LTPI ?

[catnine]

'Proper' bias point is a matter of subjectivity. I sorta worked out that the HT for V1 on the thing you described is ~365VDC give or take. A 12AY7 modelled load line for that HT with a 100k load and a 1k5 Rk is as follows:

[ATTACH=CONFIG]21342[/ATTACH]

You got about 6V P-P of grid swing before it starts grid-current limiting/clipping. (Each one of those grid curves on that chart is 1V apart). That is more than enough to keep most guitar p'ups clean through V1 without a pedal in front. However if you go to Rk = 2k7 you got about 8V P-P grid swing clean:

[ATTACH=CONFIG]21343[/ATTACH]

Is that any help?

certainly it helps . I will study the charts you supplied . The build I have after the 22K node resister is 312VDC which I assume is the HT fro V1 . One V1 the 12ay7 I have a plate of 151.5 VDC and a cathode of 2.37 VDC .

[tubeswell]

The build I have after the 22K node resister is 312VDC which I assume is the HT fro V1 . One V1 the 12ay7 I have a plate of 151.5 VDC and a cathode of 2.37 VDC .

No matter. With the HT at 312 and with RK = 2k7, it would still be biased in about the mid point of the load line

[catnine]

No matter. With the HT at 312 and with RK = 2k7, it would still be biased in about the mid point of the load line

tubeswell : I don't quite understand . Since the 5E3 uses a 820 ohm bias resister for both triodes and I used a 1.5K ohm for one triode using 2.7K is almost three times the resistance of the original 820 ohm . You say I will get more of a midpoint bias if I use the 2.7K P-P meaning peak to peak? And if I do this I will have more clean headroom ? So the 1.5 K will cause the 12ay7 to clip sooner? I know a 12ay7 or 12ax7 even a 5751 don't share the same cathode, on the 5E3 they are just tied together . does this have something to do with not using the plate and grid of one triode ie no VDC on one plate and no signal on one grid = no current draw ? I ask this because many circuits use one triode for say the second gain stage and on for the PI yet they are not 12ay7 but 12ax7 or at7's . I always thought they called dual triodes were two tubes in one envelope .

[tubeswell]

tubeswell : I don't quite understand . Since the 5E3 uses a 820 ohm bias resister for both triodes and I used a 1.5K ohm for one triode using 2.7K is almost three times the resistance of the original 820 ohm . You say I will get more of a midpoint bias if I use the 2.7K P-P meaning peak to peak?

With a 12AY7 at HT = 315 volts with a 100k plate load - yes.

And if I do this I will have more clean headroom ?

In terms of what you have feed to the grid - yes. The more that the bias point is in the 'middle' of the grid curves on the load line, the more optimally biased the stage will be in terms of clean headroom.

So the 1.5 K will cause the 12ay7 to clip sooner?

With a HT = 315V and a 100k plate load - yes - if the voltage swing at the grid is sufficienty large enough to cause the stage to go into grid current limiting. On the example shown with HT = 365, the bias point with Rk = 1k5 is about -2.8V. (This comes down to about -2.5V with HT = 315V). This is closer to the Vg = 0V end of the load line than if you chose a Rk = 2k7, therefore the stage would go into grid current limiting sooner (if the swing at the grid was large enough). These are theoretical load line and bias points based on an 'idealised' 12AY7 set of grid curves. Real world operating conditions may differ slightly depending on the actual tube.

does this have something to do with not using the plate and grid of one triode i.e.: no VDC on one plate and no signal on one grid = no current draw ?

Each triode in a dual triode is separate from the other, unless you wire the socket pins in parallel.

[catnine]

With a 12AY7 at HT = 315 volts with a 100k plate load - yes.



In terms of what you have feed to the grid - yes. The more that the bias point is in the 'middle' of the grid curves on the load line, the more optimally biased the stage will be in terms of clean headroom.



With a HT = 315V and a 100k plate load - yes - if the voltage swing at the grid is sufficienty large enough to cause the stage to go into grid current limiting. On the example shown with HT = 365, the bias point with Rk = 1k5 is about -2.8V. (This comes down to about -2.5V with HT = 315V). This is closer to the Vg = 0V end of the load line than if you chose a Rk = 2k7, therefore the stage would go into grid current limiting sooner (if the swing at the grid was large enough). These are theoretical load line and bias points based on an 'idealised' 12AY7 set of grid curves. Real world operating conditions may differ slightly depending on the actual tube.



Each triode in a dual triode is separate from the other, unless you wire the socket pins in parallel.

