Express in single ended format?

[geoffrod]

The clean to mean has nothing to do with exceeding the rated limits of the PT. No way in hell are two EL34s ever going to average 300mA. Even if you could manage to pull that much current, it's not going to sound cool. It's going to turn your power transformer into a chimney and probably trip a circuit breaker.

The clean to mean has just about everything to do with that 1K dropping resistor in the power supply between the plate and screen nodes. When you hit the power tubes with a large signal the screen current spikes, causing a voltage drop across that resistor that propagates to the preamp, changing the operating point of the PI and 3rd clipping stage and causing them to compress/limit more. You could even think of it as a weird sort of negative feedback mechanism.

You cannot completely replicate this in a SE design, because the output tube(s) are biased in class A and will be drawing too much current at idle. In a push pull power amp, both output tubes are biased near cutoff, so you are drawing far less current at idle, which means that the drop across that 1k resistor is comparatively small. That's really what the clean-to-mean boils down to; for smaller signals seen by the output tubes, the headroom of the preceding clipping stages (PI + 3rd stage) is relatively high. Increase the signal strength enough, and the headroom of those stages falls, leading to a voltage limiting compression effect.

hi dave, thanks for your input,
so if the clean to mean is a product of the sag from the dropping resistor, how can i get this to be repicated as near as possible in a single ended amp.
can i measure how much the voltage drops as i play???

[bigbeck]

Geoff, I think you're beating a dead horse as SE doesn't sag. Perhaps it's time for you to build an Express?

You know,an amp doesn't have to sound like a Trainwreck to have excellent tone, but it helps!

[Firestorm]

The clean to mean has nothing to do with exceeding the rated limits of the PT. No way in hell are two EL34s ever going to average 300mA. Even if you could manage to pull that much current, it's not going to sound cool. It's going to turn your power transformer into a chimney and probably trip a circuit breaker.

The clean to mean has just about everything to do with that 1K dropping resistor in the power supply between the plate and screen nodes. When you hit the power tubes with a large signal the screen current spikes, causing a voltage drop across that resistor that propagates to the preamp, changing the operating point of the PI and 3rd clipping stage and causing them to compress/limit more.

I'm sure daveg is right about the importance of the 1K in the "choke" position. But I beg to differ about the impact of current limits imposed by the PT. You can't really run a power transformer above the rated limits of its HT winding. Plates and screens don't "pull" current quite the same way heaters do, so you're not going to turn it into a chimney, or trip the breaker (we're still talking milliamps here, not amps). If the tranny is wound to provide 300mA before the copper loss kicks in, it won't really try to provide any more (unless there's a short). This is why you can put 6L6s in a Deluxe Reverb and still only get about 20 watts. As long as the PT isn't junk, it shouldn't overheat.

As to how much current a 2 x EL34 amp can use: according to Philips' typical operating values, at max signal, 400V B+, two EL34s will expend 200mA on the plates and another 50mA on the screens. A 12AX7 will use about 16mA (both sections), so in an Express, that's another 40mA. We're at 290mA and haven't accounted for the impact of the 1K yet.

So all I'm saying is if you can get into that territory with the volume controls on less than full, turning up won't provide more output power, but will increase signal level to the preamps.

There may be something to this. Or I may just be full of crap

[geoffrod]

Geoff, I think you're beating a dead horse as SE doesn't sag. Perhaps it's time for you to build an Express?

You know,an amp doesn't have to sound like a Trainwreck to have excellent tone, but it helps!

hey bigbeck
i know i wont get a trainwreck in a se amp but close to the tone would be good
and i was just curious if i could get the clean to mean without too much volume increase.
and sh!t yeah i will be building a trainwreck express, as a matter of fact i want to build 2


as for voltage sagging, if i am measuring the other side of the dropping resistor while i play, with the 150ohm resistor in there there was no voltage drop when playing hard.
with the 1k resistor there was a slight drop of 10 - 15 volts
with a 1k5 resistor there is more drop when playing hard at full volume, 25-35v
not sure what this proves or does to the amp.
i was just interested if i could create some sort of a drop as i understand it this is what the real trainwreck is doing????

[dave g]

I'm sure daveg is right about the importance of the 1K in the "choke" position. But I beg to differ about the impact of current limits imposed by the PT. You can't really run a power transformer above the rated limits of its HT winding. Plates and screens don't "pull" current quite the same way heaters do, so you're not going to turn it into a chimney, or trip the breaker (we're still talking milliamps here, not amps). If the tranny is wound to provide 300mA before the copper loss kicks in, it won't really try to provide any more (unless there's a short). This is why you can put 6L6s in a Deluxe Reverb and still only get about 20 watts. As long as the PT isn't junk, it shouldn't overheat.

