Adding Preouts to Stock Head Unit

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
My stock head unit does not have preouts and I want to add them to be able to connect external amplifiers. I know it is possible to use signal from speaker terminals and feed them into high level input or use special line level converters, but this is not as good as having dedicated preouts.
I have already done some investigation and found out that signal goes directly from radio/audio processor IC (TEF6901A) to amplifier IC (TDA8569Q). I connected a scope to input and output of this amplifier and see that amplifier starts to clip after 6V RMS. Input at this point is about 310mA. This matches 26dB gain stated at the datasheet of TDA8569Q.
While 310mA is enough for preout and my amplifier can accept this voltage (in fact it accepts 200mA to 5V) I would like to increase it to about 4V.
I don't have much experience with opamps, but this is a good place to start. What opamp could you recommend for this purpose? Because I have only single power source (car battery voltage), it will probable have to be rail to rail opamp, otherwise I can't achieve 4V RMS on preouts.
 
How about using something like a JL Cleansweep?

I know it is possible to use signal from speaker terminals and feed them into high level input or use special line level converters, but this is not as good as having dedicated preouts.

You'd be pretty surprised. JBL used their MS-8 processor and a stock BMW head unit, entered it in some SQ judging competitions next to all-aftermarket stuff, and came out doing very well :)
 
I would find the input electrolytic DC blocking caps for the output chip and then just install one of the MANY after market line drivers that are for sale.

The after market line drivers have up to 12 volts output, and are adjustable and have isolated DC to DC converters delivering Bi-polar power to the line drive ICs, and some have very impressive distortion spec's VIOLA !!! instant RCA out and very high quality all in a blister pack on the wall at most any sound shop..... Hope this helps some...:)
 
My opinion, for what it's worth... Use a good quality LOC. The 10% THD is at hard clipping and is only used to get the high output power rating. The distortion before clipping is going to be inaudible.

On paper, using op-amps may look simple but to get it right without ground loop noise isn't likely to be as easy. A good quality LOC will give you the output you need and make the job much simpler.
 
Here is a two channel unit , they make 4 channel units also, and there are other brands available. I am very familiar with this one though

Phoenix Gold LD22 - eBay (item 320527285038 end time May-14-10 17:29:54 PDT)

Any auto sound dealer can supply you one of these, names like SoundStream, Audio control < they make a 6 channel unit > Phoenix gold, etc.. all have line drivers available that can be used to bypass the output chip entirely. You just need to find the input pins of the power chip, and these usually have a electrolytic cap inline with them. Just tap in front of the cap and the line drivers internal cap will do the rest...VIOLA sweet and simple high level line drive outputs for about 20 to 30 bucks, and ten minutes of your time with a soldering iron...Ground loop issues may even occur with any deck, If this causes issues with this setup just connect the shields of the RCAs from the input to the output then everything will be common grounded to begin with just like any high quality Head Unit would be....
 
My guess is that these line drivers are nothing more than a simple opamp circuit. And maybe additionally some DC conversion circuit to be able to get higher output voltage. So why not build it myself? If I make the board compact enough, I will be able to leave it inside the head unit. There is plenty of free space inside my 2DIN unit.
Yes and I see electrolytic capacitors between processor and amplifier ICs so I know where to tap in.
One thing however which I don't quite understand is why there is DC voltage after mentioned caps? There are about 2,2V DC if I measure between signal ground and one of the channels. I am able to see input signal in my scope if input is AC coupled, which blocks DC.
 
Your correct nothing more then a couple op-amps and a DC to DC converter, but its all put together SMD on the one I linked to and is quite small if you remove it from its metal case.
I have several of the Pro line drivers and Tbats here for my own use. The Tbats are true balanced line drivers that require balanced receivers to restore the signal back to RCA. The pro line drivers by PG have up to 12 volts drive level out without resorting to balanced line drive methods. Just higher rails on the op-amps from the supply inside them. but these will all do the trick and for cheap.

I myself try to avoid the time sink issues of doing custom work when off the shelf gear will do the job without much hassle and time loss... But if time in on your side I say yes build your own. The bi-polar supply and any noise it may impart might be a issue with internal install on the head unit. but who knows till its built and running.. Hope it works out for you which ever way you go....:)
 
Because I only need 4V preouts (my amp can accept max 5V) I am planning to use a single supply voltage with virtual ground. I'm looking at LME49740. As I understand, it's voltage headroom is only 1V below/above supply voltage, so If I have 0..14V supply with virtual ground in the middle, I can get peak to peak voltage of about 12V which is a little bit more than 4V RMS. And LME49740 is specially designed for HiFi audio use with ultra low THD (0.00003%). And because my circuit will not use DC-DC converter, noise level will be much smaller as well. Here's a datasheet of LME49740 http://www.national.com/ds/LM/LME49740.pdf
 
All of the over the shelf line drivers have adjustable input sensitivity and output drive level so any excess could be turned down like one of our other members here likes to say..:)
I am familiar with the LM super op-amps. I must admit I never considered running them other then bi-polar though.. Tag us back with your finished work, I would be interested to see how it turns out....:)
 
here's something I did before. my old Pioneer HU did only 2V on the line outputs.

An externally hosted image should be here but it was not working when we last tested it.


I used basic non inverting op amp circuit using NJM4560 to buffer it. you can use the 8V (8.5V in other HU's) that power the electronic volume IC. (look for the supply lines for the chip connected to the inputs of the power IC) that is clean DC meant for audio sections. I also used LTC1044 (ICL7660 can also be used) to get -8V from +8V so you get plus and minus 8V supplies. that way, my HU got about 5 volts high current outputs (capable of driving 600ohm loads) which was originally only 2V and limited output current.
 
If I create a virtual ground using resistors, do I understand correctly, that in this case supply voltage must be very stable, because any ripple in supply voltage will be reflected in created midpoint and that will be amplified by opamp?
I will try to see, what voltages I can get inside the head unit and how good this voltage is (ripple mainly).
 
not sure about switching converters. the LTC1044 is a charge pump using capacitors so no EMI associated with SMPS's.

for clean supplies, (if the radio is Fujitsu ten or eclipse based) they use 8V supplies for the analog audio sections called COM8V. they also have separate 8V supplies for the tuner and deck. you don't want those as they are noisy. other radios also use around 8V but not sure what they call them.

for virtual ground, that's the reason you use a bypass cap across the 1/2Vsupply bias and audio ground so that noise and ripple from the supply will be rejected and the bias supply will be clean with respect to audio ground.
 
Just checked and my head unit uses TDA3683 voltage regulator. There are 8.5V available on regulator 4 with 350mA output current. This output is stated to reject 70dB of supply voltage ripple, but I'm not sure how much ripple is introduced by regulator itself. I will have to measure it I guess. Maybe it would be better to use a separate linear regulator instead.
There is also output 7 which can be set anywhere between 2.4 and 10V. I'll see what voltage it is set to.
 
not sure about switching converters. the LTC1044 is a charge pump using capacitors so no EMI associated with SMPS's

Charge pump has an oscillator which means there will still be ripple in the output voltage.
By the way, using LTC1044 I will be able to get -8.5V from +8.5V so I then use these voltages to power the opamp, right? But what happens if -8.5V has voltage ripple and +8.5V does not. From what I understand, this ripple will be amplified by the opamp.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.