| Author |
Message |
flipper
Guest
|
 Posted:
Fri Sep 16, 2005 5:34 pm Post subject:
Re: 12au7 as SE output |
|
|
On Fri, 16 Sep 2005 07:22:34 -0400, John Stewart
<jh.stewart@sympatico.ca> wrote:
| Quote: | flipper wrote:
On Thu, 15 Sep 2005 20:30:10 -0400, John Stewart
jh.stewart@sympatico.ca> wrote:
flipper wrote:
I don't follow. Why would it rectify in one case but not the other?
Not sure what you are referring to as the other case.
The other case is, of course, direct coupling. The output impedance of
the CF is the same so I'm asking if grid rectification is going to
'clip' the RC couple then why doesn't it 'clip' the same signal
through the same impedance when DC coupled?
Rectification occurs since when the
diode formed by the grid-cathode combo conducts the resultant negative charge builds on the
6V6 G1 side of the coupling cap. Since the RC time constant needs to be relatively long to
accommodate the lowest frequencies you intend the amp to pass there is insufficient time
for the cap to recover. This property is used to bias some kinds of Class C transmitter
(TX) circuits. You will also find it in TV flyback circuits.
Yes. I understand rectification but there's an input impedance to the
grid too. Low as hell, by comparison, when driven positive but still
some impedance. Looks to be about 5200 for the 6V6 according to the
spice simulation.
With direct coupling there is no cap to take on charge. Another way to run Class A2 is with
an Interstage Transformer (IT).
Yes. I see the bias shift but it seems to settle at an acceptable
point.
Not that I entirely trust spice as I've seen it magically create
negative voltages where they couldn't happen in real life..
Or, more to the point, if I have enough drive why isn't it ok?
As a triode where the screen is connected to the plate the 6V6 has rather low mu. In that
mode it is not a good candidate for SE Class A2 operation. That usually requires much
higher mu, probably 50 or more. No problem at all in PP.
Why is it so different for SE vs PP? Doesn't the grid rectify in both
topologies?
From what I read it was an even bigger problem in PP as the bias drift
pulled them away from crossover increasing crossover distortion.
I also don't see how DC coupling will work as there'd be no voltage
swing below ground. Well, not without another supply anyway.
Yes, you would need a -ve supply. Unless you were using one of those high mu triodes are
referred to in an earlier post. Did you look them up? The 6AC5 triode has a mu of 125!
Problem is, I don't have a high mu power triode and it's rather
problematic to run them pentode ;)
The point was to have the amp switchable between the two modes.
For example, my model is showing at 2.3 watts and a self bias of 11.3V
the grid is going 3.7V positive
3.7 volts +ve with respect to what? Common or the 6V6 cathode?
The cathode. That's what the grid going positive means, doesn't it? Or
am I not ruing the term correctly?
Best to try it in the real world. Shut off the computer & get the soldering iron warmed
up! Cheers, John Stewart
|
I will but I'm still waiting on some parts.
|
|
| Back to top |
|
 |
Jim Anable
Guest
|
| Posted:
Sat Sep 17, 2005 4:45 am Post subject:
Re: 12au7 as SE output |
|
|
Chedomir Jokanovich wrote:
| Quote: | I am going to use a litlle transformer with selectable primary
impedance (8k to 22k into 8Ohm)
The point of experiment is to use it as an interstage in a guitar amp,
when switched in, it would act as overloaded output stage, possibly
into reactive dummy load, and voltage from the secondary would be fed
to the normal phase inverter of a bigger amp (normally attenuated a
little). That way I am hoping to get that "crunchy" tube overloaded
sound at any volume level.
Also I would make an speaker output from that stage so I can use it
for practise at home, I recon half a watt (even more when overdriven)
into a pair of celestions with 97dB sensitivity will make nice practice
volume
thanks for your help
|
Ever heard of the Hughes and Kettner Cream Machine?
