Re: Re: Re: Re: Contentious stuff indeed!
Short of a complete "off issue" steerage, then why does my latest Windows Xp still suffer hiccups and multitasking seizures ? I've been informed the fault lies in the CPU.
Since I live in a high natural gamma radiation area then I suppose it's an excuse that CPU's glitch. Plausible.
Philips did specially design a 10W EL84 amp using a poor quality UL o/p tranny made in 1959. The name/model long eroded but the text mentioned "can accept an extra 6dB global nfb before instability sets in (that's tight) . Obviously designed for a single full range speaker. Put in other vendors EL84's and it oscillates ! It is well known that UL mode can take off if a-g2 zobels are omitted, but a well designed UL amp should be able to take another 15dB global nfb on top of existing value before instability is detected. This is a good and easy test with a resistance decade box and scope/dummy load.
richj
phn said:
Short of a complete "off issue" steerage, then why does my latest Windows Xp still suffer hiccups and multitasking seizures ? I've been informed the fault lies in the CPU.
Since I live in a high natural gamma radiation area then I suppose it's an excuse that CPU's glitch. Plausible.
Philips did specially design a 10W EL84 amp using a poor quality UL o/p tranny made in 1959. The name/model long eroded but the text mentioned "can accept an extra 6dB global nfb before instability sets in (that's tight) . Obviously designed for a single full range speaker. Put in other vendors EL84's and it oscillates ! It is well known that UL mode can take off if a-g2 zobels are omitted, but a well designed UL amp should be able to take another 15dB global nfb on top of existing value before instability is detected. This is a good and easy test with a resistance decade box and scope/dummy load.
richj
I wouldnt go looking at Intel as the first place to lay blame on a Windows XP issue...
Or have they issued an errata?
Or have they issued an errata?
Re: Re: Re: Re: Contentious stuff indeed!
The arguement that 'some other engineer didn't like it, or use it' just doesn't hold any water with me. It makes no sense, until you get something like 'those engineers didn't use it because ____'.
BTW, McIntosh and their multi-filar outputs ran the finals pentode mode.
cheers,
Douglas
phn said:
The arguement that 'some other engineer didn't like it, or use it' just doesn't hold any water with me. It makes no sense, until you get something like 'those engineers didn't use it because ____'.
BTW, McIntosh and their multi-filar outputs ran the finals pentode mode.
cheers,
Douglas
Re: Re: Re: Re: Re: Contentious stuff indeed!
Now you don't even try to make sense. Now you sound exactly like an UL nuthugger.
By your logic, if advanced devices like a CPU or a space shuttle have glitches or blow up, the people who made them are incompetent. Maybe they should hire UL fans, who have the superhuman ability (or maybe it's just their scopes) to see what no mortal engineer can see.
I don't know what you try to imply with the Philips amp. But I have made it more than clear that when I refer to Altec, Klangfilm and Telefunken I refer to their pro gear.
I don't take your claim that any engineer that hasn't discovered the brilliance of UL is lazy or stupid serious. I don't take you or anything you have written serious.
The only thing I have heard in favour of UL, which I mentioned in my first post (I believe), is that UL is easier. To quote Homer Simpson: "If something's hard to do, then it's not worth doing." But at least I'm not going to tell Ken Shindo to stop doing what he's doing just because I cannot do it. I cannot design a microprocessor or space shuttle either. Yet people do it. What gives?
True, if it was my only argument. The reason I don't take UL serious is that I have yet to hear a valid argument for UL. Sales figures is NOT an argument.
With all else equal, I rather put my faith in the Altec, Klangfilm and Telefunken engineers than in some home audio outfit.
richwalters said:
Now you don't even try to make sense. Now you sound exactly like an UL nuthugger.
By your logic, if advanced devices like a CPU or a space shuttle have glitches or blow up, the people who made them are incompetent. Maybe they should hire UL fans, who have the superhuman ability (or maybe it's just their scopes) to see what no mortal engineer can see.
I don't know what you try to imply with the Philips amp. But I have made it more than clear that when I refer to Altec, Klangfilm and Telefunken I refer to their pro gear.
I don't take your claim that any engineer that hasn't discovered the brilliance of UL is lazy or stupid serious. I don't take you or anything you have written serious.
