Push-Pull 845 Tube Amplifier

(General) Forum about other items than listed below

Push-Pull 845 Tube Amplifier

Postby rogerk8 » Tue Apr 24, 2012 12:46 am

Here is my preliminary 845 push-pull amplifier. I will however need to revise the transformers data but it is almost complete (with exception of the cossor). I am about to let a transformer manufacturer wind my transformers for me even if my fellow swedish specialist Lundahl will be hard to compete with. So feel free to comment. Best regards, Roger
http://knopp.se/knopp/845_simple.PNG
rogerk8
 
Posts: 24
Joined: Wed Apr 11, 2012 8:35 pm

Re: Push-Pull 845 Tube Amplifier

Postby Doc10 » Fri Oct 19, 2012 8:30 am

Hi Roger, what is the function of this potmeter? (See attached file)
Attachments
845_simple.PNG
845_simple.PNG (27.33 KiB) Viewed 13590 times
Doc10
 
Posts: 20
Joined: Wed Oct 03, 2012 3:44 pm
Location: Belgium

Re: Push-Pull 845 Tube Amplifier

Postby DanL » Sat Nov 03, 2012 10:24 pm

Roger

Are you sure you want the 6SN7 to drive the 845?
The 6SN7 is a questionable driver for the 300B.
The 6SN7 has an rp of over 11K at that operating point.
Add another 1K with the gate stopper which is not needed IMO.
Then with under 5mA going through the 6SN7,
trying to drive the miller capacitance of the 845.
The 845 needs a much better driver.
ECC99 or 12BH7 at about 15-20mA would be my choice.
Or trioded 6W6/6V6/6L6 family are really nice too.

DanL
DanL
 
Posts: 2
Joined: Sun Jul 15, 2012 7:37 am

Re: Push-Pull 845 Tube Amplifier

Postby J. Leven » Sun Nov 04, 2012 10:50 am

Hi,
is the plate dissipation for two 845 with approximate 190 Watt not to high ?

J.L
J. Leven
 
Posts: 170
Joined: Fri Jan 13, 2012 12:27 pm

Re: Push-Pull 845 Tube Amplifier

Postby rogerk8 » Mon Nov 12, 2012 12:24 am

Doc10 wrote:Hi Roger, what is the function of this potmeter? (See attached file)


Hi Doc10!

The purpose of that pot is to be able to balance the currents in the OPT (Output Transformer). This need do however depend on what type of OPT you use. But most OPTs prefere not to be DC-magnetized (except for single-end designs with air-gaps).

It is a kind of crazy design which I think I will never succeed in building. :)

Best regards, Roger
PS
To J.Leven I want to say that my data sheets for the 845 says that it can work with 100W plate dissipation. But I'm not sure when it comes to other types than the original RCA.

To DanL I want to say that I do think the 6SN7 is a better tube than many want to acknowledge. It has an amazing linearity for at least a +/- 100V swing (RL=47k, B+=450V). I do however recognize that it will not be able drive the 845s fully into maximum output power. On the other hand, do we really need that full output power? I don't know if I've calculated correctly but 50W for a pair I think is more than enough. And if we don't like 6SN7s we could always use 300Bs :)
rogerk8
 
Posts: 24
Joined: Wed Apr 11, 2012 8:35 pm

Re: Push-Pull 845 Tube Amplifier

Postby rogerk8 » Thu Nov 22, 2012 9:26 pm

DanL wrote:Roger

Are you sure you want the 6SN7 to drive the 845?
The 6SN7 is a questionable driver for the 300B.
The 6SN7 has an rp of over 11K at that operating point.
Add another 1K with the gate stopper which is not needed IMO.
Then with under 5mA going through the 6SN7,
trying to drive the miller capacitance of the 845.
The 845 needs a much better driver.
ECC99 or 12BH7 at about 15-20mA would be my choice.
Or trioded 6W6/6V6/6L6 family are really nice too.

DanL


Hi Dan L!

I have now done some preliminary calculations regarding the Miller capacitance problem you mentioned above.

Cag=14pF
u=5
rp(0)=1k5
RL=5k/tube
AvDM/2=u*RL/(RL+rp(0))=4
rp(6SN7)=7k

Which gives

Cm=(AvDM/2+1)*Cag=70pF

and thus

fh=1/(2pi*Cm*rp(6SN7))=325kHz

Which is no problem whatsoever!

Best regards, Roger
PS
The reason for omitting the (u+1)Rk-factor is that only differential mode (DM) signals will propagate through the OPT.
rogerk8
 
Posts: 24
Joined: Wed Apr 11, 2012 8:35 pm

Re: Push-Pull 845 Tube Amplifier

Postby rogerk8 » Thu Nov 22, 2012 11:17 pm

Hi!

