• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Review and Measurements of Hypex NC400 DIY Amp

6speed

Active Member
Joined
Nov 7, 2018
Messages
128
Likes
84
Location
Virginia, USA
I believe hypex goes into standby, so possibly different in this case.

There is a standby toggle available, but I have it wired so when it's powered, it's on; I bet most DIYers do the same thing. Instead, I use a power sequencer that turns outlets on after I hit power on my remote.
 

rebbiputzmaker

Major Contributor
Joined
Jan 28, 2018
Messages
1,099
Likes
462
There is a standby toggle available, but I have it wired so when it's powered, it's on; I bet most DIYers do the same thing. Instead, I use a power sequencer that turns outlets on after I hit power on my remote.
I do not use standby either.
 

rebbiputzmaker

Major Contributor
Joined
Jan 28, 2018
Messages
1,099
Likes
462
They need to be called on all their published specifications if they are found to be inflated. Hypex are clearly not expecting typical DIYers to have the gear or the knowledge to actually confirm their numbers, let alone challenge them.

I'm sorry, I see it as a bunch of followers of a 'guru' lapping up everything as gospel and never daring to question the general consensus for fear of being labelled a 'non-believer' and kicked out of the audiophile playground. Thank goodness none of us have that fear around here.

Based on what I see, the SMPS needs active cooling across the flag heatsinks and the PSU as a whole. ATX PSU style. Even that may not be enough for the main transformer. Those spot temps are ridiculous after <4mins, especially as the thing took another 2 mins+ to reset. Output devices on an adequate heatsink, no problem, they can sit on 110 degrees above ambient and not blow up, but none of those heatsinks are adequate for just one NC400 are they?
I think you need to tell your hobbyist/manufacture friend his new amps need to ship with a fire extinguisher. Put another amp on the barbie.

 
Last edited:

Sancus

Major Contributor
Forum Donor
Joined
Nov 30, 2018
Messages
2,923
Likes
7,616
Location
Canada
They need to be called on all their published specifications if they are found to be inflated. Hypex are clearly not expecting typical DIYers to have the gear or the knowledge to actually confirm their numbers, let alone challenge them.

Based on what I see, the SMPS needs active cooling across the flag heatsinks and the PSU as a whole. ATX PSU style.

I agree, and the comparison to the PC market is an apt one. Over there, people typically run synthetic benchmarks to test their new PC builds. Those benchmarks produce far higher power consumption and hence heat output than any kind of "normal use", yet if my CPU in stock configuration overheated after <5 mins of prime95 benchmarking, I would be unhappy, and reviewers absolutely would call out a manufacturer for that kind of poor performance.

The primary difference seems to be that in the PC world, anyone can download and run stress test software, whereas for amplifiers you need special gear and knowledge. So I suppose there is less pressure on the manufacturers to properly test to spec when it is so rare that the end user will actually do so.

It's not like this power supply is cheap, either, hypex sells it for €180! You get a whole lot of well-engineered ATX PSU for that much money. The addition of an (optional, given that it's DIY!) plug-in temperature controlled fan would cost maybe $5-$10 to add.
 

March Audio

Master Contributor
Audio Company
Joined
Mar 1, 2016
Messages
6,378
Likes
9,317
Location
Albany Western Australia
Well I was getting all excited to FTC test the P252 after the dummy load resistors arrived.........until.....

r1.jpg
 

Purité Audio

Master Contributor
Industry Insider
Barrowmaster
Forum Donor
Joined
Feb 29, 2016
Messages
9,049
Likes
12,147
Location
London
Does anyone ( else) test amps for four/five minute at max?
Keith
 

Armand

Member
Audio Company
Joined
Jun 25, 2018
Messages
93
Likes
321
The tape Amir was using was acting like a blanket.
No. With thermal imaging one must know the emissivity of the object. Bare metal has very low emissivity and will appear much cooler than the actual value (many tens of degrees). The tape on the lower part has emissivity of about 0.95 and will give close to correct temperature. Since the tape is in direct contact with the metal and very thin it will have the same temperature as the heatsink. This applies of course also to the metal top of the capacitors that will appear cool. The reading must be done on the plastic.
https://www.flukeprocessinstruments...-center/infrared-technology/emissivity-metals

Like I said earlier, it is the coils that get warm on both the PSU and the amplifier. It is the capacitors closest to these coils that have the most difficult life.
 

rajapruk

Senior Member
Forum Donor
Joined
Sep 22, 2018
Messages
472
Likes
489
Would using a linear PSU with nc400 reduce all this heat a lot?
If so, what would be the approximate cost and size/weight of a linear PSU driving dual nc400 to full power?

Or what kind of heatsinks would be needed for the SMPS600 to have 15 years longevity, you think (driving 8 ohm woofers with 120W peak)? This kind of alu-case would be enough?
http://www.modu.it/pesantedeng.html
Or it has to have fans as well?
 
Last edited:

Armand

Member
Audio Company
Joined
Jun 25, 2018
Messages
93
Likes
321
Here is an image of a SMPS1200 just before it went into shutdown.
1546514231832.png


123 degrees celcius on the wire and 96 degrees on the bobbin. The closest capacitors are somewhere around 70 degrees (i dont have the exact reading)


Here is an image of a NC500 module after pretty hard load of about 200W without active cooling and abient temperature of 35 degrees
1546514771760.png


Coil wires are 98 degrees, coil bobbin is 86 degrees and the circuit board between the row of capacitors to the right is 74 degrees. The capacitors closest to the coil are a bit warmer and maybe close to 80 degrees.
 

rajapruk

Senior Member
Forum Donor
Joined
Sep 22, 2018
Messages
472
Likes
489
This is not good, right? (caps being close 80 degrees hot)

Can you show this picture/measures when playing "normal" music on "normal" level?
And also after idling a long time?
 

