• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

807 E-Linear

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
A while ago I promised my brother to help him build a tube amp.
Since we were on a budget, we decided to use those 18 bucks Edcor OPT´s and a pair of 807´s.
After a few detours I finally got a prototype running yesterday,
but to my disappointment it didn´t sound as good as I had expected. The output tubes were running in UL mode with CFB from the OPT secondary, and obviously this wasn´t enough to get the output impedance low enough.

Since my brothers speakers are quite insensitive triode connection was never an option and I didn´t want to surrender to global NFB, I came up with the (seemingly) great idea to connect the input tubes plate resistors to the UL taps rather than to B+. I believe this is called E-linear. I´ve read that this topology works best with high impedance driver stages, so I removed the cathode bypass caps for the input stage.

The bass response immediately got much better, not "floppy" as it was before.
This drawing shows how the amp is wired today:

Any obvious mistakes here?
The amp sounds really good, but I haven´t done any measuerements yet.
 

Attachments

  • 807e-l.jpg
    807e-l.jpg
    33.6 KB · Views: 1,917
Hey-Hey!!!,
Looks good to me. I prefer pentodes for the input stage...actually cascodes but your ears are the final arbiter on that one. The pentode plate returns more NFB as it doesn't really care what the voltage is( like a triode ).

It is indeed called E-Linear. That circuit is different enough to deserve its own name.
cheers,
Douglas
 
What I´ve come up with here seems to be a mix between Pete Milletts E-linear amp and Alex´s RH807.
Cost and simplicity has high priority, that´s why I use a dual triode on the input. (Well, actually it´s because I didn´t plan to use this kind of feedback from the beginning).
Can I expect huge improvements by changing to pentodes (6AU6 or something)?

I´m going to run some measurements on the prototype tonight, I´m not sure that my driver stage is capable of pushing the output tubes to full power.
 
AX tech editor
Joined 2002
Paid Member
Fuling said:
A while ago I promised my brother to help him build a tube amp.
Since we were on a budget, we decided to use those 18 bucks Edcor OPT´s and a pair of 807´s.
After a few detours I finally got a prototype running yesterday,
but to my disappointment it didn´t sound as good as I had expected. The output tubes were running in UL mode with CFB from the OPT secondary, and obviously this wasn´t enough to get the output impedance low enough.

Since my brothers speakers are quite insensitive triode connection was never an option and I didn´t want to surrender to global NFB, I came up with the (seemingly) great idea to connect the input tubes plate resistors to the UL taps rather than to B+. I believe this is called E-linear. I´ve read that this topology works best with high impedance driver stages, so I removed the cathode bypass caps for the input stage.

The bass response immediately got much better, not "floppy" as it was before.
This drawing shows how the amp is wired today:

Any obvious mistakes here?
The amp sounds really good, but I haven´t done any measuerements yet.


I think if you decouple the 10 ohms cathode resistor you double the output power.

Jan Didden
 
AX tech editor
Joined 2002
Paid Member
Fuling said:
The 10 ohm resistor is only there to measure the voltage drop (bias current) across, it shouldn´t affect the output power...?


The 10 ohms doesn't know this... ;)

Fact is that the signal voltage across the 10 ohms is about equal (even more) then across the speaker load because they carry the same signal current. So you lose half your power. Just put a cap across it, you still have the DC bias but now all the signal is across the speaker.

Jan Didden
 
arnoldc said:
Is this topology specifically made for plate loaded input tubes?

What if I have a CCS on the input tube plate? Will that make a difference?

Hey-Hey!!!,
The CCS will not allow the circuit to operate as intended. It takes advantage of the near zero PSRR of a pentode. A voltage is changed by varying the current through the first stage, and the tapped plate winding of the final takes some of that change back. Triodes don't yeild much FB, and even the difference between bypassing its cathode R makes a difference. I like cascodes, mostly because I build this circuit PP and use the input stage asthe PI too. No parallel path to the cathode like a pentode's g2 offers in the cascode.

In particular, I have tried the 6AU6 and like it much! Try a 30k plate load and a g2 voltage of ~100.
cheers,
Douglas
 
Initial measurement results (my tone generator is the crappiest thing I´ve ever seen so I will redo this once I get me a proper one):

Output power: 4W into 8 ohms, probably more if the tone generator had been able to deliver more signal...
Gain: Just over 2x from input to output. Painfully low I must say.
Frequency response seems to go past 40kHz, this will also be investigated further later on.
 
Hey-Hey!!!,
As you've probably seen in Pete's article, higher gm in the driver pentode is the Way to gain. gm x Rl eh? Anyway, I like the 6AU6 but the 12HL7 has done well with 6L6's...Take one of Pete's OP for that valve, perhaps with a 15k plate load? 20k maybe? g2 at 75-90V I think...:)

I did name it. E got chosen for a few reasons for setting this U-L based circuit apart. It is voltage FB on a gm device that makes it work so well.

One of my fave's is a triode-MOSFET cascode. No need to deliver g2 current at the equivalent node point of the upper gate. The plate lines are about horizontal as well...maximum FB.
cheers,
Douglas
 
AX tech editor
Joined 2002
Paid Member
Fuling said:
I agree with Circlotron, the 10 ohm resistor is in series with the primary winding rather than with the secondary winding.
It drops about 0,65VDC at idle and perhaps 1Vpp at max output power, a very small fraction of the total plate swing.


Ahh yes, I didn't catch that there actually was a transformer, I thought that this was a cathode follower type. My bad...

Jan Didden
 
Hey-Hey!!!,
The cascode is not a gain machine any moe than a pentode is. Depends on your plate load of course gm x plate load or there-abouts...:)

I think my primary attachment to the cascode is due to its horizontal plate lines and their effect on FB level. An ECC88 and a FQP1N60 MOSFET should work well. Put the gate where the upper grid would normally go.
cheers,
Douglas
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.