• 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!

Opinions on Hypex Plate (FusionAMP) Amplifiers?

Depends on the bit rate used when sampling the analogue signal.

I would guess they directly sample with 93.75 kHz.

They do. This is why the latency is actually lower on the analog input, as apparently the resampling on the digital input takes some time.
 
Since I at some point used to prefer the analog inputs of the FusionAmp, I wanted to find out why. Are there measurable differences, or was in all in my head? My findings support the latter I'm afraid, although the digital input seems to have slightly more distortion than the analog. Not that I think it could ever be audible...

I measured this using REW. The source was a WAV-file measurement sweep played back from my Logitech Transporter streamer. I used a Focusrite Scarlett 2i2 Gen3 USB interface to capture the speaker output signal from an FA252 FusionAmp, which was connected to a 4 ohm dummy load. The output level from the FusionAmp was set at -18 dB, or somewhere around 4W @ 4ohms. The Scarlett is not exactly on the level of an AP analyser, but for this purpose good enough. The graphs below show the analog out from the Logitech Transporter as well for reference, which is way below the FusionAmp.

The rise in distortion at high frequencies is a bit alarming to me. It's certainly not due to the NC252 amp modules, so it must come from the DSP board or what? I wish Amir would take the time to measure a Hypex FusionAmp :)

EDIT: It might also be my measurement setup....


EDIT2 : will get back to this. Something is very odd with my measurement setup it seems.
 
Last edited:
Since I at some point used to prefer the analog inputs of the FusionAmp, I wanted to find out why. Are there measurable differences, or was in all in my head? My findings support the latter I'm afraid, although the digital input seems to have slightly more distortion than the analog. Not that I think it could ever be audible...

I measured this using REW. The source was a WAV-file measurement sweep played back from my Logitech Transporter streamer. I used a Focusrite Scarlett 2i2 Gen3 USB interface to capture the speaker output signal from an FA252 FusionAmp, which was connected to a 4 ohm dummy load. The output level from the FusionAmp was set at -18 dB, or somewhere around 4W @ 4ohms. The Scarlett is not exactly on the level of an AP analyser, but for this purpose good enough. The graphs below show the analog out from the Logitech Transporter as well for reference, which is way below the FusionAmp.

The rise in distortion at high frequencies is a bit alarming to me. It's certainly not due to the NC252 amp modules, so it must come from the DSP board.

View attachment 419699

The THD is best at around 0dB gain I think. A bit strange that the digital is that much worse. Could there be something in the test setup causing it?

EDIT: I misread the graph, I see that they are nearly identical.
 
The THD is best at around 0dB gain I think. A bit strange that the digital is that much worse. Could there be something in the test setup causing it?

I tested at different levels, sweep lengths, samplerates etc, but they all returned much the same results.

Next up is to measure the NC252 alone without the DSP board in the chain to see if that looks different.
 
I tested at different levels, sweep lengths, samplerates etc, but they all returned much the same results.

Next up is to measure the NC252 alone without the DSP board in the chain to see if that looks different.

Not sure I understand the comparison to the Logitech Transporter, that's not an amplifier. So you didn't have any speaker output signal on that? Are you comparing a high level and low level signal?
 
Not sure I understand the comparison to the Logitech Transporter, that's not an amplifier. So you didn't have any speaker output signal on that? Are you comparing a high level and low level signal?
It's used as measurement source, both analog and digital. I measured it to show that any distortion I measured did not come from itself or the inputs of the Scarlett.
 
The DAC outputs 3 and 4 are set in parallel for the tweeter channel to get better performance. So we do use all 4 outputs but only for a 3 way setup indeed.
Have you tested it? Does it really provide better results at high frequencies compared to mid frequencies?
BTW, do you have any plans to release a DSP with only digital input and 3 digital outputs, so it can be used with other DACs and amplifiers?
Thanks
 
I have AB tested the former external DLCP vs a miniDSP Flex with level matching. The external DLCP is a unit Hypex sold many years ago and the measuring data seems to be the same as today's internal DLCP (I'll check the datasheet later to confirm).