I have now improved my Push-Pull 845 amplifier somewhat.

I have set the optimum bias-ripple resistance to 12k (activelly cancelling all ripple, regardless of mains ripple).

I have further more corrected the Rk-resistance for the autobiasing of the 845s. It shold now be perfect.

I have also adjusted the power dissipation of the biasing network for the output tubes (setting it to lowest possible).

But most of all, I have incorporated a couple of "Vacuum State Rectifiers" (VSR) for the supply.

It is not just because I love tubes, it is mainly because you get a smooth onset of voltage when you use cossors.

This might not seem important you may think, but when the tubes are heating up the supply is delivering a very high voltage (due to not actually being loaded) which in turn means that the caps might be overloaded.

Using cossors for the supply will effectively eliminate this problem. And at the same time yielding a nice timer feature.

This is now my favorite design:

http://knopp.se/knopp/845_simple_vs.PNG

Take care!

Best regards, Roger
PS
The 10V-winding is designed for 3A only. But I have designed the Chinese transformer to be able to withstand more than that. The only problem is that the voltage will not be exactly 10VAC but more like 9VAC. But I do not think that will be a problem. I do actually think that the heater voltage may vary as much as +/- 20% without any problem. The only problem with a too low heater will be the thermal limit of the current. Which is often way beyond used current.

The old SSR-version and somewhat revised version is visible here:
http://knopp.se/knopp/845_simple.PNG
rogerk8
 
Posts: 24
Joined: Wed Apr 11, 2012 8:35 pm

Re: Push-Pull 845 Tube Amplifier

Postby J. Leven » Fri Nov 23, 2012 6:59 pm

Hi,
845 Driver Stage 6SN7
fh=1/(2pi*Cm*rp(6SN7))=325kHz

325Khz, -3dB Point, never :?
J. Leven
 
Posts: 170
Joined: Fri Jan 13, 2012 12:27 pm

Re: Push-Pull 845 Tube Amplifier

Postby rogerk8 » Sat Nov 24, 2012 12:25 am

Hi J.Leven!

Why not?

Push-Pull means that you only use differential mode signals (DM). This in turn means that the common cathodes are virtually grounded. And this means that the driving impedance is less than rp (or rp//Ra actually). And the (u+1)2Rk-part of the equation is eliminated.

Just so others can follow:

The voltage gain of the normal (single tube) common-cathode stage is:

Av=-uRa/(rp+Ra+(u+1)Rk)

where

u is the voltage amplification factor (=gm*rp)
Ra is the load (in the anode circuit)
Rk is the cathode resistance
gm is the transconductance (i.e amps out for voltage in)
rp is the plate resistance (at the operating point, obs)

When it comes to push-pull circuits you should use 2 times the Rk that is actually sitting there if you want to analyze a single tube's behaviour.

In push-pull circuits only DM-signals will propagate thru the circuit and the OPT.

In this special case we have that the differential mode voltage gain is:

AvDM/2=uRa/(rp+Ra)

The /2-part comes from the fact that we have two tubes amplifying an equal amount of the differential signal. This means that total gain (AvDM) is actually twice as much.

The common mode signal (CM, only hum in this case) has the equation:

AvCM=uRa/(rp+Ra+(u+1)2Rk)

Which indeed supresses hum quite much (due to the (u+1)2Rk-part).

Actually, this quotient (AvDM/AvCM) is called CMRR or Common Mode Rejection Ratio (which is a well known and wanted feature for OP-amps. Here we could however use it to supress hum).

It should be pointed out that this do not matter in our case. All that happens is that both of the output tubes anodes ripples exactly the same. This CM-signal will never propagate to the speaker!

Using the above equations and the earlier equation for the miller-capacitance yields a banwidth of better than 325kHz.

Unless I am totally misguided, this is the case.

Best regards, Roger
PS
I have thought it over and I now think that the heater supply may vary +10%/-20%. The (+)-part will however probably make the tube age faster, but the (-)-part won't do the tube any harm whatsoever. This is because the thermal current limit is often way beyond used current.
rogerk8
 
Posts: 24
Joined: Wed Apr 11, 2012 8:35 pm

Re: Push-Pull 845 Tube Amplifier

Postby J. Leven » Sun Nov 25, 2012 5:32 pm

Hi Roger,
have you also measured this result for the frequency response at the driver stage ?,

best regards
J.L.
J. Leven
 
Posts: 170
Joined: Fri Jan 13, 2012 12:27 pm

Next

Return to HiFi Equipment Forum

Who is online

Users browsing this forum: No registered users and 1 guest