hi,
a couple of years ago i was building a 6w+6w amp using tpa1517, and it kept burning the chip every time i'd turn it on.
after blowing a couple of them i changed the power switch from "after the PS" to "after transformer, before diodes". it stopped burning, i never knew exactly why this happened.
see figures 1 and 2.
this new era finds me building a small sub woofer plate amp using TDA2040 and a switching power supply (the one HP printers use, 16v@650mA + 32v@950mA).
apparently i made the same mistake at my first desing. as a result of this, the chip is history (short circuit between output and Vs).
this time, as we are talking about a SMPS, i don't know if the same change i did before will fix it. any way, it would be quite a mess trying to place an interruptor between transformer and diodes inside the SMPS.
the first design (the one burning the chip) is fig. 3.
will interrupting the mains save me ?see fig. 4.
or should i replace the 220uF for a 1000uF or 2200uF instead?
i don't know. you tell me...
i'd appreciate a lot your helping me with a solution, just a piece of advice, or an explanation of this burning chip problem.
thanks a lot in advance!
PS: the amp circuit is ok. it's the one found in the TDA2040 datasheet as a typical application for single supply and i've seen it working well with a linear PS interrupted "after transformer, before diodes", and with another one interrupted at the mains.
a couple of years ago i was building a 6w+6w amp using tpa1517, and it kept burning the chip every time i'd turn it on.
after blowing a couple of them i changed the power switch from "after the PS" to "after transformer, before diodes". it stopped burning, i never knew exactly why this happened.
see figures 1 and 2.
this new era finds me building a small sub woofer plate amp using TDA2040 and a switching power supply (the one HP printers use, 16v@650mA + 32v@950mA).
apparently i made the same mistake at my first desing. as a result of this, the chip is history (short circuit between output and Vs).
this time, as we are talking about a SMPS, i don't know if the same change i did before will fix it. any way, it would be quite a mess trying to place an interruptor between transformer and diodes inside the SMPS.
the first design (the one burning the chip) is fig. 3.
will interrupting the mains save me ?see fig. 4.
or should i replace the 220uF for a 1000uF or 2200uF instead?
i don't know. you tell me...
i'd appreciate a lot your helping me with a solution, just a piece of advice, or an explanation of this burning chip problem.
thanks a lot in advance!
An externally hosted image should be here but it was not working when we last tested it.
PS: the amp circuit is ok. it's the one found in the TDA2040 datasheet as a typical application for single supply and i've seen it working well with a linear PS interrupted "after transformer, before diodes", and with another one interrupted at the mains.
I don't know about SMPS but I guess the reason why your chips blew in the first scenario was the sudden voltage and current introduced to them: The potential will shift very fast from 0 to whatever and I guess the chips just could not take the transient. If you place the switch before the rectifier it will introduce a delay since the supply voltage will have to rise ("smoothly") from 0 to whatever as the reservoir capacitors charge.
In my opinion it's always best (safest, less power consuming) to just switch off the mains as in figure 4.
In my opinion it's always best (safest, less power consuming) to just switch off the mains as in figure 4.
thanks a lot!
i thought something like that was the problem, but then another question arises in my ignorant brain.
as SMPS oscillation is way faster than 50Hz, won't tension go up fast anyway? maybe not fast enough to kill the chip. maybe frequency has nothing to do and the time required to fill the caps is related to the energy required. so. it's related to the power delivered by the transformer (amplitude of voltage and current delivering capability, not frequency).
thanks a lot indeed.
i thought something like that was the problem, but then another question arises in my ignorant brain.
as SMPS oscillation is way faster than 50Hz, won't tension go up fast anyway? maybe not fast enough to kill the chip. maybe frequency has nothing to do and the time required to fill the caps is related to the energy required. so. it's related to the power delivered by the transformer (amplitude of voltage and current delivering capability, not frequency).
thanks a lot indeed.
in your original 6W amp, with the linear power supply, how close was the power supply voltage to the voltage limit of the chips?
