Neurochrome LM3886DR Build

TI list the lm3886 as a Class D amp (link),. Would it be very difficult to explain how it "turns into" a class AB in this application?

The LM3886 is and has always been a Class AB amplifier. This is clearly a typo. I'll get my buddies at TI to look into it.

The folks who update the product databases at TI must have really fat fingers these days. About a month ago the LME49600 went up to $50/each ($32 in QTY 1000 pricing). That turned out to be a typo too and the price has now been corrected. Normal price is $5-6/each.

Tom
 
Last edited:
I've had the LM3886DR amp for a few month and I'm really happy with it. Thanks, Tom for a great design!

Now I have to revisit this thread because I got some more questions:

1) Earlier on I mentioned that there was some audible noise from the amp. The bad news: it's back. The good news: I think i know what increases or decreases the noise.

amp on, nothing attached to the input: very faint noise
amp on, phono cable plugged in (no source): same noise but louder
amp on, phono pre plugged in and turned on: even more noise.

I do know that there's some grounding issue with the phono pre, but why does the noise increase just by plugging the cables in (again, no source, no signal).

I built myself a shorting plug and if I plug it in it's suddenly dead quiet. Gathering from what Tom wrote earlier i suspect that all of this would mean that there's some kind of ground loop. But how do i get rid of it?

2.) second question - much easier i suppose. the input impedance of the LM3886dr is 47k yet it seem that it works fine with a 50k pot. shouldn't the impedance of the load (the impedance that the amp "sees") always be much lower that the amp's own impedance?

thanks,
Michi
 
I've had mains transformers fail. I haven't personally experienced an SMPS failure. See. This is the problem with basing design decisions on personal anecdotes. You don't get the full data set.

Considering the number of switching power supplies in service world wide, you will see some failures even if the failure rate is in the ppm. I think blanket statements like SMPS = BAD are based on emotions rather than facts.

The SMPS modules I use in my SMPS-86 are well designed and very rugged. They include protection against various fault conditions, including over-current and over-voltage. On top of that, I added surge protection and inrush current limiting.

Tom
I agree. Most all of my amps use SMPS PSUs and I've never had one fail. They are of very solid design with all the protections you mention. Generally, the SMPS costs more than the amp modules I'm building.
 
amp on, nothing attached to the input: very faint noise

That's the noise of the 475 kΩ on the "outside" of the AC coupling cap in parallel with the 47.5 kΩ on the "inside" of the AC coupling cap.

amp on, phono cable plugged in (no source): same noise but louder
amp on, phono pre plugged in and turned on: even more noise.

That's the noise of the phono stage, amplified.

I do know that there's some grounding issue with the phono pre, but why does the noise increase just by plugging the cables in (again, no source, no signal).

Because the noise current of the LM3886 flows out the input of the amp and into the output impedance of the phono stage, which is probably pretty high when the phone stage is off.

I built myself a shorting plug and if I plug it in it's suddenly dead quiet.

That's because the shorting plug shorts out both the noise current and the thermal noise of the 475k||47.5k.

Gathering from what Tom wrote earlier i suspect that all of this would mean that there's some kind of ground loop. But how do i get rid of it?

It sounds more like a noise phono stage to me.

2.) second question - much easier i suppose. the input impedance of the LM3886dr is 47k yet it seem that it works fine with a 50k pot. shouldn't the impedance of the load (the impedance that the amp "sees") always be much lower that the amp's own impedance?

We need to agree on who's the load here. The input impedance of the amp is the load on the volume pot. So you have a 50 kΩ pot loaded by the 47 kΩ input impedance. That's a somewhat heavy load for a 50 kΩ pot, but as you point out, it works fine.

Tom
 
I've really enjoyed reading this thread, so much so that I've ordered and received 2 LM3886DR boards and a Power 86 board from Tom and am almost ready to start the build.

I've a (very) small amount of prior electronics experience but I'm basically a novice. It will be 100% to Tom’s design - I don't anticipate any tweaking, I wouldn't know where to even start.

Happy to document the journey if there is an appetite.

Neil
 
I've really enjoyed reading this thread, so much so that I've ordered and received 2 LM3886DR boards and a Power 86 board from Tom and am almost ready to start the build.

I've a (very) small amount of prior electronics experience but I'm basically a novice. It will be 100% to Tom’s design - I don't anticipate any tweaking, I wouldn't know where to even start.

Happy to document the journey if there is an appetite.

Neil

Sweet! You won't be sorry. Tom offers the best build support around. Thanks for sharing your build.
 
So, I've got almost everything. Unfortunately, in the BOM (handily linked from the documentation), there were a couple of parts out of stock with Mouser. They should be here in a couple of weeks or so, so I'll continue as far as I can regardless. If I have to put the electronics on hold, I'll start on the chassis (which also arrived today).

