Oh? Why?
I think Anand is pragmatic and science-minded enough to realize that the heat spreader has no negative impact on the performance of the amp and that it makes mounting the modules much easier.
Tom
I'm building a short stack of three Modulus-686 in the Safe-n-Sane configuration. 125 W into 8 Ω, 200 W into 4 Ω. 340 x 330 x 150 mm (WxDxH), 6 kg.
Two of them I'll sell. One has sold already. The last one I'll keep.
You can watch me build the amp here:
Tom
Two of them I'll sell. One has sold already. The last one I'll keep.
You can watch me build the amp here:
Tom
Attachments
Very nice, you never get tired of seeing cool gear like this. Are those the Mean Well RPS-400-27-C PSU's without cages? And is not having the cages any disadvantage at all? Aesthetically I like them, but you can see the benefit this setup has without the cages with regard to layout and available space.
I'll drop it off with UPS on my way to hockey this afternoon. You should see the tracking update tonight.
You're one lucky bugger. I listened to the amp for a few hours Wednesday night. It's a great amp. I need to build one for myself so I can enjoy it. It drives my KEF R700 very well.
Yep. Those are indeed RPS-400-27. Not the -C (caged) version. The purpose of the cage is to prevent fingers and other objects from getting at the high voltage. It doesn't do anything electrically.
In my first Safe-n-Sane build I used the caged version and mounted them vertically. The cage has a couple of PEM nuts for this. I did ask Mean Well if this was the intended mounting method. They hemmed and hawed and gave a wishy-washy answer that wasn't exactly a no. So I went ahead and tried it. Those were the amps I had with me to Burning Amp 2018.
Unfortunately the amps got banged around enough on my return flight (and also in transit to customers) that the cages on the power supplies got bent. This meant the supplies were wiggling a bit inside the chassis. In most cases this was solved by tightening the screws. In one case, the customer had to open the amp and bend the cage back into its original shape. There was no damage to the electronics, but obviously this is not the high-end experience I aim to deliver. So I had to find another solution.
In this build I stacked the two supplies using threaded standoffs. I got the 6 mm hex version for rigidity. Those supplies aren't going anywhere.
The power to the supplies is controlled by an ISS. Mike also requested speaker protection, so I added the Guardian-86 (Rev. 2). 'Tis a nice amp.
Tom
I am realizing now how big my 686 is! So this Safe N Sane build has its own special niche. Most don’t need more power than what it provides. Not only is it a great looking build, it’s compact and quite light to boot. Despite the voltage rails being lower, the bias current remains the same so those heatsinks will probably warm up a little, which isn’t a bad thing from a psychological standpoint (depending on which side of the psychology you lean on). I happen to like warm amps. Even with my gargantuan 686 in a 400mm deep chassis, the sinks were settling at a balmy 30-32 degrees C - not hot at all!
Hope you sell a ton of them. I think there is a good portion of folks that are happy to pay for a fully built, fully certified, guaranteed product that works from the get go. No fuss. Set it and forget it!
Best,
Anand.
Hope you sell a ton of them. I think there is a good portion of folks that are happy to pay for a fully built, fully certified, guaranteed product that works from the get go. No fuss. Set it and forget it!
Best,
Anand.
From a personal standpoint, I’m glad Tom is willing to do full builds of his amps. You could certainly make the case through reading, practice and trial-and-error, one could develop the skills to competently build an amp like this. However, the added value you get when Tom builds it is enormous. He knows the design and how it should perform, knows how it should be pieced together, has tremendous experience building these products, applies his engineering mindset to every facet of the build and has very high standards for his finished products. I could certainly learn to build it, but it would not be anything like a finished product from Tom.
Heh. Yeah. I love the "KallaxFi" form factor. I recently received a pair of amps from a client in the 'normal' 430 mm form factor and was reminded how massive that is.
Peace of mind has value. The Modulus-686 would not be my first choice for a first-time builder. I'm sure a savvy first-time builder could (with some help) figure out how to put it together, but the cost of a mishap can be really high. As I point out in the video, some have had the misfortune of reversing the power supply to the Modulus-686. That's a $500 mistake. $1k if you plug in both modules (as some have done too). A mistake like that instantly turns all the silicon on the Modulus-686 into slag. It is possible to replace all the ICs and diodes, but it's very tedious and time-consuming. It also carries the risk of pulling out the through-plating on the LM3886 pins, and then you're out both your time, parts, and PCB. Yay... (Not!) Those types of support emails usually make me throw up in my mouth a little.
Tom
I've added the two remaining Modulus-686 to my website: https://neurochrome.com/collections/fully-assembled/products/modulus-686-safe-n-sane
Yes. I know I said I'd have one left for my own use. And I will. But if someone really, really wants to buy it I won't turn down their money. I know a guy who builds amps...
Mouser didn't have my favourite red/green power switch available, so I went with blue for the second build. So the first buyer will get to choose between red/green for standby/on indication and off, blue for standby and on indication, respectively.
Tom
Yes. I know I said I'd have one left for my own use. And I will. But if someone really, really wants to buy it I won't turn down their money. I know a guy who builds amps...
Mouser didn't have my favourite red/green power switch available, so I went with blue for the second build. So the first buyer will get to choose between red/green for standby/on indication and off, blue for standby and on indication, respectively.
Tom
No problem just my 2 cents. If you have no powercon in pro audio equipment you do not sell any products.
