Thanks for the example anyway! I think at this point, the "common" requirement still has the Williamson ahead by a mile......
Dave
Dave
Examples of Class A push/pull
- Thread starter Donnie2112
- Start date
Yeah I missed the common bit and just saw push pull class A
I'd suck at naming what is common in the US as well, but I think you guys have a lot more older stuff floating around, very handy indeed
maxhifi
Addicted Member
Both of those claims are false.
PakProtector
AK member
There is no, 'starts AB...somewhere'. Its the max power condition that determines the Class. Both conducting is not the single determining characteristic.
As to the transition, sum the gm and see what happens.
cheers,
Douglas
Dandy
Addicted Member
I don't disagree, but I have seen the first in Eico ads of the time.
I was puzzled when I first saw it, in an ad mentioning their range. However, in the 1963 catalogue, the HF12 and HF 20 are explicitly described as Williamson designs. See here: http://rtellason.com/catalogs/Eico-1963.pdf
Last edited:
maxhifi
Addicted Member
The Williamson amplifier has several main components.
1. Input stage low mu triode direct coupled to phase inverter
2. Split load style phase inverter.
3. Intermediary driver stage between phase inverter and output tubes
4. Class A triode output tubes (triode connected KT66)
5. Very high levels of negative feedback, and a special output transformer to support it.
6. Choke filtered power supply
The EICO has
1. Input stage has a high mu triode direct coupled to phase inverter
2. Split load style phase inverter
3. No intermediary driver stage
4. Class AB beam power output tubes
5. Reasonable level of negative feedback, apparently a good transformer.
6. No choke in power supply
I think that at the time, Williamson was recognizable to hifi enthusiasts as high quality - but really, there isn't a lot of common ground between the two designs past the fact that they're both push pull tube amps.
Dandy
Addicted Member
Yes, there's a good article on the design here, that I read recently. https://dalmura.com.au/static/VinoAnalysis1954.pdf
I suspect there is a problem with the use of the term Williamson. E.g. I recently saw a schematic showing 'Williamson' mods to an ST70 circuit (Eico or Dynaco, I forget). It seemed to be mainly a change resistor values. And, looking at the HF12 and HF20, neither have choke filtered power supplies, nor triode connected KT66s.
I suspect there is a problem with the use of the term Williamson. E.g. I recently saw a schematic showing 'Williamson' mods to an ST70 circuit (Eico or Dynaco, I forget). It seemed to be mainly a change resistor values. And, looking at the HF12 and HF20, neither have choke filtered power supplies, nor triode connected KT66s.
Last edited:
Might be confusing to readers not familiar with the definition. Class AB is a class unto itself; it is not sometimes a class A amp and other times a class B amp. Rather it is its own definition of a class, called AB.
===
It gets confusing because people will say things like "at the point where the output stage enters class B." We all sorta know what is meant, but for accuracy, I think it should be stated as "at the point where only one tube in the push pull pair is conducting."
So in other words, an AB amp operates as an AB amp all the time; sometimes both tubes (in the push pull pair) are conducting, other times only one tube (in the pair) is conducting, depending on output power, biasing conditions, and etc. But regardless, it's still an AB amp, operating as class AB all the time.
So we might reword 6DZ7's previous statement for accuracy:
"Our [class AB push-pull] amp has both tubes conducting up through 50% of max power, theirs up through 70% max power."
Perhaps an even easier way to think about this is to treat the output stage as a black box, where you cannot see inside as to how the stage operates. So you would use some other measure to distinguish your amp from others:
"Our amp puts out 40 watts, theirs 30 watts."
Or
"Our amp produces less heat, theirs more heat."
or whatever measure you use to distinguish your output stage from their output stage.
Granted no one is going to say it just that way, but it helps hit home the notion that class AB is a thing unto itself.
My opinion anyway, in how I interpret the definition.
I totally agree. What I was taught at varsity was that an amplifier is class (X) judged at full output. As for being "class A up to 'y' watt", that is already denoted by the term class AB, the definition of which includes that both tubes are conducting for part (usually the low signal part) of the signal, but not the full signal.
