Marking almost always happens in assembly before test. Factories are confusing enough without having unmarked parts floating around in test. The exception is parts that are graded but even in this case the basic marking is usually done at assembly and the grade-out mark added after test.
Borninthefifties: I don't believe the concern about pirate parts is whether some spec is off due to them being final test rejects. The more serious concern is whether the die inside is actually an LM3886 die. There are several threads on this forum with scary pictures -- like the guy who bought power transistors but the die inside was a small-signal transistor -- blew up the first time he powered up his amp.
Borninthefifties: I don't believe the concern about pirate parts is whether some spec is off due to them being final test rejects. The more serious concern is whether the die inside is actually an LM3886 die. There are several threads on this forum with scary pictures -- like the guy who bought power transistors but the die inside was a small-signal transistor -- blew up the first time he powered up his amp.
While searching for chips I found this article. I could be wrong but counterfeit ICs may be the ones rejected by quality control. So virtually they are free to manufacturer.
I live in a small town. At the outskirts there are godowns for multinationals companies. My friend who has his own godown for printing purpose visited one godown where speakers from some company (Probably from the west. Not sure) were stored. There he came to know from owner that there were strict rules and directions that speakers that were damaged or not up to the standard were supposed to be destroyed. Same can be done to ICs which do not meet QC.
Regards
I live in a small town. At the outskirts there are godowns for multinationals companies. My friend who has his own godown for printing purpose visited one godown where speakers from some company (Probably from the west. Not sure) were stored. There he came to know from owner that there were strict rules and directions that speakers that were damaged or not up to the standard were supposed to be destroyed. Same can be done to ICs which do not meet QC.
Regards
Last edited:
To be honest, what would be the point? Of going to the considerable expense of manufacturing a fake LM3886 in this case, pin for pin compatible with the original and then selling it for pennies in a kit???? That just doesn't make any sense.
The thing is that the so called fake works, and to most peoples ears.... satisfactorily as well.
It is only in the advent of the "offer a competing product for sale: Neurochrome LM3886 Done Right" that everyone is suffering hysteria at the thought of a fake. Sure its got ultra low THD and looks a very fine product!
Can the majority of users appreciate that THD compared to the XY.....?????
This is a DIY forum , surely if people want to buy and modify the XY amp then take a punt..... its DIY cheap after all😕😕
Sure you can buy an XY PCB, I did.
I can buy 2 XY pcbs AND the big PSU pcb for ~$6 post included.
Then spend some time working out where the missing components need to go and cutting the Signal Return/ground to Power Ground bridging links to allow HBRR/HBRL to be added into a stereo amplifier.
I can buy 2 XY pcbs AND the big PSU pcb for ~$6 post included.
Then spend some time working out where the missing components need to go and cutting the Signal Return/ground to Power Ground bridging links to allow HBRR/HBRL to be added into a stereo amplifier.
Andrew, cutting the signal return/ground to power ground is not going to cut it. There is a better way to keep the power ground from signal ground. I will post pictures + explanation shortly. Together with somewhat larger decoupling caps (180 uF vs 100) and an appropriate wiring scheme, the XY pcb will get the manufacturers specification out of the LM3886. For 20 times less per pcb, compared to the DR.
I also look forward to the measurements. Now 20 times less means $2 per board. So $38 saving. That's about 20 minutes of my time per board before its more cost effective to go for one that works out the box. Just a data point for those who think board cost is everything.
But people's interests differ and some like to do it for the smallest $$$. Vive la difference. Choice is good. But don't bash something where as long as you use the BOM given you get the performance measured guaranteed.
But people's interests differ and some like to do it for the smallest $$$. Vive la difference. Choice is good. But don't bash something where as long as you use the BOM given you get the performance measured guaranteed.
2 x LM3886-DR plus a Power-86 pcb, including shipping, will only cost you $159.00.
What's the big deal. You just run a separate speaker return "ground" wire back to the PSU, bypassing the XY amp PCB completely. No need to "cut" anything. This will keep the speaker current out of the PCB ground trace. Make sure to run the supply wires for rail, PCB ground and speaker return ground as a tight bundle.
Ha ha will cost you only 159!!!!!?!?!? You are funny !
With this money you can buy a kit , I repeat , a kit with all the components and instructions from updatemydinaco.com, stereo amplifier based on lm3886 chip and stabilized power supply.
They are in business for a few good years and their kits are awesome , they even accept it back if you don't like it , even if u put it together.
I assume you meant Akitika Store
You are a factor of 2 out on prices.
EDIT: some unusual design choices there. wrapping the output inductor around the coupling cap is certainly novel. not sure about the single ended supply but I DO like choice.
You are a factor of 2 out on prices.
EDIT: some unusual design choices there. wrapping the output inductor around the coupling cap is certainly novel. not sure about the single ended supply but I DO like choice.
Last edited:
The people who sell fake ICs aren't just selling them to kit vendors. They're passing them on as real parts to unsuspecting customers, including big customers. If you're interested in the impact of fake parts, I suggest picking up one of the trade magazines (EDN, Printed Circuit Design & Fab, etc.) They run articles on the industry impact of fake parts on a regular basis.
My main selling point is that I deliver products that are well engineered, well designed, well implemented, well characterized, well documented, and well supported. I cater to those who appreciate these qualities. If you buy from me, you have the confidence that when you have assembled the product, it will work as designed and work as shown in the performance characterization measurements on my website. All this adds value, which is why I charge more than the eBay vendors.
