That's obvious. There is no differential amplifier in a SE amp.
In general, nothing comes for free. If one fine tunes the operative conditions of a pentode and adds local N-fbk or uses very linear triodes for the job they are supposed to do there is no need of anything else, IME.
Last edited:
Morgan Jones give an example of distortion cancellation for a given topology, and proves that in SE circuits distortion cancellation mechanism works from H2 to H10.
It is more a matter of transpiration than inspiration, H3 cancellation (and H5 if memory serves) was fortuitous for a given topology, i.e. mu follower+source follower+300B, high linearity valves on the mu folower give very low H2 and high H3 and H5, when lower linearity valves are used, the driver stage produces more H3 and H5 which helps to reduce them for the whole amplifier, artisanal tryal and error.
The goal is to obtain an harmonic pattern similar of Békésy experiments, so far I couldn't whatsoever.
A matter of differing interpretation I presume, on what was called a Full Null shown on Figure 3.15 page 182 of Morgan Jones "Valve Amplifiers" 4th Edition, I can only see reduction (blue bar extend lower compared to yellow bar) of H2, H5, H6, H7, not on others. Let's just drop the matter and agree to disagree.
Considering the voltage and current requirement, I assume that you use a vertical mosfet as the source follower on your build, not a lateral mosfet used by RN Marsh, correct?
Again, Morgan Jones give ONE example with ONE topology, if you look at red bars (Part null), there is a reduction on H3, ... , H10 with respect to blue bars (Full null); it is not just a matter of interpretation, but the harmonic distribution you want.
Yes, that is correct.
Indra1, as Popilin didn't answer and you seem to have the book, can you tell me what circuit is this, please?
A standard common cathode + cathode follower or a bootstrapped cathode follower (see post #55)?
Thanks
In such case I can believe that all harmonics are reduced but nothing to do with harmonics cancellation. It's just the CCS that makes them work more linearly respect to normal resistors. This might require some tweaking of the operative conditions but it doesn't mean that cancellation is taking place.
Rest assured that if it were possible it would have ben patented a long time ago. Moreover 2H cancellation only happens using stages of the same type.
Rest assured that if it were possible it would have ben patented a long time ago. Moreover 2H cancellation only happens using stages of the same type.
Let's advise Morgan Jones
Yes. It's a mistake especially if there is no mathematical explanation. Read the old articles by people that knew one or two things....and see if you find something. It just doesn't happen.
The presence of the cathode follower has it's importance because of the load given to the previous stage which is enormous in value, of the order of several Mohms...
Last edited:
Thank you.
A CCS loaded common cathode + CCS loaded cathode follower. Morgan Jones shows that there is a maximum distortion cancellation when the CF is loaded with a certain AC load. He shows spectra of no null, partial null and full null.
The spectra diagram.
Attachments
It's 1-2 mA. At worst output is a bit reduced. Being generous I can't see a problem. This is not nickel.
I noticed it too and asked. They responded that for this type is more difficult to get the same extended FR and so they prefer to use input electronic splitter....
....but they might give it a go in the future.
There is no cancellation! Cancellation is that of a PP or (the less ideal) two cascaded stages of the SAME type.
More in general, how can one cancel at the same time odd and even functions with two stages???
If one changes or tweaks operative conditions can reduce/minimize or increase/maximize distortion.....
And yes, I agree with you. The yellow one has the lowest H3, H4 and H9.
Last edited:
I wouldn't take a whit of notice about what recording engineers think is good sound. To paraphrase Alan Parsons, "We have different priorities to hifi purists."
I got to work on a renowned - even legendary - handmade desk a few years ago and was horrified at how bad the design was. Crappy unregulated power supplies and the channel boards were full of electrolytics and greencap polyesters. Ugh.
Having low H3 is not guarantee for anything. THD has to be lower first. If you decrease some harmonics but overall THD isn't reduced it might not be a good result. The reason why THD should be lower is not just because you can hear it but because of IMD. If it were for THD only we could live with rather high levels of THD (level depend) without being able to detect anything...
P.S.
If you want a good transformer like that have a look at the Hashimoto A-107 or the A105 for 1:0.5+0.5. The apparent low inductance is because of very low input level of just 5V to measure it. So if you use it as an interstage up to higher level it will be somewhat 25-30% higher. If I remember correctly the max primary voltage with windings in series should be about 70V rms down to 30Hz.
The sweet spot for me is as low as possible. Among my current amplifiers the one having that kind of distortion is the PCL86 SE triode with zero fbk. It's a 2W amp and so 0.5% THD @1W with no fbk is quite good, being 1/2 of max power.
It's been easy to get such low value with some 2H harmonic cancellation. The input preamp has the desirable feature that under higher load (i.e. anode resistor of 68K which is about the same as its plate resistance!) produces a spectrum where there is moderate 2H and very little else. The 2H really dominates up to about 10 V RMS output and the over THD is not bad at all. The swing necessary to make the power tube clip is about 8.5V RMS.
Until the onset of clipping, the spectrum at the output looks like a normal one although there already is some 2H cancellation at lower output. This gives rather low distortion at low power up to about 1.2-1.3W. Cancellation becomes evident when approaching max Pout where the 3H grows just above the level of the 2H (and all the other harmonics grow as well). Of course, like everything else, there is price to pay and this is a harder clipping but I can live with it. I only use it as back-up amp on my bench to listen to the radio or, occasionally, with sensitive speakers.
P.S.
To go back to the previous discussion on interstage coupling this is fwhere the RC coupling is convenient, being an economical amp.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.