It uses a 12AX7 in the preamp and a 12AU7 with half as the driver and
half as the output, into a tiny OPT. When you pull the speaker cable
from the jack, it engages a dummy load.
It's a stand alone amp, a recording amp (or direct to PA with Red Box
speaker cabinet simulation) or can be used as a preamp (AFTER the output
transformer). In practice, it sounds much like a cranked Marshall into
a 4x12. It's a bit gain happy for my tastes, I sometimes use two 12AU7.
It also lacks EQ. But it does what you want to do, and more. |
|
| Back to top |
|
|
Jim Anable
Guest
|
| Posted:
Sat Sep 17, 2005 4:45 am Post subject:
Re: 12au7 as SE output |
|
|
flipper wrote:
| Quote: | On Tue, 13 Sep 2005 17:44:18 GMT, Patrick Turner
info@turneraudio.com.au> wrote:
flipper wrote:
On Tue, 13 Sep 2005 09:20:17 GMT, Patrick Turner
info@turneraudio.com.au> wrote:
Chedomir Jokanovich wrote:
Hi, first of all, I am new to this group (infact this is my first
post). I really enjoy reading it, and it helped me so much, so many
times in the past. I'm glad to be here.
I am contenplating a little experiment of my own:) which may prove
usefull to me.
I was thinking of using a 12au7 as an output tube in a SE amp. Driving
around 8-10k load, with 250v supply. I figure it will give at least
half a watt that way. Now I'm not sure how to bias it propertly.
I was thinking maybe a 1kOhm bias resistor with a 10uF bypass cap. That
way it will dissipate around 2w idle. Is that ok?
Thanks
Chedomir Jokanovich
The load should be coupled via an OPT woth 8k:8 ohms.
Perhaps an old radio transformer will do.
You won't get more than about 1/5 a watt in class A1.
Maybe the tube would be good for headphones or a horn speaker...
I'm curious. Why do you say 1/5 of a watt is about all? Is it
predicated on a particular distortion level?
Its a rough guess based on typical triode efficiency.
If you have 2 watts of idle plate dissipation in an average triode,
then maximum plate efficiency in SET class A1 and limited by
3% thd is about between 20% and 30%.
EL34 can have a rated Pda = 28 watts, so expect a maximum
7 watts in triode, A1, 3% thd.
OK. So my guess about a distortion factor wasn't outlandish. I presume
you could increase that with some NFB?
What about positive grid drive?
|
This is GUITAR application, so the guy wants distortion. |
|
| Back to top |
|
|
flipper
Guest
|
| Posted:
Sun Sep 18, 2005 10:37 am Post subject:
Re: 12au7 as SE output |
|
|
On Fri, 16 Sep 2005 11:42:33 GMT, Patrick Turner
<info@turneraudio.com.au> wrote:
| Quote: |
flipper wrote:
On Fri, 16 Sep 2005 06:40:29 GMT, Patrick Turner
info@turneraudio.com.au> wrote:
flipper wrote:
snip,
A couple of mosfets would really drive it well.
A couple? The model I did was simply a FET follower just like the
tube.
I meant one mosfet for a single outpyt tube.
Maybe too well, because there is a limit for how much +ve voltage one would want to apply
to an output grid, and if its too much the grid overheats and begins emitting
electrons and the tube thermals away.
I understand. But you saw the grid currents. Are they too high?
Obviously if you force the grid to rise to a higg +ve value the
current becomes excessive since the grid has a low power dissipation limit.
Well, yes, obviously. But I don't see any grid dissipation guideline
so I was asking if you knew.
some tubes may specify max grid dissipation but most don't because they don't get used where
the Pg1 is high; it usually is quite insignificant in most apps where the g1 is biased negatively,
and rarely if ever driven positive.
But if you just set up a tube as normal and raise Eg1 you find that the grid starts to glow red,
orange,
and then it melts, and you have a stuffed tube.
|
It's ok if you don't know either ;)
Upon re inspecting the plate curves I notice they do give expected
grid current for up to +15V drive.
| Quote: | snip,
if you have a CF driving a tube where the Rout of the CF is about
1/10 of the minimum value or Rin to an output tube then the
operation will be substantially linear.
