Modulus-86 build thread

No comment. ;)

Okay. One comment: Higher output power. The Modulus-286 is known to the public already.

I hope to have the MOD286 out by Christmas.

Tom


Would it be possible to use two Mod-286 boards per channel, in bridge parallel mode, and use the THAT1203's in them (so that overall gain isn't too high, approx. 26dB) for very high output power into 8 and 4 ohms? Or should we wait for your discrete solution slated for Spring 2017?

Thanks,

Anand.
 
Would it be possible to use two Mod-286 boards per channel, in bridge parallel mode

The LM3886 is not really all that well suited for bridging. I know that it's theoretically possible, but that doesn't mean it's a good idea. Here's why:

If power is what you want, run the LM3886 on ±36 V rails. Recall that in a bridged amp, each amp half sees half the load impedance. Thus, for 4 Ω load, each amp half sees 2 Ω. So:

Iout_max = 36/2 = 18 A

Each LM3886 is guaranteed to be able to provide 7 A. 18/7 = 2.6 -> three LM3886es in parallel are needed.

So now you're talking six LM3886es. Three of them need to play together nicely in parallel. Then the parallel combo needs to play nicely in a composite amp which will then need to play nicely with another composite amp in a bridge configuration. That's a research project!

Or should we wait for your discrete solution slated for Spring 2017?

If you want more than 110 W into 4 Ω, you should look at solutions outside the chipamp space in my opinion. Sadly that means some level of discrete circuitry, probably with an opamp performing error correction. That design will be mostly SMD components, so it won't be as beginner friendly as the Modulus-86, unfortunately.

Tom
 
Last edited:
Well, finally got around to stuffing the Mod86 into a box :D

Yes, I'm one of the few (I've seen one other) using the Mod86 as a headphone amp. Works quite well for the application, actually!

XLR and power inputs on the back. 4 pin XLR and TRS on the front panel along with the volume control. Volume control is a 100k Khozmo 48 step stepped attenuator.

Chassis is the Dissipante 2U

Pics attached...
 

Attachments

  • IMG_1590a.jpg
    IMG_1590a.jpg
    195.1 KB · Views: 633
  • IMG_1591a.jpg
    IMG_1591a.jpg
    191 KB · Views: 621
  • IMG_1596a.jpg
    IMG_1596a.jpg
    166.9 KB · Views: 589
That's pretty cool! Nice build. Are you running ±28 V on that? Or did you lower the supply voltage to a somewhat saner ±20 V?

±28 V would give you around 10 W into 32 Ω and 1 W into 300 Ω. That's **A LOT** of power for a headphone amp. Careful on the volume knob. ;)

±20 V would give you more like 5 W (32 Ω), 500 mW (300 Ω).

Tom
 
Tom,
Thanks :) Actually, the rails are sitting at ± 35 V. I was looking for ± 30 V, but transformer choices... So, I ended up with a bit higher. No worries on the volume. LCD 3's are about 9:30 - 10:00 with highly compressed rock (Muse, for example). AKG K7XX actually want a bit more at 10:00-11:00.
The absurd headroom is to attempt to drive the HE-6 to their potential.
Probably could have gotten away with 28 V rails with no problem, but, hey... Why not splurge? :)

HE-6 sits at about 2:00 with Muses' Supermassive Black Hole...
And a solid 3:00 with thier Supremacy... Very happy :)
 
Last edited:
Hi Tom,
With all this talk about new projects a couple of observations if I may.
For cack-handed builders like myself the move to SMD's could rule out future Neochrome projects. One of the beauties of the Mod-86 is that it is doable for newbies like me and it sounds fabulous.
I've also got the differential pre but had to get help with the surface mounts. Also the fact that the XLR sockets and volume control are mounted on the board means that a high degree of precision is required to cut holes in the case so that everything lines up - again this makes the project that much more difficult for folk like me who have limited tools - i.e. hand held only.
Your stuff sounds so good so it would be a shame if the build requirements limits your potential market.
As we are talking about new projects, have you had any more thoughts on producing a phono stage?

Cheers, Ian
 
The opposite here. I'd rather load an SMD board than through hole. It's quicker and easier. Given the choice I'd go SMD every time.
Same for me, SMT is 5x quicker and easier to stuff (and rework) than THT. Invest in good magnifier glasses, proper tweezers and a good soldering tip I would recommend to those who have problems with SMT because of the smaller size. Once you get used to it you won't look back...
 
People also need to understand Neurochrome's design philosophy. It's not only to design the best sounding but best measuring circuits and there is a correlation there even if the subjective pundits say otherwise. Many of them do not know or understand which measurements correlate to great sound and once they give up on that pursuit, they tend to state that measurements shouldn't act as a guide to design. Which is absolutely wrong. Don't forget that the procurement of excellent devices that produce accurate measurements is an expensive hobby in and of itself. This is before you have even mastered the methodologies involved in measuring!

