[fiddler]

Okay, no nut in that case.

Well after some more listening, I've found that the level of detail in the Corda, if you can believe it, is even greater than before the mod. Also, the attack seem to be much quicker. The result: saxophones sound more saxophony, violins sound more violiny, pianos sound more pianoey.. pianoish.. whatever..

I find that the beginning of the note (the attack) really defines what a given instrument sounds like, and the modded Corda makes these attacks even clearer. Pianos and percussion sound much more articulate and now you can really hear the friction between the strings and the hair on bowed stringed instruments. I was listening to Brubeck's Take Five and I noticed much more "breathy-ness" in the sax at the beginning. The sound really seemed to come alive, giving each instrument more character. But despite all this added detail, it never sounds harsh or coarse, although it's definitely NOT what I would call smooth.
So, what kind of improments could I expect by swapping out the opamps? I'm dying to hear it.

NOTE: I haven't done any direct comparisons between the stock and modded Corda. However, I made the above comments after listening to familiar music that I've been listening to on my stock Corda + HD600 combo for about 6 months, so I have a good idea of what the Corda used to sound like.

NOTE #2: The above mod cost me virtually nothing, except for some solder, a bit of desoldering wick, a couple hex keys (METRIC ONES! I stripped a couple of the screws on the back plate using an imperial one... damn you Americans and your inches), and a bit of time. If you have a Corda and a soldering iron, this easy and almost-free mod is a must!

 

[kelly]

fiddler
It's awesome that this has worked out so easily for you. I guess there is a bit more investment to go all the way with the FETs and opamps but I'd be eager to hear your opinions of that.

To anyone else interested:
I had initially said something to the effect of "The Headmaster is better in almost every way possible but they're still different personalities and there's a quality about the Corda I like better that I can't quite put my finger on" or something like that. After some conversation with Dparrish and a day of listening to the Headmaster, I now know what that quality was.

The Headmaster is extremely smooth--easily the smoothest solid state headphone amp I've heard so far. But... at a price. The very top end of the Sugden sounds homogenized. That is - some sounds that I believe should sound distinct from one another sound a little too similar. The Corda sounds very smooth and delineated without crossing that threshold in the upper frequencies.

I've only had the Headmaster for a day and I'll certainly spend more time with it, but for those wondering what I could have possibly been talking about when I said there was something better about the Corda Blue, that's what it is.

 

[KurtW]

Since several people have mentioned the META42 amp, I decided to make a comparison with it and the Corda Blue, since I have built both. A few people heard my blue META42 at the SF HeadRoom show, although since then I've updated it with double buffers. It uses the AD8620, and like the Corda Blue has no input cap and has matched cascode FET current sources.

I used the Linn Ikemi as a source, and W2002 phones. Both amps were volume matched to each other. The cables were of course different due to the different types of inputs. For the first test the META42 was using a freshly charged 9.6v battery, measuring around 10 volts.

A friend volunteered to a blind test, using music he was familiar with. He thought the META42 had a little more lower midrange, making it sound a little heavy at times. He thought the META's soundstage was a little more incoherent than the Corda Blue. The Corda Blue sounded more open and refined to him. He preferred the Corda Blue, but thought they both sounded very good.

My impressions, not tested blind, were similar. The META does have a fuller lower midrange which can be viewed as better or worse depending on the recording, and of course the choice of other components. For the phones I usually use with the META (Koss KSC35), this is nice. The Corda Blue is more transparent overall, especially in the midrange and high frequencies where is is very open and airy sounding in comparison.

I later repeated these tests but used an external regulated power supply feeding the META's power splitter, and ran it at 20 volts. This may have improved the META's sound slightly, but didn't affect my overall impression.

Of course the META42 can be built to use the same op amp as the Corda Blue. There is also a big difference in the power supplies. I'm planning on building a META42 with dual regulated supplies and single channel op amps, and this will make a interesting comparison.

I should also mention that the Cordas offer three levels of crossfeed, whereas the META42 doesn't have crossfeed. All tests were conducted without crossfeed.

 

[The Quality Guru]

Quote:

I'm planning on building a META42 with dual regulated supplies and single channel op amps, and this will make a interesting comparison.

KurtW,

If one wanted to make a META42 with a pair of single channel op-amps, they could use two Analog Devices AD8610's, correct? Would one need a Browndog adapter as well?

Thanks.

Regards, Luke

 

[KurtW]

These AD devices come in an SOIC package, so you need a SOIC to DIP adapter, and Brown dog makes these. If you're using the META42 board from Tangent you need a dual adapter (two sided) so they fit in one DIP socket. If you have your own board you can use two single sided adapter boards to go into seperate DIP sockets.

 

[fiddler]

Kurt, how does that META42 compare to the stock Corda?

 

[Jan Meier]

Dear Kelly,

A very well written and thourough review indeed

And also a very flatterous one :-D (VERY BIG SMILE!)

As you rightly noted the CORDA isn't a perfect amp. It never was intended to be. The CORDA has been designed to offer a decent and most enjoyable sound for the lowest price and I'm glad that this has been recognized. You convinced me, that the Sugden sounds better (I never heard one myself) but it's no shame to loose from an amp that costs so much more. You can't build a Ferrari for the price of a Volkswagen.

I started building amps as a hobby, and am therefore very pleased to see how people like Kurt start improving my products. To share ones hobby is to share joy. To share joy is to double joy.

Actually, I'm learning a lot from you guys!

A few items have been discussed that I wanted to comment and a few suggestions:

1/8" jacks - Oh well, this one has been discussed on other places.

