Cathode Bias PushPull

Cathode Bias PushPull

Hi

I want to do a chatode bias low power push pull stage , how do I calculate
the
cathode bias resitors ?

Thanks
ec

Re: Cathode Bias PushPull

On Apr 22, 3:24=EF=BF=BDpm, "RealInfo" <therighti...@yahoo . com > wrote:
> Hi
>
> I want to do a chatode bias low power push pull stage , how do I calculate=

> the
> cathode bias resitors ?
>
> Thanks
> ec

Hi RATs!

To solve the problem, we need to know the spects of the transformer,
the power supply voltage and current the drive voltage and what tubes
are being used. Start with 500 ohms ;)

Happy Ears!
Al

Re: Cathode Bias PushPull

On Wed, 23 Apr 2008 00:24:43 +0200, "RealInfo"
<therightinfo@yahoo . com > wrote:

>Hi
>
>I want to do a chatode bias low power push pull stage , how do I calculate
>the
>cathode bias resitors ?
>
>Thanks
>ec

Assuming grid is referenced to ground and a traditional single, common
cathode, Rk to ground,

Rk = Vbias/(2 x idle)

Where Vbias is the desired bias voltage and idle is the per tube idle
current at that bias voltage, with the 2x because there are two tubes.

Just a simple case of R=E/I

Picking idle current is the bigger trick.

Re: Cathode Bias PushPull


>
> Rk = Vbias/(2 x idle)
>
> Where Vbias is the desired bias voltage and idle is the per tube idle
> current at that bias voltage, with the 2x because there are two tubes.
>

A problem is that if one of the pair of the tubes hogs most of the
current, which can lead to red plates.

Re: Cathode Bias PushPull

On Tue, 22 Apr 2008 23:42:01 -0600, Robert Casey
<wa2ise@ix . net com . com > wrote:

>
>>
>> Rk = Vbias/(2 x idle)
>>
>> Where Vbias is the desired bias voltage and idle is the per tube idle
>> current at that bias voltage, with the 2x because there are two tubes.
>>
>
>A problem is that if one of the pair of the tubes hogs most of the
>current, which can lead to red plates.

True, but I wonder how often that happens because a single Rk was
quite commonplace.

Re: Cathode Bias PushPull



flipper wrote:

> Robert Casey wrote:
> >>
> >> Rk = Vbias/(2 x idle)
> >>
> >> Where Vbias is the desired bias voltage and idle is the per tube idle
> >> current at that bias voltage, with the 2x because there are two tubes.
> >>
> >
> >A problem is that if one of the pair of the tubes hogs most of the
> >current, which can lead to red plates.
>
> True, but I wonder how often that happens because a single Rk was
> quite commonplace.

I've seen it in a VOX AC30 for example. FOUR EL84s sharing a single Rk. I
modified it to use 2 resistors.

Graham

Re: Cathode Bias PushPull

On Wed, 23 Apr 2008 09:59:56 +0100, Eeyore
<rabbitsfriendsandrelations@hotmail . com > wrote:

>
>
>flipper wrote:
>
>> Robert Casey wrote:
>> >>
>> >> Rk = Vbias/(2 x idle)
>> >>
>> >> Where Vbias is the desired bias voltage and idle is the per tube idle
>> >> current at that bias voltage, with the 2x because there are two tubes.
>> >>
>> >
>> >A problem is that if one of the pair of the tubes hogs most of the
>> >current, which can lead to red plates.
>>
>> True, but I wonder how often that happens because a single Rk was
>> quite commonplace.
>
>I've seen it in a VOX AC30 for example. FOUR EL84s sharing a single Rk. I
>modified it to use 2 resistors.

Well, that's interesting because the original comment was about one
per two tubes and that's what you changed it to.

>Graham
>

Re: Cathode Bias PushPull



flipper wrote:

> Eeyore wrote:
> >flipper wrote:
> >> Robert Casey wrote:
> >> >>
> >> >> Rk = Vbias/(2 x idle)
> >> >>
> >> >> Where Vbias is the desired bias voltage and idle is the per tube idle
> >> >> current at that bias voltage, with the 2x because there are two tubes.
> >> >>
> >> >
> >> >A problem is that if one of the pair of the tubes hogs most of the
> >> >current, which can lead to red plates.
> >>
> >> True, but I wonder how often that happens because a single Rk was
> >> quite commonplace.
> >
> >I've seen it in a VOX AC30 for example. FOUR EL84s sharing a single Rk. I
> >modified it to use 2 resistors.
>
> Well, that's interesting because the original comment was about one
> per two tubes and that's what you changed it to.

