Re: SE fixed bias VibroChamp works; toob math question


"flipper" <flipper@xxxxxxxx> wrote in message
news:kalp63taq3m18q9it1a2i5e9iat4f9pq59@xxxxxxxxxx
On Sun, 10 Jun 2007 21:47:08 -0400, "Phil S."
<psymonds_no_spam@xxxxxxxxxxx> wrote:


"Phil S." <psymonds_no_spam@xxxxxxxxxxx> wrote in message
news:BqKdnZJSYONlHPHbnZ2dnUVZ_sSmnZ2d@xxxxxxxxxxxxxx

"Phil S." <psymonds_no_spam@xxxxxxxxxxx> wrote in message
news:SaudnYRaLqKRHPHbnZ2dnUVZ_h-vnZ2d@xxxxxxxxxxxxxx
Ok, I fixed the bias supply and the amp seem to be working on the
bench.
I haven't plugged in the guitar yet for a taste test. Not redplating
<grin>. Most voltage readings appear reasonable, but screen is still
above plate by about 2V.

I'm feel iffy on the toob math for a change. Here is the data from the
one random GE 6AQ5 I plugged in. There are two readings, bias pot
moved
from one extreme to the other:

G1: -13.25; -14.10
A: 266; 269
Across Rk, Ia : 41.27mA; 38.33mA

This tells us Pout:
10.98W; 10.31W
Is that right? Couldn't be right.

Duncan Munro's TDSL says for Class A, SE:
G1: -12.5
A: 250
Ia: 45
I calc this to be 11.25W, but the TDSL says Pout = 4.5W

I'm not getting this.

Also, bias voltage range seems to be wrong. Should I change the bias
pot
to give a wider range and more R, like from 50z to 100z?

Help, please!

Thanks.
Phil


I said 50z and 100z pots above. I meant 50K and 100K. There is a 68K R
in series with a 50K pot.

Well, OK, I managed to settle down and answer a few questions on my own.
I
pulled up the GE spec sheet which is loaded with decent information. For
starters the toob math is correct. Pout is different from max plate
dissapation,

Bingo. Right, they aren't the same thing.

which for this tube is rated at 12W. Therefore 70% of 12 =
8.4W. I suppose, in SE operation, we might see it go a bit higher, but
the
10+W I'm seeing is probably too high.

You're mixing apples and oranges, PP and SE.

With PP you typically operate Class AB because, since you have two
tubes, one can be conducting, and providing output, while the other is
OFF.

Ok, so what's the point? Well, as you've noted, there is a maximum
plate dissipation so, if you have the tube OFF for some period of time
you can have it 'more on' during the other and still have the plate
power average to a safe value. The reason for doing so is you can get
more power out than if you operated Class A, with both tubes always
conducting.

What that means, however, is that maximum plate power doesn't occur at
idle, it occurs at some power output level above idle. That's where
the 70% rule comes from. It's to leave room for plate power to hit
max. without exceeding the rating.

SE is different in that you cannot let the tube go 'off', the signal
clips at that point because there's nothing else to provide it. And
since the tube is always conducting, plate power remains close to
constant (slight increase at max power from distortion effects).

So your SE idle power is set at near max, say 5% under to allow for
some increase due to distortion.

Check your 6AQ5 datasheet and you'll see the Class A amp at 250V is
set to 45mA idle for 11.25 Watts. And it says at max power output the
max plate current hits 47mA, for 11.75 watts.

Your 10 watts is actually low and it'll cut your power output because
the max available current swing is less.

Ok, to the 'output power' vs plate power confusion. Take the SE case
now that we've talked about idle current. As noted a sine wave swings
equally above and below 'idle'. Idle is the 'mid point' so the current
swings from 0 to twice idle. Working toward an RMS number your sine
peak is 45mA because it *has* to idle at the mid point to be always
conducting.

Now convert that to RMS current by dividing by 1.414 and you get
0.0318mA. Simple ohms law and power coming up. We have a 5k OPT in the
datasheet so current squared times impedance is power, and that gives
you 5 watts. In real life things aren't perfect though so you get a
bit less, 4.5 watts, but it's close enough to show the principle.

The absolute *best* you can ever get from SE Class A, in theory and if
everything were perfect, is 50% the plate dissipation because idle
*must* be half the maximum output swing in order to stay conducting.
In practice you can never get there because, for one, a tube will not
pull all the way to 0 volts so there is always some 'offset' to the
voltage swing and that offset is in the positive direction, which
increases average plate dissipation, robbing it from what's available
to produce output power. So figure 40% as a more practical upper limit
and, using that, you come up with 4.8 watts from a 12 watt plate,
close to the 4.5 Watt figure in their example.

