Okay, here's the thing . . . I had the plate amp section out of a really shitty computer 2.1 system. One of the satellite channels did nothing but hum so I guessed it was an output transistor. (Note: this exercise was purely for educational value). I found the amp board had three TDA 2030s so I changed all three of them for TDA 2050s. The original transformer offered 2 x 11v so I replaced it with one providing 2 x 16v.
It sounded great for about 90 seconds until HUM. BOOM. SMOKE.
I am ignorant (literally) . I know I should have upgraded the heatsink but that cannot have been the problem. I'm guessing I blew another transistor (likely the one on the right).
The unit's dead now - hopefully an in-line fuse somewhere.
What happened?
If boredom dictates I look at this again what board components should I change?
It sounded great for about 90 seconds until HUM. BOOM. SMOKE.
I am ignorant (literally) . I know I should have upgraded the heatsink but that cannot have been the problem. I'm guessing I blew another transistor (likely the one on the right).
The unit's dead now - hopefully an in-line fuse somewhere.
What happened?
Your mistake is to replace the transformer with a higher voltage without taking into account the capabilities of the rectifier and electrolytic capacitors.Check everything, replace the faulty one.
Use suitable parts from trusted suppliers.
Electrolytics didn’t explode, the fake IC did.
Fake TDA2050s are usually re-marked 2020’s or 2030s. They work on lower voltage but blow up at full voltage. And there is no such thing as a real 2050 anymore.
And the mistake was putting in a bigger transformer and expecting more power.
Fake TDA2050s are usually re-marked 2020’s or 2030s. They work on lower voltage but blow up at full voltage. And there is no such thing as a real 2050 anymore.
And the mistake was putting in a bigger transformer and expecting more power.
UTC 2050 are available in Europe, licensed and genuine.
2030 is enough for most use.
1875 will also work.
Just buy another plate amp if this is expensive to fix.
Or put the old chips and transformer back, use it with better speakers.
You may have to replace other parts.
You seem to be confused between chips and transistors.
Can you be a bit more clear?
2030 is enough for most use.
1875 will also work.
Just buy another plate amp if this is expensive to fix.
Or put the old chips and transformer back, use it with better speakers.
You may have to replace other parts.
You seem to be confused between chips and transistors.
Can you be a bit more clear?
Last edited:
https://www.diyaudio.com/community/...ip-between-psu-and-speaker-resistance.384479/
https://www.diyaudio.com/community/threads/order-of-resistors.382638/ " I'm trying to do thins from free spare parts. I've put a used plate amp into a different cabinet. At higher volumes the sub pops and grunts."
https://www.diyaudio.com/community/threads/order-of-resistors.382638/ " I'm trying to do thins from free spare parts. I've put a used plate amp into a different cabinet. At higher volumes the sub pops and grunts."
Last edited:
10$ Recycled Sub
I get that this site is dominated by engineer types who seek perfection but sometimes the goal is improvement as opposed to perfection.
You can’t really improve a TDA2030 amplifier. Putting in 2050’s and upping the supply used to work until TDA2050 was discontinued. Now you get a ticking bomb. LM1875 will “work”, but tends to put out less watts than either 2030 or 2050, on equivalent supply volts.
Not transistors at all - chip amps. Paralleling those is more trouble than it’s worth.
Playing with low voltages, some class D chip is as good as anything else.
With 12 transistors per channel you can make something that sounds nice and you can’t burn up, but it won’t put out one milliwatt more power.
Playing with low voltages, some class D chip is as good as anything else.
With 12 transistors per channel you can make something that sounds nice and you can’t burn up, but it won’t put out one milliwatt more power.
I was burnt by a bunch of fake TDA2050s, visually perfect copies, bought from a supposedly "serious" distributor.
They overheat on their own, no signal applied, no load connected, so clearly they breakdown/avalanche at TDA2050 voltages (+/-22V) but work "reliably" (so far) at TDA2030 voltages (+/-14V) so I guess that´s what they have inside.