Bass loss with the crossfeed filter - The effect indeed is a psychoacoustic one. If you would listen throught the CORDA to a monosignal and engage the crossfeed filter no difference would be heard. The frequency-response of the crossfeed filter in itself is completely flat. More conventional crossfeed filters will show a number of distinct frequency dips at the higher regions (Comb-effect) and thereby accentuate the lower and the middle frequencies. They therefore sound more warm and bass-heavy. I do very well understand that certain people prefer these conventional filters. However, my intention was to alter the shape of the signal as little as necessary while implementing the crossfeed. At the end it's a matter of taste. Some like it this way, others like it that way.

Different opamps like the OPA627 or the AD843 indeed might bring the HA-1 at a higher level and it's nice to see people experiment. (Note: do not use the AD8610 and the like as they will not survive the +/- 15V voltage supply!). Again, for me the OPA627 never was an option because of its very high price.

To use parallel buffers BUF634 in the output stage is an interesting idea. Suggestion: Instead of using plain buffers people might consider placing these in the feedback of a control opamp. The BUF634 has a rather high output impedance of 7 Ohm (3.5 Ohm if two buffers are placed in parallel)and thereby the damping factor using low impedance headphones will be rather low. A control opamp will lower the effective output impedance considerably.

RCA/Cinch jacks. The prices of a single Cardas or WBT jack
are rather high. Moreover, assemblage is more labour-intensive. and I'm personally not convinced that different jacks make different sounds, unless something is definitely wrong with your cable of with the input characteristics of the amp.

The round nut (stator nut) was used to make it invisible behind the dial. On request people can have an amp with a conventional nut. The proper tool to remove the nut can be obtained from Farnell.

Resistor to -15V instead of +15V. When I did design the amp I wasn't aware of the fact, that PNP transistors are faster than NPN transistors. (Connecting to -15V opens the PNP transistor at the output stage, +15V opens the NPN transistor). I connected one pair of opamps to -15 and one pair to +15 to guarantee proper symmetric loading of the power lines.

The use of FET's to increase the load impedance indeed is a most interesting one. I suggest people to take a look at the specs of the LM334. This is a floating current source that might work even better.

The question whether all suggestions made by Kelly and Kurt (and all other people) will ever be implemented in a new version of the CORDA will be left open. To tighten the output to the -15V voltage rail is a small one and is well considered. However, using the OPA627 or a BUF 634 would simply increase costs, and thereby the price of the amp, immensely. By laws of economics an increase of $100,- in part costs implies a much higher increase in price of the amp. It's sad but true.

Cheers,

Jan

 

[Gluegun]

Hmm... this seems quite interesting....

So, I guess the simplest way to do this would be to buy the DT931/Corda package from Jan, and then reship the corda to KurtW for the blue mods?

Hmmm....

eenteresting...

*thinks*

Wait, I need a component CD player. D'oh!

 

[Nightfall]

Jan,


Thanks for your extremely informative response to this thread. How are things coming with the preamp that you teased us all with the picture of a while back?
I'm in a quandary as to whether to wait for that or have my Corda HA-1 upgraded to "Blue" status by KurtW.
Any estimates on when the new preamp will be available and projected cost? And it WILL have a 120 ohm output for the DT-931's wont it? Thanks.


JC

 

[Jan Meier]

Nightfall,

The PCB's for the PREHEAD will be collected tomorrow. Parts of the enclosure (hopefully) will be ready next week. All other parts have been received. The amp should be ready in approximately 2~3 weeks.

Costs for customers outside the EU (no VAT) approximately USD 750~800. I haven't made any exact calculations yet.

One of the jacks indeed has an increased 120 Ohm output impedance.

Cheers,

Jan

 

[Gluegun]

So, Jan, from your post, can I assume that these statements:

"To use parallel buffers BUF634 in the output stage is an interesting idea. Suggestion: Instead of using plain buffers people might consider placing these in the feedback of a control opamp. The BUF634 has a rather high output impedance of 7 Ohm (3.5 Ohm if two buffers are placed in parallel)and thereby the damping factor using low impedance headphones will be rather low. A control opamp will lower the effective output impedance considerably."

"The use of FET's to increase the load impedance indeed is a most interesting one. I suggest people to take a look at the specs of the LM334. This is a floating current source that might work even better. "

That those would be a way to improve on even the Corda Blue mods?

 

[Jan Meier]

Yep, the new PREHEAD has 3 parallel BUF634's per channel and each opamp (LM6171) has a LM334 current source (7mA). I can't compare with the CORDA Blue, but I find the result quite satisfying.

We did listening tests with various opamps (AD825, AD8610, OPA627, LM6171) and it was truly astonishing how well the differences between these opamps could be distinguished in this setup.

Cheers,

Jan

 

[Serow]

Jan, a couple of questions about the PREHEAD. Are the buffers operated in wide-bandwidth mode? Are they inside or outside the overall NFB loop?

Thanks,
Eric

 

[Jan Meier]

"Are the buffers operated in wide-bandwidth mode? Are they inside or outside the overall NFB loop?"

Yep, and they're placed inside the FB loop.

 

[KurtW]

Hi Jan,

It sounds like the PREHEAD is coming along nicely!

Doesn't placing the buffers inside the FB loop interfere with the crossfeed circuit? It sounds like maybe you have a new crossfeed circuit, since you also got rid of one stage in the Porta Corda and the crossfeed normally went between the two stages.

Regarding the LM334, this looks like a handy solution for biasing into class A. I noticed the capacitance is much higher than using discrete FETs like the 2N5485, but I assume its still low enough not to be a problem.