DIYHiFi.org

I am glad to finally see a class D section here.
As I said a while back, I have very limited knowledge of electronics and so I need help from you guys to get my car-audio-computer project going.
Oh, by the way, I have found a very reasonable priced board manufacturer in the States. Their name is .....(don't know if this is allowed). They manufactured me five boards with each containing two different Tripath amps, an active 3way crossover and a USB Dac on it, all for $128 and the shipping to Germany was even included. The only problem is that the minimum order is five boards.

But now to my questions:
I have designed the Tripath TA2020 amp to function without coupling caps.
I fully trusted Leo on this one. He posted the schematic of his Waltham Integration Amp in the Klipsch forum. He sold a few amps, so I assume they work just fine without the caps.

Here is his post:

“I have designed, produced, and distributed (including a few sales) a small stereo amplifier based on the Tripath TA1101B. My design is taken directly from the Tripath application schematic in the TA1101 data sheet. I have made one very important change in the input, and learned a few things about which components should be optimized.
My application is driving Klipsch Forte, RB-5, and RF-7 loudspeakers. I generally use a fraction of a watt .. never more than two or three Watts. I have done some work with tubes: currently I have a design that uses KT66s, PP, triode, no feedback. The tube amp is nice, but the Tripath is very very nice.
The input:
see Tripath's TA1101B Data Sheet:
"Application/Test Circuit" at Tripath.com
There is a 2.2uF electrolytic (C1) on each input because the input amp is biased at something like 2.5V. That cap has a huge impact on the sound. To get rid of it:
Observe static electricity protection rules. Until this circuit is complete, the Tripath input is floating and very vulnerable.
1) Use a floating, non grounded, 12V power supply. They are readily available.
2) Isolate audio input ground from the 12V ground and connect it directly to the TA1101B-pin 14.
3) Connect the audio signal directly to the resistors: R1. Now the audio is connected differentially to the input amps.
4) Bypass the audio ground (now connected to pin 14) to the circuit 12V ground. I use a 22uF electrolytic and 0.1uF ceramic. + connects to the audio ground. This step limits common mode noise which otherwise totally swamps the input circuit.
5) Insert a 7 or 8 volt zener diode between the audio and 12V grounds, cathode to audio gnd, to protect the chip input, esp pin 14.
6) Put big rf beads on all power, input, and speaker wires (you really should do this anyway with this amp). A single bead for each +/- pair is fine .. that's 5 beads total.
That's it. it makes a BIG difference.“

Since my USB dac is based on the AKM AK4396 I thought I might try to run fully differential, but I don't know how to accomplish this. I had the idea because he made the following statement in his post:
“Now the audio is connected differentially to the input amps“.

Do I need to have two of the TA2020 amps to be able to run them differentially ?
How do I have to connect the signals coming from the dac ?
How are the outputs connected to the speakers since the signal is not referenced to ground ?

I attatched Leo's schematics for clarification.

Thanks in advance for your help,

Klaus

Sorry I do not have time to consider this circuit - you are asking someone to design it for you, maybe someone is interested in helping, but I would suggest getting your feet wet with a tried design first and learn a little more electronics.

It is good that you are checking, something that jumped right out at me was:
.
Now it is pretty well excepted that ceramics are not good for audio so it worries me that he is using a ceramic to connect the "gnd" (which is the input's ground) to the "agnd"

Good luck, oh and what was the name of the board maker?

Bill

I got my feet wet with five diy amps and lots of other gear that I have build. All with trial and error and the help of some informative members on forums like this.
There must be someone that is willing to tell me how to hook up my differential dac to this kind of amps.

Greets,
Klaus

Amplifier shown on the shematic has two channels, single ended inputs and bridged outputs. You connect one speaker between PAD3-PAD4 and the other between PAD5-PAD6. You connect one output of your dac to J1 and the other output to J2. DAC should have no DC offset voltage between GND and signal line.

