KJF Audio MA-01 Review (Multi-channel Amplifier)

[Koeitje]

This is no different referring to an ESS DAC chip or AKM DAC chip and saving time.

We have enough data points for DACs to show that the RCA/XLR analogue outputs need to be measured, if possible.

No different to amps in my opinion.

That is not true, because these modules are pretty much completely self contained solutions. While a DAC chip is just a chip and nothing else.

This is key though, we've seen examples where this buffer messes up Hypex module performance. Would would we know if someone wasn't measuring?

These modules already contain a Hypex designed buffer. This thing literally contains 3 pre-build PCBs.

 

[Matias]

Finally, there is a bit too much disturbance when we change frequency and sweep for power than I like

Distortion starts to rise only in the 10 kHz lines and above, whose 2nd harmonics already fall out of the hearing range. This is a deliberate design choice: focus on reducing distortion where it matters, and shift the dirt under the carpet, that is, leave the distortion and noise out of band.

Also, since crosstalk is between channels within a board, I would (and will) use front left and surround left in the same module, front right and surround right in another module, and center in the 3rd module. This way I would get maximum separation between left and right speakers.

 

[SimpleTheater]

Seeing how the amp modules are stereo anyway, this should be representative of what you get even if turned on the rest of them at full power.

Are you saying each stereo module has its own power supply, or do all 6 channels share the same power supply? Thanks.

 

[SimpleTheater]

There's more places where things go wrong in DACs than a couple of wires IMO.

According to this article you can pick a DAC just by looking at the box for the options you need.

 

[Matias]

Are you saying each stereo module has its own power supply, or do all 6 channels share the same power supply? Thanks.

See the teardown: there are 3 identical Hypex NCore NC252MP modules, each module has 1 power supply and 2 channels. So the performance measured in 1 module in this review applies to all 3 modules, total power just multiply by 3.

 

[milosz]

This amplifier is quite a bit better than the DACs in the typical Pre-Pros, and even the EXPENSIVE Pre-Pros.... what would you use to drive this thing for home theater? Is there a multichannel DAC out there that can equal this in SINAD and the multi-tone bit-depth test? I seem to recall reading about one but the name escapes me now.

 

[KJF audio]

Thanks for doing this review Amir, I'd almost completely forgotten one of your members was sending it over to you. I'm glad it measured as well for you here as it did when I was putting it all together. As this is a technical forum, I've got some extra measurements here that members might be interested in looking at? I'll post some less technical things in the teardown thread you also started.

When I was getting tests done for european CE marking 3 or 4 years or so ago now I had to get independent measurements for the conducted and radiated emissions. I posted the graphs at the bottom here. What is interesting about these measurements from an end user perspective is that they pass at all. It is testament to the engineering work of Hypex in these modules (and of course me not throwing all their good work away by spraying emissions everywhere with bad implementation ) Hypex specifically state that they won't support integrators using these modules in a multi channel environment. This is mainly because switching amps powered with switching supplies are pretty bad at high frequency radiated and conducted crap. I was one of the first to start to develop a multi channel amps with the modules and Hypex engineers thought I was crazy . The amp that went off for test was a 6 channel one based around 3 x NC502MP modules a 3 kW potential. The test graphs below are run at full power over the test duration. The conducted graph runs from 150kHz to 30mHz and passes, it is tight but it passed. The radiated runs from 30mHz to 1gHz, again it is tight but it passes. Now the interesting thing about this is that the test agency had never tested even a stereo class D amp above 500W let alone 3 x that passed at all, they were flabbergasted that Hypex had managed to produce modules that were this good. Most high powered class D amps have to be sold as professional amps not intended for domestic environments because they can't pass these tests. Some manage to just squeeze through but generally need some remedial work to do so. So hats off to Hypex here.

Stefan

PS, i s it OK to put a cheeky link to the amp page ? I'll delete if not.

https://kjfaudio.com/product/ma-01-flexible-amplifier/

 

[Koeitje]

This amplifier is quite a bit better than the DACs in the typical Pre-Pros, and even the EXPENSIVE Pre-Pros.... what would you use to drive this thing for home theater? Is there a multichannel DAC out there that can equal this in SINAD and the multi-tone bit-depth test? I seem to recall reading about one but the name escapes me now.

Well the Octodac does 8 channels at extremely high fidelity. The current problem in the market would be that its doable for independent designers to come up with high quality products for open standards, but licensing HDMI, Dolby or DTS is a completely different ball park. Its simply not possible for small independent designers to get those licenses and keep up with the constant change in standards. The bigger manufacturers have no incentive or pride to make an actual high fidelity product. Plus creating a surround processor also has a huge software aspect, which is an entirely different skillset to have (and I'd argue that most mass market producers haven't mastered that either).

I really hope the Chinese manufactures can get their shit together and start producing surround processors. For now I think its only Tonewinner that makes some and I don't know what their quality is.

 

[Andrej]

I have a general question about testing power amplifiers to drive loudspeakers, relating to real world performance. How much difference is there when the amplifier load is resistive vs reactive (a complex load as most loudspeakers with passive crossovers have, or active crossovers where amplifiers are driving a coil based drive unit, for example)? How much difference, in general, is there relating to this particular question between different classes of amplifiers: AB vs class-D, solid state vs tube, etc. Is there any benefit to making loudspeaker loads be resistive by making clossovers more complex, or, in active speakers, adding passive components to make the load less reactive?

 

[Chrise36]

Hi Amir could you test for interference with the DACs?

 

[B4ICU]

What's the Amp's DF?

