New Class D amp NAD 3050 getting HOT

[JungleXray]

How hot is too hot?

I should start by saying over the last couple of months my vintage 1980's era McIntosh was finally starting to fail.

A MC502 power amp to be exact. A barebones (compared to the rest of their catalog) class AB, Direct coupled, 50 watt into 8 ohms beauty/beast of an amp. One that paired really well with my vintage Klipsch Cornwalls. But alas, the speaker channels were starting to fail. Getting quieter and quieter until producing no sound. A flip of the speaker on/off switch would be a temporary remedy. But quickly burdening me with getting my ass off the couch to flip it again. No amount of Deoxit would fix it. After some quick research, some guy on the internet said that that speaker switch is a common failure point. A cheap fix I'm sure, but being that the amp is approaching 40 years old I thought maybe it be better to replace it. Also I'm sure popping the hood on that thing would reveal other time bombs that need to be fixed. Looking at you capacitors.

So I thought I'd replace it with a new class D amp. It's the efficiency that really intrigues me. Modern technology. State of the art.

But black boxes from China just don't really do it for me. Just my preference. I want it to look as good as it sounds. Enter the NAD 3050.

Every thing I could want in a an amp. Modern and vintage. Beautiful, practical, and "economical". This could replace my entire system if I wanted it too.... or so I thought.

The first bummer: I planned on using this as a power amp only. Pluggin' my Mini DSP SHD into the main in. But, if I do that the meters don't work. And let's be honest, the meters are 99% of my purchase of this amp. Bummer. I really like the versatile of the SHD with full Dirac. To get Dirac bass management on this would cost me another 500 bucks for the modular card BluOS thing.

Okay, so I'll just use the SHD as a transport and plug the digital out into the s/pdif input of the NAD. But I don't know how well this integrated NAD performs. Soundstage Network did measurements, but to me that look just OK in some areas like SNR and dynamic range. But I'm not the best at deciphering all those numbers sometimes.

Maybe I could experiment and put the SHD in the pre-out to main in loop?

The second bummer is this thing get pretty hot. Not like burn your hand hot but I would say past warm! I thought this thing would be cool and efficient! What's getting hot in there? I could leave the McIntosh on all day and it would barely get warm thanks to its massive heat sinks.

SO. Whats going on here?

 

[Doodski]

It is a decent 100 W/ch integrated. The dynamic and peak power specs are respectable as is the 20 Hz to 20 kHz specs. NADs are usually a med hot runner. Not hot nor cool or warm but to the hotter zone. If you have trouble touching it for several seconds it most likely exceeds 85C and that's pretty hot. Any chance we can see a pic(s) of the insides? I Googled for images but had very poor success in that search.

 

[MachOne]

The NAD D 3020 runs hot too. I have one of the Hypex modules used in it here, and it was the output inductors producing the majority of the heat.

 

[JungleXray]

So it is pretty normal for it to get a bit heated? Im just not use to that. The McIntosh never got more than barely warm and know that cool running efficiency was the point of Class D in general.

It's a very nice integrated. Inputs from HDMI eArc to MM phono. Dirac, Bluetooth, airplay, AND a remote . Oh! and big meters. A lot to like. But I don't know if I love it yet.

Any chance we can see a pic(s) of the insides? I Googled for images but had very poor success in that search.

Possibly! I'll have to take a look. If its easy enough to get inside I could get some pics.

 

[JungleXray]

Here is just a few more pics fresh out of the box.

 

[Doodski]

Here is just a few more pics fresh out of the box.View attachment 297454View attachment 297455View attachment 297456View attachment 297457

It's a handsome design. Looks very nice with that retro layout and meters.

 

[dougi]

Hmm, I have the LE version and have not noticed it getting hot. I will "have a feel" though the next time I run it.

Yeah, measurements are certainly "good enough" but won't be breaking any SINAD records. Yes unfortunately the meters in output mode actually rely on the pre-out so won't work with an external pre input into the main-in. If you are interested in the metering, I did some tests in this post.

 

[JungleXray]

Hmm, I have the LE version and have not noticed it getting hot. I will "have a feel" though the next time I run it.

Yeah, measurements are certainly "good enough" but won't be breaking any SINAD records. Yes unfortunately the meters in output mode actually rely on the pre-out so won't work with an external pre input into the main-in. If you are interested in the metering, I did some tests in this post.

