Jump to content

miniDSP SHD Power & Hypex NC400 Class D Amps


Langston

Recommended Posts

Following Claude's strong suggestion for a volume control preceding an N-Core amp I ordered a pair of "FV3 SMD Dact 21 Stepped attenuator 50K Passive Pre-amplifier preamp HIFI audio" to connect between the outputs of the miniDSPs and inputs of the Hypex Ncore NC122MP amps in my tri-amp setup. I chose these attenuators on the theory that being stepped they would be reset-able (by counting the steps from off to the desired volume) with more precision than a standard potentiometer.

 

Using REW I measured one K-Horn, removed the L-Pads from the mid and hi DCX464 in the Top Hat, cabled the volume control in between the miniDSP and the N-Core amps, set the stepped vol to maximum and measured again. Clicked the vol down one, measure, repeat until the SPL of the mid and hi were exactly the same as with the L-Pad. I expected the SPL to be 1/2 a db or so out and require some dsp gain adjustment but SPL was spot on.

I swapped in the second volume control, set it to the same attenuation as the first and ran another REW sweep. Exactly the same SPL in the mid and hi, so I installed the stepped vol control in the second channel.

 

I had used L-Pads with the DCX464s to lower their sensitivity and therefore lower the miniDSP noise.

 

Following on from the opinions on N-Core Hypex amps, the next day I re-arranged the tri amping to use my NAD C725BEE as the mid amp and the N-Core amps for Bass and Hi. After getting the gains adjusted the music sounds a bit warmer, not as dry and sterile as it did with the N-Core amps for Mid range. Time will tell.

  • Like 2
Link to comment
Share on other sites

  • 2 weeks later...

Below are new measurements of my modified NC400 as discussed here.

 

I originally purchased the AHB2 for loudspeaker measurement use due to its amazing specs, sufficient power, convenient size and weight. It didn't have enough power for the loudspeakers I was using at the time, but the Klipschorn changed that. The modified NC400 is better than the AHB2 in every way that matters for measurement use, so it gets that duty now and my Klipschorns get the AHB2.

 

The better specs of the NC400 include more power, equal (and in some cases lower) noise, very similar CMRR and (in some cases lower) distortion, better transfer function with DC coupling, lighter weight, lower output impedance and lower cost.

 

I added a high-current version of the NL4 connector on the back because they are superior and will directly interface with my cabling. BTW, I noticed that the screws on the OEM binding posts that clamp the 12AWG cable from the amp module needed tightening even though I torqued them as much as I dared during the build. Not an ideal design. If you built one of these amps I'd suggest you check on those screws.

 

120V was used for all measurements.
 

Plenty of room for the NL4.

1779392992_NC400NL44.thumb.jpg.9aeb1a4d9b8e3324c02509590c26f781.jpg

 

Also plenty of room for additional wires on the amp module screw terminals.

333966315_NC400NL43.thumb.jpg.d4bd08cc6b26eb498f7ba47b8dd51fb2.jpg

 

Gain and phase response for each measurement. The near zero output impedance results in overlapping gain traces regardless of load impedance.

608407113_GainandPhase.thumb.png.d58da04f232a5c7ade08d6f8d28ef760.png

 

This little amp is a beast. It performs best on the most difficult loads. The following IMD distortion plot is a little busy, but the take-away is that the 1W distortion decreases at higher load impedances as expected, but at 100W and max output (the level required for 1% THD from 10Hz-20kHz) decreases with lower load impedances. This makes me suspect that the input stages are approaching clipping with the higher drive voltages allowed by the higher load impedances.

1759455641_DFDIMDAllMeasurements.thumb.png.25a2ec12fa60a906c0fc8192c4925889.png

 

1W, 100W and max output at 2Ω. (-40dB = 1% THD)

1051724124_THD2.thumb.png.78f9888911ed9849fce1a06f367cf8b4.png

 

1W, 100W and max output at 4Ω. (-40dB = 1% THD)

91743897_THD4.thumb.png.2a6f4a44cb995014878914188925e6b6.png

 

1W, 100W and max output at 8Ω. (-40dB = 1% THD)

1914966690_THD8.thumb.png.201e9238441c902a98149cbadb044214.png

 

An amazing amp. : )

 

God bless you and your precious family - Langston

  • Like 1
Link to comment
Share on other sites

In other news..

 

I learned something about resistors recently that overstated all my amplifier distortion measurements prior to the modified NC400 post above. I'm seeing a reduction in IMD and THD between about 3dB and 10dB using a new set of well-designed power resistors. The good thing is that my past distortion measurements are, of course, comparable with each other.

