[aphex944]

I recently built a CK²III DIY amplifier. I decided on this one for it's relatively cheap part cost, and interesting design. I was also encouraged by the few impressions I've read on various forums, including this one.

On the bench, it measured to be very quiet. However, with a capacitive load(used a 2200pF cap), the amp oscillates. I'm sure it's fine with every headphone out there, but it's still disconcerting that the amp doesn't have much headroom at all(like if it was driving a very long headphone cable). It's also thermally unstable, and the output transistors really need heatsinks. We had to keep on going back and adjusting the trimpots as it never really settled down.

This was disappointing, but I was in for a real shock once I listened to it. In it's stock form, the amp is positively horrid sounding. It has a very "dirty" sound to it: voices, cymbals, etc. have a gritty texture and almost offensive balance. There is NO smoothness whatsoever, and the bass is very poor. Overall, the sound is very upfront without much air, and generally lifeless and boring. A friend that was listening as well came to the same conclusions. It reminded us of a headphone output on a cheap CD player. We were using AKG K 271s, Beyer DT 880s, and DT 990s. The ONLY good thing I can say about it is that the noise floor is very low; the background is very quiet.

We're going to try swapping out the JFETs for some BJTs. As far as audio is concerned, we've always found that JFETs have a "dirty" sound to them. I hope this really changes the character of the amp and at least makes it listenable, because right now it's not

It's always disappointing putting money and labor into something that doesn't turn out right. Has anyone else had a negative experience with the SOUND of this amp? I'd like to highlight though that AMB did an excellent job with the board layout and parts recommendation. The board fit the case perfectly, and overall it was a decently easy build.

Pictures to follow!

 

[Rise To The Top]

Pics are a must. I heard that the CK²III is up there next to the B22.

 

[Veniogenesis]

Could this be a problem with this particular build as opposed to the amplifier design?

By the way, Rise to the Top, who said the CK²III is up there next to the B22? It would prove extremely interesting if this were true.

 

[rb67]

Hm, this description certainly doesn't remind me of my CKKIII. I just completed a Bijou and for the price, I can say it stacks up very well against an amp of higher caliber. I wouldn't call it a gritty amp, but it certainly is cold IMO. Bass response is pretty snappy though.

 

[Rise To The Top]

Quote:

Originally Posted by Veniogenesis /img/forum/go_quote.gif Could this be a problem with this particular build as opposed to the amplifier design? By the way, Rise to the Top, who said the CK²III is up there next to the B22? It would prove extremely interesting if this were true.

I've heard from quite a few that it is like a sibling to the B22, and isnt far off from it.

Cotdt told me not too long ago a balanced CKK3 would be about equal to a B22, just a bit colder.

 

[aphex944]

UPDATE: We replaced the input transistor JFETs with BC550 and BC560 BJTs. What an astonishing difference!!

Upon first listening with this new setup, the dirt was completely gone. It had a slightly strident sound at first, but ended up mellowing out quite nicely. The overall sound is slightly shifted towards the higher frequencies, but it is NOT bright. It's actually quite "happy" sounding, and I couldn't be happier with the improvement

It would be very interesting to compare it side-by-side with a JFET version, as I think the differences would be jawdropping. I'll add more impressions once I get to know this new setup.

This circuit was originally designed with BJTs anyway!

Moreover, there is actually less distortion now. With the JFETs at 1khz-- .065%. With the BJTs, .043%

At 20khz, there is an even bigger difference. The other large change was the addition of second harmonic distortion, and less third, even though overall it was less. You're not supposed to be able to hear anything this small, but something made the difference. This is curious, too, since on AMB's page it mentions the JFETs actually lowered distortion.

Below are some photos. A slightly different power and ground scheme was used so we were able to use a standard wallwart. Otherwise, a toroid would've been necessary. This is why some parts are missing, and the addition of a jumper wire. Most of the recommended parts were used, but there are a few different capacitors.

