I am just guessing, but the Mod286 should sound similar to the Mod686. Possibly Tom will be the only one to answer this.
There a complex maze of interaction involved. For unbalanced circuits used in most amplifiers, the return current driving the load always goes through the ground, so how this return current interacts with other parts of the circuit and the power supply is one major issue. Tom seems to have addressed much of the grounding considerations explained in his web site, however, interface to the real power supply is always left out of much of the studies.
Then there is the issue of interconnects and speaker cables. Having spent time on both, I ended up doing my own interconnects which works fine above 40 Hz.
Then you have the complex loading of the speaker and enclosure which introduces back EMF. Current drive circuits seems to reduce the effects of back EMF. But the different feedback scheme in lots of amplifiers seem to be effected in different ways. I have looked into this when playing around with the MyRef circuit changing the current gain and looking at the effects, measurements were posted in the original thread.
With Tom’s expertise, I am sure he will look into this sooner or later. As I currently am more focused on speakers now, certainly hope someone looks into all this in a persistent manner.
There a complex maze of interaction involved. For unbalanced circuits used in most amplifiers, the return current driving the load always goes through the ground, so how this return current interacts with other parts of the circuit and the power supply is one major issue. Tom seems to have addressed much of the grounding considerations explained in his web site, however, interface to the real power supply is always left out of much of the studies.
Then there is the issue of interconnects and speaker cables. Having spent time on both, I ended up doing my own interconnects which works fine above 40 Hz.
Then you have the complex loading of the speaker and enclosure which introduces back EMF. Current drive circuits seems to reduce the effects of back EMF. But the different feedback scheme in lots of amplifiers seem to be effected in different ways. I have looked into this when playing around with the MyRef circuit changing the current gain and looking at the effects, measurements were posted in the original thread.
With Tom’s expertise, I am sure he will look into this sooner or later. As I currently am more focused on speakers now, certainly hope someone looks into all this in a persistent manner.
I suggest you compare the measurements of the DG300B with those of the MOD686. You'll find that the THD of the tube amp is several orders of magnitude larger than that of the Modulus amps. I don't understand how you arrived at the conclusion that they measure the same.
The DG300B measures well for a single-ended 300B tube amp. The Modulus amps measure exceptionally well (i.e. world class well). Period. You can compare the Modulus amps against just about any other solid state amp and reach this conclusion.
As I've said again and again: I enjoy the DG300B. I like the presentation. But it is certainly not precise. I would call it a "300B effects box". It's a nice "effects box", but there is no question that it adds its own signature to the music. Some like that. Others don't.
If you want precision, the Modulus amps are where it's at. I have documented this with measurements. If you are having trouble putting my plots into context, I suggest reading some of Stereophile's reviews (specifically the measurements section) and compare the plots of commercial gear to my Modulus amps. You can find the reviews online. It should be fairly obvious that the performance of my Modulus amps is significantly better than much of the available commercial gear (Benchmark ABH-2 excluded - I'm neck and neck with that one).
What does that level of precision sound like? Nothing! That's the point! The Modulus is an amplifier not an effects box. The term "straight wire with gain" is overused. As is "transparent". My Modulus amps are both of those and I back up this claim with DATA (what a concept!!)
If you are looking for a better understanding of what the different measurements do, I suggest having a look-see on AP's website. Their technical library contains lots of good articles. You may have to create a login to view the articles.
I suggest developing some context around the measurements rather than trying to memorize some arbitrary level as "good/bad". If you need a couple of reviews to read, you can have a look here:
Vinnie Rossi LIO Stereophile Measurements
Benchmark ABH-2 Stereophile Measurements
GamuT Di150 LE integrated amplifier Stereophile Measurements
I have no affiliation or vested interest in any of those products. I'm just throwing them out there as examples. For some of the products it is interesting to compare their specifications with the measurements as not all the products meet their published specs.
Let's keep this thread about the Modulus-686. While I do find the discussions about tubes vs solid state and the interpretation of measurements to be interesting, they're honestly better served in a thread elsewhere - or in an offline discussion.
Tom
Gee Tom, relax... I am not trying to question your design nor your decisions nor do I want to start a tube vs. chip amp discussion, nor highjack your thread. Admittendly I did not research your tube amp specs and therefore my question was probably misleading and I apologize for that.
