Neurochrome HP-LOAD Review (Headphone Dummy Load)

[amirm]

I sincerely hope @amirm will use this going forward in all your reviews.

I have actually been using it for the last six months or so for almost every review! Maybe even longer. I have just been too busy to do the review of it and hence lack of mention. So yes, of course I will be using it moving forward as well.

 

[amirm]

If this dummy load is resistive then why go through this trouble, why not just wire up a resistor and be done with it ?

It is not that simple. First, you need almost a dozen different loads. Second, if you want fast testing you want automation that this unit provides. Third, you have no less than four different headphone connections to interface to, either differential ("balanced"), or not. And if you want to test for stability, you need to provide the combination with different capacitances this load presents. Importantly, you want 4-wire Kelvin measurements.

The simplest solution is to build separate loads for each use which is what I did. But as Tom and I have noted, it is a pain to manage and you wind up with different accuracies.

Part cost is significant whether you build your own or buy Tom's box. I had purchased the same Caddock resistors for my load. You need them or you get hit with VCR induced distortion (voltage coefficient of resistance). At the level of precision we are testing these state of the art amplifiers, you need absolute precision. Not so easy to do with just throwing some resistors together.

Note that there is value in your time. It took me a lot of trial and error to build optimized loads myself. Buying and testing a lot of different resistors is not free. I could be using that time to test and review products.

 

[levimax]

This is a review and detailed measurements of the Neurochrome HP-LOAD "dummy load." It was kindly donated to me by its designer, and our valued member, @tomchr. It costs US $1,499 from him direct.

I don't know why but for some reason, companies sell you instruments for tens of thousands of dollars to measure audio devices but they don't sell a necessary dummy load. A dummy load is what simulates a headphone or speaker so that we can have a consistent target to use and compare. I guess they assume if you are going to measure audio, you know how to build one. That is true but building a high performance dummy load is not easy. The load itself can contribute distortion and noise to the device being tested.

A year ago Tom approached me and asked if I was interested in a properly built dummy load for headphone testing. I said yes as like him, I had cobbled together my own and was not happy with it. I actually used three different ones which I used for different circumstances. He went to work and sometime later sent me a box. I must confess, I expected to see a hobby box and instead, I received this work of beauty:

View attachment 144424

Wow! Power it on and the LEDs light up and you can push a nice feeling button to cycle between them. You can not only control the load resistance but also capacitance. Two cables, one for balanced and another for unbalanced, connect to that DB-9 socket and provide proper Kelvin-type measurements.

Looking at the back, you see a very nice automation interface:

View attachment 144425

Audio Precision has an AUX bus that you can control through automation. This port is compatible with that so you automate testing of the amplifier across the full set of loads with just one click of the mouse! I am not yet using this but should do to save some work.

As you see, power is provided by a switching power supply. I should note that in one or two cases, I somehow created a ground loop that did not exist in my simpler setup. Disconnecting the power supply solved that as the load continues to represent the default impedance even when powered off. I could never replicate this problem and it has not happened for a long time.

Back to the box, all measurements interface with Audio Precision using balanced differential output whether the headphone amp is being tested in balanced or not. So you get noise immunity that way.

As nice as the outside is, the inside is even nicer:

All load resistors are head sink mounted and are the nice high quality ones from Caddock. These things are not cheap. Add them up and all the relays in there and such and the cost travels to hundreds of dollars. And this is for parts alone.

Dummy Load Measurements
Audio Precision APx555 has its own internal loads. Because they eliminate all wiring and are of high quality, for 300 ohm testing I have been using it. I always worried about damaging them though as they have limited power capability. Let's compare that load to HP-Load using Topping A90 as the test platform:

View attachment 144427

As you see, I can measure down to SINAD of 120 dB with zero deviation from APx555.

I have not found a use for the reactive load setting. Here are the two extremes:


View attachment 144428

Will have to find an amp that misbehaves to use this feature.

I should note that you have been seeing the results of this load box for a while. Anytime you see a graph like this:

It is produced using HP-LOAD box.

Conclusions
If you are doing any high performance headphone amplifier measurements, you need a high quality dummy load. A poor one can introduce its own noise and distortion which ruins the reliability of your measurements. If you are a company producing headphone amps, you need to get the HP-LOAD box. It will help standardized our measurements so that we can better compare them with each other. Furthermore, the automation will allow you to speed up your testing and breadth of what you check in production or engineering phases.

I am happy to give a strong recommendation to Neurochrome HP-LOAD. Puts a smile on my face very time I use it!

-----------
As always, questions, comments, recommendations, etc. are welcome.

Any donations are much appreciated using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/

Very nice! Makes my dummy load look dumb

 

[CtheArgie]

An excellent "dummy load". Compared to the loads of dummies commenting (not here, of course).

 

[gvl]

Environmentalists will be disappointed. The released heat energy should be captured and put to good use.

