Aldous Voice

At time of first purchase, both the Ferrari 360 and F430 could be specified with an optional “Hi-Fi sound system (with subwoofer)” (code SND1) for an additional GBP 1,100 / USD 2,000. There is very little technical information available on this Hi-Fi system – its audio and electrical connections are not even shown in the 360 or F430 wiring diagrams.  This article attempts to shed some light on its characteristics.

The Hi-Fi system is composed of a subwoofer enclosure, an amplifier, and uprated versions of the mid-range speakers and tweeters located in the doors, as well as an additional wiring loom (part number 200853) to connect all these parts. All the active components (i.e. all speakers and the amplifier) are manufactured by Italian company ASK Group which builds audio equipment for OEM applications for the Fiat group and many others.

The subwoofer enclosure is situated between the seats, flush against the panel that separates the cabin from the engine compartment.

It houses a 130mm dual-voice-coil subwoofer speaker (bottom aperture) and a passive resonator (top aperture), which are protected by identical metal grilles. The speaker’s two coils each have a nominal impedance of 8 ohms, and are connected to the Hi-Fi wiring loom through a 4-pin Molex connector (Molex p/n 39-01-2040 on the wiring loom side, 39-01-2046 on the speaker side), which happens to be the same as the P4 power connector on ATX PC motherboards. The red and blue wires correspond to the positive terminals of the two coils, while the black and the yellow wires correspond to the negative terminals. I have no information on the power rating or quality of these speakers, but I will note that this specific unit blew during a high-volume in-system test and that the ~£100 price that Ferrari itself charges for it is suspiciously low (the door 160mm midrange speakers are even cheaper, at less than £50 each).

The amplifier is located inside a leather-finished cradle which sits underneath the central part of the dashboard, directly below the head unit. It should be noted that the coupe and spider tipos have different part numbers for the amplifier, but I have no reason to believe that they be significantly different (the other components are the same). However, for the avoidance of doubt, this article examines the spider version.

The cradle is held in place by two short metal rods protruding from its back, which slot into support holes in the bulkhead, and is secured by a single Phillips-head self-tapping screw which screws upwards into the bottom of the dashboard through a hole in the leather. In order to remove the cradle from the cabin, the 26-pin connector at the back also needs to be detached.

The cradle has metal grilles on its bottom side and on its right-hand side (when viewed from the top) to aid ventilation. The amplifier can be removed from the cradle by undoing the four M4 Phillips screws that hold it to the cradle metal frame.

The amplifier enclosure is aluminium with a steel cover, and measures 203mm by 135mm by 55mm (LxWxD) excluding the metal flanges and the green connector housing, the latter of which adds another 21 mm to its length. It weighs 1.32 kg and its build quality appears solid but it is let down by the fact that all its screws (both internal and external) crudely self-tap into the grooves of the aluminium enclosure. Removing these gives access to the internal components.

The core of the device is four TDA1562Q monolithic amplifier ICs, which are rated at 70W @ 4 ohm each, albeit with a substantial 10% of total harmonic distortion (THD) at peak power. These are held firmly against the wall of the enclosure (which acts as heatsink) with clips and heat-conducting paste.

The rest of the PCB is taken up by the passive components needed by the ICs (mainly large electrolytic capacitors – see p. 14 of the datasheet) and a few operational amplifiers, presumably part of an active filter (more on this later). This is the extent of the information that can be gathered by visual inspection – what follows was obtained with measurements and tests on this specific unit.

The first thing to determine is the pinout of the connector. I was not able to identify if this is a standard connector type, so I have to assume it is a proprietary design by ASK Group (if you recognise it from elsewhere, please do let me know). The amplifier side has 26 spade-type male pins arranged over three rows, numbered as in the schema below. 24 of three pins are 3mm wide, while two of them (no. 9 and 26) are 6mm wide.

The wiring-loom side is connected to the rest of the system as follows:

In essence, the amplifier receives the four pre-amplified channels (front and rear, left and right) from the head unit, amplifies them further (hence it is sometimes referred to as a ‘booster’), and drives the door and subwoofer speakers. Presumably this is possible as the Hi-Fi wiring loom probably intercepts the union of connectors 42D and 43D, which are normally connected to each other in a car without the Hi-Fi option (where the head unit drives the door speakers directly).

Further investigation reveals that the right- and left-hand side door speakers are driven independently (i.e. using two separate ICs) by the amplifier, using the front right and front left channels from the head unit as inputs, respectively. Incidentally, it should also be noted that there is only one pair of audio wires going into each door, which is used to drive both the mid-range speaker and the tweeter, with no crossover in between. On the other hand, the amplifier drives the two coils of the subwoofer speaker independently, but it uses the average of the rear right and rear left channels from the head unit as inputs, which means that the two coils are driven by identical signals.

In order to test the performance of the amplifier the following setup was built:

• a fully-charged Bosch S5 battery as power supply
• both a 12 MHz function generator and an iPhone with a function-generator app as signal input
• a 100 MHz digital storage oscilloscope (DSO) as measuring equipment
• four 4.7 ohm, 100W, wirewound power resistors mounted on aluminium heat sinks as dummy loads
• a Becker Traffic Pro BE6100 head unit as pre-amplifier when the iPhone signal was used

The amplifier was connected to its signal source and to the dummy loads using individual female spade connectors, as shown in the photo below.

The first test was to determine the frequency response of the four channels of the amplifier. As above, the two subwoofer channels use the same average signal as input, so it was no surprise that they have identical frequency response. They also appear to have a band-pass filter peaking at 62 Hz, with approx 50 dB/decade slope on both sides and -3 dB points at 21 Hz and 173 Hz.

On the other hand, the frequency response of the two front channels, apart from a common high-pass filter at 31 Hz, appear neither identical nor particularly flat. The test was repeated with two separate signal sources, with and without the head unit, and the data was confirmed in all cases. It is also interesting to note that the high-frequency roll-off does not happen until well past 1 MHz (way beyond what the human ear can respond to).

Another test worth performing is a measurement of the max output power the amplifier can operate at. This was done by using a sine wave as input and increasing its amplitude until the output of the amplifier, measured across the load resistor and viewed on the oscilloscope, started to distort or “clip”. At that point, the output power on that channel is given by the square of the root-mean-square (RMS) amplitude of the output wave, divided by the load resistance value. As we’ve seen in the plots above, the amplifier gain is frequency-dependent on all channels, so the maximum power will depend on the frequency. For the purpose of this test, the front-right channel was used, with a frequency of 1 kHz (where the two front channels happen to have identical gain). The measured amplitude was 11.6 V (RMS), which gives (approx) 28 W of peak output power per channel. Below is a snapshot from the oscilloscope with the output wave starting to distort (evidenced by the bottom of the wave starting to square off).

A few caveats need to be given around this measured value. First of all, a real loudspeaker with a nominal impedance of 4 ohms would draw more current than our test 4.7 ohm resistors. Secondly, although care was taken to perform the measurement with the resistors in “cold” state, they do heat up very quickly at peak power (hence the heatsinks) and resistance increases with temperature, which was not factored into the measurements above. Lastly, and perhaps more importantly, the supply voltage here was not regulated, and dipped as low as 12.2V under heavy load. This is representative of a car with its engine off, but lower than you would have in a car with the alternator running – measuring output power in that setup would have probably yielded a higher value.

Related Posts:

• Ferrari 360 & F430 Radio Removal and Aux/iPod Cable Fitting
• Ferrari Becker Radio Current Leak

Other posts in this category: Interior

If you enjoyed this post why not subscribe to this blog for further updates? Simply enter your email address in the box at the bottom of this page.  Your details will not be shared.