DAC project

[dirkwright]

Originally Posted by Ken Newton

Yes, I realized that. The Sabre DAC chips, and nearly all others today, utilize differential outputs which are referenced not to ground but rather, to a D.C. voltage midway between ground and the chip's positive supply rail. The DAC chip industry made this move mostly as a cost saving measure. Only a unipolar power supply is required, rather than the bipolar supply which used to be required.

I'm using the Twisted Pear COD DAC. It uses a TI chip, the PCM1794A. I don't completely understand it but it doesn't appear to have the kind of ground you are talking about.

[dirkwright]

Originally Posted by qusp

it really doesnt matter, the resistor still forms part of an RC filter across the primary and performs the IV conversion, or rather in the case of sabre the impedance causes the sabre to perform its own. not the transformer in either case 1865 or 9018

I believe the RC filter on the primary and secondary side of that circuit is to limit the bandwidth to reduce ringing in the transformer, not for I/V conversion. If the input impedance of the transformer were infinite, then yeah of course the resistor would be doing the conversion, but that's not the case here. The impedance is finite and virtually zero at DC, so I don't see how very much current is going to flow through that resistor. If there was a dead short across the resistor, I don't think any voltage would be developed across it.

I don't claim to be an expert in this by any means, so I could be wrong.

[dirkwright]

Wait. I meant to say "I don't know how very much voltage is going to be developed across that resistor". sorry.

I don't mean to argue with anyone. I'm just trying to understand all of this stuff. I'm a ME, not an EE.

[qusp]

what size is this imaginary resistor you arent going to use? if its more than 680R then ALL of the current is converted to voltage

[dirkwright]

Originally Posted by qusp

what size is this imaginary resistor you arent going to use? if its more than 680R then ALL of the current is converted to voltage

You'll have to ask Audio Note that. I don't have a clue.

So you're saying that if the resistor was infinite (as in not there), then all the current would certainly be converted to voltage? I'm trying to understand this here, not criticizing.

[Ken Newton]

Originally Posted by dirkwright

You'll have to ask Audio Note that. I don't have a clue.

So you're saying that if the resistor was infinite (as in not there), then all the current would certainly be converted to voltage? I'm trying to understand this here, not criticizing.

The importance of that resistor in the AN schematic is that it primarily sets the source impedance (in parallel with the DAC's typically high-ish output impedance) as seen by the transformer primary AND limits the voltage swing at the DAC chip output. In that process of limiting the voltage swing it necessarily and dominantly performs the i/v conversion. Yes, the transformer primary is in parallel, so together they form an i/v network, but is the transformer really helping the resistor much? Let's find out.

Okay, let's do some back-of-the-napkin calculations using that Jensen JT10KB-D. First, let's calculate the transformer's reflected impedance as seen by the PCM1794A, and with that op-amp circuit with the 2k43 ohm resistor load on the secondary side. The net secondary load would be: 2,430 ohms + the secondary winding resistance of 92ohms = 2,522 ohms. Now, we multiply this by the square of the turns ratio so, 16 * 2,522 = 40,352 ohms. However, we're not yet done. We have to add the primary's winding resistance. Therefore, 40,352 + 2,500 = 42,852 ohms! Great for connecting to a tube anode, not so much for connecting to a DAC chip's current output pin it looks.

Now, let's calculate the voltage swing which would theoretically be produced by the current output of a PCM1794A, assuming it had the output voltage compliance. It's datasheet says that the 1794A will output 7.8mA peak-to-peak at full scale. Which means, each differential output pin will output half that, or 3.9mA pk-pk. To calculate the resultant pk-pk voltage at each pin we simply multiply 3.9mA * 42,852 ohms = 167 Volts!!

Alright, I think we can conclude two things. One, given that the output voltage compliance of the PCM1794A is probably a few tenths of a volt, although it may be upwards of a volt or so, I'm not sure, an i/v resistor several hundred times smaller than 42,852 ohms would likely be required. Which also means that, while the transformer primary and the i/v resistor do form a parallel network, the current through the i/v resistor will be several hundred times greater than the current through the transformer primary. Two, the JT-10KB-D may not be the best transformer for this application.

[dirkwright]

The transformer is not designed to connect to a tube annode. It's a balanced line receiver.

I seriously doubt that the TI DAC will put out that kind of current into 40kohms. It'll just put out less current if it's loaded like that.

[Ken Newton]

Originally Posted by dirkwright

The transformer is not designed to connect to a tube annode. It's a balanced line receiver.

I seriously doubt that the TI DAC will put out that kind of current into 40kohms. It'll just put out less current if it's loaded like that.

You seem to be working hard at avoiding my points. In general, you seem less than receptive to guidance for someone who has essentially said that they don't really know much about electronics. Best of luck with your project, just the same.

[qusp]

not arguing though...... Ken said it all, this is too hard work for what was a really simple query. i suggest doing some basic reading on electronics that doesnt involve audiophile manufacturers datasheets.

either way i'll be leaving you to it as well

[dirkwright]

Originally Posted by Ken Newton

You seem to be working hard at avoiding my points. In general, you seem less than receptive to guidance for someone who has essentially said that they don't really know much about electronics. Best of luck with your project, just the same.

I didn't have time last night to thoroughly read your post, sorry.

I agree that the Jensen transformer I picked may not be the ideal one for this application. Probably a simple 600 ohm repeat coil would have been better.