Assuming you bought those transistors from RS (HK), fakes are unlikely. It takes a heavy load to burn out even small power transistors so, assuming the supply is not > 38V and the audio is not loud, you still have a construction fault in the amplifier, I think. Check the driver transistors for failures too.
Why not measure the bias current as before by calculating the current in the emitter resistors, to be certain you have stability. The same bias of 15-25mV across each emitter resistor is OK with any of these power transistors.
You should also compare the temperatures between your split heatsinks and make certain the bias control transistor is properly fixed to but insulated from one output transistor heatsink. This is still a very poor arrangement so you will need to monitor temperature and current carefully to know what is happening.
Take care with temperatures as you have no other way of knowing whether oscillation may be occurring. You will need to do this anyway when you fit new transistors, but I still suggest using cheaper MJ15003/4 whilst there are problems as the other types are expensive and have unnecessarily high ratings.
Why not measure the bias current as before by calculating the current in the emitter resistors, to be certain you have stability. The same bias of 15-25mV across each emitter resistor is OK with any of these power transistors.
You should also compare the temperatures between your split heatsinks and make certain the bias control transistor is properly fixed to but insulated from one output transistor heatsink. This is still a very poor arrangement so you will need to monitor temperature and current carefully to know what is happening.
Take care with temperatures as you have no other way of knowing whether oscillation may be occurring. You will need to do this anyway when you fit new transistors, but I still suggest using cheaper MJ15003/4 whilst there are problems as the other types are expensive and have unnecessarily high ratings.
Last edited:
RE: Ian. I had finished all the checking while you post the second post. I think the reason for pointing out the fake output transistors was because these were not brought from the RS. RS is only for choose the production company. And as i mention, only one side of the amp is turns down, this means the power supply do not have any problem. The only problem is come from the amp. I think the transistors burns out is because i wanna measure the voltage difference of each test point. While this action goes on, i short out one of the diode biasing, so i just replace the diode bias [ diode bias = two diode series together and a self bias transistor. ].
I am not sure as well, but i check the preamp circuit, and not transistors was fail. i check those transistors by the diode test. All are in the right range of the measurement. And all resistors are good, no burns out sign at all.
As you mention the heat sink might be one of the facts as well. But i am sure all things runs below 40 deg!
I am not sure as well, but i check the preamp circuit, and not transistors was fail. i check those transistors by the diode test. All are in the right range of the measurement. And all resistors are good, no burns out sign at all.
As you mention the heat sink might be one of the facts as well. But i am sure all things runs below 40 deg!
40C is fine, you could not measure much difference at that level. My concern is that even fakes should be able to operate at such low dissipation.
I have not tested this, but adding diodes to the bias string changes the adjustment range and thermal compensation ratio drastically, unless all the diodes and the bias transistor are fixed to the heatsink. You also have adjustment over only 1/3 of the design range, asssuming you now have 3 junctions in series. I don't think this was part of the original design so I think you should reinstate the original circuit to be certain it will compensate properly.
Finally, check that no part of the circuit like the output or power rails are connected accidentally to the heatsink. Sometimes this causes disasters when other parts are connected too.
I have not tested this, but adding diodes to the bias string changes the adjustment range and thermal compensation ratio drastically, unless all the diodes and the bias transistor are fixed to the heatsink. You also have adjustment over only 1/3 of the design range, asssuming you now have 3 junctions in series. I don't think this was part of the original design so I think you should reinstate the original circuit to be certain it will compensate properly.
Finally, check that no part of the circuit like the output or power rails are connected accidentally to the heatsink. Sometimes this causes disasters when other parts are connected too.
Re: adding diodes to the bias string changes the adjustment range and thermal compensation ratio drastically, unless all the diodes and the bias transistor are fixed to the heatsink.
< I think you have misunderstand it. The diode biasing is the EB of the BC557 which recalls 1.4V . And i measured two boards, which i found out that it is not the same, as i noticed that diode will be saturate at 0.7. Normally is not 0.7, it is about 0.67. And this might affect the gain or current flow to the differential inputs (pre amp stage). Remarks all circuits and design do not have any changes !
But during the measurement, my meter shorted out the diode and cause spark. After that i measured it again, and found out the diodes were down. So i got the board out and re solder those diodes. Re open the amps and two channel were so different. So i designed to let is be, and keep operating for 4hrs. The result was one of the output stage were all went down< this board is the shorted ones.
