[Joe Bloggs]

Is it true that putting a resistor between the ground and voltage rail of the Porta Corda is all it takes to get it to run in Class A instead of Class AB? If so will the chips on the PC be able to take it? Can somebody point out for me where the ground and voltage rail are? Or what should I look for?

 

[slindeman]

Not between ground and voltage rail, rather a resistor from the ouput pin of each opamp to one of the voltage rails.

The circuit can probably take it fine, but check with Jan to be sure. It is done on the regular Corda and I don't believe the opamps require a heat sink or anything, and they are the same opamps in the Porta Corda. The Corda uses a 1.5Kohm .6 Watt resistor for each opamp. For your reference, the notch is considered the top of the opamp.

Connect the resistor from the output pin of the opamp (the third pin down, right side) to either the V+ pin (2nd pin down, right side) or the V- pin (4th pin down, left side). I'm not sure which opamp you connect to the V+ and which to the V-. I know in the Corda, the two first-stage opamps connect the resistor to V+ and the second-stage opamps connect the resistor to V-. Not sure why though.

Before doing this I would fire an e-mail off to Jan just to be sure. I assume you are using an AC power supply, this would not be good for batteries.

 

[tangent]

Joe, you can see what Jan's doing in his HA-1 in the schematic in the article on HeadWize.

I've gotten the LM6172 (dual channel version) to run at full operating power and it does get pretty hot, but I'd say it could tolerate not having a heat sink. The single-channel LM6171 used in the Porta Corda would put out only half the heat as the 6172 at full power, so I'm sure running it without a heat sink is fine. Jan does, too, since the HA-1 kit doesn't include heat sinks for the op-amps.

It appears that Jan uses an 0.6W resistor here just in case 30V happens to appear on one side of the power supply all of a sudden, since Ohm's law says a 1.5K resistor will dissipate exactly 0.6W in such a situation. The datasheet says an LM6172 can handle all 30V at one rail with 0 at the other, so I guess this is a reasonable design choice. If you're using a smaller power supply, like 24V, 0.5W would be plenty for this purpose. If you assume the power rails will remain split, 0.25W will be fine for this purpose.

By the way, I apologize for the misinformation about the resistor being between ground and a voltage rail -- that's plainly bogus to even an electronics tyro like me. I'm glad you asked before trying it, and Slineman caught the error!

 

[Joe Bloggs]

Don't worry tangent--I don't think I'll be budging for a long time yet

 

[Joe Bloggs]

Quote:

Email from Jan: The output stages of the opamps indeed can be driven into class-A by connecting these outputs to one of the power lines. The principle is shown in the schematics of the CORDA HA-1 (see my home-page). The improvement in sound to my ears is substantial although not as big as the step of going from the normal headphone jacket to the Porta Corda. Last week I moved to a new place and most of my papers are still in boxes. I therefore do not have the exact information available yet where you may place the resistors. Please send me a note to remember next week.

Well, what do you think?

'The improvement in sound to my ears is substantial although not as big as the step of going from the normal headphone jacket to the Porta Corda.'

How good does that sound to you? Would the step of going from the PC to the big Corda be bigger than going from the normal headphone jack to the PC

How close would this take the PC towards the big Corda? I suppose the main difference would be the LM6171 vs the LM6172. What's the difference between single and dual channel? Also what's the voltage running through the Corda?

Would driving the PC at Class A help it to deliver more power to the headphones so that it can drive the HD580 better (of course with better signal fidelity any headphone should sound better. But what about areas of improvement typically associated with more power--e.g. more tighter bass and transients?)

 

[tangent]

Quote:

Would the step of going from the PC to the big Corda be bigger than going from the normal headphone jack to the PC

Here's the full set of diffs, as I see them:

1. HA-1 runs in Class A

2. HA-1 has a bigger power supply with more filtering, bigger reservoir caps, and more voltage. (+/-15V vs. +/-4.7V with a fresh battery)

3. Bigger, better volume pot on the HA-1. This may be an audible benefit, or perhaps just a tactile one.

4. Only one crossfeed setting on the PC. There are three on the HA-1.

The first two you can approach by hacking the PC. The latter two aren't attainable with this design.

Quote:

I suppose the main difference would be the LM6171 vs the LM6172.

No, both amps use four LM6171s.

Quote:

What's the difference between single and dual channel?