I rather imagined that comment referred to an output stage using 2 tubes rather
than four.

Graham

Re: Cathode Bias PushPull

On Wed, 23 Apr 2008 12:10:45 +0100, Eeyore
<rabbitsfriendsandrelations@hotmail . com > wrote:

>
>
>flipper wrote:
>
>> Eeyore wrote:
>> >flipper wrote:
>> >> Robert Casey wrote:
>> >> >>
>> >> >> Rk = Vbias/(2 x idle)
>> >> >>
>> >> >> Where Vbias is the desired bias voltage and idle is the per tube idle
>> >> >> current at that bias voltage, with the 2x because there are two tubes.
>> >> >>
>> >> >
>> >> >A problem is that if one of the pair of the tubes hogs most of the
>> >> >current, which can lead to red plates.
>> >>
>> >> True, but I wonder how often that happens because a single Rk was
>> >> quite commonplace.
>> >
>> >I've seen it in a VOX AC30 for example. FOUR EL84s sharing a single Rk. I
>> >modified it to use 2 resistors.
>>
>> Well, that's interesting because the original comment was about one
>> per two tubes and that's what you changed it to.
>
>I rather imagined that comment referred to an output stage using 2 tubes rather
>than four.

Yes, that's what I said and what you changed the 4 to: a pair of
deuces on a common Rk.

>
>Graham

Re: Cathode Bias PushPull

flipper wrote:
> Yes, that's what I said and what you changed the 4 to: a pair of
> deuces on a common Rk.
>
>>

nick picking

Re: Cathode Bias PushPull

On Wed, 23 Apr 2008 21:44:59 -0500, Restless Fingers Syndrome
<rfs@itch . com > wrote:

>flipper wrote:
>> Yes, that's what I said and what you changed the 4 to: a pair of
>> deuces on a common Rk.
>>
>>>
>
> nick picking

Do you have anything of value to contribute toward answering the
question?

Re: Cathode Bias PushPull


"Eeysore"
>
> I've seen it in a VOX AC30 for example. FOUR EL84s sharing a single Rk. I
> modified it to use 2 resistors.


** What I do is double the value of Rk to 100ohms, add 3 x 5W 4.7 volt
zeners in series ALL in parallel with Rk to limit the cathode bias
voltage to about 14 volts.

Cools the poor, over-heated EL84s off and increases power out by moving the
stage further into class AB.

Works a treat.




... Phil

Re: Cathode Bias PushPull



Phil Allison wrote:

> "Eeysore"
> >
> > I've seen it in a VOX AC30 for example. FOUR EL84s sharing a single Rk. I
> > modified it to use 2 resistors.
>
> ** What I do is double the value of Rk to 100ohms, add 3 x 5W 4.7 volt
> zeners in series ALL in parallel with Rk to limit the cathode bias
> voltage to about 14 volts.
>
> Cools the poor, over-heated EL84s off and increases power out by moving the
> stage further into class AB.
>
> Works a treat.

Yup, good idea.

Graham

Re: Cathode Bias PushPull



Eeyore wrote:
>
> Phil Allison wrote:
>
> > "Eeysore"
> > >
> > > I've seen it in a VOX AC30 for example. FOUR EL84s sharing a single Rk. I
> > > modified it to use 2 resistors.
> >
> > ** What I do is double the value of Rk to 100ohms, add 3 x 5W 4.7 volt
> > zeners in series ALL in parallel with Rk to limit the cathode bias
> > voltage to about 14 volts.
> >
> > Cools the poor, over-heated EL84s off and increases power out by moving the
> > stage further into class AB.
> >
> > Works a treat.
>
> Yup, good idea.
>
> Graham

In a guitar amp like the Vox AC30, Phil's idea does
work where distortion does not matter and power does.

However, I have had to service quite a few amps with 4 x EL84
with a variety of cathode biasing methods, all wrong,
because none used individual tube biasing.

In all cases the disgruntled muso with his unhappy amp has turned up
here
and an examination reveals the amp has one or more EL84 going into
thermal runaway before the others,
and the usual huge amount of overdriving has unmatched the quad of
output
tubes so badly that some run with 15 watts at idle and some with 5
watts.
All over the fucking place.
Installing individual Rk&Ck regulates the bias of each tube, and the
problem is much
reduced.

All this about Vox AC30 has been discussed at length before here within
the last 4 years.

