New NCDX(i) module from Lars Clausen

[Lars Clausen]

Hi EVA nice to see you. Yes in most cases you are right. However the MCU solves one very important function for us. If there is a slight DC on the input, and no music signal, the power supply will start pumping from one rail to the other. This can result in failure due to overvoltage, and damage the power supply.
With a MCU we monitor both rail voltages, and in case of imbalance, we shut down, thus keeping everything safe. The same function cannot be performed with just a simple DC sensing circuit.
The only other way to solve this problem (while keeping a wide supply voltage range) is by a DC blocking capacitor in the signal path. In this module we find the MCU is the better of these two options.
Since a little bit bigger MCU cost almost the same as a really small one, we can also add a lot of other nice features to the amp, without any significant cost.
Like pop noise eliminators just to mention one.
And now also a nice little LCD.

[soongsc]

The original NCD with optical volume control is amazingly good. I really wonder what further improvement in audible performance can be expected with the NCDXi.

[teflon1]

This begins to seem like Spy Vs Spy in a circle dance or a game of twister.

Eva, by your own example, you're not doing "MCU assisted Class D", either.

For that type of base functionality as described, the MCU is overkill. There's other methods of achieving the same in each case, and some even simpler. They still all require a certain level of interfacing, sensors and triggers etc. It's not really as though you're saving on much of anything, certainly nothing that will win any design awards.

Whichever means are used to implement such base functionality should be completely transparent to the end user. To use it as advertising is a gimmick to make people think there's some degree of intelligent control happening and that's not the case we see here. You may be making better use of available options in a completed product, but none the less it is not MCU assisted class d which is something else entirely.


I think it also calls into question the validity of the design itself if you need it simply to ensure base functionality, as seems to be the case here.

With regard to their additional noise, it will certainly be higher than were it not there. Closer to the truth though is you can't say it won't have a deleterious effect unless you're directly looking at the final product of autoroute with fingers crossed as a design methodology. That is, MCU supply rail referenced to the negative power rail, spanning the length of the PCB, violating false clean zones and so on. OK so maybe the MCU isn't the problem, and maybe that's also the point, but neither could it be the solution.

I would like to see Lars explain how his output stage, or anything, is protected in the case of a fault though, when all it can do is disable what's apt to be an already blown output stage. So much for intelligent control.

Would a very good amp need a pop noise eliminator too? How much code does it take to iron out that hardware flaw, oh but you'd still need the relay in the signal path wouldn't you. I'm sure it will sound great though.

Also, as the modules no longer undergo the lengthy three hour nano alignment trademark process, which was also the main and only reason for the cost of it being twice that of anything remotely comparable, should the price not have undergone a drastic reduction at some point, or is it more profitable to keep marketing it as being extra ordinary? This fire sale offer seems to hint to answer.

Best of luck though Lars, I hear you are big in Japan, says so on your website at least, and I know you're honest guy.

[soongsc]

I believe the protection scheme had been discussed already in the original NCD thread.

[Lars Clausen]

For that type of base functionality as described, the MCU is overkill.

A petrol engine also does not strictly need a microprocessor control, to run, they did cars without back in the 1950's. And they were working fine, you just add a lot of thermostats, regulator shafts and pegwheels instead.
And the same goes for a washing machine, aircraft, ship engine, toaster, central heating boilers, etc etc.

They can all be made without a MCU, however. In todays world a MCU just happens to be the most flexible, reliable and effective solution to controlling the processes in all of the above tasks, and Class D modules too :-)
Whether it's an overkill is an interesting thought. Today a minimum MCU system can be made up for like 50c, or the price of a few passive components. Even if it only flashes the LED for looks, you can hardly call that overkill. Not alone would it be very difficult to mimic the same functions with old-school CMOS chips, voltage comparators and 555 timers. Especially at a comparable cost.

The MCU is simply better in every aspect.

Lastly we did of course continue nano alignment og NCDX and NCDXi series, as it is at the core of making ultra high performance Class D technology, that is well worth the extra cost of the modules. (If you count by sound quality).

[Eva]

A small MCU is not overkill, it's just that once there is one it can do many things... Those modern MCU are quite powerful, even the SO-8 ones (from the assembler programmer point of view)...

[teflon1]

Point received Eva. My use of the term overkill should have been restricted to this particular application, otherwise I'll agree with you up to the assisted part.

It's not just about the number of possibilities, but the ones that get employed and how well they weren't. When you have protection that offers false security, isn't robust, isn't adaptable, it is useless, especially on a module destined for hobbyists. Hence, complete overkill because you can do useless with even less. Just think, if you buy two modules without an MCU, you'd save a dollar. Then, you could go down to the dollar store, and put a downpayment on something else, like monster cable.

Yes, if all it does is blink a light I do consider it overkill because I doubt you'd need to worry about anybody stealing it, what with that heavy transformer and all, and it's just not an application that benefits from a blinking light, which just makes people nervous these days.

Newer cars aren't often as sensibly diagnosed as the standards, aren't. They flash dummy codes at you that indicate which of a possible set of systems may have failed, then narrow it down by guessing which sensor to bill you with first. According to you however, computer control obviates the need of any transducers, so they must really be getting ripped off as MCUs use ESP now. Possibly as a result of the questionable effectiveness of ESP, it's interesting to note that many vehicles on the road these days can't run at all decent until the computer bails out and reverts to failsafe settings, where the fuel economy and all that goes out the window tenfold. Is it really doing its job then, or is it too a security blanket that suffocates? Why do people get pulled over and fined for failing emissions tests? Didn't the ever powerful MCU know already? It's not like we all know someone who's experienced a problem with a computer, program or peripheral anyway, right.

Cars however do offer adaptive, or "assisted" control, with the intent of increasing efficiency, reducing pollution, ensuring robustness with different fuels and mothballs or nail polish remover, at different times of year, in all feasible conditions, per cylinder if you like. You can even have full on race setups with street mode at the touch of a button and full real time graphical output of all monitored variables and parameters.

Even in that scene there's such a thing as too much and it's starting to cost lives by removing the feel of the road from the driver who'll then lack the ability to control the vehicle properly and bailout simultaneously with the control system. Passenger jets and fighters alike have fallen out of the sky for the very same reasons.

People tend to overlook the toys longterm, however, and focus more on primary functionality. Lacking that, you get manufacturers with more cars on their lots then they can throw matches at. All that leads to is good, public, fire insurance.

None the less, if their MCUs just flashed the hazards and turned hard right at the first sign of any trouble, I'd question that new technology as well. Although they're doing more with it than imitating 50's tech, you're not, so the line drawn from that to your class d is a broken one, much like the ground plane.

There may be higher expectations for it one day, but one day, I would like an amp with butterfly wings. I suspect I"ll be seeing that first. Having an MCU doesn't automatically equate to having bettered every aspect, you've also got to use it wisely.

[Cavok]

Originally Posted by teflon1

This begins to seem like Spy Vs Spy in a circle dance or a game of twister.

Teflon1, - In my view it must be a Competitor Vs Competitor game where I bet you are one of the participants....

Based on your cryptic posts - you are definately not interested in a succes for Newclassd..

Cheers

C

[Cyan]

How much do you think a stereo amplifier with the new module would cost? With high quality components that is.

[ttako]

Hi Lars,

What are the tech data of the new NDX modules? Are they capable to drive lov /difficoult impedance speakers? What is the current limit (rms)? what is operating voltage range of the modules?

Thanks,

Tamas