Re: Bias-wise
- From: "Dr Gang" <zorglub@xxxxxxxxxxxxx>
- Date: Sat, 14 Jan 2006 16:43:31 +0100
Your whole post could be summed up by the pic you gave the link to : biasing
an amp with a white shirt and tie as if it mattered as much as project
Manhattan, for the sake of technology. I like that too, I think people like
that are necessary because they're the one who are passioned enough to
create, invent, innovate. I don't have that kind of passion in me. I don't
have the skills for it.
I've tried to read your post with attention and, well I'm pretty sure
there's always more knowledge, more understanding, more in-depth designing
of tube amps levels that can be reached. I've witnessed numerous technical
"wars" about what is better, what is worse. I've also seen many electronic
technicians telling me that any component had a tolerance, a slight
variation between its rated characteristics and the actual ones. I
personally think that first because of this tolerace thing, going too deep
into technical matter is useless, there's a point where the precision of
your design will be limited by the components themselves. Secondly, I didn't
have, when I first considered this project, the ambition of making the
ultimate tube power amp. To be honest, I had no idea what type of sound I
wanted, no idea how would plate voltage, caps values and stuff like that
influence the final sound. People talk about the Marshall sound, I don't
think that I've the experience to distinguish it from my lawn mower !
Some think that only result matters (how the amp sounds), some build for the
sake of making the technically perfectest amp ever. I can't even choose my
category because I have no ear and certainly not enough knowledge.
This debate goes past over my head. I'm trying to find answers technical
enough to produce a decent result yet not so technical that a sunday amp
tech like me couldn't fathom and apply.
Anyway, no matter where I set the bias pot, my tubes don't seem to melt so i
guess i'll leave it a little bit after the max volume and that will be it
because my multimeter measured no current at all when I tried a "cleaner"
method...
Regards.
"Rich Koerner" <richk@xxxxxxxxxxxxx> a écrit dans le message de news:
43C8385E.4A26249A@xxxxxxxxxxxxxxxx
>
>
> Ant Moore wrote:
>>
>> Hi there,
>>
>> "Dr Gang" <adra.aganga@xxxxxxxxx> wrote in message
>> news:43c63051$0$18558$626a14ce@xxxxxxxxxxxxxxx
>> > Hi,
>> >
>> > As I've posted a few days ago, I've finished my tube power amp project
>> > and
>> > had only the bias to set.
>> > Since the bias is some sort of setting of middle point for the tubes,
>> > considering those have a maximum voltage swing, if the bias is too low,
>> > there'll be a discontinuity between the positive and negative parts of
>> > the
>> > waves -> low volume and distortion. If the bias is too high, each tube
>> > will work not only on its half of the wave but also on the most part of
>> > the other tube's wave -> lowering the global power yield and constant
>> > dissipation cutting on the tubes' life expectancy.
>> > If this is correct, then you probably can set the bias only by ear :
>> > will
>> > turning the bias pot, you should start by having no sound at all, then
>> > a
>> > bad quality low sound will come out, increase in volume and quality.
>> > You
>> > should reach, at some point, a maximum volume when the middle point is
>> > set
>> > for exactly each half the wave to be amplified by each tube. Going
>> > further
>> > will bring you into the AB class domain : the middle point goes up, the
>> > voltage swing capability lowers -> volume lowering and each tube will
>> > work
>> > increasingly on the other tube's half of wave. Going even higher will
>> > make
>> > your amp a pure A class with 2 tubes amplifying exactly the same wave,
>> > with half the power they could theorically develop.
>> >
>> > If this is correct, wouldn't it be enough, to set the bias, to simply
>> > turn
>> > the pot until you reach the maximum volume and then go a notch higher ?
>> >
>> > Yeah, i'm lazy !
>>
>> When biasing your output stage, there's two limits you need to
>> consider - too cold and too hot. Anywhere in between is OK.
>>
>> If you bias too cold - too close to class B - your amp will have
>> crossover
>> distortion. Sounds bad (unless you like that sort of thing.)
>>
>> The upper limit is that above which your output tubes' max dissipation
>> will be exceeded *during operation*. Note that this is NOT the same as
>> no signal bias: depending on your plate loading and voltages, this may
>> increase Tthis is where it gets more difficult. The average
>> dissipation
>> will decrease with increasing signal if you've designed your plate load
>> for
>> a class A bias point. But if not, your average dissipation will likely
>> increase from the static point with input signal, reaching a maximum
>> at some point before decreasing again.