For those, such as yourself, who are content with buying a product knowing that you have to hack the PCB with a knife, buy new components, etc. to get it to work well can buy on eBay.
There is no way I can compete with the eBay sellers on price, so I choose not to. I can, however, compete with them on quality. Those who appreciate a quality product buy from me. Those who want cheap can buy on eBay. I'm at peace with this.
In the better, faster, cheaper matrix, I'm solidly in the better category.
Absolutely. After all, the purpose of a hobby is to consume your time and money while providing some enjoyment, right.
I disagree that DIY has to be cheap (= of poor quality), though. To me DIY is about being able to build a quality product for a fraction of the cost of a comparable commercially available product. My Modulus-86 can be built for $400 including a chassis and competes well agains commercially available amps in the $5-10k range.
Are you cherrypicking your shipping option for effect there Mark? The boards are $130. Airmail shipping to Europe is $9. That's ... um, lemme see .... 1-3-0 plus 9 = $139.
Exactly!
Many people spend over $10/day at Starbucks without batting an eye. They have $100/month cellphone plans and $150-200/month cable TV bills. Would you be willing to drink drip coffee rather than fancy lattes for a week to buy my boards? Or how about giving up cable TV for a summer to get a complete stereo amp that competes in the $5-10k range? Would you?
That sums up my value proposition nicely. Thanks.
Tom
One word: Bingo!
The speaker return really does need to go to the local decoupling on the PCB. I've measured the difference in THD and it is substantial - in particular at frequencies above a few hundred Hz. You can find the data in the LM3886 P2P vs PCB thread.
Tom
The output inductor is neat from a manufacturing standpoint. I do wonder what sticking an aluminum can inside the inductor does to the linearity of the inductor, though.
AC coupled output wouldn't be my first choice. It does simplify the power supply design, though. That's particularly important in the Akitika amp as they're using a regulated supply to get decent PSRR.
Tom
Also for Korpberget (great Netflix series made in the place you live!), here it comes.
Before sketching the mod I made, lets first have a look at what can jerk around the signal ground. There are two forces at work.
The first is the assymetrical current drawn from the positive and negative rails. If ground is to be kept stable, matching currents will have to flow to and from the ground connection. Even with zero ohmic resistance, this is still impossible because of inductance. Conclusion, we have to cut the connection between the grounding point between the caps and signal ground.
The second current draw impacting on signal ground is the speaker return to ground. Two things about it. We cannot have it jerk around the signal ground, so it has to be isolated from it. But at the same time, it is also the reference level required for the feedback loop to determine the error signal. So, it has to be isolated from signal ground, but must be connected as the earth connection point to the 22uF cap that connects the feedback loop to ground. In short, it will look like this.
The line that was Dremmeld away disconnects the midpoint between the reservoir caps from the bottom ground plane. I am working on a 3 channel amp with 7 LM3886's per channel, most of them bridged. For bridged amps, this is al the mod you need. The small current drawn by the ground pin of the LM3886 is serviced through the shield of the input (does anyone know if it has an internal compensation cap connected to earth?).
For the high channel, I use a single 3886 and in this case further minor work has to be done. For you who are familiar with Bruno Putzey's ideas about 'following the ground' this is nothing new, but it may be a bit counterintuitive at first. Because it looks like introducing a ground loop, but it isn't. See the red lines. You have to cut all the connections between ground of the attachment point of the 22 uF cap, and connect it to the ground point of the speaker output terminal. Now you are done. The currents flowing through the earth connection of the midpoint between the reservoir caps will still impact on this point, as will the speaker return current, and so this introduces an error component. However, since it is inside the feedback loop, it will be corrected.
Now for the other mod I have seen:
This is not advisable. What happens is that you now create a reference point for the feedback loop that relates the output signal to signal ground. As we have seen above, the reference point for the feedback loop should be the earthing point for the return current of the speaker. The reason is that this speaker return earthing point is pushed around by two factors (reservoir caps and loudspeaker return), and so has to be included in the feedback loop. Read Bruno, he is a better tutor than I am.
Hope it jives and always prepared to answer questions. Just a word of caution. While doing all of this, please keep in mind that even in standard form, the XY board is highly unlikely to produce any audible distortion at all. So keep the wife happy and mow the lawn instead.
Last edited:
Does the $9 option include tracking? I thought at $130 it would be best to use an option with tracking. What happens if the boards don't arrive?
I'm pretty happy and consider my amp to work well. I think it's a bit of hyperbole to say a ebay kit won't work well.
It was a good kit for getting my feet wet - I didn't worry that it might not work or I would do something wrong because it was so cheap. I get going the other way and not wanting to spend time debugging and never knowing if what you got was performing like it should. It really depends on what you want/trying to achieve with the project. For me at the time, money was more of a concern.
The next LM3886 I build will most certainly be a kit from someone here.
Usually, if the seller doesn't primary sell non electronic stuff and just happens to have a kit, they are pretty good about providing a schematic. A couple times on aliexpress I have asked another seller for a schematic and they provided one - most of the time though the pcb silkscreen is enough though. I have also got support when things didn't work and the seller was able to help me out.
I had one bad case with a DAC kit someone bought for me where the silkscreen was useless and later when I got a schematic for it, it was clear that some values had been substituted. A lot of the components were clearly salvaged and of questionable quality (rusted leads, solder left over, etc)
There are some horror stories out there, but it's not all doom and gloom either. You just have to set your expectations and be ready to be burned with crap (though usually you can get a refund) no support, sketchy parts, etc. It's extra hassle and risk that a proper kit won't have.