Actually, one thing you can do is put a ground loop isolator between the HU and a single-ended input amp. I did and it resolved the issue of grounding the one output. You can debate the merits of this, but it will work.

Reach: I switched to an aftermarket deck and forgot about this potential solution of using a GLI, otherwise I would have mentioned it sooner.

 

Looks like its available for $170 - $220 shipped from various sources. It also looks like an excellent line driver, which is its primary purpose. It claims to easily support up to 10v out per each of its 6 outputs, and can have up to 7 amps driven off each of its outputs. Quite impressive and far more of a device than I really need, especially since the PG amp is only rated to 3v input.

I need to see if there are any cheaper devices that just perform the balanced to unbalanced conversion easily.

Still, something to consider mercman (and others), if I were to buy the Audio Control Matrix unit at $200 and add it to my setup, the whole setup would still cost less than purchasing a JL 500/5 at the cheapest rate I have found it online. I don't consider this a total loss, plus I have learned a lot in the process. I hope others will as well.

BTW... why is it that before you were hopeful the balum will work, and now you seem to indicate that it will not, and an interfacing device is necessary?

 

I find it doubtful that a simple GLI would solve this issue, given all the discussion above. But... I have one handy so I might as well give it a try. I will have to check that out tomorrow, but I'm not too hopeful.

 

The TSX has a Pioneer Audio system not a Bose.

I’m not turning it in to rocket science just stating well known electronics fundamentals. My goal is to take the mystery out of the TSX audio system. Everyone that has followed my advice with this setup has been successful. The ones that didn’t (and used the wrong type of amp, sorry Reach) all have had problems with noise. Again, compare the TSX wiring to the TL. You will see the extra wire for each channel.

Did you read my post on the correct way to bypass the TSX amp? If there is something you don’t understand I will be happy to give more detail. I’m not being funny or sarcastic, read my post.

jeff

 

It doesn't state.

https://resources.myeporia.com/company_38/matrix_OM.pdf

It seems you can (de)select the balanced input, however, it doesn't state (or I missed) whether it converts a balanced signal to unbalanced. My guess is that it is a pass through box. That AC unit isn't BestBuy type material and many amps used by people who they are marketing to do have balanced inputs.

 

uncald4 has a new Accord, which I believe to be similar to the TSX & 3G TL, also having the balanced differential outputs (note: I could be wrong about this and would love to know for sure). This is why I am so confused. I'm sure he's got access to all the diagrams. He is utilizing a LC6i, which must be doing the conversion, as he states he has no noise on the lines. I know he's is a great installer, there is just some confusion in this area of line technology.

 

 

Reach,
The pre amp you found most certainly would have unbalanced outputs since it is designed to feed aftermarket car audio amps. If it had balanced all the way through it would be useless as an OEM interface. I think most of the confusion stems from the fact that some installs work and some don’t. A quick investigation into the ones that work would revel the use of some interface/EQ device (like the three sixty) with balanced inputs or an amp with balanced inputs. The key word here is balanced and as long as the TSX HU sees a balanced termination it will be happy and noise free.

As far as what systems each car has, I only know the TSX since that’s what I have.

The TSX uses balanced audio lines to feed the remote amp and that is the end of the story.

If you recall our PMs I said that amp would not work with your system with out a device to convert the balanced signal into an unbalanced signal.

So it comes down to this, listed from expensive to not so bad.
Sell your PG and get a JL or Rockford with diff in.
Install the pre amp with balanced diff inputs.
Install the BALUN.

jeff

 

I second mercman's words... his engineering has been solid and Reach should pursue the balun as the cheapest and most effective solution.

Unbalanced amp users may have just gotten lucky due to the what the amp's input circuit looked like. USE THE BALUN or get a balanced input amp... period. if you think the pg amp will sound as good as you want, then I'd support mercman by purchasing his balun and saving money. spending $200 i think is rediculous for an aftermarket bandaid.