You'll end up with less than that to the speaker because there's OPT
losses too. Like from the DC winding resistance, coupling, and core
losses they don't show in their example because they don't know how
good an OPT you'll be using.


I will need to either change the 68K
to something higher (or is it lower?) or put in the 100K pot to get the
bias voltage more negative and achieve lower plate dissapation. I looks
to
me like I don't have enough range in the bias supply, but maybe that's not
the case? BTW, it will be much easier to change the R than the pot due to
the layout.

You actually need to go the *other* way because your max idle limit is
41mA and best would be 45mA, assuming you've got 250V B+.

Btw, an interesting 'problem' is that having a bit 'more' B+, than the
spec doesn't help but, rather, hurts because it forces a lower idle
current to keep within max plate watts. And a lower idle current means
less current swing to put out power with.


The bias supply is a 12.6v filament winding. Will I be able to squeeze
out
something on the order of -15v or 16v?

Don't need to. You've already got too much negative bias on it.

The bias supply ladder (using JTM45
as a model) consists of:

12.6vac > 1N4007 > 10uf > 15K > 10uf > 200K grid leak > power tube grid
|
68K + 50K pot > ground

That didn't line up quite right on my end but if I guess correctly you
need to lower the 68k a bit to get the bias range closer to ground.
Less negative.


Really, now I'm at the spot where I could benefit from the advice of
someone
with more experience than me. Otherwise, I've got to go fool with this
and
I'm not sure where I end up with it. It would be much more efficient to
skip the trial and error part.


Thanks.
Phil


Flipper,
This is a very helpful post and I really appreciate it.

The "1 ohm" resistors on the cathode is not exactly 1 ohm. I'm not saying
where I purchased it, but it was advertised as being made to very close
tolerance, or else my cheap meter isn't worth a damn at this range. I
realized this as soon as I stared getting reading in the 60mA range. I
failed to test it prior to putting it in circuit. I then checked the other
1 ohm R's from that batch and found they varied a bit, so I was using 1.5
ohms to correct for the error. When I got inside the amp, I found the
culprit is actually 1.3 ohms, making this whole exercise a PITA, and for
which I quickly made a conversion chart. Anyhow, this changes all of the
math and equipped with the lesson you so kindly provided, I changed the 68K
to 100K. With the pot maxed, bias voltage is -14.99, plage voltage is 271,
current is 41.85mA and static plate dissapation is 11.34 watts. I notice
this will decline a little after the amp has been on for a while. (This is
with one random sample tube.) I will now consider adding a 15K to the 100K
as I think this will tweak it into the 11.0-11.5 watt range, but before I do
that, I will cycle several tubes through the power hole to get a better idea
of what the need really is.

Regards,
Phil


.

Relevant Pages

  • Re: SEA 112 radio advice / manual / schematic needed
    ... power levels, ... Variac to control plate voltage. ... SSB don't understand that they can run a LOT more power than 150 watts, ... Operator's License for over 350 watts. ...
    (rec.boats.electronics)
  • Re: Rebuttal to Richard Clarks comments on my Chapter 19A
    ... RF power window at the input of the pi-network, resistance RLP at the ... plate, and the slope of the load line. ... the amplifier in this condition will later be shown to be 50 ohms. ...
    (rec.radio.amateur.antenna)
  • Rebuttal to Richard Clarks comments on my Chapter 19A
    ... RF power window at the input of the pi-network, ... plate, and the slope of the load line. ... the amplifier in this condition will later be shown to be 50 ohms. ...
    (rec.radio.amateur.antenna)
  • Re: More on PEP, AM, average power, etc.
    ... power output in an AM transmitter), I've been doing more scrutinizing of tube Ip/Vp characteristic curves. ... 1 kV on the plate gives 175 ma plate current ... a sharp cutoff tetrode that only at zero grid volts was there a near linear relationship between plate current and plate voltage (meaning zero current at zero voltage, and a straight line with some slope. ...
    (rec.radio.amateur.boatanchors)
  • Re: 833 Amp
    ... >>> quiescent plate dissipation vs. filament wattage? ... Perhaps it can, but to get that 150 watts of output power, ...
    (rec.audio.tubes)