 

We all know that resistors increase in value with heating, though some are designed with reduced thermal drift, the better/cheaper way to deal with this is to oversize the resistors and/or provide thermal transfer (cooling). Thermal drift tends to affect all frequencies equally in the audio spectrum and it's gradual and predictable.

 

I accidentally discovered something entirely different recently while trying to buy 200W power resistors from Parts Express to "upgrade" the 100W versions I previously bought. For some reason the 200W version showed much higher IMD and THD readings than the 100W version even at very low levels - less than 1V. I thought something was wrong with the amp at first, but found it was entirely due to the resistor and it's nonlinear distortion derived from something called the Voltage Coefficient of Resistance (VCR). Even the 100W power resistors exhibit significant VCR distortion, but I didn't know it until I spent real money on resistors that act like resistors. VCR apparently is present to a minor degree in most resistors, but particularly in the cheap crap that is so hard to avoid these days. VCR distortion is independent of temperature, varies with frequency and can occur in the millivolt range. This isn't something simple like parasitic inductance or capacitance. It generates IMD and THD and probably other bad actors. I thought you could relax when it came to resistors, but you cannot.

 

I sent the 200W resistors back to Parts Express and got some good ones, ARCOL NHS300 4Ω 1%, through Mouser per the recommendation of the ASR guy (Amir). These things are linear to 2,500V. I should have asked John Siau at Benchmark what he used because that's apparently how the Amir figured it out. I also designed a jig for (4) of these resistors that allows me to switch between 2/4/8Ω without removing cables. Very convenient. : )

 

God bless you and your precious family - Langston

 

Edit: here's a link to the 1966 paper that Amir referenced. The authors think measurement of the 3rd harmonic distortion component can reveal many of the evils of passive components (L,C,R), though they spend most of the time discussing resistors. I'm always late to the game. : )

 

Edit 2: here's a link to Bruce Hofer's take on passive component distortion. His math supports the 1966 paper above in that the 3rd harmonic is the tell-tale. So does my experience at this point - keep an eye on the 3rd harmonic and keep the resistors cool and all will be well.

  • Like 2
Link to comment
Share on other sites

  • 1 month later...

I jumped and got me the DIY NC400 monoblocks two months ago, figured if I am going for them than it better be fast before prices go up. So far so good. I could be imagining but after some time they got a bit more polite overall, not so edgy as they were at the beginning. Will have to try them with some more efficient speakers. They seem to be somewhat brighter sounding than my other AB class amps.

  • Like 1
Link to comment
Share on other sites

On 10/19/2021 at 12:45 AM, parlophone1 said:

They seem to be somewhat brighter sounding than my other AB class amps.

 

I think that's a good sign, my guess is that the NC400's near zero output impedance has removed the slight bass boost that typical AB amps provide, and that tube amps significantly provide. The load impedance (loudspeaker) independence of the NC400 may be the best available. It's also a magnificent laboratory amp.

Link to comment
Share on other sites

On 8/30/2021 at 8:23 PM, Wirrunna said:

ollowing on from the opinions on N-Core Hypex amps, the next day I re-arranged the tri amping to use my NAD C725BEE as the mid amp and the N-Core amps for Bass and Hi. After getting the gains adjusted the music sounds a bit warmer, not as dry and sterile as it did with the N-Core amps for Mid range. Time will tell.

So, what has time told you?

 

In my specific case (are they not all specific to each user with unique setups?), it was, primarily, to optimize Yamaha's room correction on the CX-A5200 PrePro with 11.2, but used as 2.2 for music.

 

So while we have similar reasons for using multi channel attenuation, the basic PRINCIPLE is to increase the Voltage SWING in the Pre-Amp, while reducing the Effective Gain of the power amps. This has the net effect of having the signal work further away from whatever noise floor might exist in the chain of events leading to the loudspeaker terminals.

 

If you look across the world as to the mean/average/typical/whatever SPEAKER Sensitivity numbers, it is about 88 db/2.83 volts power amplifier output. I sincerely believe that the operating "window" of all room correction firmware in AVR's and Pre-Pros uses this worldwide loudspeaker INEFFICIENCY that dominates majority of customers as their design target. So basically, those feedback electronics Freak Out and don't know how to handle it as well as something they expect to hear back at a much lower level.

 

So when the test signals are sent our and fed back into the respective Microphone port on these types of Control Units, they "hear" about 12-16 db LOUDER sounds from their bottom to top frequency sweeps. The net effect of this is it PUSHES the operating window of the PEQ's and System Gains way OUTSIDE of the "comfort zone" for which, they were designed.