This means that the 5E3 was not biased properly . I once tried the 12ay7 in parallel using the 820 ohm cathode resister and didn't care for that sound much but then that's using both triodes for a preamp so that's different all together or is it? Since they are wired in parallel and not series like a second gain with a coupling cap (plate to cap to grid) the 5E3 uses two inputs per triode I was using one . I can't recall ever seeing any amp with a preamp with one input wired in parallel. I know what resisters and caps do in seres or parallel but not tubes. I know some 5E3 users jump the inputs but I don't know what this creats sound wise , seems like it would breakup sooner .

[tubeswell]

This means that the 5E3 was not biased properly .

"Properly" depends on what you're setting out to achieve. Its okay for a V1 stage that usually doesn't see more than 1VP-P at the grid (like e.g.: with a SC p'up sans pedals).

[barfoden]

I must say that I do not have the knowlegde to say why the 12ay7 should have more headroom than a 12ax7 in V1 if the input gain was compensated for the difference, but on some amp balancing the gain between stages are important for best clean headroom.

But all clips I hear of 5e3's with a 12ax7 in V1 seem grainy and not so tonally pleasent as the ones with a 12ay7 in V1. Maybe the bias point is of with a 12ax7 in that position. But a Dearmond R15 has a 12ax7 in V1 with similar voltages, bypass cap and resistor, but the bypass cap on V2 (12ax7) is omitted and that balances things out. Also the Dearmond has 350K volume pots in relation to the 5e3's 1M pots. The Dearmond uses 0.047µf coupling caps but so did the Fender 5C3 and 5D3.

So the Dearmond R15 has a higher gain first stage and a lower gain second stage (total max gain factor of 92)
The 5e3 has a lower gain first stage but a higher gain second stage (total max gain factor of 93.5)

The two amps are fairly similar but the Dearmond is a bit stronger with larger transformers, smaller coupling caps and a large 40µf main filter cap to keep the bass more tight. Suppose to take humbuckers better than a stock 5e3.

Leonc build a clone of the Dearmond R15t, with the only exception of using 500k pots as the 350K (volume pots) and 650K (tone pot) are hard to come by. With a loud eminence swamp thing it is a great loud amp... check Youtube

[tubeswell]

... why the 12ay7 should have more headroom than a 12ax7 in V1 if the input gain was compensated for the difference, but on some amp balancing the gain between stages are important for best clean headroom.

But all clips I hear of 5e3's with a 12ax7 in V1 seem grainy and not so tonally pleasent as the ones with a 12ay7 in V1...

The answer lies in the transconductance and plate resistance characteristics of both types of tubes:

[Alf]

The answer lies in the transconductance and plate resistance characteristics of both types of tubes:

That's info I've been trying to get for quite some time !

Could you please explain the relation ?

Correct me if I'm wrong : what I have always thought is that the higher the transconductance the faster a tube goes into distortion when given too high a signal.

Maybe I'm completely wrong !

[jazbo8]

The basic relationship is: μ (gain) = gm * rp, so gm and rp are inversely proportional, and dependent on the bias setting.

[eschertron]

That's info I've been trying to get for quite some time !

Could you please explain the relation ?

Correct me if I'm wrong : what I have always thought is that the higher the transconductance the faster a tube goes into distortion when given too high a signal.

Maybe I'm completely wrong !

The way I understand it, transconductance (gm) is the amount of current the valve (er, tube) can push from an increment of voltage swing. a 12AT7, for example, has a high gm, so can provide much more current to the plate than the 12AX7. Better in certain applications. What I noticed from the chart above, is that the 12AY7 and 12AX7 have nearly the same gm, but their mu is different, so that the gain of each tube will be different in the same circuit.

If I would generalize, I'd say that mu is related to the ability of the tube to be overdriven in a specific circuit. Note that jazzbo8 points out that with these two parameters (mu and gm), the tube's resistance also changes. When tubes are swapped with different models, the designer's assumptions about the circuit may be totally thrown out of whack, so headroom and bias may need to be calculated and adjusted to achieve desired performance.

Again to generalize, it's OK to swap any of these tubes and observe the changes to the circuit's behavior. Note the tube's voltages, plot the bias point on a grid-curve chart, and see what has happened to headroom, etc.

[tubeswell]

barfoden's question was why the 12ay7 should have more headroom than a 12ax7

The answer is that the transconductance and plate resistance determine the tube's amplification factor. (jazbo8 explained how gm and rp get mu).

All other things being equal, if you change the plate resistance and/or transconductance, you will change the possible gain. A 12AY7 has a lot lower plate resistance than a 12AX7, even though the transconductance of both types is similar.