oxidized copper wire
This may not be the proper thread for talking about wire but I'll be wiring my parallel86 soon.
I think I may have mentioned that I have an old nakamichi speaker selector box. About fifteen years ago I cleaned and treated the stranded copper wires and clamped on banana plugs. I just checked it out, removed the banana plugs and found the the wire strands were almost black. The oxidation continued under the jacketing for about 1/4".
That makes me wonder what happens to copper wires used inside an amp over time. I've never opened any of my amps to check out or clean wires.
If skin effect is real i'm thinking that amps that use copper wire for internal connections must lose some sonic quality over time.
I have lots of the mil spec teflon coated silver plated wire, and it's a pain to strip, but why wouldn't it be better for internal wiring over the long haul?
This may not be the proper thread for talking about wire but I'll be wiring my parallel86 soon.
I think I may have mentioned that I have an old nakamichi speaker selector box. About fifteen years ago I cleaned and treated the stranded copper wires and clamped on banana plugs. I just checked it out, removed the banana plugs and found the the wire strands were almost black. The oxidation continued under the jacketing for about 1/4".
That makes me wonder what happens to copper wires used inside an amp over time. I've never opened any of my amps to check out or clean wires.
If skin effect is real i'm thinking that amps that use copper wire for internal connections must lose some sonic quality over time.
I have lots of the mil spec teflon coated silver plated wire, and it's a pain to strip, but why wouldn't it be better for internal wiring over the long haul?
Skin effect is very real at RF, not audio. That Teflon wire you have is excellent in a fixed install such as inside an amp. Only issue is it is prone to cold flow, the insulation will thin and then fail under sustained pressure, so don't tightly twist your wires. I've just wired up my Mod-86 amp using 16AWG teflon wire. Ran the high current pairs/triples in heatshrunk bundles.
The "Ideal Custom Stripmaster" is an excellent tool for stripping this wire with no damage to the conductor. Otherwise you can use thermal strippers.
The "Ideal Custom Stripmaster" is an excellent tool for stripping this wire with no damage to the conductor. Otherwise you can use thermal strippers.
If skin effect is not a concern in audio, would that be why old amps with oxidized copper wire still sound good. I have three amps, from the mid 70s, 80s and 2000, and they still sound very good. I also have some OFCH copper 14 awg and that is very easy to strip. I'm now thinking the only benefit of using the teflon coated wire maybe the dielectric value. if that also has no effect at audio frequencies I might as well use the copper wire and toss the teflon coated silver plated wire. Chalk it up to lack of knowledge and move on.
I've been quietly humming along putting together my SMPS-86(System check is a go!) and my 2 Modulus-86s. I'm at the final step of the modulus-86's and I have't soldered the LM3886 yet. I don't have my heat sinks picked out yet because I'm still pondering some casing ideas. I also saw in this thread the advice of mounting the heat sink in order to make sure the LM3886 perpendicular to the board so I don't stress it when I mount it. I can't tell if this is a necessary step or if it is being a bit... well, you pick a word. I figured if it was necessary it would be in the incredible thorough and revised instructions. Also, is the heat sink necessary for doing the final test steps 1-7? Obviously I'm not going to actually run the amp without heatsinks for long, but for the quick sanity check, is it ok? Alternatively, I have a couple random blocks of aluminum Perhaps I could tap a hole to them?
Do I have questions about heat sinks. If a 12" profile from heatsinkusa is C/W/3" .85. What is the resistance if it is cut in half. In other words it becomes a 6" profile with all other dimensions the same.
Also, what happens to the 12" profile resistance when two of the same chips are mounted on it, lets say 6" apart?
I ask because I'm thinking of using the 12" profile, not cut in half, for my Parallel86.
I'm not looking for "it's ok". I'd like to know what happens to the resistance value. Maybe I should be asking heatsinkusa?
Also, what happens to the 12" profile resistance when two of the same chips are mounted on it, lets say 6" apart?
I ask because I'm thinking of using the 12" profile, not cut in half, for my Parallel86.
I'm not looking for "it's ok". I'd like to know what happens to the resistance value. Maybe I should be asking heatsinkusa?
That's a question for Heatsink USA. No doubt the thermal resistance will increase. I just don't know by how much.
If anything it may become a gnat fart lower as the heat source now covers a larger area of the heat sink.
It's OK.
Seriously. It is.Tom
Sweet!
One note: I learned yesterday that if you're into really loud music and running 4 Ω drivers that dip well into the 3 Ω range, you need to run one SMPS-86 per channel to ensure that you have enough peak current. The SMPS-86 delivers 2.5 A continuously and the current limiter trips at about 3 A.
I haven't personally had any issues (using either 4 Ω or 8 Ω drivers) even at critical listening levels but a builder reported this and I was able to reproduce the issue with some very bass-heavy music cranked to 11.
When the current limiter in the supply trips, the supply shuts down and the amp mutes. No drama. The music just turns off.
I'll update the SMPS-86 website to reflect this.
You can mount the LM3886 without having the heat sink picked out. I do that all the time. Just make sure that the back of the package is parallel to the edge of the PCB and perpendicular to the top plane of the board. The most critical bit is to get the IC parallel to the edge of the board so it doesn't have to rotate when you mount it to the heat sink.
You can test the board without a heat sink. Without a speaker connected you should be able to complete all steps in the final test procedure. The LM3886 will get hot during testing. That's normal.
Note that to drive a speaker or resistive load you will need a heat sink.