 

When using high efficiency speakers with horns ranging from 100-112 db sensitivity, we end up confusing these units by creating sounds that are too loud for them to handle. On the deeper, technical side, I could be off base on this, but I have REACTED to it all based on the USER side of the equation.

 

Besides, ganged Potentiometers are relatively CHEAP vs. the rest of all this HiFi STUFF!!

 

 

  • Like 2
Link to comment
Share on other sites

14 hours ago, Langston said:

 

I think that's a good sign, my guess is that the NC400's near zero output impedance has removed the slight bass boost that typical AB amps provide, and that tube amps significantly provide. The load impedance (loudspeaker) independence of the NC400 may be the best available. It's also a magnificent laboratory amp.

 

I guess that might be true and confirmed by my ears also.

What bass is left here, it is tight and articulate. The rest of the spectrum ranging from mid to high freqs might be described in audiophile words as "clinical" and more "accurate". So, that would be a confirmation of NC400s being a laboratory amp, they reveal the true nature of a speaker.

 

  • Like 2
Link to comment
Share on other sites

On 10/21/2021 at 3:00 AM, ClaudeJ1 said:

So, what has time told you?

That I have a lot to learn.


When I swapped the miniDSP outputs with Bass to the N-Core amp and Mid to the NAD I noticed two things immediately,
1. Bass tighter and maybe even a bit louder on the N-Core than it was on the NAD
2. Residual noise in the horn was different and a bit louder with the Mid on the NAD amp. No more frying bacon, rather a machine like hum. Only noticeable with my ear within a few inches of the Eliptrac horn mouth.


I fixed the noise problem by adding an L-Pad to cut the DCX464 by 6db and re measured the levels and adjusted the miniDSP output channel gains.


When tri-amping you connect each driver directly to its amplifier, no crossover network in between knocking a few db off, so when that driver is 110db at 1m for 1w and is mounted on a horn it is a very efficient transducer so unless the amp chain noise floor is very low it is going to be audible.


But the biggest change came later when I redid the EQ and rePhase settings, here are the curves

https://community.klipsch.com/index.php?/topic/195572-bc-dcx464-as-the-mid-hi-drivers-for-a-k-horn/&do=findComment&comment=2586957

 

  • Like 1
Link to comment
Share on other sites

Very interesting thread I'd like to follow by way of commenting.  I believe I first spoke with @Langston a few months ago when you kindly supplied your listening impressions of the ABH2 as I was pondering a purchase.  Further research to appease the chief accountant/wife led me to the ASR review of the Benchmark and then on to the Hypex amps.  An enterprising engineering student about an hour up US RT 2 from me had started Buckeye Amps out of his home focusing designs around the Hypex modules.  The design was basically crowd-sourced on the forum and the sincerity (and low cost) of the endeavor was too much to resist and I purchased my first solid state amp in quite some time for the princely total of $500 delivered.  The CFO was pleased.  2 months later (the kid has gotten quite busy with orders) the NC252MP based bread box arrived at my doorstep. So far the experience has been extraordinarly positive.  It mates exceptionally well with my Minidsp SHD and I left the damn thing on all weekend and it barely got warm.  Listening impressions are equally positive.  It's preternaturally quiet at idle and smoothly transparent at volume on my Volti/BMS/Fostex t500a MK3 khorns.  Anyway, thank you all for being smart and enthusiastic (and mostly) objective.  

  • Like 3
Link to comment
Share on other sites

31 minutes ago, Godataloss said:

So far the experience has been extraordinarly positive.  It mates exceptionally well with my Minidsp SHD and I left the damn thing on all weekend and it barely got warm.  Listening impressions are equally positive.  It's preternaturally quiet at idle and smoothly transparent at volume on my Volti/BMS/Fostex t500a MK3 khorns.  Anyway, thank you all for being smart and enthusiastic (and mostly) objective.  

I first heard a Hypex based stereo amplifier about 7 years ago at Axpona, Chicago. The custom amp was based on the Hypex Tech. a few generations ago. It was hooked up to the incredibly sensitive Sadurni Stacatto Horns (who got best of show) and the designer, Mr. Sadurni himself told me they only needed Milliwatts to operate.

 

Nonetheless it's the only speaker at the show that literally brought tears to my eyes when I listened to "Come Away With Me" by Nora Jones, along with a few other demo songs. The micro detail, atmosphere, air, and Imaging was the best I have ever heard.

