink.
Some help on what the circumstances are would be useful.
Hi,
Sorry for the delay, too many works.
For wg_ski and john_ellis, I begin to draw the schematic of the amplifier, (not finished) and it seem to me that the BD441-Bd442 are pre-driver, not VAS.
For Nareshbrd, the bias are ok, but the four output transistor of the amp are place between the heatsink and the PCB and it have a very poor heatsink,
too thin. Also I leave them always on.
It is possible for me to have some BD139-16-BD140-16, but the BD441-BD442 are a no go. I need to find a good substitute for the BD441-BD442,
so, in the same time why not to change the BD139-16-BD140-16 for something more actual.
Sorry for the delay, too many works.
For wg_ski and john_ellis, I begin to draw the schematic of the amplifier, (not finished) and it seem to me that the BD441-Bd442 are pre-driver, not VAS.
For Nareshbrd, the bias are ok, but the four output transistor of the amp are place between the heatsink and the PCB and it have a very poor heatsink,
too thin. Also I leave them always on.
It is possible for me to have some BD139-16-BD140-16, but the BD441-BD442 are a no go. I need to find a good substitute for the BD441-BD442,
so, in the same time why not to change the BD139-16-BD140-16 for something more actual.
Attachments
Put an AC driven panel fan, biggest that will fit.
AC because they are more reliable than DC fans.
That should reduce heat related issues, it is a hack, a partial and not very elegant solution.
AC because they are more reliable than DC fans.
That should reduce heat related issues, it is a hack, a partial and not very elegant solution.
The BD139/40 are in the VAS position, and run at around 20 mA. That will be hot enough that you will sit up and take notice when you put your finger on one of them. With that much emitter degeneration it’s not likely to run away, but at almost 1 watt I’d put a small heat radiator on each.
The 441/2 are in the driver position - no “pre drivers”. With that 18 ohm bias resistor between them they will run HOT at over 2 watts apiece just idling. They often do this with BIG amps to keep switching distortion low when driving a big bank of outputs, and when they do they put them on the same heat sink with the outputs. If I was going to run them that high I’d be using MJE1503x and using pretty good size heat sinks or on the output heat sink. MJE243/253 is certainly possible but I’m pretty sure the 1503x are going to be easier to get.
The 441/2 are in the driver position - no “pre drivers”. With that 18 ohm bias resistor between them they will run HOT at over 2 watts apiece just idling. They often do this with BIG amps to keep switching distortion low when driving a big bank of outputs, and when they do they put them on the same heat sink with the outputs. If I was going to run them that high I’d be using MJE1503x and using pretty good size heat sinks or on the output heat sink. MJE243/253 is certainly possible but I’m pretty sure the 1503x are going to be easier to get.
Thanks for the circuit. I agree with wg_ski that sometnhing like the mJE1503x would be better. The BD441/442 are only 3MHz type devices and not really suitable as drviers in a hifi amp anyway.
Due to the high dissipation as mentioned the devices really should be on the main heatsink. IF they're not, the high dissipatoin will mean they get very hot and that means the bias stabiliser can't do it's job properly. On the other hand I have found the bias current of a EF2 to be reasonably stable when the drivers and output are on the main heatsink, along with the bias stabiliser. I would also recommend the non-insulated version and mica+silicone grease or one of the new highly thermally conductive silicone washers as the thermal resistance can be, with good mounting, lower than those devices fully insulated.
Due to the high dissipation as mentioned the devices really should be on the main heatsink. IF they're not, the high dissipatoin will mean they get very hot and that means the bias stabiliser can't do it's job properly. On the other hand I have found the bias current of a EF2 to be reasonably stable when the drivers and output are on the main heatsink, along with the bias stabiliser. I would also recommend the non-insulated version and mica+silicone grease or one of the new highly thermally conductive silicone washers as the thermal resistance can be, with good mounting, lower than those devices fully insulated.
The thermal glue used for LED mounting can also be considered, but it is a type of silicon, it must be uused up quickly once the tube is opened, as it cures on exposure to air.
Hi,
Yes you're right. The Bd139-16-BD140-16 and the BD441-BD442 are installed on the same heat sink as the outputs.
MJE243/253 and MJE1503X are available at mouser, Arrow, Digikey etc. What one would you considerred to be the best choice for that location ?
And for BD139-140 replacement, do the TTA004-TTC004 could be ok ?
Yes you're right. The Bd139-16-BD140-16 and the BD441-BD442 are installed on the same heat sink as the outputs.