You wil see in a view years all the better consumer products will also have powercon.
Further using big heatsinks as side walls is a nice design option but they radiate a lot of heat back in the housing. So isolating the spare surface inside the housing can prevent that.
You wil see in a view years all the better consumer products will also have powercon.
Further using big heatsinks as side walls is a nice design option but they radiate a lot of heat back in the housing. So isolating the spare surface inside the housing can prevent that.
If you're willing to pay me to make a custom amp for you I'm happy to include whatever power connector you want as long as it'll fit in the chassis.
99% of my customers use my products in residential settings. If I don't have an IEC320-C14 on the chassis I won't sell any amps. Similarly, if I don't have binding posts on the chassis I won't sell any amps. It doesn't matter that the speakON is vastly superior to any binding post. Know your market!
I'm pretty conservative in my heat sink sizing. I ran the amp for about 90 minutes at 2x40 W output (about the worst case for power dissipation in the heat sinks). The heat sinks reached 58 ºC. That's very comfortable for all the components involved. The chassis is well ventilated so the components will last a long time.
BTW: Some have expressed concern about the thermal interface added by the mounting bracket/heat spreader. So I measured the temperature of the bracket in the middle between LM3886 #3 and #4 and the temperature of the heat sink at that spot. The difference was 0.8 ºC. Not enough to worry about.
Tom
Oh. OK. So you want the local decoupling for the LM3886 to be as far away from the heatsink and, thereby, the LM3886 as possible. That will cause the LM3886 to oscillate. Maybe a power amp that oscillates is your idea of a good design but it sure isn't mine.
I'm well aware of the effects of temperature on reliability. I don't know where you get your 50% number from. As far as I know the probability of failure stays fairly low until near the rated operating temperature of the part where it starts to take off exponentially. I use electrolytic caps rated for at least 105 ºC, especially for the capacitors near the heat sink. If a capacitor is rated for, say, 5k hours at 105 ºC it'll last orders of magnitude longer at 60 ºC and even longer at the 35-40 ºC where it's likely to operate in the residential settings I cater to.
Insulating the heat sink on the inside won't change the steady-state temperature internally in the chassis either. Think about it. The thermal system can be modelled as a simple RC circuit. Increase R and the charging time will increase but the final steady-state voltage across the capacitor is the same. It could work well in a fan-cooled PA amp, though, which is where I suspect you have the idea from.
Oh, so now you want me to spend tens of thousands of dollars on compliance testing for an amp that I will sell 2-3 of? Are you for real?!
There are many vendors, including marquee names, who sell products that have never seen the inside of a compliance testing lab. You may just have to live with that.
Half fabricates???! Because I don't use your favourite power connector? Because I don't insulate the heat sink? Or because I won't screw up the decoupling and turn the amp into a power oscillator? Seriously!? Get over yourself!
I think we have someone who got his feelings hurt by my "Engineering Done Right" tagline. It seems, to him, his solutions are the only right ones. All other solutions are wrong. That's not something I can help him with.
To me, "Engineering Done Right" means choosing a good optimum solution to a multivariate optimization problem. Most of the engineering problems I work on have multiple solutions. And many of those solutions may be valid ones. I believe that the solution I have chosen is the best solution available with the constraints I have imposed on it. One of those constraints is: Must deliver state of the art performance. Another constraint is: Must be manufacturable. Yet another is: Must be reasonably priced. I will not use a $30 opamp when a $3 will deliver the same performance, for example. There are many other considerations that go into the design.
I make it plainly clear what you are getting. Unlike other players in the field I document the performance of my circuits with many measurements. I have nothing to hide. What you see is what you will get. I also push the limits on the instrumentation and testing. I was among the first to make multi-tone measurements available for my circuits. I was also among those who nudged Amir at Audio Science Review to start making multi-tone measurements.
I will not hide behind marketing babble or some sappy story about how I conceived of the product in a dream or whatever. You know why? Because I'm honest. And also because I suck at writing such babble.
I back my circuits with near-immediate email support. I often respond within minutes, and certainly within a few hours, should you get stuck on your build. Your success is my success.
If what I offer isn't your cup of tea or doesn't suit your needs, that's fine. Just move along. Buy something else. I won't pretend to satisfy everybody because in doing so I would end up satisfying nobody.
Tom
Oy. I really liked this page, until the derail. @groovybassist made a great point about the Neurochrome product becoming much more accessible with a fully assembled option. An aspect that I particularly appreciate of having DIY boards available as well as completed products is that the modules are available to be implemented in whatever manner the builder/integrator chooses. On the other side, if someone just wants an amp, they can get an amp. There are so many folks out there who are passionate about music, audio gear, etc, but don't have the DIY inclination. While it does make owners of Mod 686s an exclusive club, it doesn't have to be limited to DIYers.
I own a number of Neurochrome circuits (Modulus 286, 686, Power 686, Guardian 686, ISS, Meanwell Control, and possibly others.) They have all worked faithfully. My Modulus 286 that I built in January of 2017 has years of runtime. It rarely gets turned off (because I'm lazy). I hope we can get this thread back on track. Cheers.
I own a number of Neurochrome circuits (Modulus 286, 686, Power 686, Guardian 686, ISS, Meanwell Control, and possibly others.) They have all worked faithfully. My Modulus 286 that I built in January of 2017 has years of runtime. It rarely gets turned off (because I'm lazy). I hope we can get this thread back on track. Cheers.