The only thing I have heard in favour of UL, which I mentioned in my first post (I believe), is that UL is easier. To quote Homer Simpson: "If something's hard to do, then it's not worth doing." But at least I'm not going to tell Ken Shindo to stop doing what he's doing just because I cannot do it. I cannot design a microprocessor or space shuttle either. Yet people do it. What gives?
Bandersnatch said:
True, if it was my only argument. The reason I don't take UL serious is that I have yet to hear a valid argument for UL. Sales figures is NOT an argument.
With all else equal, I rather put my faith in the Altec, Klangfilm and Telefunken engineers than in some home audio outfit.
Well, my liking for UL is linked to the good amps I build in that topology. Nobody else has to like anything about them. So far I have seen no valid reason to abandon UL, just as you claim there is no reason to support it. Seems a fool's errand to prove it is worthwhile to you...and I am not trying.
cheers,
Douglas
cheers,
Douglas
Some historical information:
From: http://web.telia.com/~u43200663/blocks/ulaudioamps.htm
"Patent 496,883 dated 5 June 1937 is one which is considered to be among the most important, which later became known as the "Ultra - Linear" amplifier. With so many of Blumlein’s circuits the design is deceptively simple. "
From:http://www.webace.com.au/~electron/tubes/ul.html
"The ultra-linear amplifier concept is beautifully described by David Hafler and Herbert Keroes in their 1952 US Patent Application 2710312. "
I believe that UL is good.
Pete B.
From: http://web.telia.com/~u43200663/blocks/ulaudioamps.htm
"Patent 496,883 dated 5 June 1937 is one which is considered to be among the most important, which later became known as the "Ultra - Linear" amplifier. With so many of Blumlein’s circuits the design is deceptively simple. "
From:http://www.webace.com.au/~electron/tubes/ul.html
"The ultra-linear amplifier concept is beautifully described by David Hafler and Herbert Keroes in their 1952 US Patent Application 2710312. "
I believe that UL is good.
Pete B.
I believe that despite it is good it was not so good to pay money for it's usage; the patent made other people to search for other ways for the same results. The same may be said of Peter Walker's patent about "Current Damping". The world was happy without this patented topologies so when patents expired nobody rushed to implement them. For example, I'm satisfied by pentodes loaded on twice higher than needed for UL resistance and with lower well regulated G2 voltage. Yes, I am getting less power from particular tubes, but they serve longer than would live in UL giving more power and heat. And I don't need to regulate the whole B+ for similar results, it is enough to regulate G2 voltage only.
I'm tired of having to write everything twice. The title is "Why do some people dislike ultralinear?" Which I stressed in a previous post.
I have not once said UL is bad. I have not once mentioned sonic qualities or taste. But I get upset when people accuse other people of being lazy or stupid for not agreeing with them. That's not an argument. That's a lack of argument.
"All of life is a dispute over taste and tasting."--Friedrich Nietzsche
I have not once said UL is bad. I have not once mentioned sonic qualities or taste. But I get upset when people accuse other people of being lazy or stupid for not agreeing with them. That's not an argument. That's a lack of argument.
"All of life is a dispute over taste and tasting."--Friedrich Nietzsche
Hi phn,
Perhaps you remember me? I also tried a few times to motivate why I think UL is an option. I quoted manufacturer's data and referred to graphs. I tried to sum up from a scientific point, like the following excerpt from my post #172.
Yet you say you have yet to see proof that UL is what some of us says it is. With greatest respect, you come pretty close to calling us liers, and even so I do not take offense. But I do think it is then only fair to ask: He who alleges, must prove...? (And for the record, I am not just naively grabbing tube manufacturers' data, even though, as I said before, I cannot see what they stand to gain by cheating. I myself have checked some of these results with a spectrum analyser in my own designs.)
So perhaps you see my problem with your stance? I do not ask that you like it, but that you at least either acknowledge what is scientifically offered, or disprove it! I really think that is a reasonable position?
[I hear what you say regarding firms' choice; maybe they also do not "see" the advantage; it is after all, not a big one. Maybe economics were involved - who knows! But over 130 000 Quad IIs were sold, many to broadcasting studios ..... so?