 

[Bsinger]

My D

One thing I forgot to mention. The gain is a bit low. Usually I target 29 dB and here it is 25 dB. To get full power, I had to get about 1.6 volts or so. If you pair these up with an AVR, make sure they can produce their optimal output there.

My Denon AVR X4000 puts out 4.2 clean volts

 

[sq225917]

Just checked the website, there's a fair bit of flexibility on build options, thats nice to see.

 

[bigguyca]

Hardly state-of-the-art measurements and not inexpensive - pass...

 

[AndreaT]

Wonderful review. Interesting to compare with Buckeye amplifier with the same Hypex module. I own one (Buckeye) and it has behaved flawlessly and competently delivering the emotions of well recorded Music stretching my disposable Dollar to great value.

 

[milosz]

Thanks for doing this review Amir, I'd almost completely forgotten one of your members was sending it over to you. I'm glad it measured as well for you here as it did when I was putting it all together. As this is a technical forum, I've got some extra measurements here that members might be interested in looking at? I'll post some less technical things in the teardown thread you also started.

When I was getting tests done for european CE marking 3 or 4 years or so ago now I had to get independent measurements for the conducted and radiated emissions. I posted the graphs at the bottom here. What is interesting about these measurements from an end user perspective is that they pass at all. It is testament to the engineering work of Hypex in these modules (and of course me not throwing all their good work away by spraying emissions everywhere with bad implementation ) Hypex specifically state that they won't support integrators using these modules in a multi channel environment. This is mainly because switching amps powered with switching supplies are pretty bad at high frequency radiated and conducted crap. I was one of the first to start to develop a multi channel amps with the modules and Hypex engineers thought I was crazy . The amp that went off for test was a 6 channel one based around 3 x NC502MP modules a 3 kW potential. The test graphs below are run at full power over the test duration. The conducted graph runs from 150kHz to 30mHz and passes, it is tight but it passed. The radiated runs from 30mHz to 1gHz, again it is tight but it passes. Now the interesting thing about this is that the test agency had never tested even a stereo class D amp above 500W let alone 3 x that passed at all, they were flabbergasted that Hypex had managed to produce modules that were this good. Most high powered class D amps have to be sold as professional amps not intended for domestic environments because they can't pass these tests. Some manage to just squeeze through but generally need some remedial work to do so. So hats off to Hypex here.

View attachment 125953

View attachment 125954

Stefan

PS, i s it OK to put a cheeky link to the amp page ? I'll delete if not.

https://kjfaudio.com/product/ma-01-flexible-amplifier/

Wow, that is super cool as well as super interesting.

 

[milosz]

Well the Octodac does 8 channels at extremely high fidelity. The current problem in the market would be that its doable for independent designers to come up with high quality products for open standards, but licensing HDMI, Dolby or DTS is a completely different ball park. Its simply not possible for small independent designers to get those licenses and keep up with the constant change in standards. The bigger manufacturers have no incentive or pride to make an actual high fidelity product. Plus creating a surround processor also has a huge software aspect, which is an entirely different skillset to have (and I'd argue that most mass market producers haven't mastered that either).

I really hope the Chinese manufactures can get their shit together and start producing surround processors. For now I think its only Tonewinner that makes some and I don't know what their quality is.

Yeah the OCTODAC that's what I was thinking about. But then how do you decode the surround format, you'd have to have some digital domain gizmo to take the audio embedded in the HDMI bitstream, pull out and decode the audio and output it in digital form. Nothing exists like that, AFAIK . Maybe you could build one.... there are Dolby 5.1 and Atmos drivers for Windows... maybe someone clever could take the 5.1 or 7.2 or whatever channels of digital bitstreams in the WIndows sound realm and output them from a PC somehow as AES/EBU streams that the Okto product would accept.

Or we can just wait for Topping to build a Pre/Pro....

 

[mgood]

Yeah the OCTODAC that's what I was thinking about. But then how do you decode the surround format, you'd have to have some digital domain gizmo to take the audio embedded in the HDMI bitstream, pull out and decode the audio and output it in digital form. Nothing exists like that, AFAIK . Maybe you could build one.... there are Dolby 5.1 and Atmos drivers for Windows... maybe someone clever could take the 5.1 or 7.2 or whatever channels of digital bitstreams in the WIndows sound realm and output them from a PC somehow as AES/EBU streams that the Okto product would accept.

Or we can just wait for Topping to build a Pre/Pro....

The best device to output 7.1 PCM would be the Apple TV 4K running Infuse, IMO. I get great results connected to a Denon X4500.

EDIT: Sorry, didn’t realize the limintations in terms of inputs on the Octo.

 

[milosz]

The best device to output 7.1 PCM would be the Apple TV 4K running Infuse, IMO. I get great results connected to a Denon X4500.

EDIT: Sorry, didn’t realize the limitations in terms of inputs on the Octo.

Yeah I don't see any digital audio outputs at all on the Apple TV 4k

 

[Koeitje]

Yeah the OCTODAC that's what I was thinking about. But then how do you decode the surround format, you'd have to have some digital domain gizmo to take the audio embedded in the HDMI bitstream, pull out and decode the audio and output it in digital form. Nothing exists like that, AFAIK . Maybe you could build one.... there are Dolby 5.1 and Atmos drivers for Windows... maybe someone clever could take the 5.1 or 7.2 or whatever channels of digital bitstreams in the WIndows sound realm and output them from a PC somehow as AES/EBU streams that the Okto product would accept.

Or we can just wait for Topping to build a Pre/Pro....

Yes, obviously. All the licensing issues surrounding surround sound it what makes it impossible for smaller manufacturers to compete.