Ok, time for my view on this one. The OP isn't trying to sell ten million of this design and get filthy rich. Maybe he will, but I doubt it. The point of this offering is to provide a product a price level where the inventor can provide the following:
Tech support and expertise from a single inventor being provided abundantly.
Advice on how to ensure authenticity of the complete package, even if not sourced from the OP.
Inside information and additional OEM supply chain management history and definite operating practice detail and clarification provided publicly, pro Bono!
Additional diy-centric help, debate, advice, hand-holding, etc.
If you're looking for a Walmart or Best buy grade diy jobber, go ebaying or sign on with the forty thieves. Quit trying to defend the lowest common denominator and pony up the dinero to properly diy! It's one thing to stand on Goliath's shoulders, quite another to stand behind little ol' David.
My hat is off to the OP for continuing to endure and remain throughout the pokes, jabs, and stabs of the 99 percent.
Tech support and expertise from a single inventor being provided abundantly.
Advice on how to ensure authenticity of the complete package, even if not sourced from the OP.
Inside information and additional OEM supply chain management history and definite operating practice detail and clarification provided publicly, pro Bono!
Additional diy-centric help, debate, advice, hand-holding, etc.
If you're looking for a Walmart or Best buy grade diy jobber, go ebaying or sign on with the forty thieves. Quit trying to defend the lowest common denominator and pony up the dinero to properly diy! It's one thing to stand on Goliath's shoulders, quite another to stand behind little ol' David.
My hat is off to the OP for continuing to endure and remain throughout the pokes, jabs, and stabs of the 99 percent.
Last edited:
If getting rich was my ultimate goal in life, I wouldn't cater to the DIY market. It's an incredibly small market and few people within this already limited market are willing to spend much money. I would make a 5-6x as much if I worked in an engineering job (but I'm having 10x as much fun now than I was in the engineering job I left).
I would like to make something that resembles an engineering income, so I'm now branching into retail products (Tom Christiansen Audio: Hear the Music – Not the Amp. World Class Audio by Tom Christiansen.). I plan to keep both the DIY and the retail brands going.
Well put. Thank you. I appreciate it.
Tom
First of all, please excuse me if my question is very primary. Despite having assembled some amplifiers, I only took a chance with a much simpler chip than the LM3886. I've actually been using the NXP's TDA2616. Well, let's go to the questions: I have found schemes where the signal input goes through the following path:
POT -> C1 -> R1 -> pin 10
From the point between C1 and R1, R2 leaves for the ground. The values that have been common are:
POT - 10k ~ 50k
C1 - 2u2 ~ 4u7 and up to 10uF
R1 - 1k in almost all. But I already met 470ohm.
R2 - 10k ~ 22k. I believe that this resistor will determine the input impedance of the chip.
As I have 2 PCBs of these XY, I would like to use them. But R2 does not leave the point between C1 and R1 but connects AFTER R1. In practice R2 is directly between pin 10 and ground. Is the difference between most schemes and what XY does too big? Would it be better not to connect R2 on the board, but connect this resistor from the outside, directly between the ground and the point between C1 and R1?
Another doubt that I have is that I recalculated the components to obtain a very low output power in relation to what I have found out there. I would like to use the LM3886 because of its low distortion (well below the TDA2616) but I intend to use it with 15w @ 4ohm speakers. It was with surprise that I realized, if I didn't make a mistake in the calculations, that the only changes resulted in a lower voltage source and much smaller sinks. That is, in the filters of the sound path and in the feedback for the gain little has changed in relation to the schemes for 40w of output or more. Is it okay to supply the LM3886 with a 13v-0v-13v power source (which results in approx. 18v-0v-18v, after rectification and filtering)? Despite being within the parameters of the datasheet, does it not affect the sound output in feeding practices with such a low voltage?
Thank you.
POT -> C1 -> R1 -> pin 10
From the point between C1 and R1, R2 leaves for the ground. The values that have been common are:
POT - 10k ~ 50k
C1 - 2u2 ~ 4u7 and up to 10uF
R1 - 1k in almost all. But I already met 470ohm.
R2 - 10k ~ 22k. I believe that this resistor will determine the input impedance of the chip.
As I have 2 PCBs of these XY, I would like to use them. But R2 does not leave the point between C1 and R1 but connects AFTER R1. In practice R2 is directly between pin 10 and ground. Is the difference between most schemes and what XY does too big? Would it be better not to connect R2 on the board, but connect this resistor from the outside, directly between the ground and the point between C1 and R1?
Another doubt that I have is that I recalculated the components to obtain a very low output power in relation to what I have found out there. I would like to use the LM3886 because of its low distortion (well below the TDA2616) but I intend to use it with 15w @ 4ohm speakers. It was with surprise that I realized, if I didn't make a mistake in the calculations, that the only changes resulted in a lower voltage source and much smaller sinks. That is, in the filters of the sound path and in the feedback for the gain little has changed in relation to the schemes for 40w of output or more. Is it okay to supply the LM3886 with a 13v-0v-13v power source (which results in approx. 18v-0v-18v, after rectification and filtering)? Despite being within the parameters of the datasheet, does it not affect the sound output in feeding practices with such a low voltage?