>>
>> There are several excellent technical papers by Randall Aiken on his
>> web-site that discuss this, amongst other things. If you look at
>> www.aikenamps.com/, in his advanced technical articles, the
>> "Idle Current Biasing - why 70%?" paper is one you should take
>> a look at. Also "Cross-over notch biasing - why it should be avoided"
>> is worth a read too; your "do it by ear" method is kind of the same
>> thing!
>>
>> (FWIW, I tend to bias my AB amps slightly cooler than the 70%
>> that Mr. Aiken suggests. They sound fine to me at around 60%.)
>>
>> Cheers, and regards,
>>
>> Ant. (Tamworth, UK)
>>
>> --
>> "Make me one with everything." The Dalai Lama at a hot dog stand.
>
> I had recently had posted some things in another thread, that has great
> relevance here.
>
> Let me paste some of that here, to establish some ground work, and address
> some issues and concerns
> of mine for a long time here in AGA.
>
> Please read closely the following, as the answers have been with us all
> the time. Yet, failed to
> see the obvious. The reasons why, are found in the following.
>
>
>
> ========================================================
>
>>
>> Stewart Ward wrote:
>> >
>> > Hi Rich,
>> >
>> > I am a foreigner and I may not understand your phraseology and tone too
>> > well. So apologies in advance, if I've got it wrong. I *never* set
>> > out to upset anyone.
>>
>> LOL, no need. There are times when I don't understand myself either.
>>
>
>> > >> Remember, every aspect of how an amplifier produces the end result
>> > >> sound is
>> > >> decided by the *designer* - a person. A hitherto, overlooked
>> > >> factor.
>> > >
>> > > Even when the case in point is operated with user error, component
>> > > drift, a vacuum
>> > > tube of another characteristic, etc,....
>> >
>> > Within the operating tolerances of all the components employed. I
>> > would not
>> > expect to have to say that, for fear of making posts seem like legal
>> > documents.
>
>
>
>> > >> Every product in the world is designed by a person who decides how
>> > >> technology is
>> > >> used to make it work the way it does.
>
>
>
>> > Not at all. The point is related to the previous comments. The
>> > designer is able to
>> > use the technology in a number of ways, by tinkering with its various
>> > attributes, to
>> > produce an end result - whilst keeping within safe operating limits.
>>
>> Tinkering is the operative word. We tinker with the pre-existing, till
>> we get it right for the
>> musicians taste, and now we are going to call this engineering.
>>
>> That's the problem with the amp companies today. They are staffed with
>> tinkerers, and the likes of
>> an Ed Jahns is not likely to be seen working on staff at a company as
>> large as Fender today. There
>> is a major difference of the pedigree between then, and now.
>
>
>> But, there is one basic fact you have to NEVER lose sight of.
>> DISTORTION, was something that was
>> never wanted. These amps were all designed with CLEAN undistorted sound
>> in mind.
>>
>> What happened when they were over driven, was NEVER a design
>> consideration, for the flavor of
>> distortion they are known for today.
>>
>> Just listen to the music they produced at the time they hit the
>> marketplace. That is what they were
>> designed to produce. The clean sounds.
>> >
>> > Nah, not back then. They were just churning out cheap amps that would
>> > go as loud and
>> > clean as possible for the lowest cost.
>> >
>> > Bias control as a tone control? Only in recent years. Talk about
>> > biasing valves
>> > even ten years ago to a musician (here in the UK) and you'd watch their
>> > eyes glaze
>> > over and roll backwards. The idea would go right over their heads.
>
>
>
>> With the selling of polished turds for tubes, that were to *replace* the
>> likes of the Sylvania, GE,
>> or RCA 6L6GC, which these amp *designs* were *originally* based upon and
>> would NOT bias up in the
>> same place as these original tubes,...... real KNOWLEDGEABLE and
>> EXPERIENCED old time amp techs,
>> such as myself, would OBJECT to these OFF spec tubes, would RESIST using
>> them, and passing them off
>> on their customers as EXACT REPLACEMENTS for the ORIGINALS.
>>
>> So, in order to sell them, they had to be SOLD to the amp owners with a
>> shit pile of snake oil, and
>> a biasing method to by pass the amp tech's who'd protest the use of these
>> tubes.