Ok.
What about Stewart's argument that it just plain can't be done
capacitive coupled?
When you you try to drive the g1 +ve with CR coupling, then it g1 current flows for the top
part of the cycle and the coupling cap charges so that the difference between the
preceeeding driving tube's anode voltage and the bias voltage at the driven tubes' grid becomes
graeter.
This results in an increasing DC voltage across the cap used for coupling, and it remains
during the AC cylcles due to the slow discharge RC time constant so the tube is working
as if someone had increase the bias voltage.
This is a mechanism which protects output tubes from excessive currents since the tube
with atificually high bias voltage is effectively cut off.
Serious overdrive in guitar amps results in the PP output stage running in class C.
Lotsa odd order distortion.
|
Right. I understand that.
| Quote: |
snipo,
Building and measuring is always going to be different to a simulated
outcome.
I roughtly calculate then build and measure.
I neve use spice programs; I don't need to.
I have a brain.
Well, you have a few decades of experience with them to draw on that I
don't.
I only started 10 years ago.
|
Ok. 10 years then.
<snip>
| Quote: |
Maybe say it's a 2W
amp with some hefty headroom ;) That's closer to the truth anyway as
it's obvious the NFB is 'working at it' to keep it in line at 3W.
If there is 4W at the tube, expect 3W at the OPT outlet because many little
OPT have maybe 25% winding losses.
Yeah. I took note of your earlier mention of that. I've got the OPT
winding resistance in the model but not capacitance leakage as I had
no convenient means to measure that. At any rate, I get about 20% loss
that way.
The Lp, LL nor Cshunt will affect what happens at 1 kHz.
Is that a neither will or an all will?
You need to understand the model for an audio transformer.
|
Thank you. I think I do. I was simply asking for confirmation of it.
<snip>
| Quote: | If I'm guessing the sentence right I take it you're saying the losses
are due to the internal resistance I've got in the model.
Winding losses = I squared x R where I is the current flow and R is the DC resistance of the wire.
At 1 kHz, other losses are low due to current flows in the inductive and capaitive elements
shunting windings or in series with them.
|
So, yes.
<snip>
I'm looking at what power supply changes I'd need to make to do DC
coupling, vs what parts I have at hand, because it would also help the
OPT issue at the available B+ I've got. Not quite so simple as just
adding some negative V because it's then 40V-50V over the max 300V
plate on the 12SL7.
Think it would be ok to just run it 350V plate if I keep the watts in
line or is it worth it to pop in perhaps a SS series pass regulator or
a glow tube shunt regulator for a lower B+ on that section? Maybe a
zener in series? |
|
| Back to top |
|
|
Patrick Turner
Guest
|
| Posted:
Sun Sep 18, 2005 8:48 pm Post subject:
Re: 12au7 as SE output |
|
|
flipper wrote:
| Quote: | On Fri, 16 Sep 2005 11:42:33 GMT, Patrick Turner
info@turneraudio.com.au> wrote:
flipper wrote:
On Fri, 16 Sep 2005 06:40:29 GMT, Patrick Turner
info@turneraudio.com.au> wrote:
flipper wrote:
But if you just set up a tube as normal and raise Eg1 you find that the grid starts to glow red,
orange,
and then it melts, and you have a stuffed tube.
It's ok if you don't know either ;)
Upon re inspecting the plate curves I notice they do give expected
grid current for up to +15V drive.
|
Its unusual that any average octal output tube will need more than
+15V drive to its grid during wave cycles since at such grid voltages
the anode swing down is almost as far as you'd want it
to go so no point in applying any more grid voltage.
And the grid current won't bwe sustained at the +15 volt level,
so the grid will survive.