March Audio

Master Contributor
Audio Company
Joined
Mar 1, 2016
Messages
6,378
Likes
9,317
Location
Albany Western Australia
Does anyone ( else) test amps for four/five minute at max?
Keith
Well it is part of the FTC test for comparing amp outputs, but as we have discussed it bears no resemblance to real world load conditions, where music RMS levels may be 1/5 to 1/8 of peak.

We absolutely need a basis for comparison and rating of amps but.........
 

Armand

Member
Audio Company
Joined
Jun 25, 2018
Messages
93
Likes
321
Would using a linear PSU with nc400 reduce all this heat a lot?
If so, what would be the approximate cost and size/weight of a linear PSU driving dual nc400 to full power?

Or what kind of heatsinks would be needed for the SMPS600 to have 15 years longevity, you think (driving 8 ohm woofers with 120W peak)? This kind of alu-case would be enough?
http://www.modu.it/pesantedeng.html
Or it has to have fans as well?

A linear PSU have efficiency of about 0.7 to 0.8. To be able to deliver 2x400W you would need at least a 1kVA transformer and something like at least 100.000uF or more. It will be much heavier (12-15 kg), bulkier and will also generate more heat. IMO there is no advantages using linear PSU any longer. Modern low noise SMPS does not have the noise problems associated with the older hard switching regulated SMPS. (But that is another discussion I am sure there are people that will disagree :)

To ensure a long life I would not use any SMPS for anything more that half rated continous power without active cooling.
For the Hypex SMPS600 this are the numbers in the datasheet: Peak 600W, continous 150W
My recommendation for continous power without active cooling is 75W.
 

Armand

Member
Audio Company
Joined
Jun 25, 2018
Messages
93
Likes
321
Well it is part of the FTC test for comparing amp outputs, but as we have discussed it bears no resemblance to real world load conditions, where music RMS levels may be 1/5 to 1/8 of peak.

In most situations this is true, but as I showed with the Metallica track it is in fact only 1/2...
 

Armand

Member
Audio Company
Joined
Jun 25, 2018
Messages
93
Likes
321
This is not good, right? (caps being close 80 degrees hot)

Can you show this picture/measures when playing "normal" music on "normal" level?
And also after idling a long time?
The NC500 and NC400 are not the same but their spec is similar.
Here is a shot of the inside just seconds after I took off the lid of a prototype cabinet after playing 20W continous on one channel until temperatures were stable. Temperature inside cabinet 45 degrees. The SMPS3k transformer and the NC500 output coil at 88 degrees. (43 degrees over ambient) I don't have exact readings of capacitors but from the colour scale it looks like something around 60-65 degrees. (15-20 degrees over ambient. The temperatures when idle is the same or perhaps just 2-3 degrees lower.
I am not too worried about these temperatures. Transformers and coils at 90 degrees is not a problem and capacitors at 65 degrees should not be a problem for 105 degrees rated properly designed capacitors. It is when power increases a lot the problem start. The ambient temperature inside the cabinet raises, the coils temperature raises and the capacitors starts to create heat themselves due to increased ripple current. All this combined will create unhealthy conditions inside a poolrly vented cabinet if used hard over longer period of time.
1546525604201.png
 

Armand

Member
Audio Company
Joined
Jun 25, 2018
Messages
93
Likes
321
OK, here are the results of thermal stress test.

I hooked up input to one channel and let it the unit warm up driving that one channel at 40 watts. I logged THD+N during that time and got this:

View attachment 19779

There is some equilibrium that occurs after 3 to 5 minutes but then thermal effects start to set in and distortion increases a bit (1 dB or so). An arrow to the heart of people who think they should leave the gear on 24/7. :D

I then jacked up the power to 330 watts or so and started to cook the unit. Here is the above graph again:

View attachment 19782

Here is the power measurement showing the shut down with power dropping to zero:

View attachment 19783

I had hooked up a 4-channel themorcouple and here are the results at the end of warm-up (left) and max power (right):

View attachment 19784

As you see, the large power supply heatsink was the hottest at 81 degrees. Strangely, shortly after I took the above right picture, two of the thermal probe started to glow red in the middle and completely melted the insulation!!! Cheap junk. :)

Thermal imaging showed I had probed the wrong things:

View attachment 19785

As noted, the hottest components were actually the inductors/transformers! Both in the amp and power supply. Max temp for the large power supply transformer is about 101 degrees C. Wonder what their wire insulations are rated at.

After the stress test, I ran a power versus distortion test:
View attachment 19786

Power increased to 315 watts which is not much more than when both channels were driven. Distortion performance is the same as when the unit was much cooler in the original review.

Now please me go and cry over my damaged thermocouple meter.....

Thank you for the test. From the thermal image it looks like the capacitors on the left channel is around 80 degrees. (red colour). But then again the capacitors on the SMPS are also red and the thermocouple measuring there showed 58 degrees. Also the thermocoupe on the large PS heatsink show 81 degrees while the thermal camera show 100 degrees. Which one should we trust do you think?
 
Top Bottom