BTW... the 10 to 15mV dc you measured could cause problems, but if the amp has any dc rejection it shouldn't be an issue. If its still setup, measure V ac. that would definitely be audible considering a guess of ~28dB of gain. (28dB = 20xLog (25V/V)

10mV ac x 25(v/v) = 250mV ac output to your speakers.

Electrical power = (200mV)^2/4 ohm speaker = 10mW.

Assume a healthy acoustic power efficiency of 90dB/1W at 1m. you lose about 10dB per decade of power (factors of ten). so 90db at 1W, 80dB at 0.1W, 70dB at 0.01W. Since 10mW =0.01W, then there could be about 70dB worth of crap coming out of each speaker speaker assuming 200mV input and 90dB/1w at 1m efficiency. If you had 4 speakers, then you add 3dB per additional speaker for a whopping 79dB. That's definitely audible. Hope i didn't goof up any math, but you get the point.

 

The math and underlying physics are honestly a little beyond me. Much past checking continuity, measuring voltages and checking against specs I get a little fuzzy, but I believe you are trying to reassure me that my measured difference is enough to make a significant impact on the audible output. That I hear loud and clear. (pun intended).

So, it sounds like the balun is the most reasonable answer to my riddle, and one I will gladly await Jeff's produced solution for.

 

I found this neat explanation of balanced audio signals, http://www.bcae1.com/ballndrv.htm

Reach, A GLI may or may not work for you because it is designed to isolate the ground wire only and not provide CMRR (common mode rejection or canceling of the noise). In order to get the most noise rejection you need a device to work like a balanced input.

With any luck I’ll get the Blu installed tonight and start my real word tests.

jeff

 

Reach,
Sorry if the math was a little overkill, but you were right... i was trying to convey that the small voltages you measured were potentials to cause you issues.

You measured dc (direct current) Voltage, but ac (alternating current) Voltage would be the measurement to do. It may be unnecessary at this point, but if you do measure ac V (V with ~ on top is ac, V with --- on top is dc) i would guess you'll see some voltage that indicates the problem sound you hear. Its always good practice to measure ac and dc V when troubleshooting.

hope it helped... good luck!

 

Reach,
The BALUN works and I was able to confirm that you will get noise with an unbalanced amp (the Bluapunkpt). I have some photos that I will upload when I get some time. I had planned to drive around with the Blu and BALUN but I changed my mind because I had no way to safely do it. I any case, the O’scope photos tell all.

One more thing,
Since you have a DVM would you set it to ohms and measure the resistance between the amps RCA input connecters. Just disconnect the RCAs from the HU and measure the outside (ground) from connector to connector (on the amp). Then measure from one the amps RCAs to the ground wire. Let me know what you get. Don’t forget to read the “k” or “m” on the display.


Thanks.
jeff

 

The following photos are taken from the speaker terminals of a blaupunkt 4 channel amp wired to the TSX HU balanced outputs. The test CD was a -20 db anti phase test. This test puts a signal on the L/R channels out of phase. I used it because when you tie the (-) side outs of the HU to a normal amps RCA jacks they all get connected. An anti phase test will so a reduced signal output level under this condition. The amp I used (as most amps) ties the RCA grounds together. You can see the result, you will hear bit as a whine and distortion.
For the ultra techie, this GTA 4 amp measured about 900 ohms from the RCA ring to ground. Not bad for ground isolation but not a true diff input. The JL 300/4 measured over 4 megs to ground.











This is a photo of the prototype BALUN (the small box on top of the amp).



This should put the issue of balanced outputs and unbalanced inputs to bed.

If you are going to keep the TSX HU then get the right amp or interface and follow my instructions. A noise free install awaits you!

jeff

 

I see that the BALUN worked well, but did you get just as good of results with the BALUN as you did using the JL amp?

