I hope nobody takes this the wrong way...

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Your DAC may already be grounded via the computer

SPDIF optical.:D

A PC is a nasty thing to electricaly connect to an amp in my opinion.

The only thing that I'll ever connect to the amp is the Entech DAC so I'm not worried about how it will act with other sources.

There's no hum at all and the noise floor is the same level it was before the mod. The grounds of all the amps I have put together are floating and I only connect the mains ground to the chassis or in some cases nothing at all. Every time I have tried connecting an amp or source ground to the mains ground i get hums and ground loops.

May I post this nice new to Jan from 41Hz?

I'm just a noob that tried someone elses advice. Fell free to pass it around.
 
Ex-Moderator
Joined 2005
theAnonymous1 said:
Basically just wire the inputs without any caps, but instead of connecting the input ground to analog ground; connect it to the biascap pin. Then you just need to make sure the analog ground and your source ground are isolated and that the source has no DC.

By connecting the source's ground to the biascap pin you can potentially introduce a lot more noise into the input stage. Also, as was stated, the power supply for your source, your amp, or both must be floating. I still wouldn't be comfortable doing it this way because unbalanced input bias currents on the input opamps could cause noise, DC offset, etc. I had thought about bypassing the whole input stage all together a while ago but never really did anything about it. It would have been nice if Tripath had made the input opamp stage more accessible, but it is apparently tightly integrated with the other proprietary 5V digital/analog signal processing circuitry in the chip. It would be better yet if they made their chips to run off of split supply rails so they could safely do away with input coupling caps!
 
I still wouldn't be comfortable doing it this way because unbalanced input bias currents on the input opamps could cause noise, DC offset, etc.

These were warnings I understood before I tried it, but after actually doing it there is no more noise than there was before and the DC offset didn't budge, so all is good.;)

Maybe I got lucky and have a good source. Dunno, but since its the only source I will use I am very happy with how its working.

Like I said above, all my equiptment is always floating(high as a kite).
 
Does your source have a coupling cap on its output?

No sir, the output is from an OPA2134PA with only a resistor in series. Even if it did, I directly shorted the inputs to the biascap pin while everything was powered up and all it did was quiet the output noise down that was present with open inputs.

Thomas, consider power ground and analog ground of the amp the same thing in this case. The only important thing is that your source ground is isolated from both of them.

Be sure to double check voltages. There should be no voltage between the input and the biascap pin. Also once you disconnect your source ground from the board, make sure there is no continuity between the two before you attatch it to biascap.
 
Ex-Moderator
Joined 2005
theAnonymous1 said:
No sir, the output is from an OPA2134PA with only a resistor in series. Even if it did, I directly shorted the inputs to the biascap pin while everything was powered up and all it did was quiet the output noise down that was present with open inputs.

Sounds like you're good to go then! DC coupling sounds so much better doesn't it? That's the reason I haven't gone back to the Tripath stuff :)

The next step would be to not use the on chip opamps at all and run the signal from the OPA2134 right into the feedback pin attached to the output of the internal buffer opamp. I'd be willing to bet you'd get even more improvement.
 
Administrator
Joined 2004
Paid Member
BWRX said:
The next step would be to not use the on chip opamps at all and run the signal from the OPA2134 right into the feedback pin attached to the output of the internal buffer opamp.

But he's going to have to do something with those onboard opamps. There are still powered up. Don't want them adding noise or other junk to the input bus.

What do you think? Tie them to ground?
 
Ex-Moderator
Joined 2005
panomaniac said:
But he's going to have to do something with those onboard opamps. There are still powered up. Don't want them adding noise or other junk to the input bus.

What do you think? Tie them to ground?

That's an excellent point, Michael. Seems I hadn't fully thought that one through.

If the inverting inputs are tied to ground with the non inverting inputs sitting at 2.5V then it will cause the output of the opamps to swing to the positive rail IF there is no feedback loop.

Since we have access to the inverting and output terminals of the internal opamps they could be wired up for unity gain (short inverting input to the output) and then you can feed the signal into the feedback pin through a resistor to sum the audio signal with the DC bias. That will cause the internal opamps to act as a DC buffer where the outputs will sit at the DC bias on the non inverting input. You'd still have that 2.5V bias and that should effectively take them out of the equation (I think).