 

[PeteL]

But isn't the purpose of dummy load to present a real world load to the amp ? And a real headphone will have varying impedance and phase across the whole frequency band. It is this type of load (and not just resistive) that is most difficult (and the real test) for the amp.

If this dummy load is resistive then why go through this trouble, why not just wire up a resistor and be done with it ? You could have 2-3 headphone plugs with different value resistors (30 ohm, 300 ohm..) wired in each.

I feel you are mistaking trying to apply a stress test to an amp where measurment is about collecting Data to characterize an amp behavior relative to impedance. Yes one could do with your method, this unit is more flexible, and you need good resistors that behave linearly at high power, you want efficient heat dissipation, etc. Plus this unit DOES represent a dynamic load if you need. It's ok if it feels too expensive to you tough, it's lab equipment, it's heavy duty, it's precise, and yes it's expensive.

 

[Labjr]

Environmentalists will be disappointed. The released heat energy should be captured and put to good use.

Wasn't The EU going to ban inefficient amplfier designs like Class A and Vacuum Tube?

 

[gvl]

Wasn't The EU going to ban inefficient amplfier designs like Class A and Vacuum Tube?

Long overdue, long overdue....

 

[Labjr]

Long overdue, long overdue....

Are they going to have hi-fi police? I'll never give up my Class A Tubes. I have two systems hoping to get a third.

 

[gvl]

Are they going to have hi-fi police? I'll never give up my Class A Tubes.

Yes, they will make you go through a double-blind test against efficient class D amps and give you 40 lashes every time you fail to tell a difference. If you can tell them apart you will just have to pay a fine.

 

[tomchr]

Which real load? Every headphone is different in that regard. So you can't just pick one. Or two. Or even three. It is not line one or two devices dominate 90% of the market.

Exactly. A quick market survey reveals headphones ranging in impedance from 12 to 600 Ω. I choose those end points deliberately. I bet if you made a histogram of headphone impedances, you'd find three modes: 32 Ω, 300 Ω, and a smattering around the 50-80 Ω mark. But that's just the real part of the impedance. You then need to add the imaginary (complex) part. Best of luck finding a generic fit that matches the peaks and nulls of the impedance curve of a representative set of headphones.
It's not like in Speaker Land where we can use the 2-way box speaker model that Stereophile uses and call it good. I think most would agree that a 2-way box speaker is reasonably representative of a "real world" load even if it doesn't represent your speakers.

I do recall an amplifier load box that offered real and imaginary loads. I.e., both resistive and capacitive/inductive loads. I can't seem to find it now. Maybe the company that made them went under. I seem to recall it being around $10k. Maybe $8. Yes. $8-10k for a dummy load.

For a while I tested with real headphones. They generated totally bogus data due to back EMF. They can even act as microphones and inject noise into your measurements! At the end, it was wasted effort producing garbage and misleading measurements.

Gotta love those, eh? Another day sunk into chasing some obscure issue that turns out to be caused by your test setup.

Tom

 

[tomchr]

Environmentalists will be disappointed. The released heat energy should be captured and put to good use.

I usually do, actually. It just so happens that I do much of my amplifier development during the winter months, so the heat from the dummy loads goes to heating up my basement lab. In the summer I take advantage of the cool lab location and sleep on a futon in there when the upstairs gets too hot for comfort.

Torture testing a stereo speaker amp does a wonderful job of heating up the lab.

Tom

 

[jtwrace]

You sleep?

Hmm, learn something new everyday

 

[tomchr]

It's been known to happen. I aim for a nightly sleep schedule but don't always make it.

Tom

 

[amirm]

Environmentalists will be disappointed. The released heat energy should be captured and put to good use.

No worries. I sit on mine in winter.....

 

[amirm]

I do recall an amplifier load box that offered real and imaginary loads. I.e., both resistive and capacitive/inductive loads. I can't seem to find it now. Maybe the company that made them went under. I seem to recall it being around $10k. Maybe $8. Yes. $8-10k for a dummy load.

Yes, the powercube. They are still around but interest seems to have died off.

 

[jtwrace]

Yes, the powercube. They are still around but interest seems to have died off.

https://www.audiograph.se/

 

[JensH]

In the frequency response figure you write "100 pF and 10 uF load settings". I assume you mean "100 pF and 10 nF load settings".

Regarding inductive load: Can an inductive load be expected to trigger (test for) instability or is it only the capacitive load, which is of concern here?

 

[amirm]

In the frequency response figure you write "100 pF and 10 uF load settings". I assume you mean "100 pF and 10 nF load settings".

Yes. Added a note.

Regarding inductive load: Can an inductive load be expected to trigger (test for) instability or is it only the capacitive load, which is of concern here?

From what I have read, headphones become capacitive in very high frequencies where instability may occur.

 

[Dan Clark]

If they added a 100MOhm option with 100pF load you could even create a dummy electrostatic. Easy mod I suspect...