Thank you.
Thanks. That's what I took from the discussions as well. My question is a more general one, whether no real 3886 would have a "PM" marking because simply no factory/subcontractor in China could/is allowed to make the chip. Asking because having given away a few 3886s to a friend, he came back recently telling me that one had a "PM" marking, which he decided to throw away before his house gets burned down (a joke of course). I only ordered all my 3886 kits from the old chipamp.com quite a few years ago (maybe even a decade ago), though I am not sure whether it was before or after the ownership change.
Nothing substantial here, only curiosity.
Btw, I am seriously considering your Modulus-86, though I am also tempted to build a STA540 first simply because I have a spare linear regulated power supply and all the passive parts -- never heard a STA or TDA chip amp before, and so again curiosity is at work.
From what I vaguely recall, the PM marked parts were deemed fake.
I always order parts through authorized distributors to minimize the risk of receiving fake parts. I will occasionally order directly from TI - especially if I need a reasonable amount of ICs (like 100+). They have better prices than Mouser, charge $7 flat-rate shipping, and will happily ship you one IC. TI even accepts paypal for payment.
Tom
I always order parts through authorized distributors to minimize the risk of receiving fake parts. I will occasionally order directly from TI - especially if I need a reasonable amount of ICs (like 100+). They have better prices than Mouser, charge $7 flat-rate shipping, and will happily ship you one IC. TI even accepts paypal for payment.
Tom
Thanks Tom for the great set of measurements. That was helpful indeed.
A couple of years ago I have built four channels based on XY boards for a 2-way system and left that project unfinished.
Now, encouraged with your measurements I have measured my XY boards and was very happy with rhe results. I have used the first class components and genuine LM3886 chips. In addition to that I have purchased last week additional 4 LM3886T and plan to expand the project to a 3-way system. Fortunately, one of my reliable suppliers, Reichelt, had them on stock. I didn't hesitate for a moment - the number of available chips was limited.
I mean, why should one want an amplifier that messures better at the 6th significant digit if the difference in meassured parameter has no real impact in the real world?
Regards
Tom, actually I aint no miser and I'm not looking for a "Hi-Fi" amplifier at the price of breakfast. Currently I am building both Wolverine and Alpha Nirvana and both are considered as expensive. I am building 4 channels each. This will cost me quite a few thousands but I don't give a stuff. My average monthly expenditures just on this hobby are in average 250 EUR.
Like you I'm a professional in technical sciences and I earn by selling know-how. A long time ago I've decided to abandon low-cost projects and focus exclusively on wealthy customers. In this department we differ.
However, serious projects require substantial savings of the customers money and this means you have to trade-of prudently costs vs. benefits when dealing with design alternatives. That's the essence of good designer, otherwise, knowledge is worthless.
It is common knowledge that marginal benefits of alternative design can not compensate for higher price. This is the case here: I really don't care whether the THD values differ at the 3rd significant digit. In the real world that makes no difference whatsoever.
That is why I have decided to finish that XY project. At modest pricing by the modest but diligent and capable designer. I do indeed appreciate modesty, when appropriate.
Regards
Last edited:
And that's your choice, which you are free to make.
I believe a circuit board for a particular part should at least reproduce the data sheet performance of the part. The XY board doesn't do that. If you want to leave performance on the table that's entirely up to you. Maybe the additional performance isn't meaningful to you, but it is to many. They have the right to their opinions and choices too.
Tom
I believe a circuit board for a particular part should at least reproduce the data sheet performance of the part. The XY board doesn't do that. If you want to leave performance on the table that's entirely up to you. Maybe the additional performance isn't meaningful to you, but it is to many. They have the right to their opinions and choices too.
Tom
Indeed, but its even worse: he is selling his variant by spreading negative information about the XY board which is otherwise well established and highly regarded amplifier. XY board meassures very good and sounds great when built with genuine components. This is a clear example of unfair competition.
This is my last post in this thread because I don't want enable thread-bumping tactics in this thread.
NB: this thread drags on for 6 years exactly because of thread-bumping. And now, it isn't anymore about the XY, it is exclusively about Toms product.
Bye
Last edited:
Then use the (!)Report button to bring it to the attention of a Moderator. Before it bumps again.
I really don't understand where you're coming from with that comment. Anybody can look at the measurements on the first page of this thread and draw their own conclusions. Unless you're somehow arguing that the Audio Precision APx525 I used to measure the XY board and compare it with my LM3886DR is somehow biased and will measure my circuits more favourably than others.
Is measuring and publishing the performance of a circuit "selling by spreading negative information"? I don't need to spread negative information about the XY board. I think its performance, measured objectively, speaks for itself.
My comments regarding the lack of decoupling on the XY board are based on the application section of the LM3886 data sheet, simulation, and my roughly one decade of experience in the lab with the LM3886.
And, ironically, you gave it more life with your post.
Tom