Jan says it's better for crosstalk, but I don't understand why. NatSemi says the LM6172 has -110 dB of channel crosstalk rejection through the audio range. In other words, with a 1V input signal in one channel and silence in the other, you'd expect as much as a 1/330,000V signal in the other channel. 3 microvolts!

The real benefit, I think, is that when you're running the chips in Class A each chip only has to dissipate half as much heat as if you'd used dual channel chips. That can mean the difference between using heatsinks or risking an early chip death, and being able to get away without heatsinks safely.

Quote:

Would driving the PC at Class A help it to deliver more power to the headphone

No. Lots of voltage is good for swinging high-impedance drivers, and lots of current is good for moving low-impedance drivers and handling big bass. Running in Class A provides different benefits I don't fully understand, but I think it has to do with eliminating the on-off switching of running in class AB or B -- you keep the chips active devices turned on all the time, so they react quicker and contribute less noise to the output signal.

 

[slindeman]

I would make the mod. The reason? With your 24V AC adapter, you would use 1.2k ohm resistors to bias the opamps into class A. That would sink a constant 10mA per opamp (12V/1200 ohms), the same as is done in the Corda. This assumes your AC adapter is regulated and so the V+ is really 12V and the V- is -12V. Measure to be sure. When you switch to using a 9V battery, then it will only be sinking a constant 3.75mA per opamp (4.5V/1200 ohms), for a total of 15mA. That would probably only cut the battery life in half, which is reasonable. With the 9V battery it would only be in class A up to 3.75 mA, but that will still get you to 100 db SPL with the HD600. For even better battery life, increase the resistor, say to 2.2k or so. It would only be in class A to 5mA or so with your 24V adapter, but that might be a reasonable compromise if you still plan on using it with batteries.

 

[Joe Bloggs]

Quote:

No. Lots of voltage is good for swinging high-impedance drivers, and lots of current is good for moving low-impedance drivers and handling big bass. Running in Class A provides different benefits I don't fully understand, but I think it has to do with eliminating the on-off switching of running in class AB or B -- you keep the chips active devices turned on all the time, so they react quicker and contribute less noise to the output signal.

Ya, I know--it's called intermodulation distortion, right? (The difference between AB and A) I might be horribly wrong too, of course

I was thinking along the lines of, 'well, since running the PC in class A makes it dissipate so much more power, the might be more extra power fed to the cans too, right?' I should have known better

So, the resistance of the resistor between the opamps and the voltage line determines what current level the opamp can go up to in Class A?

I have used it with batteries for only once or twice. It might have something to do with the fact that I only ever take my EX70's outside

 

[taoster]

old post revisited!


Hi, I'm planning to mod my Porta Corda to Class A as I myself use the Porta Corda quite alot at work, running off at 24V regulated supply.

Has anyone else(beside Joe, ive pm him already) tried the mod too? I'd be most interested in the result!

Taoster

 

[KurtW]

I've done this by putting a cascade FET current source between the output and the V- rail, which is fairly accessable in the PC. I selected the FETs so the total current for each channel is less than 2mA so it still is fine for battery operation. This is for the first stage only, not the output stage which would required much more current. I thought it produced a noticable improvement but it still isn't as good as the Corda.

 

[ppl]

KurtW Is corect I have stoped using resistors to Bias Opamps into Class A in the late 70's In 1979 the resistor was a neet trick nowdays it is outdated and restricts the dynamics.

 

[taoster]

KurtW,

ive read your post about your FET mod but you didnt get into detail on how it was done and im nothing more than a soldering monkey thus i was opting on the 4 resistor mod instead.

I assume the FET current source mod is the better solution to turning the porta corda from class AB to class A amp? or would it be pretty similar? how do one go about putting a cascade FET current source between the output and the V- rail?? that's a chip rite?

damn, and i was all fired up all day, waiting to get home and try the 4 resistor tweak!

Taoster

 

[zbuddah]

so is it advisable to keep R10 in the meta in if you already have a set of jfets in? does the same thing here apply with the meta?

 

[Dreamslacker]

Quote:

Originally posted by zbuddah so is it advisable to keep R10 in the meta in if you already have a set of jfets in? does the same thing here apply with the meta?

Yes, you would keep the resistor there if you already have one.

 

[ppl]

See this http://www4.head-fi.org/forums/showt...current+source For futher info