If a grid cops a continuous +15V due to a leaky coupling cap, the plate current and grid current
will often cause a tube to fail.
| Quote: |
I'm looking at what power supply changes I'd need to make to do DC
coupling, vs what parts I have at hand, because it would also help the
OPT issue at the available B+ I've got. Not quite so simple as just
adding some negative V because it's then 40V-50V over the max 300V
plate on the 12SL7.
Think it would be ok to just run it 350V plate if I keep the watts in
line or is it worth it to pop in perhaps a SS series pass regulator or
a glow tube shunt regulator for a lower B+ on that section? Maybe a
zener in series?
|
You should be able to place an RC filter to
drop the anode B+ supply voltage to say +200V for the CF tube
whch has no RL in its annode circuit.
It should be possible to create a negative supply of say -200V for the
cathode resistor of the 12SL7, and to create a
fixed bias supply for the 12SL7 grid so that the output tube
has its own cathode grounded and its grid
is biased from the negative SL7 cathode because it is directly connected.
There is no need to use glow tubes or zener diodes.
Input voltage is applied to the CF grid via a CR coupling.
Patrick Turner. |
|
| Back to top |
|
|
flipper
Guest
|
| Posted:
Mon Sep 19, 2005 3:58 pm Post subject:
Re: 12au7 as SE output |
|
|
On Sun, 18 Sep 2005 15:48:22 GMT, Patrick Turner
<info@turneraudio.com.au> wrote:
| Quote: |
flipper wrote:
On Fri, 16 Sep 2005 11:42:33 GMT, Patrick Turner
info@turneraudio.com.au> wrote:
flipper wrote:
On Fri, 16 Sep 2005 06:40:29 GMT, Patrick Turner
info@turneraudio.com.au> wrote:
flipper wrote:
But if you just set up a tube as normal and raise Eg1 you find that the grid starts to glow red,
orange,
and then it melts, and you have a stuffed tube.
It's ok if you don't know either ;)
Upon re inspecting the plate curves I notice they do give expected
grid current for up to +15V drive.
Its unusual that any average octal output tube will need more than
+15V drive to its grid during wave cycles since at such grid voltages
the anode swing down is almost as far as you'd want it
to go so no point in applying any more grid voltage.
And the grid current won't bwe sustained at the +15 volt level,
so the grid will survive.
If a grid cops a continuous +15V due to a leaky coupling cap, the plate current and grid current
will often cause a tube to fail.
|
Yes. I'm not driving it over 15 volts. Max is under 12.
| Quote: | I'm looking at what power supply changes I'd need to make to do DC
coupling, vs what parts I have at hand, because it would also help the
OPT issue at the available B+ I've got. Not quite so simple as just
adding some negative V because it's then 40V-50V over the max 300V
plate on the 12SL7.
Think it would be ok to just run it 350V plate if I keep the watts in
line or is it worth it to pop in perhaps a SS series pass regulator or
a glow tube shunt regulator for a lower B+ on that section? Maybe a
zener in series?
You should be able to place an RC filter to
drop the anode B+ supply voltage to say +200V for the CF tube
whch has no RL in its annode circuit.
|
That was the first thing I did but the spice model shows it increasing
distortion even with some pretty impressive uFs in there. And ,of
course, the cap charge varies because the tube current isn't
symmetrical.
It is as much as 7dB on the third harmonics at 3W and 2 or 3dB at 1W,
depending on the RC selected, but maybe I'm just obsessing.
| Quote: | It should be possible to create a negative supply of say -200V for the
cathode resistor of the 12SL7, and to create a
fixed bias supply for the 12SL7 grid so that the output tube
has its own cathode grounded and its grid
is biased from the negative SL7 cathode because it is directly connected.
|
Right. No sweat. Just complicates my parts count a bit.
| Quote: | There is no need to use glow tubes or zener diodes.
