36 Watt with EL-34

[hanskwijhe]

Is there someone with experience with this:

Making a 36 W amp with 2 EL-34's in PP.

I am wondering how this would sound.
As an electrical engineer I think this is a better approach than 4 EL-84's in PP. Since the 2+2 EL 84's are never exactly equal and one of each pair will be stressed more than the other. If you take two EL-34's you do not have the matching problem and they are not stressed to the max to deliver 36 W output and it is also a bit cheaper probably.

Feed back welcome !

[morcey2]

Do a search for '18watt heavyII' (I think). It's a 18-watt derived 6L6 amp.

As for the matching/stress issue, you're right from an engineering approach. Now knock it off. Take the engineer hat off for a while.


Mismatched power tubes aren't a bad thing in a guitar amp, as long as they're not insanely mismatched. In a PP amp like the 36-watt, the mismatch leads to a warmer, more pleasant sound. In a perfectly matched PP output stage, all even harmonics are canceled out and only odds (and the fundamental) remain. Mismatching allows many of these even harmonics through and creates a very rich distorted tone that these amps are known for.

Another part of the sound is pushing them to their limits and beyond. It's another piece of the psychotic puzzle we call tone.

I think you're approaching this from the perspective of the amp being a sound-reproduction device. For most hi-fi and audio amps, that's true. When you get into guitar amps, it's a totally different ball-game. Most non-guitar amps are designed to color the sound as little as possible (0.0001% THD @ 5,000,000W etc.). With guitar amps, we don't just want the sound colored, we want the amp to grab the sound, give it 20-coats with a Wagner power-painter, strip half of it back off, stomp on it a few times, roll it in spackle, sand it smooth, and throw it out into the street. It's as much an instrument as the guitar is.

Now, as for an EL34 based 36-watt, I think it would sound awesome. Crunchier than EL84's and not as creamy as 6V6's. They all sound great to me.

Matt.

[ur12]

Is there someone with experience with this:

Making a 36 W amp with 2 EL-34's in PP.

I am wondering how this would sound.
As an electrical engineer I think this is a better approach than 4 EL-84's in PP. Since the 2+2 EL 84's are never exactly equal and one of each pair will be stressed more than the other. If you take two EL-34's you do not have the matching problem and they are not stressed to the max to deliver 36 W output and it is also a bit cheaper probably.

Feed back welcome !

The Trainwreck Express uses 2 EL34s and it puts out 36 watts. It runs about 390-410 volts on the plates and you can also swap in 6V6s without rebiasing for even less watts. I have built a few of these and they sound phenominal but are still really loud.

[sub]

Since the 2+2 EL 84's are never exactly equal and one of each pair will be stressed more than the other.

This is what we called: sweet, warm sound

Anyway i've a old 50W Marshall OT, a low voltage 50W PT (310-0-310) and four brand new SED EL34......

[hanskwijhe]

Do a search for '18watt heavyII' (I think). It's a 18-watt derived 6L6 amp.

As for the matching/stress issue, you're right from an engineering approach. Now knock it off. Take the engineer hat off for a while.


Mismatched power tubes aren't a bad thing in a guitar amp, as long as they're not insanely mismatched. In a PP amp like the 36-watt, the mismatch leads to a warmer, more pleasant sound. In a perfectly matched PP output stage, all even harmonics are canceled out and only odds (and the fundamental) remain. Mismatching allows many of these even harmonics through and creates a very rich distorted tone that these amps are known for.

Another part of the sound is pushing them to their limits and beyond. It's another piece of the psychotic puzzle we call tone.

I think you're approaching this from the perspective of the amp being a sound-reproduction device. For most hi-fi and audio amps, that's true. When you get into guitar amps, it's a totally different ball-game. Most non-guitar amps are designed to color the sound as little as possible (0.0001% THD @ 5,000,000W etc.). With guitar amps, we don't just want the sound colored, we want the amp to grab the sound, give it 20-coats with a Wagner power-painter, strip half of it back off, stomp on it a few times, roll it in spackle, sand it smooth, and throw it out into the street. It's as much an instrument as the guitar is.

Now, as for an EL34 based 36-watt, I think it would sound awesome. Crunchier than EL84's and not as creamy as 6V6's. They all sound great to me.

Matt.

OK Matt, Thanks.
One question about the mismatch: I think we have to make a difference between the kind of mismatch. You are possibly referring to the mismatch between the upper and lower tubes but I meant the mismatch between the 2 upper tubes or the 2 lower tubes. Since they are not equal one of them will be stressed more than the other. Do you think this type of mismatch will also give a different sound? I understand that when the upper side and lower side are exactly matched the even harmonics will be cancelled out by the OT. That is why in most cases you see different resistor values in the PI (e.g. 82k versus 100k)

[hanskwijhe]

Is there someone with experience with this:

Making a 36 W amp with 2 EL-34's in PP.