Not to junk them without fighting, I am using 2 of them bridged into 8 ohms, (at +/-14V rails) to roughly replace what I would have gotten with one true 2050 into 4 ohms, we are talking some 25-30W RMS, until I use all of them.
They overheat on their own, no signal applied, no load connected, so clearly they breakdown/avalanche at TDA2050 voltages (+/-22V) but work "reliably" (so far) at TDA2030 voltages (+/-14V) so I guess that´s what they have inside.
Not to junk them without fighting, I am using 2 of them bridged into 8 ohms, (at +/-14V rails) to roughly replace what I would have gotten with one true 2050 into 4 ohms, we are talking some 25-30W RMS, until I use all of them.
A lot to work through . . . Lots of unsolicited opinions about the TDA2050.
Again, from the top. I had this really bad plate / drop in amp from a computer 2.1 system. It was faulty. In addition to it being faulty the amp was so bad the power LED dimmed with the music when you turned the volume up. I figured one of the chips had blown. I decided to replace all the 2030s with 2050s, and to replace the 2 x 11v transformer with a bigger one 2 x 16v I had lying around. Save for the chips (4 for $6), the project hasn't cost me a penny. The two capacitors accompanying the rectifier are 3300uF @25v. I surmised the power supply would tolerate the increase in voltage.
Sidebar: the 'technical' people need to understand this exercise is like doing the crossword - it taxes the mind, but when I'm done with crossword I'll throw the paper away.
Please understand, there are basic concepts the I do not comprehend: 3 identical chips, output quoted as 10w + 5w + 5w. I'll assume the variation is governed by the input signal. At the point of failure the bass signal was being driven into 8 ohms and the satellites 6 ohms - (I believe) it was the chip driving the sub that failed. It could have been down to the failure of a crappy chip. I have a spare, I can replace that chip.
Again, from the top. I had this really bad plate / drop in amp from a computer 2.1 system. It was faulty. In addition to it being faulty the amp was so bad the power LED dimmed with the music when you turned the volume up. I figured one of the chips had blown. I decided to replace all the 2030s with 2050s, and to replace the 2 x 11v transformer with a bigger one 2 x 16v I had lying around. Save for the chips (4 for $6), the project hasn't cost me a penny. The two capacitors accompanying the rectifier are 3300uF @25v. I surmised the power supply would tolerate the increase in voltage.
Sidebar: the 'technical' people need to understand this exercise is like doing the crossword - it taxes the mind, but when I'm done with crossword I'll throw the paper away.
Please understand, there are basic concepts the I do not comprehend: 3 identical chips, output quoted as 10w + 5w + 5w. I'll assume the variation is governed by the input signal. At the point of failure the bass signal was being driven into 8 ohms and the satellites 6 ohms - (I believe) it was the chip driving the sub that failed. It could have been down to the failure of a crappy chip. I have a spare, I can replace that chip.
The sub is probably lower impedance than the satellites. Lower Z is more watts off the same power supply. If this isn’t the case, somebody is lying about power. 10 watts at 4 ohms and 5 at 8 seems reasonable. That would also make that chip run harder than the other two and hence easier to blow Up.
The unsolicited opinions about the 2050 are NOT without justification. If yours we’re REAL, they would NOT have blown up with the 16 volt transformer. Period. If you managed to blow a real one on that voltage, it would have run for many hours and gotten crazy hot if the heat sink was too small, not fail instantly. Instant fail means there is something wrong with it. You never could get real 2050’s at 4 for $6 so that should tell you something too.
The unsolicited opinions about the 2050 are NOT without justification. If yours we’re REAL, they would NOT have blown up with the 16 volt transformer. Period. If you managed to blow a real one on that voltage, it would have run for many hours and gotten crazy hot if the heat sink was too small, not fail instantly. Instant fail means there is something wrong with it. You never could get real 2050’s at 4 for $6 so that should tell you something too.
Really?
But YOU posted:
also that they exploded
YOU called this very thread:
and posted it in a public Forum, which means you are soliciting opinion and answers.
WTF?????