I'm thinking of the fact that linear supplies, when powered up and unloaded, will reach a much higher voltage in some cases than when loaded. When you throw that output switch, this high voltage appears at the chip until the chip discharges the capacitor down to a normal level. If the chip was rated for less than this higher voltage, it could have fried.
The above is unlikely, but then again, it's seemingly unusual for circuits like yours to fry the way yours did.
I'm thinking of the fact that linear supplies, when powered up and unloaded, will reach a much higher voltage in some cases than when loaded. When you throw that output switch, this high voltage appears at the chip until the chip discharges the capacitor down to a normal level. If the chip was rated for less than this higher voltage, it could have fried.
The above is unlikely, but then again, it's seemingly unusual for circuits like yours to fry the way yours did.
I suppose
it is your TDA2040 chip that is blowing up.
You would imagine to draw PCB for such a simple
but not bad amplifier as the good old TDA2040
wouldn't be difficult - not be at all critical with Layout.
But I have read several TDA2040 projects at websites (Search)
as well as read posts from Chip Amps builders,
saying you have to be careful when layout boards for TDA2040.
You should study this PDF ( I guess you have it ).
Why do you think they have bothered
to show you layout suggestion - for EVERY CIRCUIT in that PDF???
... you do not often see this is datasheets ...
My Guess it is more important with layout, than usual.
Especially you should notice:
- How the ground is drawn.
- How the connections +V, 0V and LSP-GND and Input-GND are done.
- In what order components are joined to Ground plane and rails.
- How supply bypass capacitors are put
- These Caps are in my attached diagram: C3(220uF), C5(100nF)
- Also R6 + C6 can be of great importance. To stop oscillation burn outs
christmas
lineup
it is your TDA2040 chip that is blowing up.
You would imagine to draw PCB for such a simple
but not bad amplifier as the good old TDA2040
wouldn't be difficult - not be at all critical with Layout.
But I have read several TDA2040 projects at websites (Search)
as well as read posts from Chip Amps builders,
saying you have to be careful when layout boards for TDA2040.
You should study this PDF ( I guess you have it ).
Why do you think they have bothered
to show you layout suggestion - for EVERY CIRCUIT in that PDF???
... you do not often see this is datasheets ...
My Guess it is more important with layout, than usual.
Especially you should notice:
- How the ground is drawn.
- How the connections +V, 0V and LSP-GND and Input-GND are done.
- In what order components are joined to Ground plane and rails.
- How supply bypass capacitors are put
- These Caps are in my attached diagram: C3(220uF), C5(100nF)
- Also R6 + C6 can be of great importance. To stop oscillation burn outs
christmas
lineup
Attachments
Sizzle..
Yes the switching should be at the mains on the primary side....I would bet you are indeed over the limit on the B+ supply voltage, there is really no reason for it to blow with a proven design with the switching as described.
Regards from the"Ocean of silver"___________
_______________________________________Rick.............
Yes the switching should be at the mains on the primary side....I would bet you are indeed over the limit on the B+ supply voltage, there is really no reason for it to blow with a proven design with the switching as described.
Regards from the"Ocean of silver"___________
_______________________________________Rick.............
this PS has a 32V/950mA and a 16V/650mA output, would it be the same placing in one or another outputs? or must it be connected to the output that's burning the chip?
i ask this because i'll be using the 32V output of the PS close to its limit when playing loud, but the 16V will be feeding only a filter preamp with negligible power consumption.
i ask this because i'll be using the 32V output of the PS close to its limit when playing loud, but the 16V will be feeding only a filter preamp with negligible power consumption.
finally, it didn't die anymore.
it's working!
http://www.diyaudio.com/forums/showthread.php?s=&threadid=93185
it's working!
http://www.diyaudio.com/forums/showthread.php?s=&threadid=93185
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.