I thought the difference using digital inputs on both was actually fairly easy to hear. The Flex was more transparent sounding. Obviously we're not talking about anything major, but difference was enough that I didn't feel it was necessary to blind test here (I know several of you will object to that). Since I haven't had my DLCP unit measured, it's difficult to know why. Never tried analogue in with it.
 
I’ve used FusionAmps with surround processors and always used the analog outs, since there in reality is no other option if you want to use all surround channels, and the processor’s DSP for volume control, EQ, bass management etc …. There is nothing wrong with the analog route either as long as you have a good processor.
Hi, I just saw this thread. I am about to embark on a DIY active speaker project. I imagine it being based on FusionAmps and making use of their DSP including EQ. The inputs will be fed from an AV processor with analog XLR outputs.

I have no personal experience with FusionAmps. Is there anything I should be concerned about with using them as above? Is there a better option that I should be considering?

thanks
 
OK, now I understand your intention. Of course you could hook up any other amp instead of the NC100HF or even in parallel. In theory this should not be a big issue. Gain settings could be off a bit. I am not sure. Basic fact is that we cannot support such setups and therefor cannot give warranty.
How can this be done in practice? How can I make a line output from the DSP to connect an external amplifier, for example to an active sub?
Thanks.
 
How can I make a line output from the DSP to connect an external amplifier, for example to an active sub?
Just connect the DSP line output to your external amp's input. But the Fusionamp is not prepared for this and the manufacturer doesn't support it, so you have to figure out how.
 
Last edited:
Hi, I just saw this thread. I am about to embark on a DIY active speaker project. I imagine it being based on FusionAmps and making use of their DSP including EQ. The inputs will be fed from an AV processor with analog XLR outputs.

I have no personal experience with FusionAmps. Is there anything I should be concerned about with using them as above? Is there a better option that I should be considering?

thanks
Hypex FusionAmps are very cumbersome to work with compared to other DSP solutions like MiniDSP in my opinion.
Because :
- The GUI of the Hypex Filter Design software is extremely slow and difficult to work with.
- You cannot have more than one unit connected to your computer at a time. Imagine having ten FusionAmps like I did at a time.
- You need to upload to the DSP every time you change something, and this can take as long as 30 Seconds.
- Fixed number of taps for FIR filters
- If you mute an output channel, it goes unmute again every time you upload settings.

And more !
 
Last edited:
The outputs of the Fusionamp on it's back doesn't comes from the DSP.
Yes it does. There are two flat cables where you find the signal from the DSP.

IMG_0484.jpeg
 
Hypex FusionAmps are very cumbersome to work with compared to other DSP solutions like MiniDSP in my opinion.
Because :
- The GUI of the Hypex Filter Design software is extremely slow and difficult to work with.
- You cannot have more than one unit connected to your computer at a time. Imagine having ten FusionAmps like I did at a time.
- You need to upload to the DSP every time you change something, and this can take as long as 30 Seconds.
- Fixed number of taps for FIR filters
- If you mute an output channel, it goes unmute again every time you upload settings.

And more !
Hypex should focus on improving their DSP and incorporate a wireless solution instead of inaudible improvements from new amp modules.

It's a quality DSP the market is missing now and that's been the case for a very long time. miniDSP has clear drawbacks as well and I've experienced Flex dumping DC to the speakers.
 
Hypex FusionAmps are very cumbersome to work with compared to other DSP solutions like MiniDSP in my opinion.
Because :
- The GUI of the Hypex Filter Design software is extremely slow and difficult to work with.
- You cannot have more than one unit connected to your computer at a time. Imagine having ten FusionAmps like I did at a time.
- You need to upload to the DSP every time you change something, and this can take as long as 30 Seconds.
- Fixed number of taps for FIR filters
- If you mute an output channel, it goes unmute again every time you upload settings.

And more !
Hypex should focus on improving their DSP and incorporate a wireless solution instead of inaudible improvements from new amp modules.

It's a quality DSP the market is missing now and that's been the case for a very long time. miniDSP has clear drawbacks as well and I've experienced Flex dumping DC to the speakers.
Thanks guys. That's generally very sobering..... :confused:

Noting that the ideal solution is not out there, what is your opinion on the best approach to take, today, for several 3-way and 4-way DIY active speakers with DSP on each driver, fed from an AV processor with XLR analog outputs?
 
Back
Top Bottom