< I think you have misunderstand it. The diode biasing is the EB of the BC557 which recalls 1.4V . And i measured two boards, which i found out that it is not the same, as i noticed that diode will be saturate at 0.7. Normally is not 0.7, it is about 0.67. And this might affect the gain or current flow to the differential inputs (pre amp stage). Remarks all circuits and design do not have any changes !
But during the measurement, my meter shorted out the diode and cause spark. After that i measured it again, and found out the diodes were down. So i got the board out and re solder those diodes. Re open the amps and two channel were so different. So i designed to let is be, and keep operating for 4hrs. The result was one of the output stage were all went down< this board is the shorted ones.
hi everyone ,
i want to ask that, i have a 35-0-35 vac transformer , if i replace 2n3055/MJ2955 with 2sc5200/2sa1943 can i use this transformer? and then what will be output of this amplifier at 4ohm load?
i want to ask that, i have a 35-0-35 vac transformer , if i replace 2n3055/MJ2955 with 2sc5200/2sa1943 can i use this transformer? and then what will be output of this amplifier at 4ohm load?
Hi Maroof
As explained early in the thread, The voltage rating (Vceo) of the transistors used determines how much voltage can be safely applied, like any amplifier. You don't say whether you are starting from your own PCB, using all different transistors to those specified elsewhere, as in the VAS components or driver transistors, since using the original ones means the maximum voltage should still be only +/- 40V DC max.
A 35-0-35 transformer results in ~ +/- 50V rails, much too high for the SC480, even if you substitute higher voltage parts in the VAS and driver positions because the bias currents in the input and VAS stages will need to be readjusted. I'm sure you understand that raising voltage through resistive circuits means raising current too. By choosing slightly larger emitter resistors for the current source in the input stage and that for the VAS, you can adjust their bias current to the correct amounts, though.
It will be a different amplifier because changing VAS transistors affects frequency compensation and sound quality but it will still work OK if you are careful to carry out the amendments with suitable parts. Be careful that some BC639/640, like On-semi versions, have very low capacitance and this will make a big difference when you substitute something awful like MJE340/350. The transistor types to use will vary with what is available or affordable in your region. For example, in the VAS, I now use only Fairchild KSC3503/KSA1381 but you may not find them locally, as I don't.
Assuming you use a 300VA transformer, 2 pairs of 2SC5200/A1943 and good heatsinking, you should get about 150W/4R output - perhaps a little more with a larger VA rating transformer and less sag under load.
Otherwise, there are slightly more interesting designs for 50-60V rails based on this topology, like the many versions of Symasym. This was originally introduced by Destroyer X and you can search it here on the forum.
As explained early in the thread, The voltage rating (Vceo) of the transistors used determines how much voltage can be safely applied, like any amplifier. You don't say whether you are starting from your own PCB, using all different transistors to those specified elsewhere, as in the VAS components or driver transistors, since using the original ones means the maximum voltage should still be only +/- 40V DC max.
A 35-0-35 transformer results in ~ +/- 50V rails, much too high for the SC480, even if you substitute higher voltage parts in the VAS and driver positions because the bias currents in the input and VAS stages will need to be readjusted. I'm sure you understand that raising voltage through resistive circuits means raising current too. By choosing slightly larger emitter resistors for the current source in the input stage and that for the VAS, you can adjust their bias current to the correct amounts, though.
It will be a different amplifier because changing VAS transistors affects frequency compensation and sound quality but it will still work OK if you are careful to carry out the amendments with suitable parts. Be careful that some BC639/640, like On-semi versions, have very low capacitance and this will make a big difference when you substitute something awful like MJE340/350. The transistor types to use will vary with what is available or affordable in your region. For example, in the VAS, I now use only Fairchild KSC3503/KSA1381 but you may not find them locally, as I don't.
Assuming you use a 300VA transformer, 2 pairs of 2SC5200/A1943 and good heatsinking, you should get about 150W/4R output - perhaps a little more with a larger VA rating transformer and less sag under load.
Otherwise, there are slightly more interesting designs for 50-60V rails based on this topology, like the many versions of Symasym. This was originally introduced by Destroyer X and you can search it here on the forum.
Last edited:
- Status
- Not open for further replies.