As to how much current a 2 x EL34 amp can use: according to Philips' typical operating values, at max signal, 400V B+, two EL34s will expend 200mA on the plates and another 50mA on the screens. A 12AX7 will use about 16mA (both sections), so in an Express, that's another 40mA. We're at 290mA and haven't accounted for the impact of the 1K yet.

So all I'm saying is if you can get into that territory with the volume controls on less than full, turning up won't provide more output power, but will increase signal level to the preamps.

There may be something to this. Or I may just be full of crap

16mA in a 12AX7???? LOL where are you buying your tubes from? Have you looked at a load line for a 12AX7? You realize that 8mA through a 100k plate resistor is an 800 volt drop, right? So if your supply voltage is 300 volts, then you can expect a quiescent plate voltage of -500 volts

As far as the resistive losses of copper in the PT, please explain to me how a resistive loss can "kick in" (assuming we're not talking about the thermal dependance of resistivity). If you've uncovered evidence that Ohm's Law is nonlinear, I'd love to see it. I think every scientific journal on the planet would too!

The current rating for a HT winding is simple. If you exceed it, you run a good chance of damaging the winding, likely causing a short, which will cause the primary current to spike. Which, if your fuse doesn't blow fast enough, would cause a breaker to trip. Oh, and the melting insulation will give of some of that smoke I was talking about.

If you have a Deluxe Reverb with 450 volts at B+1 and you installed two 6L6s, biasing them at 40mA each, you'd be idling them at 18 watts each. That's totally realistic. So explain to me how you'd only get 20 watts out total under full volume

I can tell you what would limit you, though - the output transformer!!! That thing is TINY! It's simply incapable of passing that much power. It will pass more than 20 watts though. Maybe 30? Depends. At what bandwidth? I don't really buy your figure of 20 watts with a pair of 6L6s anyway.

I'm not sure what EL34 datasheet you misinterpreted, but yes it's possible for an EL34 to PEAK 100mA. But to AVERAGE 100mA? At 400 volts? Sorry but try dissipating 40 watts on the plate and you're going to be seeing red.

Geoff, you are not seeing a large drop because like I said, your output stage is biased in class A and therefore pulling a large current at idle. Turn up the volume, and the average screen current draw won't change much, so the voltage across the 1k resistor won't change much.

[geoffrod]

Geoff, you are not seeing a large drop because like I said, your output stage is biased in class A and therefore pulling a large current at idle. Turn up the volume, and the average screen current draw won't change much, so the voltage across the 1k resistor won't change much.

So dave what am i seeing with the 1k5 resistor when i heit the strings hard at high volume?

[dave g]

Depends. I said the voltage across the resistor wouldn't change "much". You're probably not biased into "perfect" class A operation, and are clipping at grid conduction before absolute cutoff. This means that when the output tube clips, the resulting waveform will spend a disproportionately large portion of its cycle at the peak current draw, compared to the minimum current draw (i.e., 0). So that's going to result in a slightly larger time averaged current draw (which includes both the plates and screens), resulting in a slightly larger drop across the resistor. The effect is still there, but as I said before it's not going to be nearly as dramatic as with a PP amp, where the difference between the idle current draw and the current draw at full power is going to be far more significant.

[Firestorm]

as for voltage sagging, if i am measuring the other side of the dropping resistor while i play, with the 150ohm resistor in there there was no voltage drop when playing hard.
with the 1k resistor there was a slight drop of 10 - 15 volts
with a 1k5 resistor there is more drop when playing hard at full volume, 25-35v

You may be onto something. So how did it sound with the 1K or the 1K5? Try measuring the voltage on the "other side" of the resistor when you are playing clean, right before the onset of any real dirt and compare that to the voltages you measured while playing hard.

In the Express, the difference in screen current from idle to maximum signal is probably on the order of 25-35mA with maybe a 20-25mA difference between clean and mean. You may be closing in on that (SE amps do increase their current demands with signal, just not as dramatically as PP).

Since the voltage drop won't be as dramatic, one thing you can try to emulate what the Express does is increase the cathode resistor on V2 (the "clipping" stage). Not a lot, just dabble a bit: 12K - 15K maybe. That will reduce the available output swing voltage, which is what the voltage drop through the 1K does too. If that seems like it helps, you could try adding a little resistance to the cathode of your final "half a PI" stage, too.

[Firestorm]

16mA in a 12AX7???? LOL where are you buying your tubes from? Have you looked at a load line for a 12AX7? You realize that 8mA through a 100k plate resistor is an 800 volt drop, right? So if your supply voltage is 300 volts, then you can expect a quiescent plate voltage of -500 volts

Ah, but what you fail to realize is that it is possible for me to type without using my brain. Decimal point. Oops.

