Here is a list of all the 5pin TO-220 power op amp ICs I'm aware of. If you know of any others, please add to the list. All but the last four are obsolete parts and would likely be counterfeit if purchased from a Chinese vender and probably other suppliers unless it is honest NOS.
tda2002
tda2003
tda2006
tda2008
tda2030
tda2030A
tda2040
tda2050
tda2051
UPC1238
LM383
L165
lm675
lm1875
OPA544
tda2002
tda2003
tda2006
tda2008
tda2030
tda2030A
tda2040
tda2050
tda2051
UPC1238
LM383
L165
lm675
lm1875
OPA544
I just wish they were marked reasonably!
But, there's no need to buy the fakes on purpose, because here's some non-fake alternatives:
LM1875's baby sister, the LM675 was made for low gain, has about 2/3rds the power rating and deals out some fun with the biggie size presentation and good dynamics of lifelike audio (after a bit fine tuning, of course). The LM675 is slightly more sturdy AND has slightly higher voltage tolerance than clone/fake/remark 5pin chips and is easier to use because of less guesswork. However, the usage is very similar for when super-low gain and low voltage meet to the end result of flea power fun with big imaging.
At less power and a bit less techy performance than the LM675 is an alternative of the TDA2009 (a 2x TDA2003), providing some fun for Singleton fans, in chip amp form. Mainly what a singleton does is NOT sound like an LTP, meaning that the harmonics are a little different. Not mistakenly rejecting some useful audio signal comes at the cost of not rejecting power noise very well; so, the TDA2009, like other singleton input amplifiers, will be needing a bit cleaner power (such as tracking-pre-regulator, a fine quality supply doable with chips at low cost).
.
I sometimes wonder about the whole counterfeit parts idea. Surely Mouser and other major suppliers are aware of the possibility of counterfeits, so they handle only parts that are traceable to the factory.
This leaves eBay and such as outlets. At which point, as Hearinspace says, how much profit can there be?
.
I sometimes wonder about the whole counterfeit parts idea. Surely Mouser and other major suppliers are aware of the possibility of counterfeits, so they handle only parts that are traceable to the factory.
This leaves eBay and such as outlets. At which point, as Hearinspace says, how much profit can there be?
.
Last edited:
First of all, sorry for the necro.
So, I came across these very same "TDA2050" that look exactly the same as the one you mention, down to the silkscreen. What's interesting is that the backside looks quite the same to the real TDA2050, middle one, what doesn't is the metal tab which is rectangular and doesn't have these little indentations. Anyway, they're what you describe as such in the quoted post, and I doubt they're the real thing.
A google search for "NEC UPC 10W pentawatt" hasn't turned up anything useful so I'm necro'ing to ask if you could shed some light on what would be the appropiate voltage for these parts? Such information would be useful to have on this thread since it shows up on a google search... I suppose a 12+12V 1A CT transformer would be appropriate for these?
I mean, I've already bought them for cheap to make a replacement amp for one of those cheap edifier 2.1 setups (R133T in this case) which use a pair of YG2025/TEA2025 chips for horrible sound quality, might as well go ahead and make them work comfortably and hope they make those speakers sound better than what the crappy factory amp can do. I already have other amps built with real parts (LM3886 and the like) but they're not worthy of such speakers
Thanks.
So, I came across these very same "TDA2050" that look exactly the same as the one you mention, down to the silkscreen. What's interesting is that the backside looks quite the same to the real TDA2050, middle one, what doesn't is the metal tab which is rectangular and doesn't have these little indentations. Anyway, they're what you describe as such in the quoted post, and I doubt they're the real thing.
A google search for "NEC UPC 10W pentawatt" hasn't turned up anything useful so I'm necro'ing to ask if you could shed some light on what would be the appropiate voltage for these parts? Such information would be useful to have on this thread since it shows up on a google search... I suppose a 12+12V 1A CT transformer would be appropriate for these?
I mean, I've already bought them for cheap to make a replacement amp for one of those cheap edifier 2.1 setups (R133T in this case) which use a pair of YG2025/TEA2025 chips for horrible sound quality, might as well go ahead and make them work comfortably and hope they make those speakers sound better than what the crappy factory amp can do. I already have other amps built with real parts (LM3886 and the like) but they're not worthy of such speakers
Thanks.
Vodka, The only 5 pin TO-220 NEC IC I'm aware of is the upc1238 unless it is a die from some other chip. What they stuff inside can change. Early fakes I bought had no short circuit protection. Every single one would blow the very instant I touched the output leads together while playing a test tone. Authentic chips would never blow. The new fakes had short protection but the output swing was not up to the authentic parts.
Could be a TDA2030 or TDA2030A re-mark - these are still in production, though on the way out, while the TDA2050 is EOLed. Check the profile of the pins - they tend to be wider on genuine ST parts where they enter the plastic package, like the middle one above.
Yup, or maybe 12-0-12, 3A if you have a stereo setup.
Sorry for the crappy pictures, I only had my phone's camera at hand. Same IC different lighting
Picture 1 | Picture 2
The pins' width is the same up to where they meet the plastic package. So, that gives it away, these are not genuine parts. As you can see, in comparison to the suspicious TDA2050 above it has the same silkscreen and body with the difference being the backside, which looks quite similar to the real 2050, but if you look closely you'll see the plastic isn't grainy like the 2050 and 1875, it's shiny and smooth like the suspicious one. For comparison's sake my spare LM3886's backside have the same grainy plastic as the real 2050/1875.