Only I fear I do stop short of suspicions of dark scheming every time I don't agree with matters. When we start calling too many things lies, that quickly leads nowhere, because then all must come under the loop.]
Regards!
Perhaps you remember me? I also tried a few times to motivate why I think UL is an option. I quoted manufacturer's data and referred to graphs. I tried to sum up from a scientific point, like the following excerpt from my post #172.
Johan Potgieter said:
Yet you say you have yet to see proof that UL is what some of us says it is. With greatest respect, you come pretty close to calling us liers, and even so I do not take offense. But I do think it is then only fair to ask: He who alleges, must prove...? (And for the record, I am not just naively grabbing tube manufacturers' data, even though, as I said before, I cannot see what they stand to gain by cheating. I myself have checked some of these results with a spectrum analyser in my own designs.)
So perhaps you see my problem with your stance? I do not ask that you like it, but that you at least either acknowledge what is scientifically offered, or disprove it! I really think that is a reasonable position?
[I hear what you say regarding firms' choice; maybe they also do not "see" the advantage; it is after all, not a big one. Maybe economics were involved - who knows! But over 130 000 Quad IIs were sold, many to broadcasting studios ..... so?
Only I fear I do stop short of suspicions of dark scheming every time I don't agree with matters. When we start calling too many things lies, that quickly leads nowhere, because then all must come under the loop.]
Regards!
Phn (following),
Again please realise that we seem to have posted simultaneously. I did not see your #188 until after I posted the previous.
Yes, I also seem to have to repeat things - at least we can feel for each other on that!
And you did call UL "just another marketing stunt". Agreed, not literally saying that it is bad. My point in the previous post was that you don't seem to acknowledge that it has any merit - otherwise, respectfully, you have a strange choice of words. That only; I would agree that that is not entirely how this thread started, but it seems to have evolved to disregarding proof (amongst other things).
Regards.
Again please realise that we seem to have posted simultaneously. I did not see your #188 until after I posted the previous.
Yes, I also seem to have to repeat things - at least we can feel for each other on that!
And you did call UL "just another marketing stunt". Agreed, not literally saying that it is bad. My point in the previous post was that you don't seem to acknowledge that it has any merit - otherwise, respectfully, you have a strange choice of words. That only; I would agree that that is not entirely how this thread started, but it seems to have evolved to disregarding proof (amongst other things).
Regards.
Johan,
phn,
As to why UL was apparently unaccepted in some professional organizations, there are all sorts of possible reasons for this, e.g.:
* unaware of the UL option;
* happy with what they had already had, saw no reason to change;
* too conservative to risk what they might consider a new and insufficiently proven topology;
* not free to try UL, constrained by corporate policy;
* had tried UL and didn't like it.
The manufacturers would probably have wanted both professional and amateur audio designers and builders to believe that their tubes were good for UL, since UL had become fashionable. That is not to accuse anyone of cheating, just an example of what they might have had to gain from it.
phn,
So do I! That happens quite regularly on another audio forum I know, but not so much here.
As to why UL was apparently unaccepted in some professional organizations, there are all sorts of possible reasons for this, e.g.:
* unaware of the UL option;
* happy with what they had already had, saw no reason to change;
* too conservative to risk what they might consider a new and insufficiently proven topology;
* not free to try UL, constrained by corporate policy;
* had tried UL and didn't like it.
Gentlemen, please remember that the merits (or otherwise) of ultra-linear are not of life and death importance. Have a civilised argument by all means - but let's ensure that we are careful to attack only arguments and not personalities.
This is starting to sound like a tube VS SS debate. It is possible to build both good and bad sounding amps using either technology, and both have avid supporters. I can assume that some of the most vocal detractors of UL have never actually built a UL amp. Or they built one and it sounded bad. There are some decent sounding UL amps out there, but they are more picky about the speakers being used than an equivalent triode amp.