>>
>> The most out spoken of these polished turd tube salesman was Lord Valve
>> in this newsgroup. His own
>> word gave testament to this fact. How many times did you hear the words
>> that this 6L6XYZ sounds
>> cold and has to be biased this way in this amp, or this 6L6ZXY has great
>> tone in a Fender, etc, etc,
>> etc,.....
>>
>> You had to do none of that before the existance of these polished turds.
>
>
>> Today's generation of self schooled amp techs would be fooled by this,
>> then THEY would be lost with
>> conclusion in error. Not likely, with the old school formally trained in
>> proper electronics
>> industry practice like myself.
>>
>> You can't change my past, and life experience. As long as I exist, I'll
>> be here to REMIND you all
>> of the past, most WISH you'd forget, and like LV and others, try to
>> re-write as something else.
>
> =============================================================================
>
> Here, let me put the focus where it belongs.
>
> READ THIS AGAIN, and the we can talk about the issues in biasing output
> tubes today.
>
>
>
>> With the selling of polished turds for tubes, that were to *replace* the
>> likes of the Sylvania, GE,
>> or RCA 6L6GC, which these amp *designs* were *originally* based upon and
>> would NOT bias up in the
>> same place as these original tubes,...... real KNOWLEDGEABLE and
>> EXPERIENCED old time amp techs,
>> such as myself, would OBJECT to these OFF spec tubes, would RESIST using
>> them, and passing them off
>> on their customers as EXACT REPLACEMENTS for the ORIGINALS.
>>
>> So, in order to sell them, they had to be SOLD to the amp owners with a
>> shit pile of snake oil, and
>> a biasing method to by pass the amp tech's who'd protest the use of these
>> tubes.
>
>
> <Rich comments>
>
> Now let's talk about this a bit, by example.
>
> When an old schooled electronic engineer sets about the design of an
> amplifier, there are design
> requirements to meet. Power output and selection of the tubes to do the
> job is the foremost.
>
> It was standard to approach the tube companies directly for data and
> samples, along with any updates
> and suggestions for the tube companies. Could you think of any reason why
> this would not be common
> practice.
>
> So, let us say that this amplifier is to produce 100 watts R.M.S. at the
> load. Let's say that the
> selection for the output tube was for headroom, reliability, and long
> service life of the
> amplifier. These were common design considerations back THEN. It is not
> a NOW consideration.
>
> Let's say based on this, a tube is selected, along with a power supply
> that would yield 500 volts on
> the plates of the output tubes when the supply is loaded in the static
> operation condition of the
> amplifier. Which yields the 100 watt R.M.S. goal, with headroom.
>
> So now, this amplifier is still with us today, now it sits in front of
> you, and now you have a bias
> issue that confounds you.
>
>
> But then, you were told, this was not rocket science many times here in
> AGA. Well, if it's not
> rocket science, consider the following.
>
> First, look at this picture.
>
> http://timeelect.com/ed-jahns.jpg
>
> Look at the gear in this picture. If this is not rocket science, why is
> all that state of the art
> test gear sitting on that bench in the first place. That room is filled
> with it and you don't see
> all of it in the picture. They are not just props for the picture.
>
> You are looking at Ed Jahns, the man behind the development of the Tung
> Sol 6550 coke bottle tube
> that gave the great sounds of the first generation of a new class of high
> power audio amplifier.
> Yes,... that is a Fender 400-PS chassis on the bench too. I knew Ed quite
> well, and learned a lot
> from him.
>
> This is a little picture of that old school generation I come from and the
> pedigree of individual
> from the back THEN time period, that gave us that amplifier in our
> example, that sits in front of
> you.
>
> So, here you sit. You sit with you handy digital multi meter, and a few
> kathode resistors to put
> into the amp.
> Why, the need to install them to set the bias. There are none in there
> now. Why, the need to add
> them now in order to set the bias. That thought ever crosses your mind.
> Maybe, there is something
> missing here.
>
> OK, maybe you have your amp bias rig with the meters and tube socket
> adapters to set the bias. You
> are confused by the numbers to adjust for, and wonder which numbers are to
> ones to use. Or, now you
> wonder if there isn't a better method to set the bias.
>
> So now, you look to the internet for some answers, and you find yourself
> here in AGA. Which is the
> playground for the polished turd promotion, amp tinkers, with continuous
> beauty pageants of polished
> turds, with the AGA resident judges holding up their score cards. This,
> is what it has come down
> to.