As far as the resistive losses of copper in the PT, please explain to me how a resistive loss can "kick in" (assuming we're not talking about the thermal dependance of resistivity). If you've uncovered evidence that Ohm's Law is nonlinear, I'd love to see it. I think every scientific journal on the planet would too!

It's precisely because Ohm's Law is linear that well-made power transformers don't go up in smoke that easily. A transformer wound for a given voltage will deliver that voltage up to the rated current. When the current demand exceeds that, the transformer begins dropping voltage due to the conductivity limitations of the wire gauge used. Reduce the voltage to an output tube and the current drops proportionally and you reach a warm, but not firey equilibrium.

I'm not sure what EL34 datasheet you misinterpreted, but yes it's possible for an EL34 to PEAK 100mA. But to AVERAGE 100mA? At 400 volts? Sorry but try dissipating 40 watts on the plate and you're going to be seeing red.

Take your pick; they all have figures in roughly the same range. Not all of the current is dissipated as heat in the plate (otherwise there would be no output). It's a 25W tube right? Run a pair of them at the max plate voltage and you get output of 100 watts.

[dave g]

Ah, but what you fail to realize is that it is possible for me to type without using my brain. Decimal point. Oops.

.

Sure.

It's precisely because Ohm's Law is linear that well-made power transformers don't go up in smoke that easily. A transformer wound for a given voltage will deliver that voltage up to the rated current. When the current demand exceeds that, the transformer begins dropping voltage due to the conductivity limitations of the wire gauge used. Reduce the voltage to an output tube and the current drops proportionally and you reach a warm, but not firey equilibrium.

Pretty sure you just contradicted yourself. You claim that the transformer will deliver whatever current you want (below the rated current) at the specified voltage. This implies that the resistance of the windings is 0. You then say that the transformer windings begin dropping voltage after the current draw exceeds the maximum rating. This implies that the resistance of the windings is nonzero. Is there some magick at work here? You just claimed that the resistivity of the wire is a nonlinear function of the current passing through it...which (assuming not enough current flows to drastically increase the temperature) it isn't. Therefore, according to you, Ohm's Law is still nonlinear.

Take your pick; they all have figures in roughly the same range. Not all of the current is dissipated as heat in the plate (otherwise there would be no output). It's a 25W tube right? Run a pair of them at the max plate voltage and you get output of 100 watts.

...Huh?

Do you know how impedance matching works? The best you can ever do is to drop half your power across the source impedance. That's why plate dissipation figures tell you what kind of power output you can expect. I'm not even sure what you're trying to say here though. I think you're still missing the point that peak current does not equal RMS current.

[geoffrod]

No PI. Yes on the MV, although it's not really needed as the amp sounds best with the MV and gain turned almost all the way up to 3:00.

I also used a modified James/Bax TB stack. I didn't like the one that's on the schematic.

If I were to build another,I think I'd use the tone stack from the current Express schematic. If you don't want/like a mid control,just use a 15K resistor to ground in place of the mid pot. I did that on my Express and am very happy with it.

I used a 5K 15 watt Edcor OT. Also 500K gain pot and a .022 coupling cap on the first stage going into the tone stack.

The B+ voltages are 365 - 360 - 350 - 322 - 320. The dropping resistor on the first filter node is 100R 2 watt. That's pretty much the only changes I made.

hey guys,

thought i would report back with where i am up to with this.
BigBeck, i would have to concure, i have gone back to basics at the moment, i have the derailment circuit into a single 6V6 power, only difference with the derailment is i have the current Express tone stack, i have no MV and only 3 stages in the preamp.
with the limited playing i have done with this so far the sweet spot is opend right up, near flat stick, sounded great,
i will report back my voltages, and a revised schematic as well.

cheers
Geoff

[geoffrod]

I am in the process of adding the PI back into this build, unfortunately it isn't going to be pretty

i am attaching some pics of the cab i nade on the weekend for this amp, but be warned i am no carpenter or cabinet maker, simple design, simple finish, not too bad a result i think.

cheers
Geoff

[geoffrod]

Hi All,
these are the voltages i have for my SE-Wreck at the moment, i have put the PI back into the mix, but am a little concerned that the voltages might be a bit skewif on it. Any comments and suggestions welcome,

B+1 = 368v
B+2 = 354v
B+3 = 313v
B+4 = 295v

V1b;
pin6 = 183v
pin8 = 1.65v
V1a;
pin1 = 214v
pin3 = 2.1v
V2a;
pin1 = 276v
pin3 = 3.4v
PI V3a
pin1 = 204v
pin6 = 214v
pin7 = 53v
pin2 = 52v
pin3&8 = 79v

6V6 V4;
pin3 = 350v
pin4 = 348v
pin6 = 354v
pin1&8 = 21v

regards
geoff