Tom
I believed it would be OK. It's just curiosity. I'll ask HTUSA a few questions. Maybe they have some conversion tables or calculations.That's a question for Heatsink USA. No doubt the thermal resistance will increase. I just don't know by how much.
If anything it may become a gnat fart lower as the heat source now covers a larger area of the heat sink.
It's OK.
Tom
I suggest reading the first bit of the Thermal Design page of my Taming the LM3886 series. Follow the links I give for determining heat sink thermal resistances, length & temperature correction. Aavid makes a lot of information available on the ins and outs of heat sinks.
Tom
This won't be a problem for me. You'd be appalled at the rest of my set up and musical tastes. The amp is really going to be like a Ferrari on country roads. If I do get into the range where I'm out drawing the SMPS-86's current it is because I've moved somewhere much much nicer.
This is the warning I was looking for. It's the torque on the whole chip rather than any angle off of true perpendicular that I should look out for.
In this same vein, I've been trying to read through the build thread but its a little hard to keep threads and pick up pertinent information from the now 248 pages. I only really picked up on the heat sink idea because it is in the first couple pages and some people seemed pretty passionate about it. Obviously, you've been monitoring the thread this whole time. Are there any particularly valuable questions(both in learning or in building) that people have asked that I might have missed?
I recall a photo of the Mod86 PCB with the chip soldered to it and the PCB having stand offs screwed at the four corners. Next to it was a heatsink and they both rested on a very flat surface.
The pic showed the back of the chip vertical and slightly away from the edge of the PCB. The back of the chip and heatsink surface were very parallel to each other. With a setup like that I think the only rotational torque or twisting might come from tightening the chip mounting screw.
I think using a spring clip, like the Aavid, should allow one to easily eliminate any rotational twisting of the chip during clamping. Just a thought.
The pic showed the back of the chip vertical and slightly away from the edge of the PCB. The back of the chip and heatsink surface were very parallel to each other. With a setup like that I think the only rotational torque or twisting might come from tightening the chip mounting screw.
I think using a spring clip, like the Aavid, should allow one to easily eliminate any rotational twisting of the chip during clamping. Just a thought.
Not really. It's one of those things that are super obvious when you have the parts in hand but hard to describe without pointing and talking.
The important piece is: No gaps! When the board is mounted in the chassis, you want the LM3886 to be resting flat up against the heat sink. No pitch. No roll. Yaw is allowed but looks goofy.
There are some pictures of the LM3886 mounting in Post #73. Note that this post pertains to Rev. 1.0 and 1.01 of the board. On those revisions, the LM3886 was flush with the edge of the PCB and it was pointed out that there should be a larger gap. On Rev. 2.0 and beyond, the LM3886 is extending beyond the edge of the PCB by 0.019" (0.48 mm). Same is the case for my other LM3886 boards.
Any major lessons learned have been included in the design documentation by now. I think the major ones were musings about heat sink sizing, which resulted in that chapter in the design doc being completely rewritten for clarity and to provide more guidance on heat sink selection. Same for the power supply chapter.
Based on my experience after a couple hundred boards sold, there's actually be very few build issues. I think the very few issues fall into three categories:
- Soldering (cold solder joints or solder bridges).
- Mis-identifying pin 1 on ICs (despite the legend on the board silk screen).
- Rushing to complete the project late at night and missing something.
Take your time. Follow the build instructions. Go to bed if you're tired. You'll do fine. Should you get stuck, I'm only an email away.
I built a 4xMOD86 + MiniDSP 2x4 amp for a friend to run a pair of LXmini speakers. You can see build pictures in Post #888 and Post #805. The system is fed by a Sonos Connect. It works great. Super quiet. No hum or noise. Feeding lossless files through the system sounds fantastic.
With the media players out there, beware of their output voltages. Many run on 3.3 V rails, hence, only provide 900 mV RMS out. If you use a player like that, I suggest building the Modulus-86 with 26 dB of gain. For players with 2.0 V RMS out, the default gain of 20 dB is perfect.
Tom
Last edited:
Second anniversary!
Folks,
I just realized that the Modulus-86 turned two years old a bit over a month ago. It was on July 20th, 2014 that the first revision of boards were submitted to the board fab. I think that calls for a celebration.
Thank you to all the Modulus-86 builders out there and all who supported the project. You are what keep me going. Thanks!
Tom
Folks,
I just realized that the Modulus-86 turned two years old a bit over a month ago. It was on July 20th, 2014 that the first revision of boards were submitted to the board fab. I think that calls for a celebration.
Thank you to all the Modulus-86 builders out there and all who supported the project. You are what keep me going. Thanks!
Tom
I got a pre-populated DAC board, then built Tom's THAT Line Driver, paired them in the same chassis as a DAC/Preamp. Works awesome and feeds the Differential/XLR signal to mod86. I love it! I stream Apple Lossless through AppleTV into the DAC. I imagine that if you find a streaning devise with analog output, pairing it with THAT Line driver would be a good and simple solution.
Yep. The one with the MiniDSP is in daily use by a friend of mine on Lummi Island, WA. He uses a Sonos Connect to drive it. The one I brought with me to the DIY fest on Vancouver Island recently is without the MiniDSP. I chose XLR inputs to have more flexibility. I now drive it with a MiniDSP 4x10 in my LXmini system, but used to use a PC-based XO (Reaper w/ReaEQ).
I built the two amps at the same time. It took about a week. You can see a family photo of the eight channels built (in just over eight hours!) in Post #840.
Tom