You can first try with the shorts over C3 and C4 removed. If the amplifier works, you might measure voltage across C3 and C4 and if it is zero, then you might insert shorts over them. It would be probably wise if you use resistors as load instead of loudspeakers while you perform such measurements.

Since your amp has already bridged outputs, it can not be used diferentially.

If it takes two amps to go differential I will do that.
Do I still have to use the ground of the dac or is it just the four signals (for stereo)that will feed the amps?
It's embarrassing that I don't have a concept of what differential really means.
I would be thankful of a web accessable writing that explains it in an easy way.
Does someone know of such?

Greets,
Klaus

Wouldn't it be easier to use an amp that has differential inputs????

Like a Hypex Class D.

I'm positive it is also a much better amp than anything made with Tripath.

Jocko

1) You can not have differential input on your amp unless you add differential to single ended stage in front of it.
2) You can not combine two amps into one differential unit.
3) You still need ground wire for the DAC.
4)http://www.dself.dsl.pipex.com/ampins/balanced/balanced.htm
5) google further for balanced line

well actually you can, I fixed a Crown that did this many years ago - you just need two independant floating supplies and... Of course since the voltage swing is doubled the output current into a particular load is doubled, hence 4x the power of a single amp - the amp better be designed for it.

Of course you can't combine these two amps that are on one chip 'cause they share supplies and grounds - So Raka is correct " You can not combine two amps into one differential unit."

Those amps that used one amp to drive the "ground" of the power supply of the other to jack the power output up, do you?

I think many of us would care to not be reminded of them, thank you very much..........[joke]

Jocko

Didn't say I liked or approved of the idea...

It is possible!

never seemed like a good idea for the power supply "to dance with the signal"

while going down memory lane: remember the kenwood amp which they decided didn't need mica insulators so they floated the heat sinks ±80V if memory serves... watch where you put your hands.

or the teac cassette deck that had the 120V ac switch terminals right on side so that when you went to turn it over (out of the case) you got zapped - don't care to remember how many times that one bit me.

now back to your scheduled programing

The little charlize Tripath amp that I have modded sounds every bit as good as my SKA GB150 and that sounds
better than the Accuphase E 303 I had and better than my AKSA 55 N.
Shinobivan on the other forum compared quite a few amps and he liked the SKA better than the UCD.
One big reason that I use the little Tripath amps is that I can use a 12V Battery or two for 24V as PS.
This makes them highly portable and cheap (because I also use my own boards).
If there is one thing that I learned in HiFi, it is not to advocate a certain technology.
I thought I would never use anything else than Tube amps, but the SKA came so close to my SE 84L amp
that it is only a matter of mood to chose one or the other.
The same goes for dacs. With the purchase of a DDDac I tought non OS is the only way to listen to music,
but then I tweaked (passive out and Rubycon ZL caps) a soundcard with the AK 4396 on it and it walks
all over the DDDac and all my other tweaked players, in all aspects. A few years ago I never ever thought I would use a computer to listen
to musik, but I do now.

Greets,
Klaus

Klaus, post your current setup, output network of AK 4396 and input network of your TA2020 and I'll show you how to interconnect them in differential manner, if you don't mind to buy 2x 1:1 line level transformers

Thank you so much Dimi.
The last picture in my web-album gives the passive out description that I took from Scott Endler.
Someone suggested to me that the AK4396 does not need the two pole filter of the AK4394.
Mybe I should try to go without a filter and just use a cap.

Is there any sonic benefit to be expected by going differential?
Trading a cap for a transformer does not seem to be the better option to me.

http://picasaweb.google.de/Klaus.Freis/Audio

Greets Klaus

it is sonic benefit by going transformer coupled.
Yes you can throw away 2-nd order filter and decoupling elco. You will eliminae common mode garbage from USB port as well.
Show me your TA2020 input circuitry. Line transformers from lundahl is $50 each.