Input voltage is applied to the CF grid via a CR coupling.
|
Again, no problem
|
|
| Back to top |
|
|
Patrick Turner
Guest
|
| Posted:
Mon Sep 19, 2005 5:49 pm Post subject:
Re: 12au7 as SE output |
|
|
flipper wrote:
| Quote: | On Sun, 18 Sep 2005 15:48:22 GMT, Patrick Turner
info@turneraudio.com.au> wrote:
flipper wrote:
On Fri, 16 Sep 2005 11:42:33 GMT, Patrick Turner
info@turneraudio.com.au> wrote:
flipper wrote:
On Fri, 16 Sep 2005 06:40:29 GMT, Patrick Turner
info@turneraudio.com.au> wrote:
flipper wrote:
But if you just set up a tube as normal and raise Eg1 you find that the grid starts to glow red,
orange,
and then it melts, and you have a stuffed tube.
It's ok if you don't know either ;)
Upon re inspecting the plate curves I notice they do give expected
grid current for up to +15V drive.
Its unusual that any average octal output tube will need more than
+15V drive to its grid during wave cycles since at such grid voltages
the anode swing down is almost as far as you'd want it
to go so no point in applying any more grid voltage.
And the grid current won't bwe sustained at the +15 volt level,
so the grid will survive.
If a grid cops a continuous +15V due to a leaky coupling cap, the plate current and grid current
will often cause a tube to fail.
Yes. I'm not driving it over 15 volts. Max is under 12.
I'm looking at what power supply changes I'd need to make to do DC
coupling, vs what parts I have at hand, because it would also help the
OPT issue at the available B+ I've got. Not quite so simple as just
adding some negative V because it's then 40V-50V over the max 300V
plate on the 12SL7.
Think it would be ok to just run it 350V plate if I keep the watts in
line or is it worth it to pop in perhaps a SS series pass regulator or
a glow tube shunt regulator for a lower B+ on that section? Maybe a
zener in series?
You should be able to place an RC filter to
drop the anode B+ supply voltage to say +200V for the CF tube
whch has no RL in its annode circuit.
That was the first thing I did but the spice model shows it increasing
distortion even with some pretty impressive uFs in there. And ,of
course, the cap charge varies because the tube current isn't
symmetrical.
|
Well use a 6SN7 instead of a pissant little 12SL7
and that will lessen distortions if you are so worried by them.
| Quote: |
It is as much as 7dB on the third harmonics at 3W and 2 or 3dB at 1W,
depending on the RC selected, but maybe I'm just obsessing.
It should be possible to create a negative supply of say -200V for the
cathode resistor of the 12SL7, and to create a
fixed bias supply for the 12SL7 grid so that the output tube
has its own cathode grounded and its grid
is biased from the negative SL7 cathode because it is directly connected.
Right. No sweat. Just complicates my parts count a bit.
|
Einstein said everything should be designed to be as simple as possible,
but no simpler.
Maybe he was thinking about how he saw the universe.
But even with your amp, the same wisdom applies.
If you had only a two dimensional universe, then the design and
rules would all be a lot simpler.
But we have at least 3 dimensions, up, down, across,
so to make a box you need more bits than if you just
make postcard.
But you can carry bananas in a box, and you can't do that with a postcard,
so the extra complexity is worth it.
Patrick Turner.
| Quote: |
There is no need to use glow tubes or zener diodes.
Input voltage is applied to the CF grid via a CR coupling.
Again, no problem
Patrick Turner.
|
|
|
| Back to top |
|
|
flipper
Guest
|
| Posted:
Tue Sep 20, 2005 4:43 am Post subject:
Re: 12au7 as SE output |
|
|
On Mon, 19 Sep 2005 12:49:21 GMT, Patrick Turner
<info@turneraudio.com.au> wrote:
| Quote: |
flipper wrote:
On Sun, 18 Sep 2005 15:48:22 GMT, Patrick Turner
info@turneraudio.com.au> wrote:
flipper wrote:
On Fri, 16 Sep 2005 11:42:33 GMT, Patrick Turner
info@turneraudio.com.au> wrote:
flipper wrote:
On Fri, 16 Sep 2005 06:40:29 GMT, Patrick Turner
info@turneraudio.com.au> wrote:
flipper wrote:
But if you just set up a tube as normal and raise Eg1 you find that the grid starts to glow red,
orange,
and then it melts, and you have a stuffed tube.
