[montr]

I am upgrading my base audio (2010 E90 335d) using the Alpine kit 65-41-2-163-268. The 268 kit does not require any wires cutting (plug-and-play) but per instructions required radio recoding. There is another Alpine kit 65-41-0-445-684. The 684 kit require wires cutting but no recoding.

I studied the instructions of both kit, wiring diagrams, parts number, etc . Here is what I found:
1) Both kit used the same amplifier.
2) Both kit end up to be connected to the car the same way. The difference is that the 268 kit use connectors and the 684 kit use crimp wires.
3) For the 684 kit, you are instructed to put the switch on the Alpine amplifier to position #2 to equalize the response (undo the equalization in the radio). For the 268 kit, you are instructed to put the switch on the Alpine amplifier to position #1 (flat?) but you have to recode the radio to remove the equalization in order to get a flat response.
4) You can convert a 268 kit to a 684 kit by simply cutting the connectors on the kit that connect to the radio and speakers.
Here is my theory. I should be able to use the 268 kit without recoding the radio by putting the switch on the Alpine amplifier to position #2 instead of position #1.

I hope to verify this soon. I ordered this week a 268 kit from Denmark Schmiedmann. In the mean time I measured the audio frequency response of my car. See attached response. To measure the response, I created a CD with a filtered flat white noise (200Hz to 20KHz). I used the sound card (16 bit @ 96KHz) of my laptop to capture the output of the radio. I did a Power Spectral Density of the data using Matlab.

Note: With Bass=0 and Treble=-4, the frequency response is within -1dB and +2dB.

[kmarei]

Treble is boosted from the factory to account for the cheap ass BMW bean counter who thought it was a good idea to not have tweeters

[jacar]

Thanks for this info. I would like to do the same to my 2010 e90 with the 268 kit from the same place, so I am looking forward to how it works for you.
What a nasty EQ curve-I wonder if you can compare the non-flattened curve to the flattened one by changing the switch position on the amp.

[Technic]

Originally Posted by montr

3) For the 684 kit, you are instructed to put the switch on the Alpine amplifier to position #2 to equalize the response (undo the equalization in the radio). For the 268 kit, you are instructed to put the switch on the Alpine amplifier to position #1 (flat?) but you have to recode the radio to remove the equalization in order to get a flat response.
4) You can convert a 268 kit to a 684 kit by simply cutting the connectors on the kit that connect to the radio and speakers.
Here is my theory. I should be able to use the 268 kit without recoding the radio by putting the switch on the Alpine amplifier to position #2 instead of position #1.

I hope to verify this soon.

I would like to know your results in switch position #2 as well, as it does not make any sense to me that a different harness connector is sufficient to warrant an OEM HU coding change.

Something else must be going on here.

[montr]

Originally Posted by Technic

I would like to know your results in switch position #2 as well, as it does not make any sense to me that a different harness connector is sufficient to warrant an OEM HU coding change.

Something else must be going on here.

I will make more measuments when the kit arrive. I will post the results.

If the response with the switch set at position #2 is relatively flat, next I will probably replace the rear speakers by component speakers driven by the head-unit. Because the back speaker will not be equalized by the Alpine amplifier, I will most likely have to attenuate the high frequencies to the tweter by 3-4dB. I will see when I am there.

[montr]

I got the kit yesterday. I did more measurements. The previous data of the HeadUnit was done with a source that is flat from 200 Hz to 20 KHz (test take 3 second). For my new measurements, in order to get the response at lower frequency I changed the source to 20 Hz to 20 KHz (test take 30 second).

Here is the response of the amplifier for switch position #1 and #2. As you can see, there is a crossover in the amplifier. Frequencies below 120 Hz are sent to the woofer under the seat. Frequencies above 120 Hz are sent to the door midbass and tweeter. The purpose of the switch on the amplifier is to attenunate the woofer output, 0 dB in position 1 and -4 dB in position 2. My take is that if you have a 4 Ohms woofer then position 1, for a 2 Ohms woofer then position 2. This is to normalize the response of the woofer with the midbass and tweeter.


I did an estimate of the composite response of the HeadUnit and the Alpine amplifier. The HU is set for Bass=0 and Treble=-5. This appears to be the setting that gives me the best flat electrical response. I have not measured the acoustical performance yet.

Note: The passive crossover that will fit in the door has a lowpass for the midbass driver and a highpass with attenuation for the tweeter. I have not measured it yet.

[Technic]

Thanks...

[Imperator]

Originally Posted by montr

Note: The passive crossover that will fit in the door has a lowpass for the midbass driver and a highpass with attenuation for the tweeter. I have not measured it yet.

Are you referring to the PnP passive crossover included in the Alpine HiFi Upgrade kit? I assumed it's little more than a cap, filtering a frequency shelf but does not provide any attenuation.

[Imperator]

Originally Posted by montr

In the mean time I measured the audio frequency response of my car. See attached response. To measure the response, I created a CD with a filtered flat white noise (200Hz to 20KHz). I used the sound card (16 bit @ 96KHz) of my laptop to capture the output of the radio. I did a Power Spectral Density of the data using Matlab.