>
> So, here is Rich K, the only one calling these tubes polished turds,
> singing away the praises of the
> old days of NOS tubes,... all in the face of the flood of polished turds.
> LOL,... so who the hell
> are you going to believe. When, you are looking for the simple answers,
> and the loud voice in here
> is yelling, "It ain't rocket science. Here, use these resistors and set
> it up by the numbers".
>
> I think, once the dust settles, some discover there is more to it, and now
> search for more
> meaningful information on the internet. Which, leads to Randall's
> technical papers on the subject.
>
> In Randall's paper he gives two amplifier examples, as a reason not to use
> the cross over notch
> method. Which is technically correct for the time period of today. But
> leaves out things not in
> play, that we were aware of for the time period of the old schooled.
>
> In the case of the higher voltages amplifier, the end result stated would
> have not gone un-noticed.
>
> There would have been load current monitored along with the line current.
> Which would have given
> more information to what is happening. Just having a scope on the load,
> and a meter on the grid
> bias is to be blind and working in the dark. Knowing the line and load
> current ratios during
> dynamic operation of an amplifier is something not considered common
> practice today. I'm beginning
> to think, I maybe the only one here who places value in it, and uses it as
> everyday practice.
>
> There is no ONE method, but a blend of those that exist. It dependent on
> the knowledge of the
> dynamic operation of the amplifier as a whole, which to some, is rocket
> science. It take a firm
> understanding of dynamic operation of the amplifier, the test gear to
> confirm it, that the tinkered
> just doesn't have.
>
>
> Once normal static levels of line current is established for a given
> amplifier, line current alone
> can be used to set bias too. Yet, it is not a method for those with a
> tinker's understanding, even
> if the test gear were available. BTW, the 400 is biased by line current
> observation.
>
> One reason that Randall's paper on the notch method gave yield to the bias
> error found with the
> higher voltage amp example, is because of the polished turd used in the
> test example, did NOT
> provide the same spec as the ORIGIN test sample tubes, the designer had
> used in the original design
> of the amplifier with it's headroom.
>
> With respect to Randall's conclusion, I would have arrived at possibility
> the opposite conclusion.
> I would have concluded that the amplifier had been improperly designed
> outside of the tubes maximum
> plate specs for that yield.
>
> Another possible reason, is the line voltage may have been different for
> the original design. At
> the time many of these old school amps were designed in the past, a
> different line voltage than we
> have today, may have existed.
>
> In our example of that 100 watt R.M.S. amplifier design, was the line
> voltage at the time 110 vac,
> 115, vac, 120 vac, or higher that came from the wall socket. If the amp
> was originally designed to
> run at 110 line, and you performed this test at 125 volt line, those tubes
> would be run way over
> maximum of spec for the plate voltage found.
>
> When I sat with Aaron Newman in his design lab at Guild, he would run the
> Quild Quantum Bass amp
> proto types at 125 vac line for days on end to insure that there would be
> no high line voltage
> problems. Then he slowly lowered it to 95 volts vac to establish the line
> voltage operating range
> of the amplifier. Range of line voltage operation was in the mind of the
> designers of my day, as
> all the quality test gear found in design labs gave that spec.
>
> Which, brings me back to one thing I had said here in AGA years ago on the
> issue of biasing up an
> output stage.
>
> There is no ONE method, but a blend of those that exist. It's dependent
> on the knowledge of the
> dynamic operation of the amplifier as a whole. Which, to some, is rocket
> science. It takes a firm
> understanding of the dynamic operation of the amplifier, the test gear to
> confirm it, which the
> tinkerer amp tech just doesn't have.
>
>
>
> Regards,
>
> Rich Koerner,
> Time Electronics.
> http://www.timeelect.com
>
> Specialists in Live Sound FOH Engineering,
> Music & Studio Production,
> Vintage Instruments, and Tube Amplifiers
.
- Follow-Ups:
- Re: Bias-wise
- From: jh
- Re: Bias-wise
- References:
- Bias-wise
- From: Dr Gang
- Re: Bias-wise
- From: Ant Moore
- Re: Bias-wise
- From: Rich Koerner
- Bias-wise
- Prev by Date: Re: Help with device in Peterson Strobe
- Next by Date: Re: Time for liberal Bruce Morgen and others to put up or shut up
- Previous by thread: Re: Bias-wise
- Next by thread: Re: Bias-wise
- Index(es):
Relevant Pages
|