 

This is what sold me on the Hypex technology. I then read Bruno Putzey's White Paper on the Technology involved and I ordered Two NC-400 Kits from their site in Europe................Sound UNHEARD!!! I have only turned them off once in 6 years, when I moved to a house 3 years ago!!

 

"Neither dirt nor Fairy Dust" is a truly great motto for that amplifier. Time, my ears,  and others here have proven that to be correct!

  • Like 3
Link to comment
Share on other sites

4 hours ago, Godataloss said:

Very interesting thread I'd like to follow by way of commenting.  I believe I first spoke with @Langston a few months ago when you kindly supplied your listening impressions of the ABH2 as I was pondering a purchase.  Further research to appease the chief accountant/wife led me to the ASR review of the Benchmark and then on to the Hypex amps.  An enterprising engineering student about an hour up US RT 2 from me had started Buckeye Amps out of his home focusing designs around the Hypex modules.  The design was basically crowd-sourced on the forum and the sincerity (and low cost) of the endeavor was too much to resist and I purchased my first solid state amp in quite some time for the princely total of $500 delivered.  The CFO was pleased.  2 months later (the kid has gotten quite busy with orders) the NC252MP based bread box arrived at my doorstep. So far the experience has been extraordinarly positive.  It mates exceptionally well with my Minidsp SHD and I left the damn thing on all weekend and it barely got warm.  Listening impressions are equally positive.  It's preternaturally quiet at idle and smoothly transparent at volume on my Volti/BMS/Fostex t500a MK3 khorns.  Anyway, thank you all for being smart and enthusiastic (and mostly) objective.  

 

If it wasn't for the wait into late January now, I would seriously contemplate one of those amps in the 2 channel NC502MP config.  But I can't get past that 3 month wait.  

Link to comment
Share on other sites

I am currently comparing nc400 monoblocks to the Benchmark on passively crossed Jubilees. I’m nearly finished with a modest room remodel so I’ve had little time with either; the Hypex took their position this morning after about a week on the Benchmark. 

Link to comment
Share on other sites

On 8/11/2021 at 4:06 PM, ClaudeJ1 said:

Since you have one. Why didn't you use it to compare the AHB2 to the NC-400? 

Because then he wouldn't have been able to tell you he enjoyed the AHB2 more because he wouldn't have been able to discern the two amps. 

 

I'm entirely convinced that had he truly removed his own bias from the equation (the only way you can, blind), then he would have surprised himself even further. These amps measure so close to each other there's no way anyone could reliably discern them blindly. You'd never convince me otherwise. I've personally experienced a blind test and let me tell you, everything I thought I knew went out the window as soon as I couldn't see which shiny device was creating sound. Expectation bias is a powerful thing and it WILL affect the way you hear gear. 

  • Like 1
  • Thanks 1
Link to comment
Share on other sites

7 hours ago, Godataloss said:

His sample size is not large enough to consider the results.

 

I agree! : )

I'll get around to doing the ABX test again before too long - for the full 25 trials. I started working out with a personal trainer about a month ago that also trains soldiers at the local Naval base and the pain he puts me through reminds me of the ABX test.

 

From the QSC ABX manual. Sorry for the poor quality, but it's the best I'm aware of.

 

434900861_QSCABX.thumb.jpg.f17400eae9a5c6e54b5ef950970c1960.jpg

 

Edit: after pasting the above from the QSC ABX manual, I got interested in figuring out the probability math and it turns out that my earlier ABX test where I got 8 right out of 11 trials results in a confidence of 92% that I heard a difference between the amps. If you are interested in the math I just added it to the original post.

  • Like 2
Link to comment
Share on other sites

31 minutes ago, Langston said:

 

I agree! : )

I'll get around to doing the ABX test again before too long - for the full 25 trails. I started working out with a personal trainer about a month ago that also trains soldiers at the local Naval base and the pain he puts me through reminds me of the ABX test.

 

From the QSC ABX manual. Sorry for the poor quality, but it's the best I'm aware of.

 

 

You need to live stream that so we can all partake in your discomfort, lol.  The testing, not the workouts...

  • Like 1
Link to comment
Share on other sites

To me the beauty of ABX testing is if the ABX hardware is implemented well you can switch between 2 items you are comparing almost instantly and not know which is which. To me it is utter nonsense that after a half hour changing cables, switches, etc.etc.etc.... that you can count on your auditory memory to not fool you into believing that the gizmo you just hooked up that is shinier and prettier does not sound better.

  • Like 4
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...