Note: With Bass=0 and Treble=-4, the frequency response is within -1dB and +2dB.

I just noted your initial measurements were taken from the rear speaker terminals of the base audio output. I didn't expect such an ugly EQ curve but believe it is quite different from the front, full-range channels.

Unless I am missing something, would you mind running this exercise again with data for the front channel outputs of the BMW base audio head unit, with one graph detailing the electrical response before the Alpine amp is fitted and again with the outputs of the Alpine amp? I'd like to see what sort of normalization it provides as the HiFi Upgrade Kit only interfaces with the front channels.

[montr]

The passive crossover is more than just a capacitor. I did not want to break the heat shrink tubing and foam to see what is the design. However, I can peek inside from both end. I see an inductor for the midbass and a capacitor and resistor for the tweeter. I may try to get the response of the network later on.

I did a capture of the signal first on the rear speaker because they are the most accessible. I will get the response at the front speaker when I open the door to install the new speaker.

[Imperator]

Have you installed the Alpine HiFi amplifier yet? You'll want to take your initial electrical measurements at the branch-audio connector under the seat. The front channel outputs of the base audio system are supposed to be full-range and use the 4-pin connector at the 6.5" midbass enclosure to interface with the door speakers.

I believe the front channel speakers are wired in parallel with the underseat midbass drivers. I recall noting a cap integrated into the enclosure/connector that likely is intended for the woofer. I wish I could provide you pictures, but I sold my enclosures to Technic. He might be able to disclose where the cap connects at the 4-pin connector.

I'm most curious in your findings as I haven't yet graphed the electrical output of the base audio front channels and am really interested to see what sort of normalization the Alpine amp is responsible for with regard to installations that don't require coding.

[Technic]

Originally Posted by montr

The passive crossover is more than just a capacitor. I did not want to break the heat shrink tubing and foam to see what is the design. However, I can peek inside from both end. I see an inductor for the midbass and a capacitor and resistor for the tweeter. I may try to get the response of the network later on.

I did a capture of the signal first on the rear speaker because they are the most accessible. I will get the response at the front speaker when I open the door to install the new speaker.

The crossover circuit is an LR design, with 3 levels of tweeter attenuation (depending on the sound system installed -Business, Professional HU, iDrive):

[montr]

I have not started the installation. My BMW dealer did not have in stock the door sail to install the tweeter ($32 each). I should get them in a day or two. Then if I have the time, I will do the install this weekend.

For the E90 LCI, the signal goes from the HU to the woofer under the seat (pin 2 and 3). The signal is then bridged to the door speaker (pin 1 and 4). I do not know if there is a capacitor in series or not at the woofer location to send the signal to the door speaker. I will know when I will do the install.

[montr]

Originally Posted by Technic

The crossover circuit is an LR design, with 3 levels of tweeter attenuation (depending on the sound system installed -Business, Professional HU, iDrive):

I think my passive crossover is different than your crossover. I have the 268 kit, not the 684. The crossover has 2 wires on the inductor side and 4 wires on the other side. I can see the inductor on one side. On the other side I can see a resistor and a capacitor. Measurements with a DVM confirm that there is a capacitor in the path.

I think the crossover is a first order lowpass for the midbass (1 inductor). For the highpass to the tweeter, it is a first order (1 capacitor) with some attenuation (1 resistor).

[Imperator]

Originally Posted by montr

I have not started the installation. My BMW dealer did not have in stock the door sail to install the tweeter ($32 each). I should get them in a day or two. Then if I have the time, I will do the install this weekend.

In that case, what part of the installation is complete and how did you take your measurements with the Alpine amp?

[montr]

Originally Posted by Imperator

In that case, what part of the installation is complete and how did you take your measurements with the Alpine amp?

I did some measurements on the back speaker of the car. For the Alpine amplifier, I did the testing on a test bench. I have not started the installation, I just got the kit late yesterday.

Here is the amplifier. The crossover is on a plug-in board. This is the board with the switch. Inside the amp, there are 4 jumpers. They set the input level (High or Low). They are set by default to High.

[montr]

Here is the response of the crossover from 100 Hz to 20 KHz. For the load, I used the speakers included in the kit. I did not use a resistive load. The plots are the magnitude at the midbass and at tweeter connectors. To drive the crossover, I used a good quality audio amplifier (low impedance). The crossover midbass to tweeter is around 7 KHz.

[Imperator]

Originally Posted by montr

I did some measurements on the back speaker of the car. For the Alpine amplifier, I did the testing on a test bench. I have not started the installation, I just got the kit late yesterday.

Here is the amplifier. The crossover is on a plug-in board. This is the board with the switch. Inside the amp, there are 4 jumpers. They set the input level (High or Low). They are set by default to High.

Okay, that makes sense. I'd like to see what you find from the front channels before installation and after with the Alpine amp.