Oh well, I'll treat these as if they were 2030s, use low voltage and if they don't blow up in a test PCB I'll be extra careful when connecting and disconnecting speakers/test leads on the finished amp.
Thank you very much!
Picture 1 | Picture 2
The pins' width is the same up to where they meet the plastic package. So, that gives it away, these are not genuine parts. As you can see, in comparison to the suspicious TDA2050 above it has the same silkscreen and body with the difference being the backside, which looks quite similar to the real 2050, but if you look closely you'll see the plastic isn't grainy like the 2050 and 1875, it's shiny and smooth like the suspicious one. For comparison's sake my spare LM3886's backside have the same grainy plastic as the real 2050/1875.
Oh well, I'll treat these as if they were 2030s, use low voltage and if they don't blow up in a test PCB I'll be extra careful when connecting and disconnecting speakers/test leads on the finished amp.
Thank you very much!
I've used up to 12+12vac transformer (or less) for the power supply of the fake 5-pin chips,
or up to 14+14vac transformer for the fake TDA7294's.
Be sure to customize the gain for sound quality when undervolting--that can be rewarding!
See related project:
https://www.boxedkitamps.com/gobo-stereo-audio-amp-kit
or up to 14+14vac transformer for the fake TDA7294's.
Be sure to customize the gain for sound quality when undervolting--that can be rewarding!
See related project:
https://www.boxedkitamps.com/gobo-stereo-audio-amp-kit
Hey all. I was able to get some genuine LM1875s for this little amplifier, and turned out just fine. Great quality for bedroom listening with small speakers... that 12v-0-12v 3A transformer I'd bought with the TDAs in mind turned out to be more than enough for the task. Naturally, it has nothing to do against my LM3886 based amplifier for the bigger bookshelf speakers at the living room, but it holds its own even with the limitations it has. Pleasantly surprised.
Anyway, I remembered I'd bought these fake TDAs and decided to crack one open:
Picture
That die is even smaller than the fake's one in the OP, and seems like 1/4 the size of a genuine TDA2050. I do not believe that is a result of a more advanced, smaller process node; just a cheaper, even less capable IC for the job... I wouldn't even want these powering a portable radio speaker. I suppose a 4 ohm load would let the magic smoke out of these quite fast!
Anyway, I remembered I'd bought these fake TDAs and decided to crack one open:
Picture
That die is even smaller than the fake's one in the OP, and seems like 1/4 the size of a genuine TDA2050. I do not believe that is a result of a more advanced, smaller process node; just a cheaper, even less capable IC for the job... I wouldn't even want these powering a portable radio speaker. I suppose a 4 ohm load would let the magic smoke out of these quite fast!
Check this video out, especially the section from about 9 minutes in.
https://www.youtube.com/watch?v=35jWSQXku74
DUG's package details are quite useful here.
15.1mm < L6 < 15.8mm
6.00mm < L7 < 6.60mm
(H2 = H3) < 10.40mm
Let's take the midpoint for these ranges, and assume L6 measures 15.45mm, L7 measures 6.30mm. L6-L7 = 9.15mm which is the plastic part's length in the package, where the die is mounted. Using a ruler I measure 10mm for H2=H3, and 9mm for L6-L7. Good enough approximation, then.
Let's input these measurements into Photoshop's vanishing point:
Picture 1
Picture 2
Picture 3
10mm H2=H3 results in 9.05mm calculated for L6-L7. Excellent. Measuring the die itself, we get 2.07mm by 1.81mm, which amounts to 3.7467mm² ~ 3.75mm². Absolutely tiny. The original, real TDA2050 die looks gigantic in comparison. I think my 4 times smaller eye estimate is accurate enough.
15.1mm < L6 < 15.8mm
6.00mm < L7 < 6.60mm
(H2 = H3) < 10.40mm
Let's take the midpoint for these ranges, and assume L6 measures 15.45mm, L7 measures 6.30mm. L6-L7 = 9.15mm which is the plastic part's length in the package, where the die is mounted. Using a ruler I measure 10mm for H2=H3, and 9mm for L6-L7. Good enough approximation, then.
Let's input these measurements into Photoshop's vanishing point:
Picture 1
Picture 2
Picture 3
10mm H2=H3 results in 9.05mm calculated for L6-L7. Excellent. Measuring the die itself, we get 2.07mm by 1.81mm, which amounts to 3.7467mm² ~ 3.75mm². Absolutely tiny. The original, real TDA2050 die looks gigantic in comparison. I think my 4 times smaller eye estimate is accurate enough.
Am I detecting a crackpot conspiracy theory?
Perhaps it's just the further development in chip miniaturization and the need to keep the physical dimensions the same?
No. The chip on the left measured lower output and had no short circuit protection. Chip miniaturization is great for computers, but power transistors still need area for current and thermal dissipation. I measured these two ICs with my micrometer. The counterfeit one on the left is 2.23x1.80mm. The authentic TDA2040 IC on the right is 3.18x2.59mm which is about twice the area. The authentic one also uses double bond leads on the power and output pads.
Last edited:
Member
Joined 2009
Paid Member
Member
Joined 2009
Paid Member
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.