I am wondering how this would sound.
As an electrical engineer I think this is a better approach than 4 EL-84's in PP. Since the 2+2 EL 84's are never exactly equal and one of each pair will be stressed more than the other. If you take two EL-34's you do not have the matching problem and they are not stressed to the max to deliver 36 W output and it is also a bit cheaper probably.

Feed back welcome !

The Trainwreck Express uses 2 EL34s and it puts out 36 watts. It runs about 390-410 volts on the plates and you can also swap in 6V6s without rebiasing for even less watts. I have built a few of these and they sound phenominal but are still really loud.

Thanks for this post. I think I will go for a trainwreck express !!

[ur12]

Thanks for this post. I think I will go for a trainwreck express !!

If you are serious about building an Express you can find all of the info you need at the Ampgarage.com. It also might be cool to take Plexi's 36watt and convert it over to 2 EL34s instead of the EL84 Quad. It would definately be a different amp.

[soundchaser59]

Is it true what Ken Fischer said? That the TW Express clones will never sound anything like a real Trainwreck built by Ken himself?

[ur12]

Is it true what Ken Fischer said? That the TW Express clones will never sound anything like a real Trainwreck built by Ken himself?

That is a touchy subject with some people. (Kind of like talking about non 18watt amps on this board ) I do know of reports of people who own certain clones and also own certain real wrecks that like the clones better than the real thing. So I guess it is up to the player or listener to make up his own mind. Glen Kuykendal has done a few comparisons of his real TW and a couple of other amps (Including his two clones) and while every amp sounds a little different thay are close enough to leave his real one at home and use his clones on the road.

You can also listen to some clips of the TW Express Memorial amp at
http://www.ampbuildersguild.com/gallery2007Q1.htm Just scroll down to the bottom of the page to listen to the clips by Glen.

I would invite you to take a look at the Ampgarage.com for all things regarding Trainwrecks and Dumbles.

[Merlinb]

As an electrical engineer I think this is a better approach than 4 EL-84's in PP. Since the 2+2 EL 84's are never exactly equal and one of each pair will be stressed more than the other.

On the other hand, provided they are run within ratings, it is better from an electrical engineering point of view to use multiple devices in parallel than one, single, bigger device. This is because the crossover region becomes 'smeared' due to small differences beween devices (even matched ones) and is particularly useful in Class B SS amps where crossover distortion is the main problem!
Just to add fuel to the fire!

[stevesuk]

Tempted to defect to the other side and build an EL34 36 Watt just for the hell of it.
I have a donor just begging for it.

I am also tempted to leave the Voltmeter and Ammeter on the front plate, along with the other controls suitably labelled.

It already had two Mullard EL34s in it and two small nine pins, I even have the transformer.

I do love a challenge and it just means deciding on a 36W 4 X EL84 (of which I have several) or no drilling and a 2 X EL34 36W Lite.

Now I just have to find the idiot who cut the transformer out complete with heater wiring all loomed and connected. Uh, that was me !!

Pics attached. (good job I am retiring in 3 days time !!!!!)







Steve UK

[Phil_S]

With a pair of EL34's in a TMB, you are basically building a JTM45 without a choke.

I built a TWX recently. This is a finicky amp and is not for the faint of heart. It is easier to build a Marshall type.

If those meters are working properly, you might put them in the circuit so you can see what's happening with plate voltage and current. It is probably adequate to set bias voltage.

[Alexo]

I've been thinking about building a 2xEL34 EF86 lite for a bit now. Basically because I already have a good pt for a pair of EL34's, and more importantly, I've already done the hardest part of the build - cutting out that big hole from a steel chassis for a lay-down PT!

So the main issue with this concept, to me, is that EL34's require about a 50 to 60 volt signal, peak-to-peak, to be driven into distortion (with 450 Va, 40W output) while EL84's only require about 18VAC.

So I've been thinking, you need a little more gain in the pi; a pi tube that can swing a bigger signal than a 12AX7 with a lower plate resistance probably wouldn't hurt either.

How about building a stock 18 watt, up to and including the pi. Follow each side of the pi with a triode from a 12AU7, then hit the EL34's with the boosted pi signal. 12AU7's don't clip too easily, don't have too much gain and can source a lot of current. Plus I know I'm not the only guy with a drawer full of poor, lonely homeless 12AU7's. I'm always trying to build an amp that uses the darn things, and it never works out!


Any thoughts?

BTW, that chassis looks super-cool, and happy retirement!

[stevesuk]

I got the impression a TWK was 'finicky' from reading the posts.

I had enough of that when we built the first EL84 36 Watters over 4 years ago.

Now the 36 Watters are more stable with the shielded lead etc. mods it looks like I will try and squeeze a 36 Watter into that donor chassis.

But for the first time with just Volume and Tone (Lite) controls and probably an MV.