I still have mine and took it apart earlier, discovering two NXP TDA8920BTH ICs, one for the front door speakers and one powering the two midbass channels. I also noticed the input level jumpers but completely ignored the crossover board-- I should have looked at that to discover its function. Thanks for digging this up and I'm interested to see what you find from additional RTA sweeps!

[Technic]

Originally Posted by montr

I think my passive crossover is different than your crossover. I have the 268 kit, not the 684. The crossover has 2 wires on the inductor side and 4 wires on the other side. I can see the inductor on one side. On the other side I can see a resistor and a capacitor. Measurements with a DVM confirm that there is a capacitor in the path.

I think the crossover is a first order lowpass for the midbass (1 inductor). For the highpass to the tweeter, it is a first order (1 capacitor) with some attenuation (1 resistor).

Thanks... I can see it now in the installation documents. The new 268 crossover have a single tweeter output regardless of OEM sound system (Business or Professional). The 684 crossover has two different tweeter outputs, one for Business and one for Professional systems.

[montr]

My install is done. I did few more measurements. To get a flat response without recoding, I had to set the switch on the amplifier to #2 (reduce output to woofer under the seat by 4dB) and set bass=-2 and treble=-5.

Figure 1 is the electrical output response of the HU for the front speakers from 10Hz to 20KHz. As you can see, there is strong peaking at 60Hz. This peaking is partly compensated by switch position #2 and bass=-2. The high frequency peaking is compensated by setting treble=-5 and the rollof of the acoustical drivers. Note: The rolloff around 18KHz is not the amplifier. It is my measurement setup, I forgot to set the signal capture to 96KHz sampling rate instead of 44.1KHz.

Next I tried to measure the acoustical performance of the Alpine speakers kit. I have an expensive ($5K) calibrated microphone (Bruel & Kjaer #2238) to do the measurements. First, doing acoustical measurements in a car is tricky. You move the position of the microphone by few inches and you get different results. You open the window or the sunroff and you get totally different results.

In this test (figure 2), I tried to measure the response of the midbass and the tweeter by locating the microphone 2" from one driver and blocking the sound from the other driver by using a blanket on it. Note: The valley around 1.5KHz in the midbass is maybe due from out of phase signal from the tweeter. The distance between the tweeter and the midbass is 20cm. This is about one wavelength at 1.5KHz. These measurements are done with the door open.

I also did more mesurements from the driver seat (figure 3) with all the doors and sunroff closed. The microphone is at ears level without anyone on the seat. As I said before, these are tricky measurements. You should not pay too much attention to the sharp peak and valley, they move in frequencies when you change the position of the microphone by only few inches. What is important is the trend.

Note: The figure 3 red is for switch=1, bass=-2, treble=-5. For switch=2, the peak at 60Hz will be 4dB lower.

[markm3guy]

Montr, Thank you for such an informative post. I've been following all the board posts very closely, as I too have a 328i (just bought it for my wife) and need to upgrade the base stereo...I am currently leaning towards the alpine setup.

I'n no expert, but it certainly would seem that the audio system is much improved by the Alpine system in combination with the head unit tone settings, as evidenced by the 'flatenned' spectrum.

Independent of the measurements....How does it actually sound? Understandable, the volume levels would be much better, but its it worth the price of upgrade in your opinion, or is it just a more expensive tin can?

Thanks again.

[montr]

Originally Posted by markm3guy

Montr, Thank you for such an informative post. I've been following all the board posts very closely, as I too have a 328i (just bought it for my wife) and need to upgrade the base stereo...I am currently leaning towards the alpine setup.

I'n no expert, but it certainly would seem that the audio system is much improved by the Alpine system in combination with the head unit tone settings, as evidenced by the 'flatenned' spectrum.

Independent of the measurements....How does it actually sound? Understandable, the volume levels would be much better, but its it worth the price of upgrade in your opinion, or is it just a more expensive tin can?

Thanks again.

I am glad that I did this change. This kit is a plug and play, no wires to cut or crimp and it is easily reversible. The sound quality is much improved and it can play much louder without distortion. The total cost including tax+shipping is $438 (kit) + $74 (door sails) = $512

The original midbass are cheap paper cone that rolloff after 10KHz. High frequencies are directional and because of the location of the midbass, you cannot hear them from the driver seat. Having a tweeter in the door and a better midbass make a big difference.

Here is the response of the original midbass driven by the HU with bass=0 and treble=0. The frequencies below 200Hz are from the woofer under the seat, not the midbass. As you can see, even with the HU boost at high frequencies there a rolloff above 10KHz.

Note: The efficiency of the new Alpine midbass is identical to the OEM speaker however the Alpine amplifier has about 9dB of gain. This means that the front speaker loudness relative to the rear speakers is increased. This can be easily corrected by the fader control. Also the gong is louder. This can be corrected in the sound menu of the HU.

One more thing. As other have found, the labelling on the passive crossover is wrong. Connector B2 is labelled "to Head-Unit". This is wrong, B2 go to the the tweeter, not the HU. Connector B3 has no label, it should be connected to the HU. This info is on page 7 of the 65 41 2 163 268 install instructions (found online).