Anyone built an ef86 dumble?

[briane]

I was wondering if anyone had built a clone with an ef86 in v1?

I'm considering this myself....But I am thinking of a more clean only amp, so likely I'll omit the OD stage, and sub that in with a built in dumbleator....I really like the extra harmonics the dumbleator adds to my other dumble clones (both d'lites, off the normster layout)....My thought is the ef86 might add some extra jangly cleans, but not sure how involved the tweaking will be...I'm just going to assume that will be part of the process.

I've built another ef86 amp, a simple amp driving el84's in PP. It has some great cleans, very nice amp. I keep thinking the ef86 combined with the dumble clone topology might really be the bees nees. My current thought is to take the pre from the ef86, and just feed that into the dumbleator then to the PI and power tubes. My ef86 layout pre is very close to the ceriatone 18watt ef86 pre (currently). From that standpoint its a pretty simple idea. This will be a 6l6 amp and have either a 4k or 2k OT.

So, I know the amp will work, but does anyone have any idea how it will sound? If it will really have even more cleans than the other clones? Or am I barking up the wrong tree?

Thanks in advance,
brian

[Darkbluemurder]

Check for threads started by Svempan. He discussed this in great detail.

[Svempan]

Check for threads started by Svempan. He discussed this in great detail.

Well, I don't think I discussed it in detail... "In detail" would be a much deeper discussion...

The thread in question is this: http://ampgarage.com/forum/viewtopic.php?t=2804

The current schematic for the amp is downwards in this thread

A couple of notes since the other thread:

After I've a played the amp on a couple of gigs with a paralleled 12AY7 as the input tube, I’ve now gone back to a triode connected EF86 there. I do think that the sound is better with that. I really wanted to try the sound with the paralleled double triode, but in my ears I don’t agree with my amp tech friend that a triode connected EF86 sounds a little “life less” in guitar amps. I think the sound is more “lively” with that than with a paralleled double triode 12AY7 or 12AX7.

The amp now has a reverb transformer that matches the EL85 and the reverb tank very well, both impedance wise and power wise.

I really love the sound of this amp! It does what I want it to do! (By the way: I almost always play it together with my Fender ’64 Vibroverb copy).

My next step with this amp will probably(!) be to make it back as a piggy back. The amp back in its original top cabinet, and a separate 4x10 speaker. The output transformer will then probably be a Mercury Magnetics Blackface Twin Reverb transformer (FBFTO-85) I’ve got or an output transformer from a Dynaco Mark III I’ve also got (A-431).

Please ask questions if you want and I will try to answer as good as possible.

Take care,

Sven-Johan

[Svempan]

A couple of more things:

The sound of the amp is... clean!... In my ears, I really do hear the guitar's true sound, and what I do with the guitar is unusually audible.

It seems as the EF86 makes the sound to be as shimmering in the treble as I wanted.

The five position rotary switches "Treble Shift" and "Middle Shift" in my amp is useful, but of course you can omit them (if you really know which values you want on the capacitors in question) or use three way switches.

The 6550's and the way I've done the phase inverter is, by the way, a very important part of the sound.
If you want a clean sound, why not consider 6550 or KT88 instead of 6L6?
(I'm testing a pair of old Genalex GEC KT88 in my amp now and will compare the sound between those and the old GE 6550 I've had in the amp quite a while).

Sven-Johan

[briane]

thanks Svempan, I have seen your other thread, somehow I thought I was doing something different, but I now see this is not the case, at least not so much.

I originally was going towards 6550, and have some tubes for just that, I was leaning away only because I thought i might make for an easier build working from a well known design.

Could you describe the ef86 pentode/ triode mode related to your tubes? Is it v1 running pentode and v2 running triode?

Could you elaborate on the parallelling of the 12ay7 with the ef86? Was this paralleled with v1?

What I had originally though of was an ef86 v1 running a switable cathode follower. Do you have any comments about that idea if I used a 1/2 12ax7 right after the first ef86 (v1). I would probably take the cf design from the dumbleator since I like that very well.

Do you think this would work well without the reverb? Whats the specs on the tank that you liked? How about the reverb transformer? I might still put reverb in.

And also, I want to put in a 12ax7 dumbleator, which would go right before your master volume (I am thinking), any comments on how that might (or might not) work out?

Also, what primary imedance for an OT would you suggest?

again thanks, This is fairly close to the direction I am heading in.

brian

[Svempan]

...somehow I thought I was doing something different, but I now see this is not the case, at least not so much...

The preamp in my amp is based on the Steel String Singer (or if it is a Dumbleland) schematic with two EF86 that you can find on this forum. Look in this thread: http://ampgarage.com/forum/viewtopic.php?t=412

My amp’s preamp is almost the same, but I have a different value on the slope resistor and I have the two rotary switches Treble Shift and Middle Shift (and some other small differences). In fact I usually have the rotary switches in the positions so that the treble and middle capacitors get the same values as on the EF86 Steel String Singer schematic...

Regarding the slope resistor: I think(!) I like the sound better with 82k rather than 100k or 150k...

I originally was going towards 6550, and have some tubes for just that, I was leaning away only because I thought i might make for an easier build working from a well known design.

According to my opinion, building an amp with 6550 or KT88 would be just as easy (or difficult) as an amp with 6L6.

The 6550/KT88 draws much more heater current than 6L6 so the power transformer must be designed for that.