Remember I own two of your amps (mono 286s) and nothing has changed since that very positive (subjective) review here - I still love them and listen to them daily.
http://www.diyaudio.com/forums/chip-amps/309848-mod-286-build-thread-14.html#post5171721
And I will definetly buy also two 686 as soon as I have time (probably August) for such a project. I will then be able to compare one on one the 686 with the 286.
Until then, and I come back to your own statement earlier in this thread, where you said, that the 686 seems to benefit from the additional power over the 286, which led me to think that it sounds (slightly) differently. This surprised me, as I would not have expected any audible difference at, let´s say around 0.5-1W ca. 85-90db listening levels, where clipping most likely does not affect the sound. So both amps measure very well, and pretty much identical (e.g. thd+n 286: 0.00026 % at 55W,8ohms vs. 686: 0.00025 % at 140W,8ohms), as we all know, and still you seem to detect a difference in sound, provided that is what you wanted to say. At least, that´s how I remember your statement (I possibly misunderstood).
So again, now you have two amps (686 vs. 286, so we dont have to compare to other amps from other manufacturers) which both measure very well (and pretty much identical) into a resistive load. Yet you said, they sound different (disclaimer - maybe I misunderstood).
Now coming back to my Initial question (at least that´s how I meant it - sorry if I mispoke):
Could it be, that taking the same measurments into a speaker simulating load (as opposed to a resistive load) one could measure differences between the two amps, which are more pronounced than the current ones, which would then explain the (very slight) difference in sound beween the 286 and 686? And if not, how elso would one explain the difference in sound between the 686 and 286?
Sorry for my persistence, it is just interesting, since you seem to be the only one (or one among very few) who can explain (possible) difference in sound with facts / measurements and not with heresay. That´s why I bought the 286s from you and will buy the 686, provided you let me after this message ;-)
BR
SH
Remember I own two of your amps (mono 286s) and nothing has changed since that very positive (subjective) review here - I still love them and listen to them daily.
http://www.diyaudio.com/forums/chip-amps/309848-mod-286-build-thread-14.html#post5171721
And I will definetly buy also two 686 as soon as I have time (probably August) for such a project. I will then be able to compare one on one the 686 with the 286.
Until then, and I come back to your own statement earlier in this thread, where you said, that the 686 seems to benefit from the additional power over the 286, which led me to think that it sounds (slightly) differently. This surprised me, as I would not have expected any audible difference at, let´s say around 0.5-1W ca. 85-90db listening levels, where clipping most likely does not affect the sound. So both amps measure very well, and pretty much identical (e.g. thd+n 286: 0.00026 % at 55W,8ohms vs. 686: 0.00025 % at 140W,8ohms), as we all know, and still you seem to detect a difference in sound, provided that is what you wanted to say. At least, that´s how I remember your statement (I possibly misunderstood).
So again, now you have two amps (686 vs. 286, so we dont have to compare to other amps from other manufacturers) which both measure very well (and pretty much identical) into a resistive load. Yet you said, they sound different (disclaimer - maybe I misunderstood).
Now coming back to my Initial question (at least that´s how I meant it - sorry if I mispoke):
Could it be, that taking the same measurments into a speaker simulating load (as opposed to a resistive load) one could measure differences between the two amps, which are more pronounced than the current ones, which would then explain the (very slight) difference in sound beween the 286 and 686? And if not, how elso would one explain the difference in sound between the 686 and 286?
Sorry for my persistence, it is just interesting, since you seem to be the only one (or one among very few) who can explain (possible) difference in sound with facts / measurements and not with heresay. That´s why I bought the 286s from you and will buy the 686, provided you let me after this message ;-)
BR
SH
Last edited:
I don't recall comparing the MOD286 to the MOD686. I do recall comparing (albeit based on experiences years apart) the 86 to the 686.
I am sorry if I come across as crass or dismissive. I am really, really, really, REALLY, REALLY stressed right now. I've been spending 14-15 hours/day seven days/week on the MOD686 for the past two months. I cannot afford to spend that kind of time right now. I am really sorry. I bend over backwards to support you and everybody else. I'm at the breaking point now.