Edit: that may not work well. I can't see a good way of bypassing the internal opamps besides wiring it up for unity gain and using the external opamp with gain. Or maybe including the internal opamp inside the feedback loop of the external opamp circuit.
 
Ex-Moderator
Joined 2005
I thought about this a little more and came up with a circuit that puts a feedback loop around the internal opamp. Think differential amplifier with the internal opamp used as an inverting stage inside the feedback loop. The values shown in the schematic will give you a gain of 1V/V. Simply move the input voltage source to the noninerting input for a gain of -1V/V.

U3, V3, V4, and R6 are just for simulation purposes and would be part of the chip.

Also note that the 5V supply for the internal opamp is floating, as well as the 2.5V DC bias with respect to the external opamp ground. Either the amp's supply needs to be floating or the preamp/source supply needs to be floating. I would float the amp supply.
 

Attachments

  • tripathinputcircuit.gif
    tripathinputcircuit.gif
    60.4 KB · Views: 395
Member
Joined 2005
Paid Member
Anonymous1 --> Are you thinking of toying with this input cap removal on the DCP-501 amp Project?

The loaded/unloaded Entec DAC does not introduce any DC on it's output/amp input?

I like. I was planning on putting the DCP-501 ampboard and Entec in one case and using only the Optical Input - this will remove any possibility of upstream DC on input...
 
john65b said:
Anonymous1 --> Are you thinking of toying with this input cap removal on the DCP-501 amp Project?

The loaded/unloaded Entec DAC does not introduce any DC on it's output/amp input?

I like. I was planning on putting the DCP-501 ampboard and Entec in one case and using only the Optical Input - this will remove any possibility of upstream DC on input...

There isn't a single bit of DC coming from it outputs.:cool:

I'm going to try this on my AMP1-B today and if all goes well I do plan on doing this to my 501 boards. The 501's will get their signal from three Entechs so I should have the same great results.

This will save me from ordering about $20 in caps from parts-express too.:D

I will probably test it on the center channel first. I don't plan on using it in my final setup so if it dies its not a big loss.

I won't get around to that project for a very long time though. I want to buy two SPS80 supplies from coldamp to power the boards, but thats a serious amount of money and I don't have it right now.
 
Good news Thomas, I just DC coupled my AMP1-B with the same success.

It was as simple as removing the input ground from the boards analog ground and connecting it to pin 27 (or pin 27 side of C48).:D

In this amp I plan on leaving the caps in series with the inputs and adding a bypass switch. This way If for some reason I ever have a source on it that has DC I can just flip the switch. Doing this will also let me instantly A-B test the sound with and without the caps.;)
 
Seems I have run into a bit of a snag.:(

I was playing some low frequency sine waves and the AMP1-B's protection kicked in at a much lower level than before the mod. I checked my DAC's output and sure enough there was some DC.

At frequencies lower than about 40hz some very low DC starts to show up. It fluctuates but is around 20-40mv or so. I guess this is enough for the amp to shut off. With a digital signal of 0hz I get the full 2.5v(did this by mistake).:hot:

I don't really know anything about opamps. Any guesses why the OPA2134 would do that?

What about the pot ground?

Connect it to the new signal ground. The AMP1-B I'm playing now has a 10k pot connected between the DAC and amp.
 
Member
Joined 2005
Paid Member
Panny --> are you looking into this option (Sans Input caps) on the SI? You cannot connect the signal grounds to the Biascap input as they must be kept separate, correct? Sounds like this mod may only be done to some of the other Tripath amps...right?

Thomas --> I spent most of my childhood in Rimini - till I was five...so we are Piasans! I go there every two - three years to visit mia famiglia...Forza Italia!
 
Panny --> are you looking into this option (Sans Input caps) on the SI? You cannot connect the signal grounds to the Biascap input as they must be kept separate, correct? Sounds like this mod may only be done to some of the other Tripath amps...right?

I've modded a few SI in the past, but don't have any around at the moment. I can't think of a reason why you couldn't do this on one.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.