MJE243/253 and MJE1503X are available at mouser, Arrow, Digikey etc. What one would you considerred to be the best choice for that location ?
And for BD139-140 replacement, do the TTA004-TTC004 could be ok ?
Hi,
Here's a picture of one channel as requested. The output are under the PCB connected with the wire a the upper left of the picture.
The BD 139/140 and the BD441/442 are at the upper right of the picture.
A more recent board of the same amplifier, The TO-39 transistor BC141/161 was replaced with BD139/140, and the output transistors connections are now integrated in the PCB
Here's a picture of one channel as requested. The output are under the PCB connected with the wire a the upper left of the picture.
The BD 139/140 and the BD441/442 are at the upper right of the picture.
A more recent board of the same amplifier, The TO-39 transistor BC141/161 was replaced with BD139/140, and the output transistors connections are now integrated in the PCB
Attachments
Hi,
Actually, I made a new board for this amplifier, and I will install them in a new manner, to not have the transistor outputs placed between the heatsink and the PCB. So, TO-126 or TO-220 will not be affected by my layout, I can place the TO-126 on the top circuit or the TO-220 under the circuit, the pinout will be correct.
The only thing that bug me is the Cob value of the BD441/442. I never been able to find it, and ideally I would like to replace them with an other one with a similar or a lower value.
The MJE have a Cob of 50, but no value indicated for the MJE1503X. I read somewhere that generally the Cob value will increase with a bigger Transistor DIE, but not able to confirm this.
I have a friend who have lot of old Toshiba TO-220 2SA1930/2SC5171 but rated at 2A. Do you think it could be good for this purpose ?
Actually, I made a new board for this amplifier, and I will install them in a new manner, to not have the transistor outputs placed between the heatsink and the PCB. So, TO-126 or TO-220 will not be affected by my layout, I can place the TO-126 on the top circuit or the TO-220 under the circuit, the pinout will be correct.
The only thing that bug me is the Cob value of the BD441/442. I never been able to find it, and ideally I would like to replace them with an other one with a similar or a lower value.
The MJE have a Cob of 50, but no value indicated for the MJE1503X. I read somewhere that generally the Cob value will increase with a bigger Transistor DIE, but not able to confirm this.
I have a friend who have lot of old Toshiba TO-220 2SA1930/2SC5171 but rated at 2A. Do you think it could be good for this purpose ?
Just a detail, but your Q17 in #23 is named DB140, which could point to a german locomotive instead to a solid state bipolar.
When on steam, it can generate some heat indeed.
When on steam, it can generate some heat indeed.
Hi,
After some research, I could have the following combo to replace the BD441/442 to drive the 2SA1943/2SC5200. All original, no fake unit.
Onsemi MJE1503X (MJE15034G/MJE15035G) - TO-220
Onsemi MJE243/253 - TO126
Toshiba 2SA1930/2SC5171 - TO-220F
Toshiba 2SA1837/2SC4793 - TO-220F
Sanken 2SA1859/2SC4883 - TO-220F
Is there some model I need to avoid for my configuration ?
Is anyone have try the Sanken for this purpose ? Seems to be a very good transistor.
After some research, I could have the following combo to replace the BD441/442 to drive the 2SA1943/2SC5200. All original, no fake unit.
Onsemi MJE1503X (MJE15034G/MJE15035G) - TO-220
Onsemi MJE243/253 - TO126
Toshiba 2SA1930/2SC5171 - TO-220F
Toshiba 2SA1837/2SC4793 - TO-220F
Sanken 2SA1859/2SC4883 - TO-220F
Is there some model I need to avoid for my configuration ?
Is anyone have try the Sanken for this purpose ? Seems to be a very good transistor.
For epitaxial planar types I would lean toward Sanken. For triple diffused, Toshiba. Both were kings of their respective hills.
Hi,
Thanks for your answer wg_ski. So if I understand you correctly, the Sanken will be a better choice over the MJE ? For my setup of course.
And for the BD139-16/BD140-16, TTA004B/TTC004B or TTA008B/TTC015B ?
I ask it to be sure, because I will order a big quantity of each transistors, as I want to have enough spares parts for the rest of my life.
Thanks for your answer wg_ski. So if I understand you correctly, the Sanken will be a better choice over the MJE ? For my setup of course.
And for the BD139-16/BD140-16, TTA004B/TTC004B or TTA008B/TTC015B ?
I ask it to be sure, because I will order a big quantity of each transistors, as I want to have enough spares parts for the rest of my life.