On the meters side, towards the back if I remember correctly. Just wondering how hot is normal, I was expecting none heat lol.

I will have to take a look at your work tomorrow. I'm interested.

 

[dougi]

On the meters side, towards the back if I remember correctly. Just wondering how hot is normal, I was expecting none heat lol.

I will have to take a look at your work tomorrow. I'm interested.

In the pic below, the power supply is far right, the UcD module and output to speakers on the left of it

 

[Doodski]

That's not your moma's and papa's NAD that's for sure. It's SMPS and class D from the look of it. New technology. It's Hypex.

 

[MachOne]

Interestingly, they have fitted larger output inductors to these newer versions of the UcD module. On my V1 module from the D 3020 they use a much smaller upright ferrite encased inductor.

 

[Doodski]

Interestingly, they have fitted larger output inductors to these newer versions of the UcD module. On my V1 module from the D 3020 they use a much smaller upright ferrite encased inductor.

They could have changed the windings and used a different sized value for the donut.

 

[dougi]

That's not your moma's and papa's NAD that's for sure. It's SMPS and class D from the look of it. New technology. It's Hypex.

Yeah it seems to use the old old Hypex UcD 102 but somehow makes some changes to get more power out of it, which it definitely does. Would just the inductor change be responsible for that? Or different FETs as well?

 

[Doodski]

Yeah it seems to use the old old Hypex UcD 102 but somehow makes some changes to get more power out of it, which it definitely does. Would just the inductor change be responsible for that? Or different FETs as well?

The heatsink is a improvement right off the bat I suppose. The FETs may be upgraded but with that heatsinking alone they might just pull off more power. It's all about the heatsinking with those FETs there. The inductor value I'm not sure about. I need to up my reading on class D before I get into the component level circuitry with peeps. I have a decent idea what is happening but not the nitty gritty get right into it viewpoint. Now that I said that. That inductor I think is sized according to it's required inductor value for low pass and is not determining RMS power output.

 

[JungleXray]

So today I took a thermometer to the top of the amp. It measured 104.2 degrees Fahrenheit. Or about 40 degrees celsius.

Is this a normal operating temp?

 

[JungleXray]

Hmm, I have the LE version and have not noticed it getting hot. I will "have a feel" though the next time I run it.

Yeah, measurements are certainly "good enough" but won't be breaking any SINAD records. Yes unfortunately the meters in output mode actually rely on the pre-out so won't work with an external pre input into the main-in. If you are interested in the metering, I did some tests in this post.

Hey thanks for this. I was wondering what was happening with the meters using the digital signal from my SHD. I had figured 0 would correspond to 0dBfs. I thought I was clipping, which didn't make sense to me given the SHD is a unity gain preamp. So 2.0 v output. This cleared up that confusion and allowed me to crank the SHD (when Dirac is off).

+3 on the meters = 0dBfs

The manual is pretty sparse... at least the one I got.

 

[Doodski]

So today I took a thermometer to the top of the amp. It measured 104.2 degrees Fahrenheit. Or about 40 degrees celsius.

Is this a normal operating temp?

Average consumer silicon is stable to ~85C and the next grade is usually 115C and then 130-135C etc. So 40C at the case is most likely OK. I've handled silicon that specs out dynamically @ ~177C.

 

[JungleXray]

Average consumer silicon is stable to ~85C and the next grade is usually 115C and then 130-135C etc. So 40C at the case is most likely OK. I've handled silicon that specs out dynamically @ ~177C.

Okay so I worried for nothing. That good! I just wasn't expecting that from my first Class D amp.

Like I said before the McIntosh never got warm, but I guess that's because a third of was heatsinks!

 

[restorer-john]

Firstly they are not at all efficient at idle to low powers. Measure the idle consumption with a killwatt and consider, all that is going up as heat.

The idle losses are significant and the internal temperatures often get well past the levels for long term reliability with much of this Class D stuff.

75 to 85 degrees C (185F) is not unusual at idle on the components I circled. They basically cook the nearby capacitors. Often the auxilliary SMPS transformer runs very hot.

 

[Deleted member 48726]

In general the idle losses are much more concentrated to smaller areas and components than before. There is no escaping the law of thermodynamics.
Poor thermal design of this NAD. Yes there's a top "vent" though it is more a design feature than anything else since there are no vents on the sides or on the bottom?