Staggers me to even consider arguing with that life view .
Clearly then ALL amplifiers.... As well as 'Whiskys' are all the same 😀
When is the keyword.
The price of the xy is 1/10 of a perfect designed genuine part amp.
And people like the sound. And the measurements are pretty decent in my view.
(And, yes, my view are that all well designed amps that are not stressed, sound the same)
The whisky analogy is reading review of ib bottler whisky, and when a good batch comes on the market, you stack up. Then you stay away from the crap.
Buying fake parts supports an underground economy of fraudsters. From a moral standpoint I cannot support that. Others are free to disagree.
Tom
Tom
Absolutely!
Purveyors of fakes are parasites who thrive off of the intellectual property of others while simultaneously stifling innovation. To knowingly buy fake parts is tantamount to colluding in theft, making one just as guilty as the fraudster.
Clive
Why publish the results of a kit with fake parts? Are you not actually helping the fraudsters?
I can't see how it helps a fraudsters? Tom's kit is superior in every respect and in a view of the total price of the amplifier (power supply, chassis, etc) saving a few dollars is not worth buying a inferior product.
tomchr made a great amplifier for a modest price and as a bonus we are getting for free his deep knowledge about amps here.
tomchr made a great amplifier for a modest price and as a bonus we are getting for free his deep knowledge about amps here.
Last edited:
When someone like Tom writes "The PCB layout is decent" and posts not so bad results. I see that as promoting the product.
I didn't say the layout was good. Nor did I recommend the XY kit. I said the layout was decent. As in, not downright horrible, could be improved. That's not an endorsement.
The measurements show significant deviation from the data sheet performance of the LM3886 above a few watt of output power. That's pretty horrid in my book. Meeting the data sheet specs is not hard but it does take some paying attention and engineering skill.
The XY kit is a mediocre product that uses fake parts and offers mediocre performance. Does that really sound like an endorsement to you?
Tom
The measurements show significant deviation from the data sheet performance of the LM3886 above a few watt of output power. That's pretty horrid in my book. Meeting the data sheet specs is not hard but it does take some paying attention and engineering skill.
The XY kit is a mediocre product that uses fake parts and offers mediocre performance. Does that really sound like an endorsement to you?
Tom
From a moral view point, what is the distinction between someone purchasing fake parts to build an amp and someone purchasing fake parts for any other purpose?
If someone is convinced that the parts are fake then they should put in a Paypal claim and not use them.
If someone is convinced that the parts are fake then they should put in a Paypal claim and not use them.
Ebay and PayPal are effectively giving us a warranty for misdescribed product and that includes fake product.
If we repeatedly claim via Ebay and PayPal when we have evidence of fake product then eventually Ebay and PayPal will put the offending retailers out of business.
What we should not do is just let the fakes exist. Complain.
Tomchr when he realised that there were some fake components should have started that complaint. He should never have assembled and tested and reported on fake components. That is not a service we require, nor benefit from.
We can buy XY PCBs. two + a PSU for ~$6
Tomchr sells amp PCB.
The real test would have been to compare the two PCB's performance using the same quality components in both builds.
That way Tomchr's expertise would show in the improved performance.
I agree 100% with this post. I will add my 2 cents as well.
One could argue that comparing the XY board with Tom's product is like comparing apples to oranges. The XY board works fine but is not designed for getting the highest performance out of the chip. It's just a run of the mill commercial product out of China. On the other hand, Tom's board was developed using an iterative process, SPICE simulation, and uses additional circuits not found on the XY board. Is it any real surprise that it is better in all respects? Sure, and it costs a little bit more, too (as it should).
The XY board makes a great "intro" amp kit and is cheap. A serious amp geek is not going to bother with it, but for the masses and those DIYers who just want to get their hands dirty building something the XY kit (2 amp boards and PS) sure hits the spot. Is it any surprise that it is all over Ebay? If these things were failing and/or oscillating don't you think that demand would drop and sellers would stop carrying it? That's hardly the case.
There is a thread in this forum where modification to the ubiquitous XY board are described. What I would like to see is a comparison of the modified XY board, populated with genuine parts, to Tom's product. Both should be powered from the same PS. Then I think we would be doing a bit more of an apples-to-apples comparison.
In my view, any LM3886 circuit that does not deliver the data sheet performance of the LM3886 is a failure. The XY kit does not deliver the data sheet performance.
It is easy in retrospect to throw me under the bus for not having complained or addressed my suspicion of fake parts with the vendor/eBay/Paypal. That's called hindsight bias. You all know as well as I do that to measure the performance of the XY kit you have to assemble it.
As I've said 200 times now, I will measure the performance of the XY board with a real LM3886 in it. I will include an LM3886 and 0.15 Ω output resistor with my next Mouser/Digikey order. I forgot to include it with the order I just received (grr!). Also, paid projects come before charity work. That's business reality. My apologies for living in the real world and requiring to eat periodically.
Tom
It is easy in retrospect to throw me under the bus for not having complained or addressed my suspicion of fake parts with the vendor/eBay/Paypal. That's called hindsight bias. You all know as well as I do that to measure the performance of the XY kit you have to assemble it.
As I've said 200 times now, I will measure the performance of the XY board with a real LM3886 in it. I will include an LM3886 and 0.15 Ω output resistor with my next Mouser/Digikey order. I forgot to include it with the order I just received (grr!). Also, paid projects come before charity work. That's business reality. My apologies for living in the real world and requiring to eat periodically.