Where am I going with this? The new and exciting technologies involve SMD. The DIY 2ch world is small although it may seem enormous if you only lurk here. The demand for assembled work is high and the high end headphone market is über high. That market continues to grow as millennials continue to choose headphone listening as their primary source. It makes financial sense to supply to this growing group which also demands smaller but higher performing components, hence SMD.

Ian, I suggest you take a serious look at SMD and don't despair. There are great tutorials by Curious Inventor on YouTube to assist you. Just like through hole soldering, it can be daunting at first but with the right tools (Loupe/Flip-Focus, Wiha tweezers, small gauge solder as explained on Tom's website, i.e. 0.025 in, Liquid flux or flux pen, etc...), you should well on your way.

Surface Mount Soldering 101 Video

My very best,

:cheers:

Anand.
 
Last edited:
People also need to understand Neurochrome's design philosophy. It's not only to design the best sounding but best measuring circuits and there is a correlation there even if the subjective pundits say otherwise. Many of them do not know or understand which measurements correlate to great sound and once they give up on that pursuit, they tend to state that measurements shouldn't act as a guide to design. Which is absolutely wrong. Don't forget that the procurement of excellent devices that produce accurate measurements is an expensive hobby in and of itself. This is before you have even mastered the methodologies involved in measuring!

Where am I going with this? The new and exciting technologies involve SMD. The DIY 2ch world is small although it may seem enormous if you only lurk here. The demand for assembled work is high and the high end headphone market is über high. That market continues to grow as millennials continue to choose headphone listening as their primary source. It makes financial sense to supply to this growing group which also demands smaller but higher performing components, hence SMD.

Ian, I suggest you take a serious look at SMD and don't despair. There are great tutorials by Curious Inventor on YouTube to assist you. Just like through hole soldering, it can be daunting at first but with the right tools (Loupe/Flip-Focus, Wiha tweezers, small gauge solder as explained on Tom's website, i.e. 0.025 in, Liquid flux or flux pen, etc...), you should well on your way.

Surface Mount Soldering 101 Video

My very best,

:cheers:

Anand.

I recognize the arguments, but I cannot deal with SMDs and I choose not to. At 60 years old, neither my eyes nor my fine motor skills will allow mw to deal with SMDs. Small form factor is not an issue for me (and recognize that I live in a relatively small house, comparable in size to many condos/apartments. HPshones are nice (I have both a Stax SR5 and an Sennhsierx that would not qualify as "audiophile") but they are not ever going to replace my speakers (Vandersteen 2ce and old concentric Tannoys).

I'm not saying that Tom has to cater to me. He makes business decisions: that is his right as a vendor and I totally respect that - indexed, every vendor makes those decisions. I'm simply sharing my view as just another data point.
 
After ten months, the SMD vs leaded survey still sits at an even split. Personally, I'm somewhat agnostic when it comes to soldering SMD vs leaded, except when it comes to rework of 2-pin components which is much, MUCH easier on SMD with hot tweezers.

There are a couple of things driving my gradual shift towards SMD:
  • The more exciting parts (OPA1611 for example) are only available in SMD.
  • Precision resistors (±0.5 % or ±0.1 %) are expensive or unobtainable in the leaded world but really quite inexpensive in the SMD world.
  • Four-layer PCBs can offer stellar performance in audio circuits but are rather expensive (3x the cost of a two-layer PCB). To keep the cost reasonable, the board needs to be smaller. SMD components help here.
  • Long-term, I would like to be set up for automated assembly so I can offer assembled or partially assembled (SMD only) boards at a reasonable cost. Right now, I do the assembly by hand. I'd rather focus on product development and sales than stuff boards.

That said, I deliberately choose components which are on the large side by SMD standards so that humans have a chance of soldering them. I use 0805 size resistors, SOICs, SOT-23, etc. Even this limits my component choices. The OPA1622 is out for example.

I do enjoy catering to the DIY community but I would prefer to be able to sell into other markets as well. I want to ensure that I don't paint myself into a DIY-only corner; or to be all business-y about it, I want to ensure that my serviceable available market is as large as possible as that gives me the most room to grow.
It's one thing to run a li'l hobby side business while earning a comfortable income working for someone else. It's quite another when my ability to pay rent and put food on the table depends on my business. I have to look forward as I want to make sure Neurochrome is around for years to come. Both for me and for all of you who enjoy my work.

Tom
 
As one whose near vision is showing its age (and I 'blame' a lot of it on having to work with SMT for more than 25 years...), I understand the reticence of folks to venture into SMT. However, I find that these days I use 'OptiVisor' style magnifiers for even through-hole work. Sure, I could solder in TH parts with only my glasses, but it is hard to read component values and have good view of solder joints. SMT work is really not that hard - it just requires a little practice and perhaps a smaller tip on the soldering iron. On my own boards, I try to stick to 1206-size R's and C's as well as larger ICs (SOIC). At work I've had to work with ridiculously tiny ICs and 0402 passive components, though it is a huge pain.

Yes, it's scary at first to build SMT boards, but like anything else, you can do it with a little practice. Tom points out the reality that through-hole components are getting harder to source, especially in smaller quantities.