I have been a supporter of "the non optimized 'flip da switch' UL mode", and I find that even this has its benefits. I run most of my triode mode amps with a higher than "optimum" load impedance. I use 5K transformers on 6L6, EL34, and 6550 SE amps. This gives lower distortion and slightly better damping factor at the expense of some power. Then when you 'flip da switch' the load is not far from being optimized. Then for most music just use triode mode. When the conditions warrrant it, simply 'flip da switch'. If you don't like it, you don't have to use it, you still have a good triode amp. Cathode feedback (again defeatable) is helpful for difficult speakers. If you are building an amp that uses triode wired pentodes, and your OPT's have the tap, then all you have to loose is a couple of switches.
There are a few people working on variable UL without using a tapped transformer. If this ever works out, then there may be another option.
Try this for a low cost but good sounding design. If you can find some 307A or VT-225 tubes (they are commonly found surplus) wire it up in your favorite 300B circuit, except connect the screen to the UL tap through a 100 ohm resistor. Run it at about 300 volts and 50 mA with a 5K load. I used the TubelabSE design with the cheap Edcor transformers. You get about 5 nice sounding watts. I suspect that you could do the same thing with a 47 tube in a 45 circuit, but I have not tried it yet.
I have been a supporter of "the non optimized 'flip da switch' UL mode", and I find that even this has its benefits. I run most of my triode mode amps with a higher than "optimum" load impedance. I use 5K transformers on 6L6, EL34, and 6550 SE amps. This gives lower distortion and slightly better damping factor at the expense of some power. Then when you 'flip da switch' the load is not far from being optimized. Then for most music just use triode mode. When the conditions warrrant it, simply 'flip da switch'. If you don't like it, you don't have to use it, you still have a good triode amp. Cathode feedback (again defeatable) is helpful for difficult speakers. If you are building an amp that uses triode wired pentodes, and your OPT's have the tap, then all you have to loose is a couple of switches.
There are a few people working on variable UL without using a tapped transformer. If this ever works out, then there may be another option.
Try this for a low cost but good sounding design. If you can find some 307A or VT-225 tubes (they are commonly found surplus) wire it up in your favorite 300B circuit, except connect the screen to the UL tap through a 100 ohm resistor. Run it at about 300 volts and 50 mA with a 5K load. I used the TubelabSE design with the cheap Edcor transformers. You get about 5 nice sounding watts. I suspect that you could do the same thing with a 47 tube in a 45 circuit, but I have not tried it yet.
And two more reasons-
1. The cost and availability of properly-made transformers. That's the key to making UL work properly, but not every transformer supplier was on board for that.
2. Enforceability of then-valid patents. If you didn't get transformers from someone licensed to provide them, you were open to a lawsuit.
tubelab.com said:
This is fine but I find with hifi amps using UL 43% p-p, thd shoots up at hf audio end coupled with slackish power handling (typ of worn out tubes). Most tube amps suffer from degradation at the top end...If anything, it suggests most UL p-p output stages would benefit from slightly lower transformer Z with "todays" made tubes at the expense of higher thd. I suspect actual internal Z cum load curve isn't quite matched for throughput power. Sussing this out isn't easy, as so much depends on the dummy v.s real LS load and what is ideal.
There are various simulated dummy loads about using passive RC's and L's but If anyone can come up with a better model then I'm interested.
The bogey is the output tranny leakage capacitance acts as a cancelling shunt and sucks out vital hf audio end power marring b/w. Solution is to use parallel o/p pairs (lower Z) but ye larger the core to get LF b/w one arrives back to the original problems.....too much winding surface area (sheer size) makes matters worse. I recently designed a 150W amp using 6x 6550c's and to my horror found the optimised square response was way lower than for a medium power standard p-p with 10-50W o/ps. This was a fight with the transformer manufacturer, more sections can make matters worse. So anyone contemplating building an iron hulk, beware large amps are a law to themselves and accept -3dB down at 30Khz.
Perhaps I ought to try 20% UL taps instead of 43%.
Of all the winding configurations, the Williamson 18 sect multisectioned does seem to be manufacturers optimum.
richj
I think you got something there, Rich! That seems to explain what I hear. Without the best iron, and best tap point, it's hard to get great results. Not so difficult in pentode with any iron, if the screen grid is done right. That's the compromise, extra circuitry or getting the right UL iron. I think both ways are good, it's just a matter of doing it right.
And then there's the personal preference factor, based on whatever sound a person likes, and what amp gave that sound.
