I am re-doing the sound system that I put together for my car, in an attempt to reclaim some trunk space. At present, there is a multi-channel line level converter, 6 channel active crossover and a 5 channel class-AB amp in there as part of my tri-amped system (SB acoustics sub & mids, Seas tweeters). It sounds great as far as car systems go, but it eats up the trunk!
So, the overhaul that I am working on involves a cheap 5 channel class D amp that I plan to modify with a custom built active filter board with differential inputs that will replace all of the cheap input stage electronics that are in it from the factory. My first quandry is with the way that the input stage connects to the class D amp IC's inputs.
According to the amp IC's datasheet, here's all the guidance they give as far as input suggestions.
They recommend using the differential inputs for what are probably obvious reasons since there is going to be all sorts of EMI being kicked out.
http://www.nxp.com/documents/data_sheet/TDA8950.pdf
Now, I spent some time with a DMM and the amp guts working out what the input stages look like. Although I am going to remove/bypass all of it, I wanted to get a handle on what is in there so I can effectively integrate my own input stage. So, can anyone tell me what the deal is with C9, C10, C11, R22 and R23 on the right? Doesn't look much like the datasheet, and you lose high frequency response through the components. The IC datasheet indicates that "clean" 20kHz output is not a problem, so I am not seeing a big purpose for all those passive bits.
I measured the cap values (off-board) and got a LTspice model working that does a reasonable job of matching the claimed specs for the stock crossover and gain controls. You can see that the high frequency response is...leaving a bit to be desired. The image is linked rather than embedded since it is a 1600x1200 screen shot. The SPICE model uses differential inputs because the audio source is a stock head unit with BTL outputs. At least with the line level converter I have in there now, this arrangement does a great job of rejecting alternator and ignition noise. The signal line that runs from the dash to the trunk is CAT 6 cable with each BTL signal pair running through twisted pairs. I tried a single ended input initially and the interference was incredibly bad.
http://www.e30tuner.com/speakerbuilding/images/car2/4chinputac.png
My ideal plan is to build up an active filter input stage and run the single ended filtered output to the IC's positive input pin (with the 470nF series cap) and tie the negative input pin to ground. Maybe I would add an inverted buffer stage to the filter output to use the IC's differential inputs if noise is a concern. Am I missing something major here such that my idea is no good?
Thanks!
Oh and in case anyone is wondering which amp this is...take a look and get a laugh. Some nondescript car audio thing, made in Korea, with instructions that don't even properly describe which pins are what for the inputs, or even which terminal is +12V/GND lol. I was going to have to crack this thing open anyway. They say you can't polish a turd, but I'm going to try anyway. The ambient noise level while driving on the highway is 60dB or more, so I don't hold this audio build to the same level of scrutiny as the stuff in my living room. But it is fun to try!
Power Acoustik RZ5-2500D (rz52500d) 2500W 5-Channel RAZOR Series
And no, the amp doesn't use Linear opamps. I found the closest-spec'ed ones in LTspice to the cheap ones in the amp.
So, the overhaul that I am working on involves a cheap 5 channel class D amp that I plan to modify with a custom built active filter board with differential inputs that will replace all of the cheap input stage electronics that are in it from the factory. My first quandry is with the way that the input stage connects to the class D amp IC's inputs.
According to the amp IC's datasheet, here's all the guidance they give as far as input suggestions.
They recommend using the differential inputs for what are probably obvious reasons since there is going to be all sorts of EMI being kicked out.
http://www.nxp.com/documents/data_sheet/TDA8950.pdf
Now, I spent some time with a DMM and the amp guts working out what the input stages look like. Although I am going to remove/bypass all of it, I wanted to get a handle on what is in there so I can effectively integrate my own input stage. So, can anyone tell me what the deal is with C9, C10, C11, R22 and R23 on the right? Doesn't look much like the datasheet, and you lose high frequency response through the components. The IC datasheet indicates that "clean" 20kHz output is not a problem, so I am not seeing a big purpose for all those passive bits.
I measured the cap values (off-board) and got a LTspice model working that does a reasonable job of matching the claimed specs for the stock crossover and gain controls. You can see that the high frequency response is...leaving a bit to be desired. The image is linked rather than embedded since it is a 1600x1200 screen shot. The SPICE model uses differential inputs because the audio source is a stock head unit with BTL outputs. At least with the line level converter I have in there now, this arrangement does a great job of rejecting alternator and ignition noise. The signal line that runs from the dash to the trunk is CAT 6 cable with each BTL signal pair running through twisted pairs. I tried a single ended input initially and the interference was incredibly bad.
http://www.e30tuner.com/speakerbuilding/images/car2/4chinputac.png
My ideal plan is to build up an active filter input stage and run the single ended filtered output to the IC's positive input pin (with the 470nF series cap) and tie the negative input pin to ground. Maybe I would add an inverted buffer stage to the filter output to use the IC's differential inputs if noise is a concern. Am I missing something major here such that my idea is no good?
Thanks!
Oh and in case anyone is wondering which amp this is...take a look and get a laugh. Some nondescript car audio thing, made in Korea, with instructions that don't even properly describe which pins are what for the inputs, or even which terminal is +12V/GND lol. I was going to have to crack this thing open anyway. They say you can't polish a turd, but I'm going to try anyway. The ambient noise level while driving on the highway is 60dB or more, so I don't hold this audio build to the same level of scrutiny as the stuff in my living room. But it is fun to try!
Power Acoustik RZ5-2500D (rz52500d) 2500W 5-Channel RAZOR Series
And no, the amp doesn't use Linear opamps. I found the closest-spec'ed ones in LTspice to the cheap ones in the amp.
Last edited:
Next course of action: put the amp back together and take it to the lab at work to do some things with a signal generator to see what happens.
I'm still baffled by the input components they chose. This thing is a budget item for sure, so I cannot imagine that they put in any extra parts (cost adders) unless absolutely necessary.
I'm still baffled by the input components they chose. This thing is a budget item for sure, so I cannot imagine that they put in any extra parts (cost adders) unless absolutely necessary.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.