 

Now that’s not a fair question.
As far as sound quality the JL is way far superior amp. No knock of the Blu but the JL has more power and way lower THD and IM numbers. As far as audible noise (whines and pops) none were noticed. Without the BALUN the Blu sounded poor (no imaging and limited Freq response). With the BALUN I would say that it sounded as good as the Blu could get, certainly not objectionable, but not the JL.

Also the TSX is no sound room.

jeff

 

Yea, I just meant in terms of noise reduction, not the quality differences between the actual amp. Any idea how much the balun is going to cost?

 

He's working on putting together a production version of the balun and will have pricing and delivery for it when its ready. Give him a bit.

 

mercman,
your last oscope pic using balanced inputs looks like it likely has high f components. Although the sine wave is smoother, the waveform is thick and fuzzy. Could be just the noise floor, but try zooming in on your time scale and see if you can pick up higher freq components. You'll have to adjust your trigger level. I'm sure you already know this stuff, but trying to be clear. If you see anything, then i believe its inherently due to the amp ic. It could be similar to what i saw.

 

The hash is inherent in the HU and is a combination of the DAC and LPF of the CD. Also contributing to the noise is the aftermarket amp circuits and switching power supply. The Honda amp is not in the circuit only the HU and the aftermarket amp. The audio level was very low (-20dB test tone) and the amp gain was adjusted so that the 1khz sine wave was just loud enough for me to hear. I didn’t want a lot of power going to my speakers (for the sake of my ears and voice coils). In any case the hash is low relative to the signal and is made up of out of band components (1CM=5khz so the small divisions are 25khz) I don’t see anything that is sharp enough to fit into a small dev (under 25khz). If you look at the middle picture you will see in band noise ridding on the sine wave this noise was audible and the BALUN (or correct HU termination) clears it up (3rd photo). My goal was not to improve the factory systems S/N but to provide a balanced interface for the HU that load with CMMR. I did some tests incorporating filters in the BALUN but did not like the high freq phase shift; it’s a heavy-handed approach.

I won’t be doing any more scope tests for a while. I have to return the Blu to its owner and I need to spend time in the shop (not in the trunk of the TSX). Not to mention my Honey-Do list is getting long.

jeff

 

Thank you for your invaluable work and clarifications on this topic, mercman!

We all owe you a big

 

Mercman,
I spaced out you were using the cd player. The cd player's performance is poor so i shouldn't be surprised to see a noisy sine wave. The tests i conducted were using a calibrated and clean sine wave which obvious looked clearer on a scope. That's probably the main discrepancy.

I understand the priority of the honey-do list.... Thanks for the scope pics. Your balun certainly improved in band noise. Its too bad the system noise floor is that large, but for a car I shouldn't be that picky.

 

im actually having a hard time figuring out how to take off the seat belt cover thing. can you insist me on how to remove it? also the rear covers so i can hide my power wires. thanks

 

Jeff.....Did you get my PM? Just curious. Thanks.

 

Grab a service manual man. You have to remove both front and rear channels, then pop 2 christmas tree taps down low, and then it will hinge upwards and slide out. Basically, you have to swing it up from the bottom.

 

Although I enjoyed the balanced signal discussion as I'm still working on some noise issues of my own, I'd like to add I did manage to run speaker wire through the driver's door last weekend.

Here's how I did it.... remove door panel, remove rubber boot inbetween the door and your TSX, unplug both wiring harness, pull the white plastic piece the plugs were in (called a "molex" i guess) through the door to the speaker hole (oh yeah, remove speaker too), drill a hole in the molex, run the wire through (I used 16ga) and put it all back together. I saw two places to drill the molex, upper-right or lower-left corner.

some pictures and the basic process is here (on a Mazda3)..

http://www.the12volt.com/installbay/...TID~78731~PN~1

 

Sorry to bring this back up, but I'm also having noise coming from my speakers after installing my jl 300/4 amp with mercmans harness.

 

But nothing in this thread relates to noise, its just about running wires. We'll keep that discussion to your thread.

 

sorry, i was going through the thread and you mentioned some noise when running mercmans harness so i thought my problem might relate to yours.