• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). There are daily reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

PreSonus Firebox 24bit/96khz loopback measurments

kipman725

Active Member
Joined
Jul 6, 2020
Messages
155
Likes
141
I acquired this interface cheaply and have performed some measurements to evaluate its suitability for measuring amplifier etc. https://www.presonus.com/products/firebox This is an example of a several generations old budget audio interface.

1624785631051.png


Measurement setup is with a 3 conductor jack lead input 3 and output 1L. It's worth noting that the outputs of the firebox are not diffirential and are instead 'impedance balanced' which means that the -leg is connected to ground via a resistor instead of driven. I normally use the external AC adapter to power the interface as otherwise its flaky to initialize on boot-up without it but in loop-back I can see significant AC harmonics:

1624784892303.png


Remaining tests are conducted powered from Firewire only. The optimal level for this interface in loop-back (don't know if ADC or DAC limited) is around 5.7V . Beyond this voltage the performance utterly collapses with harmonic and non harmonic tones across the spectrum. I have conducted further testing at 4V as this is close to optimal and also used in ASR testing.

Next lets compare 96kHz sample rate:
1624785223552.png


and 48 kHz:
1624785257365.png


So noise is elevated by using 96kHz sample rate even if low and high pass filtering to 20-20kHZ is used. Further tests done at 48kHz sample rate:

Low frequency test:
1624785565907.png


High Frequency test:
1624785588167.png


Results:
22.5Hz: THD = 0.0026% THD+N = 0.0054%, SINAD = 85.4 dB (noise limited)
1kHz: THD = 0.0020% THD+N = 0.0048%, SINAD = 86.4 dB (noise limited)
7.75kHz: THD = 0.0028% THD+N = 0.0052%, SINAD = 85.7dB (noise limited)

Equivalent noise input voltage at 1kHz: 172uV

Looking at other loop-back measurements it looks to be possible to get around 94 dB SINAD from modern interfaces in the same price range so this seems OK for something so cheap and old.
 
OP
K

kipman725

Active Member
Joined
Jul 6, 2020
Messages
155
Likes
141
Here are some RMAA measurements, not sure how to adjust this software and exactly what its doing.

1624789885939.png


But THD/THD+N seems similar, extra measurements like cross-talk also look good
 

AnalogSteph

Major Contributor
Joined
Nov 6, 2018
Messages
2,181
Likes
2,065
Location
.de
Seems alright for what it is. Frequency generation seems to have a bit of a PSRR problem (and distortion worsening at 96 kHz is clearly pointing towards some issues, possibly correlated jitter which may be a layout or power-related issue), and booting requiring an external PSU may suggest some degraded electrolytics in the power section (I mean, this thing should be from the mid-2000s, when such problems ran rampant, and Presonus isn't exactly the kind of brand to use premium components).

Converter specs:
ADC: AK5384 (4ch) - DR 107 dB(A) - THD+N -100 dB - filter ultimate -80 dB - filter ripple +/-0.005 dB
DAC: AK4358 (8ch) - DR 112 dB(A) - THD+N -94 dB
So distortion seems about right, while some dynamic range was clearly left on the table.
The thing is supposedly full of MC33078 (dual) and 33079 (quad) opamps. These were common back then, but you could probably do better these days (some reverse-engineering and part value adjustment may be required to extract best performance).

Official specs for the Firebox do not seem to line up with your findings. About 5.7 Vrms is +17.3 dBu - at least 2 dB short of the +20 dBu (@0.5%) max input spec, but at the same time well above the supposed +10 dBu 0dBFS reference level. An identical SNR spec for both mic and line inputs seems suspect, too. Dodgy manufacturer specs aren't all that uncommon.
 
OP
K

kipman725

Active Member
Joined
Jul 6, 2020
Messages
155
Likes
141
The startup issue I think is more of a windows issue as its OK to start if I unplug it and plug it back in once the PC is booted, I have tried different firewire cards with no change. On boot it goes into displaying a red light instead of a blue light and doesn't show in device manager. I have already modified mine a bit to lower the insane headphone amp gain, probably not worth swapping the opamps in something so old though. I was thinking about getting a Motu M4 for measurements or a QA402 but the gains in performance don't seem worth it, I think I would be better off using this and just hiring a AP unit for a week if I need better performance.

Looking at the output on a multimeter I can increase the output voltage a few volts more than the point at which I see clipping so I think its the inputs that are clipping before the output.
 

AnalogSteph

Major Contributor
Joined
Nov 6, 2018
Messages
2,181
Likes
2,065
Location
.de
I was thinking about getting a Motu M4 for measurements or a QA402 but the gains in performance don't seem worth it, I think I would be better off using this and just hiring a AP unit for a week if I need better performance.
It all depends on what you need it for, I guess. Acoustic measurements or an average hi-fi integrated or power amp (with suitable attenuators) should pose no major issues, and the mic input (assuming the EIN spec is accurate) should provide low enough noise to gauge the output noise of most line outs. Max levels around the +18 dBu mark also betray the unit's original price tag, more the equivalent of an Audient iD14 MkII or iD22 today.

That kind of comparison also shows the kind of progress we've made since 2005... I'd think either of those would show a solid 10-15 dB higher dynamic range in loopback.
 
Top Bottom