Patrick Turner.

Re: Cathode Bias PushPull



Eeyore wrote:
>
> flipper wrote:
>
> > Robert Casey wrote:
> > >>
> > >> Rk = Vbias/(2 x idle)
> > >>
> > >> Where Vbias is the desired bias voltage and idle is the per tube idle
> > >> current at that bias voltage, with the 2x because there are two tubes.
> > >>
> > >
> > >A problem is that if one of the pair of the tubes hogs most of the
> > >current, which can lead to red plates.
> >
> > True, but I wonder how often that happens because a single Rk was
> > quite commonplace.
>
> I've seen it in a VOX AC30 for example. FOUR EL84s sharing a single Rk. I
> modified it to use 2 resistors.
>
> Graham

The ONLY way cathode biasing should be done is with individual Rk and CK
for each output tube.

Despite the use of one common Rk for 2+ output tubes,
Its a completely BS idea leading always to smoke and
bad tempered moments, and unnecessary expense.
Even in Quad-II power amps with the 3W x 180 ohm common Rk,
I have often seen one KT66 glowing red with 90mA of idle current and the
other
running too cool with 40mA. This mucks up the sound really badly.

To determine the Rk value, the best method for all octal and
nine pin tubes is start with 470 ohms experimentally.

There is a very simple formula, but its far too confusing for diyers.

So, here is how to do it right......

1.With both output tubes plugged in and grids grounded, and with Rk =
470 ohms to each output tube,
MEASURE the voltage across each 470 ohms 30 seconds after turn on.
measure the anode to 0V voltage dc.

2. Then turn the amp off!

3. Calculate Ea, the Vdc between anode and cathode.

Ea = Vdc at anode to 0V minus Vdc across Rk.

4.hen calculate Ik, cathode to 0V current flow.

Ik = Ek / Rk

5. Calculate the total tube electrode heat dissipation, Pdiss.
Pdiss = Ea x Ik, ie, Volts x Amps, and answer in Watts.

6. Is this below the sensible value for the tube?

7. Look up data for maximum pdiss in triode for the tube concerned.
Eg, 6L6 has anode diss Pda = 22 watts in beam tetrode , and 25W
for combined anode plus screen Pd for triode.
Anode + screen current = cathode current, OK?

8. For pure class A, measured and calculated Pdiss can be up to 90%
of makers data Pdiss for triode connection.

( An exception is the 300B which has a rating max of 42 watts,
but you'd be mad to allow 42 watts. )
If Pdtrated Vrkoff.

9. If the measured Pdiss is MORE than the data figure, say 32 watts for
EL34,
or 6L6, then INCREASE Rk, and recalculate and remeasure all
until Pdiss falls below max allowable.

10. Do you want class AB?

If so, and it IS nearly always wanted,
then what % of class A PO can be had with the intended design load
before the AB threshold begins?

If the % of class A PO is MORE than 50% of wanted class AB PO
at clipping onset, then increase Rk so Pdiss = 0.7 x rated Pdiss max for
the tube, or less.

11. learn load line analysis so that PO in class A and AB can be
calculated.

Adjust Rk to suit the AB conditions with the working Ea and Ia.

12. If class A PO is less than 50% of total AB PO at clipping,
then consider fixed bias.

For all about building good amps see my pages beginning at
* w w w .turneraudio . com .au/loadmatch4-pp-beamtetrodes.html

13. If you cannot understand a word of what i am saying at the website,
then assume you have less then 50% of total PO in class A for the
nominal RL
and make tube Pdiss = 1/2 of maker maximum, ie, 12 watts for 6L6 and
EL34.

This is done by adjusting Rk
until the measured Pdiss = the wanted and safe Pdiss = 1/2 design rated
maximum from tube data.

14. There are no other short cuts worth mentioning.

15. Ck, the bypassing caps should be at least 470uF and rated for at
least
3 times the working Ek. So if the cathode Ek = +31V, use 63V rated
electros.

16. Bypass the electros with a minimum of 0.47uF polyester or
polypropylene caps, rated at 400V.

17. Is for something I forgot.
If Pdiss is more than allowable data permits, tubes will expire soon,
and after maybe glowing red hot, hence the need to measure accurately
and quickly
after turn on BEFORE the overheating turns the tubes red hot.
Do not sit there like a slow witted dummy and marvel at the melt down.

Biasing tubes correctly is obviously best done by a well practised tech
because unless you are aware of all the above, chances are you'll
make an expensive mistake.

Patrick Turner.