Tom
Interesting test. I wonder if the die is authentic but is being packaged by a 3rd party? I don't know if that would save costs. The TDA2040/2050 chips I've opened up always have a much smaller die with the counterfeit parts. At a given supply voltage, the maximum power output before clipping would be about 20% lower due to the weaker output devices in the fakes. The 3886 on this XY kit does seem to be performing like an authentic part (just on a marginal board).
Tom, I too run a small business doing custom fabrication and very small production runs (non-electronics items). Every so often you get the person bitching about why it costs so much. I tell them I have machines to pay for and service, taxes, materials and bills to pay like any other person - not to mention I'd like to make some money as well. These types will say, "Okay, I'll get back to you." and they never do.
Tom, I too run a small business doing custom fabrication and very small production runs (non-electronics items). Every so often you get the person bitching about why it costs so much. I tell them I have machines to pay for and service, taxes, materials and bills to pay like any other person - not to mention I'd like to make some money as well. These types will say, "Okay, I'll get back to you." and they never do.
John, I agree on the business comments. Some people expect to get something for nothing and will remain disappointed until the come to terms with reality. Such is life.
Now for the FAKE vs REAL part of the story:
I had a buddy of mine at TI look at the part. The part was examined externally, via X-ray, and by inspection of the die after decapsulation.
The lot code is NOT a TI or National Semiconductor lot code. The markings on the package are fake, so it is now a confirmed fake part. Interestingly, there aren't any marks from removing the original laser inscribed product code and part number, so it's a bit unclear what the origin of this part is.
The die is marked LM3886A as it would be according to National's die marking standard. The guy who did the analysis noted that the buried layers on the transistors on die appeared to be misaligned relative to the rest of the layers.
There are probe marks on the bond pads, including those that aren't bonded out (used for testing). This indicates that the die has been through wafer sort. The package matches the package used by TI.
All this points to a part that failed final test and somehow made it into the outside world. A competing theory is that the part was stolen from distributor stock and rebadged to disguise its origin.
I'll drop a real, known good, LM3886 into the XY board when I get a chance. I suspect as you mention that the fake LM3886 performed well enough that the differences I measure are due to the XY board and circuit and not the fake LM3886.
Do note that just because this one fake "LM3886" contained a real LM3886 die doesn't mean that another fake "LM3886", even one bought from XY, will contain a real LM3886 die as well. If you want a real LM3886, please purchase one through an authorized distributor (Mouser, Digikey, Newark/Farnell, RS Components, etc.)
Tom
Now for the FAKE vs REAL part of the story:
I had a buddy of mine at TI look at the part. The part was examined externally, via X-ray, and by inspection of the die after decapsulation.
The lot code is NOT a TI or National Semiconductor lot code. The markings on the package are fake, so it is now a confirmed fake part. Interestingly, there aren't any marks from removing the original laser inscribed product code and part number, so it's a bit unclear what the origin of this part is.
The die is marked LM3886A as it would be according to National's die marking standard. The guy who did the analysis noted that the buried layers on the transistors on die appeared to be misaligned relative to the rest of the layers.
There are probe marks on the bond pads, including those that aren't bonded out (used for testing). This indicates that the die has been through wafer sort. The package matches the package used by TI.
All this points to a part that failed final test and somehow made it into the outside world. A competing theory is that the part was stolen from distributor stock and rebadged to disguise its origin.
I'll drop a real, known good, LM3886 into the XY board when I get a chance. I suspect as you mention that the fake LM3886 performed well enough that the differences I measure are due to the XY board and circuit and not the fake LM3886.
Do note that just because this one fake "LM3886" contained a real LM3886 die doesn't mean that another fake "LM3886", even one bought from XY, will contain a real LM3886 die as well. If you want a real LM3886, please purchase one through an authorized distributor (Mouser, Digikey, Newark/Farnell, RS Components, etc.)
Tom
Poor Tom,
He just wanted to illustrate the failings of a mediocre board using bootleg components for the education of the DIY community, and he gets a lot of grief. The old adage is true "No good deed goes unpunished".
He just wanted to illustrate the failings of a mediocre board using bootleg components for the education of the DIY community, and he gets a lot of grief. The old adage is true "No good deed goes unpunished".
OK, Wonder if buing a rejected part is regarded as buying a fake, legaly. Buying a stolen part is of course something else.
The interesting thing with this part is how cheap an amp can be and still sound good. Stolen goods tales away the fun. (Still I think they are stolen from the thrash bin. So the ethics is more grey)
Again, tanks for your time and interesting information.
I don't think the two products competes in the same market. Like comparing a fake and a real diamond. Most people won't see the difference. But, they do not regard it as the same product....
The interesting thing with this part is how cheap an amp can be and still sound good. Stolen goods tales away the fun. (Still I think they are stolen from the thrash bin. So the ethics is more grey)
Again, tanks for your time and interesting information.
I don't think the two products competes in the same market. Like comparing a fake and a real diamond. Most people won't see the difference. But, they do not regard it as the same product....
Tom,
Thank you much for your efforts. Is it possible for you to get the die sizes for authentic LM1875 and 3886 chips? I'd appreciate that greatly.
Thank you much for your efforts. Is it possible for you to get the die sizes for authentic LM1875 and 3886 chips? I'd appreciate that greatly.
You're welcome. I doubt I can get the die dimensions from TI. I can ask.
The only kind of LM3886 that I would consider to be a real, genuine LM3886 is one that comes through the TI distribution chain. I work hard to make money and want to ensure that I spend it on parts that are guaranteed to meet the specs.
That's a pretty good analogy, actually.
Tom