As I mentioned previously, I'm pairing the LM3886DR boards with a Power-86 and that is where I started this evening. This is the progress so far...
 

Attachments

  • 20180522_213308.jpg
    20180522_213308.jpg
    1.1 MB · Views: 382
Neila,

Thanks for documenting your build. I think what people don’t realize is there are a lot of non-nested feedback implementations of the LM3886 both diy and commercial but none that I know of have been brutally and rigorously measured, i.e. put to the test like Tom has of his own design, the LM3886DR. You know EXACTLY what you are getting with his designs. That isn’t true of the plethora of other LM3886 designs especially here on diyaudio.

For a first build it’s an EXCELLENT choice and frankly I am surprised that it is not more popular than the Modulus series of amplifiers. This might be because it is more expensive than many of the neophyte implementations of the LM3886 that you see here on diyaudio and as such the LM3886DR loses out in that regard. But if you look closely, you will easily see that Tom’s implementation is demonstrably superior. And his documentation, and accessibility via email for questions is excellent. Reading his posts are useful as he teases through the vagaries of diy tube and solid state electronic design that some of us find challenging but are eager to learn.

Have fun!

Anand.
 
Last edited:
Yeah... Unfortunately the heat sink is out of stock at both Mouser and Digikey at the moment. It looks like Mouser will have about 3200 of them ready tomorrow, so your order should arrive soon.

It's easy to find replacements for the diode bridge and caps. If you need help, just toss me an email.

Tom

I'm OK on the heatsink. I think I'm short about 4 parts, one is a capacitor for the LM3886DR boards and the others are all connectors. They are all on back-order with Mouser and should be here before too long. I'm in no desperate hurry so I'm happy to wait and there is always the chassis preparation in the meantime.

Neil
 
Neila,

Thanks for documenting your build. I think what people don’t realize is there are a lot of non-nested feedback implementations of the LM3886 both diy and commercial but none that I know of have been brutally and rigorously measured, i.e. put to the test like Tom has of his own design, the LM3886DR. You know EXACTLY what you are getting with his designs. That isn’t true of the plethora of other LM3886 designs especially here on diyaudio.

For a first build it’s an EXCELLENT choice and frankly I am surprised that it is not more popular than the Modulus series of amplifiers. This might be because it is more expensive than many of the neophyte implementations of the LM3886 that you see here on diyaudio and as such the LM3886DR loses out in that regard. But if you look closely, you will easily see that Tom’s implementation is demonstrably superior. And his documentation, and accessibility via email for questions is excellent. Reading his posts are useful as he teases through the vagaries of diy tube and solid state electronic design that some of us find challenging but are eager to learn.

Have fun!

Anand.

Hi Anand,

I did look at a number of the alternatives before going with Neurochrome. It is quite difficult to assess without much experience. While I cannot attest to the accuracy, there is an interesting article on circuitbasics.com about a LM3886 build (here) which covers a lot of ground and that was tempting initially. And then there are a lot of cheap ebay options which are obviously attractive if you only consider costs.

I lurked here on diyaudio for a little while before signing up. There are a lot of threads on here about LM3886 designs and builds, but one of the things I have struggled with is extracting the relevant information from a thread containing thousands of posts. The forthcoming guides here looks to try to address this.

One thing that did become very clear through looking at the posts here is that Tom Christiansen's designs are highly regarded and he is very active on the forums. Given that he trying to make a living from Neurochrome, he is extremely generous with his time and expertise. That is the kind of company and individual that I am happy to support, so LM3886DR it was.

Neil
 
One thing that did become very clear through looking at the posts here is that Tom Christiansen's designs are highly regarded and he is very active on the forums. Given that he trying to make a living from Neurochrome, he is extremely generous with his time and expertise. That is the kind of company and individual that I am happy to support, so LM3886DR it was.

Thank you. I really appreciate both your support and your feedback.

Tom
 
So, a bit more progress over the last couple of days...

Power-86 board completed. Soldering looks OK. I worried a bit that I'd have issues with being steady enough with the soldering iron. Turns out my hands are perfectly steady. My eyes on the other hand have definitely deteriorated since I last did any soldering...

The Power-86 documentation suggests Antek transformers. Since these are difficult to get in Europe, it also suggests a Hammond alternative, which I was able to add to the Mouser order. Very impressed that the documentation acknowledges that not everyone is in North America and provides options for us Europeans. :)

Powered it up this evening with no load attached. As detailed in the documentation, the output voltages are a little higher when there is no load, but well within expectations (±33V).

Only slight concern was that the transformer hummed slightly. I'd welcome thoughts as to what might be causing that.

Neil
 

Attachments

  • 20180524_210355.jpg
    20180524_210355.jpg
    687.5 KB · Views: 308
  • 20180524_205912.jpg
    20180524_205912.jpg
    351.4 KB · Views: 301