It's ok if you don't know either ;)
Upon re inspecting the plate curves I notice they do give expected
grid current for up to +15V drive.
Its unusual that any average octal output tube will need more than
+15V drive to its grid during wave cycles since at such grid voltages
the anode swing down is almost as far as you'd want it
to go so no point in applying any more grid voltage.
And the grid current won't bwe sustained at the +15 volt level,
so the grid will survive.
If a grid cops a continuous +15V due to a leaky coupling cap, the plate current and grid current
will often cause a tube to fail.
Yes. I'm not driving it over 15 volts. Max is under 12.
I'm looking at what power supply changes I'd need to make to do DC
coupling, vs what parts I have at hand, because it would also help the
OPT issue at the available B+ I've got. Not quite so simple as just
adding some negative V because it's then 40V-50V over the max 300V
plate on the 12SL7.
Think it would be ok to just run it 350V plate if I keep the watts in
line or is it worth it to pop in perhaps a SS series pass regulator or
a glow tube shunt regulator for a lower B+ on that section? Maybe a
zener in series?
You should be able to place an RC filter to
drop the anode B+ supply voltage to say +200V for the CF tube
whch has no RL in its annode circuit.
That was the first thing I did but the spice model shows it increasing
distortion even with some pretty impressive uFs in there. And ,of
course, the cap charge varies because the tube current isn't
symmetrical.
Well use a 6SN7 instead of a pissant little 12SL7
and that will lessen distortions if you are so worried by them.
|
That's one possibility, and certainly an easy one, but it seems to me
so would be providing a regulated lower supply to the 12SL7.
Cost is a consideration and I don't have a 6SN7. Not sure if I have
something suitable for regulating a lower supply either so I'll just
have to look at the tradeoffs.
Sorry if I upset you. I'm just trying to draw on your experience to
see if there was something I was missing.
| Quote: | It is as much as 7dB on the third harmonics at 3W and 2 or 3dB at 1W,
depending on the RC selected, but maybe I'm just obsessing.
It should be possible to create a negative supply of say -200V for the
cathode resistor of the 12SL7, and to create a
fixed bias supply for the 12SL7 grid so that the output tube
has its own cathode grounded and its grid
is biased from the negative SL7 cathode because it is directly connected.
Right. No sweat. Just complicates my parts count a bit.
Einstein said everything should be designed to be as simple as possible,
but no simpler.
Maybe he was thinking about how he saw the universe.
|
Yes, he was speaking of scientific theories. I think the actual quote
goes something like "The supreme goal of all theory is to make the
irreducible basic elements as simple and as few as possible without
having to surrender the adequate representation of a single datum of
experience." Which is commonly paraphrased into the condensed form of
"Everything should be made as simple as possible, but not one bit
simpler."
It's similar to Occam's Razor, which people erroneously paraphrase as
"the simpler solution is usually correct." The actual notion is
""Entities should not be multiplied beyond necessity," which does
include the 'but no simpler' aspect in "necessity" but the condensed
Einstein version is probably clearer about it.
| Quote: | But even with your amp, the same wisdom applies.
|
Of course. Which is why DC coupling to the gain stage, rather than the
6V6, seemed a simpler solution.
But there's always a devil in the details.
| Quote: | If you had only a two dimensional universe, then the design and
rules would all be a lot simpler.
But we have at least 3 dimensions, up, down, across,
so to make a box you need more bits than if you just
make postcard.
But you can carry bananas in a box, and you can't do that with a postcard,
so the extra complexity is worth it.
|
Unless all one needed was a postcard to mail ;) |
|
| Back to top |
|
|
|
|
|
|
Profile
Log in to check your private messages
Log in