I have an exam on Friday. I have a term paper due on Monday that I'm one page into (out of 10-15). I have a paper due April 7th, another 10-20 page term paper on the 12th, and three exams April 16-19th. I have done zero reading for any of my classes due to the work related to the MOD686. I haven't had a day off for over two months.
I will be more than happy to answer any direct question about the Modulus-686.
I cannot afford getting sucked into a discussion about perceived sound quality versus measurements at this time. I will be happy to try to explain my findings based on measurements and my subjective experience in late April when I'm done with my exams. You can find my thoughts on that in other threads as well.
Tom
I am sorry if I come across as crass or dismissive. I am really, really, really, REALLY, REALLY stressed right now. I've been spending 14-15 hours/day seven days/week on the MOD686 for the past two months. I cannot afford to spend that kind of time right now. I am really sorry. I bend over backwards to support you and everybody else. I'm at the breaking point now.
I have an exam on Friday. I have a term paper due on Monday that I'm one page into (out of 10-15). I have a paper due April 7th, another 10-20 page term paper on the 12th, and three exams April 16-19th. I have done zero reading for any of my classes due to the work related to the MOD686. I haven't had a day off for over two months.
I will be more than happy to answer any direct question about the Modulus-686.
I cannot afford getting sucked into a discussion about perceived sound quality versus measurements at this time. I will be happy to try to explain my findings based on measurements and my subjective experience in late April when I'm done with my exams. You can find my thoughts on that in other threads as well.
Tom
Btw Tom , I have some questions regarding 86 vs 286 vs 686 vs I7...haha ! Take it easy Tom ! Its's called LIFE! Good luck!
LIFE? I've heard of it. Where can I get one? Is there a Mouser part number for that? 🙂
Yeah. I used to cruise up to the hot springs in Banff after exams and soak for a few hours. It's about a 90-minute drive, so it would provide a good 5-hour break without internet access. Glorious! That was great stress relief. Sadly, I haven't had the time to do so for the past couple of semesters.
I'm toying with the idea of just getting away to some place without internet access after my exam on the 19th. Stay for a day or two. No devices. No temptation to work. At the very minimum, I'll play hockey in the afternoon that day and cruise up to Banff for a soak.
Tom
Yeah. I used to cruise up to the hot springs in Banff after exams and soak for a few hours. It's about a 90-minute drive, so it would provide a good 5-hour break without internet access. Glorious! That was great stress relief. Sadly, I haven't had the time to do so for the past couple of semesters.
I'm toying with the idea of just getting away to some place without internet access after my exam on the 19th. Stay for a day or two. No devices. No temptation to work. At the very minimum, I'll play hockey in the afternoon that day and cruise up to Banff for a soak.
Tom
All good - I know the feeling. Let’s stop the topic here. Sorry I could have been more sensitive to your mentioning having exams pretty soon. Also I am very grateful for all your support and patience. Please, your exams are more important than this here. I really wish you all the best.
And once you are done with your exams, and when you have some time again, I would be interested in the measurements into simulated loudspeaker loads. If there is something to be gained or derived from these measurements I know you are more capable to point them out than I would. Until then, I enjoy my 286 and you go and study hard. Good luck.
Last edited:
True that. I need to make sure my business is here when I finish too. I have rent and food covered for this month, so I'm in better shape than many. I want to do better, though. I've made a significant investment in the Modulus-686 project and would like to see a return on that investment. I generally find I catch more flies with honey than with vinegar, so being available for questions and support is key. That said, I also need to spend my time wisely and in ways that are more likely to yield a high return on investment. It's all about balance. A balance that I clearly struggle to find. 🙂
Tom
Procrastination continues. I found some time this morning to get my Modulus-686 product page updated. Those interested in some of the background for the measurements and how they impact the perceived sound quality will likely find the Performance Graphs section informative.
Neurochrome Modulus-686: 240W(8Ω); 360W(4Ω) @ 0.00025% THD+N
Tom
Neurochrome Modulus-686: 240W(8Ω); 360W(4Ω) @ 0.00025% THD+N
Tom
Hey Tom,
I think the thread could be turned down from 11 to 10 and the 686 thread and amp will just keep steaming along. Many have been waiting for it.
And don't forget that turning out top performance in your continuing education will result in better projects in the future!
Education :: Done :: Right
leads to
Engineering :: Done :: Right
leads to
Diyaudio :: Done :: Right
😉
Cheers,
Jeff
PS Don't respond, go study! Whhit-tsh!