And then there's the personal preference factor, based on whatever sound a person likes, and what amp gave that sound.
Richwalters,
Exactly so. I did touch on this earlier, but to support you: Several commercial output transformers I have tested had a leakage reactance -3dB at some 100 - 160KHz (and boasted with that), but a total frequency response of -3dB at only 20 - 40KHz as a result of equivalent inter-winding capacitance. Stability with NFB, especially with pentodes, then demanded that a Cdom network be included that in some amps started cutting as low as 3 KHz. There I would say is your prime reason for poor h.f. performance.
I try to get by [especially in big transformers - I am using some 400 KVAs (50 Hz)] with 3 secondary sections, and no more than 4. Sometimes in smaller transformers 2 is enough for a balanced transformer. (And of course the 1/2P-S-P....P-S-1/2P section proportions - some do not seem to know that). Cutting down on winding size (lower C) is another reason why I prefer C-cores. In this case there is a cost penalty.
There is still the basic rule of not cutting below 20 KHz, open loop response (or loop gain).
Regards.
Exactly so. I did touch on this earlier, but to support you: Several commercial output transformers I have tested had a leakage reactance -3dB at some 100 - 160KHz (and boasted with that), but a total frequency response of -3dB at only 20 - 40KHz as a result of equivalent inter-winding capacitance. Stability with NFB, especially with pentodes, then demanded that a Cdom network be included that in some amps started cutting as low as 3 KHz. There I would say is your prime reason for poor h.f. performance.
I try to get by [especially in big transformers - I am using some 400 KVAs (50 Hz)] with 3 secondary sections, and no more than 4. Sometimes in smaller transformers 2 is enough for a balanced transformer. (And of course the 1/2P-S-P....P-S-1/2P section proportions - some do not seem to know that). Cutting down on winding size (lower C) is another reason why I prefer C-cores. In this case there is a cost penalty.
There is still the basic rule of not cutting below 20 KHz, open loop response (or loop gain).
Regards.
Even with a bit of inductance and capacitance thrown in there, can a dummy load simulate the dynamic impedance of a speaker?
I'm not sure if compromise is really the right term any more. In these days of affordable SS devices that can safely be used at high voltages to build simple but effective regulators, it seems the extra circuitry required for pentode is not that much of a burden. In any case, some would argue that even UL and triode operation need a pretty stiff power supply to sound at their best, which is more demanding than just regulating pentode screens.
Ray_Moth,
Again in support: Regulation is also a moot thing. I use choke-input filters for high loads, and then a large output capacitor. Here I rely on the low average-peak relationship of music generally - one thus gets close to a maintained h.t. most of the time (music power, not to be confused with peak power). I take a look at the power supply "constancy" (scope) while playing different kinds of music near max. output. Power-wise, one must keep in mind that poor regulation can at most cause a power difference of some 3 dB; I doubt that will be readily audible. (I am not talking about stability problems resulting from poor regulation.)
I use a concoction of an equivalent loudspeaker load, but as you say, this will vary with every loudspeaker. About all the variations we can do nothing. I am bold enough to say that I have little sympathy with a widely varying impedance, especially those dipping low. Sorry, in these days there are enough computer programs and components to compensate for this. Yet they occur. The poor amplifier designer must bear with matters that ancillary component designers should have looked after.
Regards
Again in support: Regulation is also a moot thing. I use choke-input filters for high loads, and then a large output capacitor. Here I rely on the low average-peak relationship of music generally - one thus gets close to a maintained h.t. most of the time (music power, not to be confused with peak power). I take a look at the power supply "constancy" (scope) while playing different kinds of music near max. output. Power-wise, one must keep in mind that poor regulation can at most cause a power difference of some 3 dB; I doubt that will be readily audible. (I am not talking about stability problems resulting from poor regulation.)
I use a concoction of an equivalent loudspeaker load, but as you say, this will vary with every loudspeaker. About all the variations we can do nothing. I am bold enough to say that I have little sympathy with a widely varying impedance, especially those dipping low. Sorry, in these days there are enough computer programs and components to compensate for this. Yet they occur. The poor amplifier designer must bear with matters that ancillary component designers should have looked after.
Regards
- Status
- Not open for further replies.