Drawbacks of Parafeed?

[dsavitsk]

As an initial experiment for building a new linestage, I just breadboarded a parafeed linestage (6N6p, Jupiter 4.7uf coupling caps, solid state CCS, magnequest iron) and it sounds terrific. Really terrific. Surprisingly so.

As I understand it, the biggest benefit of parafeed over a traditional single feed transformer coupled amp is that it gets the big electrolytic cap out of the signal path, and indeed, the clarity of the amp I built is very good. But, not everyone is building parafeed amps, and in fact in the scheme if things they seem to be in the small minority of tube amps being made. So it seems there must be a drawback. Can anyone tell me what it is and what other topologies might do better?

[dhaen]

As a line stage, I see few drawbacks, and getting the PS caps out of the audio path has got to be good news.
Using a CCS as a load will necesitate a higher supply voltage for a given signal swing, but this is just a design consideration, as is the transformer's ratio.

[chrisb]

Well perhaps it depends on the application.

The very popular and excellent bang for buck power amp designs by Doc Bottlehead and Paul Joppa use SE parallel feed topology for the express purpose of elimating the B+ current from the OT. The DC plate supply is decoupled from the OT via a larger value cap (3.3mf or so), which definitely is in the signal path, so quality is very important.

Indeed some of the extreme tweakoholic upgrades to these amps often include exotica caps and output iron that can cost more than the orginal complete kit.

[analog_sa]

The way i see it parafeed allows the use of exotic cores and generally better transformers, which is great. What is not so great is the presence of a coupling cap and a resonant circuit. While it may do miracles for bass and prat, it sounds quite a bit artificial.

[Wavebourn]

The single minus of a parafeed is a power consumption by a CCS load, no more minuses, pluses only. Better sound bought for wasdte of power, everyoner can decide what is better for gim/her.
One capacitor is no way worse than a saturated core. And if CCS is very good no ripples come from a power source, and no voltage regulation is needed since a load current is highly constant.

Why everyone don't use CCS loaded triodes?

Because beliefs and fashions rule the world. And because it is easier just to solder a transformer with a magnetic gap to anode of a tube.

[dsavitsk]

Originally posted by chrisb
The DC plate supply is decoupled from the OT via a larger value cap (3.3mf or so), which definitely is in the signal path, so quality is very important.

Right, I understand that, but the point is that it is a 3.3uF (4.7 in my case) high quality cap instead of the 100-470uF electrolytic that usually bypasses the cathode bias resistor.

Originally posted by analog_sa
While it may do miracles for bass and prat, it sounds quite a bit artificial.

That's interesting. I'll just have to listen more to see what I think. So far I am enjoying it very much and am planning to next try it out with some 01A's I have around.

[tubelab.com]

Why everyone don't use CCS loaded triodes?

I am a big proponent of CCS loaded triodes. You will find one in all of my recent designs. However some tubes (the 12B4 comes to mind) just don't like a CCS load.

[bear]

Originally posted by Wavebourn
The single minus of a parafeed is a power consumption by a CCS load, no more minuses, pluses only. Better sound bought for wasdte of power, everyoner can decide what is better for gim/her.
One capacitor is no way worse than a saturated core. And if CCS is very good no ripples come from a power source, and no voltage regulation is needed since a load current is highly constant.

Why everyone don't use CCS loaded triodes?

Because beliefs and fashions rule the world. And because it is easier just to solder a transformer with a magnetic gap to anode of a tube.

My suspicion is that parafeed is liked because it makes it possible to use P-P iron in a single ended application. That's one reason.

Other reasons include: it sounds ******* fantastic on things like power amps driving horn tweeters; it is extremely minimalist; people (as you imply) can experiment all day long with caps and chokes and output iron without much theoretical background; and some people want "gorgeous beautific sound" and this seems to deliver it simply and easily.

There are all sorts of CCS approaches out there for toobes - whole bunches of sites with people doing that. I think, iirc, TubeCad has a whole series of articles analyzing different topologies using CCS...

The other things that stands out in my mind is that "everyone" doesn't do "anything the same" in audio - everyone seems to have their own mix and preferences...



_-_-bear

[jlsem]

Because beliefs and fashions rule the world. And because it is easier just to solder a transformer with a magnetic gap to anode of a tube.

Or maybe nobody else is as smart as you.

The only other problem I can see with a shunt-feed amp is that with no DC present in the magnetic core of the OPT, there exists the problem of the magnetic field collapsing as the AC signal crosses zero and the resultant nonlinearities which are easily heard. However, in the end, it's just a matter of taste and some people prefer one sound over another. I just happen to like the sound of a single-ended triode with a very large and unfortunately expensive gapped OPT. And it's not because I am lazy or fashion-minded.

John

[jlsem]

Right, I understand that, but the point is that it is a 3.3uF (4.7 in my case) high quality cap instead of the 100-470uF electrolytic that usually bypasses the cathode bias resistor.

The AC signal no longer passes through the last cap in the power supply.