 

[jinp6301]

The CKKIII I have is very good. For its price, I cant recommend anything else, especially for senns. It really tames their slowness that they have with entry level amps. But theyre not so good with grados imo. Everything is TOO tight.

 

[MrLA]

i have a ckk3, buiilt by rockcod, and to me it sounds like a baby dynahi or a gs-1 lite. it is very good, indeed. sound signature - creamy smooth.

 

[TzeYang]

i hate you guys for doing this.








(ouch, i'm going to starve for another month so i can afford to pay for the shipping and the ckiii pcb and parts with my country's tiny currency)

 

[rb67]

Excellent casework! Very clean.

I'm glad you were able to get the sound the way you like it. It's truly one talented amp in a little package!

 

[amb]

I can't explain why switching from JFETs to BJT input transistors improved anything. My own experience and measurements showed that JFETs were superior. It was because of that Alex Cavalli and I changed the input stage to JFETs from Benny Jørgensen's original Kumisa III design. Also, the OP's THD measurement results (either JFET or BJT) are quite a bit higher than mine.

Just as a side note, the CK²III input stage is not very different than some of the past designs by Erno Borbely (which graced the pages of The Audio Amateur and AudioXpress magazines), and had garnered very high praise. Virtually all of his designs employed complementary JFET input stages, and nowadays he uses JFETs and MOSFETs almost exclusively throughout the circuit.

 

[NelsonVandal]

To me this is interesting reading. Having tried many different opamps with JFET or bipolar input stages and discrete buffers using only BJT's, I think there's a different character between JFET and BJT. JFET's being a bit, like you say, gritty but are often warmer and "more musical". BJT's have a smoother/cleaner sound that could be found a bit sterile or distant. Is there really a JFET-sound, or is it just in my imagination?

 

[aphex944]

Quote:

Originally Posted by amb /img/forum/go_quote.gif I can't explain why switching from JFETs to BJT input transistors improved anything. My own experience and measurements showed that JFETs were superior. It was because of that Alex Cavalli and I changed the input stage to JFETs from Benny Jørgensen's original Kumisa III design. Also, the OP's THD measurement results (either JFET or BJT) are quite a bit higher than mine.

Did you listen to an all-BJT version and compare it to the JFET input version, or just went by measurements?

We used a high end HP distortion analyzer, a function generator, and a 40mhz oscilloscope. I noticed on the specifications page you used a sound card and RMAA to test the distortion, so maybe that's where the discrepancy is coming from.

Quote:

Originally Posted by NelsonVandal Is there really a JFET-sound, or is it just in my imagination?

In our experience, there is a common sound associated with JFETs. Every circuit *I've* heard with JFETs has suffered from the same issues to varying degrees.

 

[amb]

aphex944, if you like the BJT input stage, well, it's your build, you are free do whatever you like. However, your impressions about the CK²III with JFET inputs run contrary to just about all others I've seen.

If you think that JFETs sound "gritty" in every circuit, then so would a β22, dynahi, dynalo (or the Headamp products based on it), or any amps with JFET opamps.

 

[aphex944]

Quote:

Originally Posted by amb /img/forum/go_quote.gif aphex944, if you like the BJT input stage, well, it's your build, you are free do whatever you like. However, your impressions about the CK²III with JFET inputs run contrary to just about all others I've seen.

Indeed, which is why I was curious to see if ANYONE has had a negative experience with the amp. I haven't been able to find any, but I'm not sure what most people are comparing it to.

Quote:

Originally Posted by amb If you think that JFETs sound "gritty" in every circuit, then so would a β22, dynahi, dynalo (or the Headamp products based on it), or any amps with JFET opamps.

Never heard those amps, but I've heard plenty of JFET opamps and none have been very good(save for the OP275, but that still has bipolar parts). I'd call the JFET sound more "dirty" than "gritty" anyway.

Thanks again for taking the time to make the boards, etc.