A person I know with many years experience of designing and building tube amps told me that Solid State CCS, especially in the anode circuit, add high order harmonic distortion.
Does anybody have measurements results for SS CCS?
Does anybody have measurements results for SS CCS?
The real man's way of implementing a CCS is a huge B+ and a giant resistor, personally I'm not so much into adding silicon when a few tubes can do the job instead.
I'm still waiting for a nice pentode or double triode CCS implementation however, although I'm sure there must be one 🙂
I'm still waiting for a nice pentode or double triode CCS implementation however, although I'm sure there must be one 🙂
I've posted many here; a little searching will turn some up. Bottom line: your friend is full of it.
Check "guess the tube" thread for starters.
El156 project
It is a pitty you dont go ahead with your El156 project! Any particular reason? Costs? Power? You said to like the sound of el156 very much!
I am going on with mine,but only to finish next year,at the moment i am working with another one (EL34 PP).
Silvino
Hi Joshua!
It is a pitty you dont go ahead with your El156 project! Any particular reason? Costs? Power? You said to like the sound of el156 very much!
I am going on with mine,but only to finish next year,at the moment i am working with another one (EL34 PP).
Silvino
For example, a JFET or Depletion MOSFET with gate connected with source and with (or without) source resistor is not the best CCS one might design. It might be enough for some applications. The key is what the requirements are.
Ummm, that WAS a real circuit. Built it myself. Solid state CCS load, vanishingly low distortion, no "higher harmonics."
What is it you're trying to ask?
I think he means a load one is likely to see. But, it begs the question that, if a more difficult load led to high order harmonics when just the CCS didn't, wouldn't the blame be on either the tube, or the load?
At any rate, SY is of course correct, the CCS detractor has told you nonsense.
It is worth noting that even a poor single chip CCS is generally orders of magnitude better than a resistor load.
Your "person" doesn't know what he's talking about. It's unfortunate that there are those who view solid state as the "Spawn of Satan". It just isn't so.
If an active plate load could cause high order harmonics, then it wouldn't matter how the CCS itself were to be implemented. It's like anything else: a poor design will lead to inferior sonics, regardless of the actual implementation. Giving a triode as close to a horizontal loadline as possible reduces the resulting harmonic distortion.
Fact is, a solid state CCS will out perform a hollow state implementation. An active load implemented with discrete, cascoded, BJTs is pretty hard to beat. The gain of the BJT is much higher, and therefore, the impedance represented by that CCS will be much higher as well. Cascoding also helps to reduce Early Effect that compromises the "CCS-ness".
I've used cascoded BJT CCS's and these are sonically transparent. Use what gets the job done the best.
Bingo.
Triode distortion is minimized under constant current conditions, both in theory and in practice. Competent solid state CCS have much higher impedance than tube-based circuits, bringing the circuit much closer to ideal. The higher harmonic nonsense is... nonsense. Unfortunately, the audio world is full of that.
That person definitely knows what he/she is doing. 😀
But I do something diametrically opposing: I use CCS (actually, Gyrators) such a way all distortions go down under the noise floor. When level increases and approches clipping however harmonics start to grow one by one, starting from the 2'nd order one, but anyway the result is incomparably better than with resistive load when powered from the same B+ source.
Sure. A lot on this forum, in many threads!
Here was the basic idea which started a bunch of successful off-springs:
Here is the amp that uses them: http://www.diyaudio.com/forums/tube...-high-end-hybrid-amp.html?highlight=Tower-III
Here is the one that people enjoy:
http://wavebourn.com/forum/download.php?id=125&f=7
Also, there are many options presented by other people.
Last edited:
There are several opinions and even web sites devoted to CCS loads. I will agree that a 12AX7 running at 1 mA or less is probably a special case, but for most triodes running at 5 ma or more the CCS chip (IXYS 10M45) works very well and provides far less distortion than a resistor loaded tube. Even tubes like the 12AT7 that are known for excessive second harmonics clean up nicely with a chip CCS. The jury is still out on the 12AU7 though.
It is definitely possible to build a CCS that is better than the 10M45 or DN2540, but once you consider the entire amplifier where the output stage distortion is dominant, the choice of CCS is less important. Both of my SE amp designs use the 10M45 with great success.
Yes, I have tried several very complicated CCS circuits. Two cascoded chips or DN2540 fets works about as good as anything, but the added complexity isn't needed in most amps. See Gary Pimms web site for a full discussion of CCS circuits.
Welcome to Gary Pimm's DIY Page
- Status
- Not open for further replies.