I think the thread could be turned down from 11 to 10 and the 686 thread and amp will just keep steaming along. Many have been waiting for it.
And don't forget that turning out top performance in your continuing education will result in better projects in the future!
Education :: Done :: Right
leads to
Engineering :: Done :: Right
leads to
Diyaudio :: Done :: Right
😉
Cheers,
Jeff
PS Don't respond, go study! Whhit-tsh!
Last edited:
I'm just letting people know that they can find answers to their questions on my website. That's all... It is the Vendor's forum after all... 🙂
I hope you're right about the interest. I'm basing my business on it for the foreseeable future.
Oh, I will. I have an update planned for the stability pages of the Taming the LM3886 series. I have a couple of redesigns to do before I can tend to that, though. TI's decision to discontinue the LME49710 has not been kind to my projects. Oh, well. Life in the business.
Meh. I'm watching hockey and making full use of the commercial breaks. 🙂
My exam went pretty well (I think) and I'm about halfway through the paper that's due Monday with a good idea of what to write for the rest of it. Life is good.
Tom
Tom, how do you get any tasks done if you spend all of your time posting on the forum and watching hockey?
PS. Hockey season doesn't really start until April 11.
PS. Hockey season doesn't really start until April 11.
I have no idea, honestly. Somehow a 16-page paper on racial bias in the Canadian justice system spontaneously appeared on my hard drive within the past week. Granted, this included three pages of references and a title page. I think I had four pages of references at one point, until I reduced it so the paper wouldn't run away from me completely. Man. That was work! It's due in an hour. It's printed and ready to go. Yay!
True that. And the Leafs are in it, which makes me happy. About three hours of watching hockey is my scheduled weekly relaxation. I need to get better at relaxing.
🙂 Thank you.
Tom
I further asked around about heat sink contact, it seems that using hard anodizing of (I think) 3uM thickness should do. It seems best to press the T package using a bar rather than the hole provided.
Yeah... It's interesting what people choose to geek out about. There's a thread around here somewhere, where someone precision machined a block of brass to make contact on all sides of the LM3886 and provide some clamping force. They claimed this was superior to the standard mounting method, but true to form provided no data to back up this claim. They also completely neglected that the vast majority of the heat is conducted through the back side of the package and very little escapes through the plastic.
There are many ways to mess up a clamping solution and fewer ways to mess up when using the manufacturer's recommended solution (i.e. the mounting hole in the tab), so I went with the latter.
I just received a roll of 1000 (minimum order quantity from the distributor) of Keratherm Red (86/82) in a size that fits the LM3886. They'll be available on my website by Sunday night. To the best of my knowledge, they offer the lowest thermal resistance of any thermal pad on the market.
Tom
There are many ways to mess up a clamping solution and fewer ways to mess up when using the manufacturer's recommended solution (i.e. the mounting hole in the tab), so I went with the latter.
I just received a roll of 1000 (minimum order quantity from the distributor) of Keratherm Red (86/82) in a size that fits the LM3886. They'll be available on my website by Sunday night. To the best of my knowledge, they offer the lowest thermal resistance of any thermal pad on the market.
Tom
Attachments
Nice!
Actually the bar attachment method offers more uniform pressure and less need to align the mounting when your assemble most chassis. Right now you use a special bracket to keep alignment which would cost more and soldering the LM3886 is a bit cumbersome, the alignment would be a bit better though.
In most commercial practices, the tolerance in the PCB mounting holes would allow good alignment without introducing stress and the LM3886 legs, aligning multiple devices does take some skill and preparation though.
Another trade off is the number of contacts to the heat sink. I would go for less contact interfacing surfaces.
Actually the bar attachment method offers more uniform pressure and less need to align the mounting when your assemble most chassis. Right now you use a special bracket to keep alignment which would cost more and soldering the LM3886 is a bit cumbersome, the alignment would be a bit better though.
In most commercial practices, the tolerance in the PCB mounting holes would allow good alignment without introducing stress and the LM3886 legs, aligning multiple devices does take some skill and preparation though.
Another trade off is the number of contacts to the heat sink. I would go for less contact interfacing surfaces.
Actually, the LM3886es are easy to solder through the openings in the bottom of the bracket.
Tom