A pity to waste the Octal holes that are already there, steel chassis as well, looks like I will be making some hole reducers to mount the B9As.

Definitely stand up transformers, it is not fun cutting steel for laydowns.

Now where are my drawings and notes for the 36 Watters ?

I wondered how long ago I built possibly the second ever 36 watter ?

With lots of help from Graydon and Richie.

Graydon finished his on March 14th and mine was finished on Saturday March 20th 2004 !!

Fired up my 36 watt at last.

First attempt - 1 missing link.

Second attempt - Loud almost violent buzzing, not present on standby, low volume , but at least it works , sort of.

Had enough for today, back on it tomorrow.

Have an idea to take the 820K grid resistors off the valve base, used pin 1 as a terminal, it seemed to be much less violent without the two middle EL84s in. Bit of redplating as well, that will be the 120 ohm 5 watters I thought I would attempt to use. Replace those as well.

Used a 1650N Hammond OT which seems to be either a 4 or 8 ohm or 16 ohm only depending how it is wired ? never used a Hammond before.

I tried a choke temporarily wired in across the 32 + 32 cap, no difference.

Will try shields on the pre amp tubes, change the grid stoppers to terminals and generally poke about tomorrow. By the way it is based on a JA/Richie MV TMB. Digital Camera is also broken.

Steve Uk

[klingo]

Alexo you're right more PI voltage swing is not bad with bigger bottles.
the long tail PI loose output voltage swing because of the big tail resistor, a choke here is something you can try (big AC impedance/low DC resistance) link: http://boozhoundlabs.com/bhl-15/
also loosing voltage swing because of high ouput impedance(~twice the non decoupled single triode Zout) around 200K(EDIT i redo the math and found ~70k ) if Rp=100K/Rk=47K. considering that many octals OT seem happier with 100Ks grid load resistors, then the plate load is half the Zout...
paraphase have better voltage swing and a lower Zout
so what do you think of a floating paraphase with a DC coupled 12au7 cathode follower http://www.novotone.be/_site/projets/Projet01/Fig01.pdf

strangely i was just putting some of these ideas on paper this morning before i read your post. i attach the schem because it has a liteII input stage
not tested, may include big mistake/poor design etc...

edit: 2k may be better than 2k2 for PI cathode resistors and bigger cathode caps too. change the 22nf/220k to 47nf/100k.

[morcey2]

This is totally off the top of my head, but couldn't you create a negative voltage off the HT winding and run the PI tail to that? Just something like a -50v bias supply, but only running to the tail of the PI. I know that it's done that way when someone runs a constant-current source instead of a tail resistor. The grids are at a virtual ground anyway, so they wouldn't be a problem.

Just trying to think on a Monday morning, so my brain may not be working yet.

Matt

[klingo]

This is totally off the top of my head, but couldn't you create a negative voltage off the HT winding and run the PI tail to that? Just something like a -50v bias supply, but only running to the tail of the PI. I know that it's done that way when someone runs a constant-current source instead of a tail resistor. The grids are at a virtual ground anyway, so they wouldn't be a problem.

Just trying to think on a Monday morning, so my brain may not be working yet.

Matt

yes, it have been done in the 50's in HewlettPackard scope as vertical amplifier with an 12at7/-100v/+300v iirc. but in sound machine i just don't know

[Alexo]

I think the DC-coupled AU7 CF is a great idea. I didn't know that about the high output impedance of the LTP! Also, it looks like 12AU7's only swing a marginal amount further than 12AX7's, so they would be put to much better use as cathode followers. Maybe 12BH7's?

I don't see why you couldn't apply a negative voltage to the tail, as long as you stay within 100 volts of the heater reference. Although if you're using EL34's, you should have a plate voltage around 400 to 450 volts, which gives you plenty extra voltage for the pi already.

[klingo]

I have some 12au7 lying around too...that said i have EF86/TB tone/vox PI/KT77 ultralinear/500volt amp, (sort of RT77) that i can push into real distortion using pedals or extra gain stage and that sound great...just to say that this 2x100K-Rp/47K-Rk long tail PI can drive a lot of guitar output stage.
from what i read here http://pagesperso-orange.fr/franck.doucet/index.htm output impedance with ecc83 is 2.2 times more than the single triode. 7,5 times more using ecc82. this is in a differential amplifier and assuming LTP is a differential amplifier with one input grounded.

[zaphod_phil]

For the EL34s, all you need to do is use something like a 15k tail resistor for the LTP PI. I did this once with an amp using a Matchless Clubman-ish power stage and it worked great. The 47k tail resistor was unable to fully drive the EL34s, and it the amp lacked volume and overdrive until the value got reduced. If you think about it, that's the same effective value used in the bigger EL34 Marshall amps. I think an EL34 version of a 36W would be a pretty nice sounding amp.