I am also somewhat uncomfortable regarding the 'idealised' view of 'cutting off' at a certain value of bias. Tube Va-Ia characteristics are not linear (straight lines), particularly at low Ia, conducting more or less fully up to a 'cut-off' point and then suddenly becoming zero as per the concept 'cut off' (that almost smacks of class B) Tube contributions in the cross-over region can be made 'complementary' to the extent that cross-over distortion should these days be non-existent - that problem has been dealt with decades ago. (One even has class AB1 and AB2 for further distinction.) Not to explain further here; excellent graphs/examples can be found on the internet under the appropriate tube principles.
Then finally I always frown at the Williamson amplifier being regarded as having 'high NFB'. Heavens, it is 20 dB, quite lower than just about all transistor amps (the latter easily going over the 30 - 40 dB) In fact, I think it might have one of the lower values of NFB for tube amplifiers (the figure for Quad II is lower, but there is nested NFB.). I am unfortunately not familiar with U.S. amplifiers, but both the Leaks and Mullard 520 have some 26 dB.. And the McIntosh has a total of ??
I am not 'waving the flag' here for anything, but misconceptions are easily started; more difficult to correct.
P.S.: Apology, I seem to have momentarily disregarded the two immediately previous posts; mine is partly rather a repeat of those. (Particularly that by Kward, which somehow arrived while I was typing - however that could be possible.)
Last edited:
PakProtector
AK member
Use your own argument. There is no A in an AB amp either...
thanks for noticing there is no B.cheers,
Douglas
PakProtector
AK member
That is saying no such thing; you reason improperly.
Second, I will bet quite heavily that no such worry occurred. Class A meant one thing, AB another, and B a third, each with their grid current or not 2 or 1. And then the Marketing folks stepped in, and folks who don't know or don't care about naming convention carried it onwards.
I think kward did a fine job summing it up, AB is AB no matter what is going on at a particular signal level. The leaned out bit where both are conducting is AB, as is the part where one tube is working into a quarter of the a-a loading. Extremely light bias delivers B, more than that, and less than full 360 conduction at full power is AB, and the rest of the increasing bias is a more and more power limited Class A.
cheers,
Douglas
maxhifi
Addicted Member
Explanation of classes of amplification, from Audio Cyclopedia - Class AB is clearly shown as its own thing. The illustrations here are quite useful. Not useful is the typo in the text
Note that the class is mostly based on where a tube is biased relative to its own individual transfer curve. For this reason, calling an AB amp "Class A for xx watts", is not strictly accurate.
Note that the class is mostly based on where a tube is biased relative to its own individual transfer curve. For this reason, calling an AB amp "Class A for xx watts", is not strictly accurate.
maxhifi
Addicted Member
I'm the guilty party here.. never analyzed it in detail, just assumed so based on its reputation for being fussy about output transformers, and the excellent specs. Come to think about it, that could be partially due to use of triodes - carry on.
Pio1980
AK Member
The second looks like Mullard, simple, more direct, and better inherent stability. It supplanted the Williamson for good reasons, imo.
maxhifi
Addicted Member
Mullard has a pentode input direct coupled to a long tailed pair - it's another animal. I am not sure about which is best, I think there's room for diversity in amplifiers just like anything else.
Pio1980
AK Member
The pentode input is optional to the topology, Marantz used it in the tube amps with a pentode wired as a triode in the front end.
It is the same thing, the Mullard just had more gain in the front end to accomodate the 0.5V British line level standard.
maxhifi
Addicted Member
Mullard always uses a long tail pair, versus the split load type used by others.Speaking of Marantz, Isn't that Marantz 8 a beautiful design to analyze? So much attention to detail, it's quite a design.
maxhifi
Addicted Member
That's actually a misunderstanding - the original Williamson used triode connected tubes. The tapped screen ultralinear operation was a later improvement, and was applied retroactively to the Williamson design.