As I see it you can have the same load impedance for 6550 or KT88 in a push pull amp as for 6L6. The 6550/KT88 might give more output power, so the output transformer must be able to take that.

I think the sound sounds cleaner with 6550s or KT88s than with 6L6s or EL34s. It’s clean but still have character and are not too Hi-Fi-ish.

Could you describe the ef86 pentode/ triode mode related to your tubes? Is it v1 running pentode and v2 running triode?

V1 is triode connected and V2 is pentode connected.

As you can see on the schematics the screen grid is connected to the plate on V1. That makes the screen grid more as an added part to the plate than a screen grid. On V2 the screen grid has its own connection to the B+ thru a screen grid resistor.
By the way: In my opinion it’s not totally correct to call connecting a pentode in the way as V1 as a triode. The tube still has five active parts instead of the triode’s three. Connecting the screen grid to the plate makes the tube act more like triode. Therefore this type of connection is technically and commercially accepted to be called “triode connected”. But the screen grid and suppressor grid probably makes a difference in sound even in a triode connected pentode.

You can find information for each tube type and how tube works, for example here:
http://www.ibiblio.org/obp/electricCirc ... MI_13.html
http://www.qsl.net/dl7avf/roehren/eintro.html#B.4
http://www.jacmusic.com/html/articles/e ... works.html
Or on several other web sites.

[Off topic: I’m very much a Fender blackface lover but I fell in love with the sound of EF86s several years ago. After servicing and tweaking a couple of VOX AC10 (that has EF86 in one of its channels) I tried a version of the EF86 channel in Matchless C30 in one of my blackface Fender amps. I modified the Normal channel. Used the six way rotary “Bass Control” as on the C30 and used a simple Treble cut control. There where no negative feedback over the power amp, the phase inverter was as in the C30 (or in a VOX AC15/AC30) and the channels was mixed together in the phase inverter as on the Matchless/VOX. This amp was a really good distortion amp! I do understand why people are building there own versions of Marshall 18 Watt amps with one EF86 channel].

Could you elaborate on the parallelling of the 12ay7 with the ef86? Was this paralleled with v1?

Look at page 1 on the schematic for my amp version 1.5 in the first thread about my amp. Or here: http://ampgarage.com/forum/files/clean_ ... es_391.pdf

The 6072/12AY7 tube is a double triode in the same glass bulb (just as 12AX7, 12AT7, 12AU7 and several other types). The most usual way to use these are as two separate triodes. But you can also parallel the two triodes (connect the two plates together, the two cathodes together and the two control grids together) and use it as one single triode. In that way the amplification stage in question gets more head room than with a single triode. The amplification factor doesn't get higher, but there is more head room. (The components around the tube should be different with a paralleled triode than with a single triode of the same type in the same position).

What I had originally though of was an ef86 v1 running a switable cathode follower. Do you have any comments about that idea if I used a 1/2 12ax7 right after the first ef86 (v1). I would probably take the cf design from the dumbleator since I like that very well...

... And also, I want to put in a 12ax7 dumbleator, which would go right before your master volume (I am thinking), any comments on how that might (or might not) work out?...

Sorry, but I don’t think I have experimented enough with cathode followers (and not at all with the Dumblelator) to have any comments.

In an amp with the V1 triode connected as in my amp, the amplification factor is in the same area as with a 12AY7 or 12AX7 triode. I guess it would be OK to do as you suggest. I haven’t calculated the impedances and how they match, though.

Do you think this would work well without the reverb? Whats the specs on the tank that you liked? How about the reverb transformer? I might still put reverb in.

It would probably work just fine without the reverb (connecting the preamp directly to the power amp).

The reverb tank is a long Accutronics six spring tank. I call it a three spring tank but Accutronics calls it six spring. I haven’t the amp here now but I think the tank code is 9AB3C1B (more information here: http://www.accutronicsreverb.com/prodspec.htm ).

That tank has an input impedance of 10 ohms. The usual blackface Fender tank (4AB3C1B) has, by the way, 8 ohms input impedance. The reverb transformer in my Clean King is an output transformer from an old single ended Hi-Fi amp. The EL85 tube wants a load impedance of about 11 kohms. With the transformer and tank I use it gets about 10.2 kohms.

If I should use the more commercially available EL84 as reverb driver tube, I probably would use an output transformer for a single ended EL84 amp and 8 ohms speaker load. For example this one: http://store.triodestore.com/tfchxfwi48oh.html

Also, what primary imedance for an OT would you suggest?

I suppose you mean the output transformer in the power amp. Not the reverb transformer.

The load impedance of the output tubes is a big subject with many answers. I also think that you can start many of the answers with “It depends on...”.

Se my next post in this thread for a litle more information.

For a pair of 6L6, 6550 or KT88 in a push pull power amp with plate voltages of about 450 V I would suggest a load impedance for the tubes of about 4000 – 4500 ohms, and half of that with four tubes.

With the 2.7 ohms speaker impedance and the output transformer in my Clean King amp, the load impedance to the tubes is about 5.4 kohms. That’s probably not optimized for the best clean sound. As I’ve mentioned earlier I will do changes here. Probably go to a piggy back design with a separate 4x10 speaker. The output transformer will then (again, probably) be a Mercury Magnetics Blackface Twin Reverb transformer (FBFTO-85) I’ve got, or an output transformer from a Dynaco Mark III I’ve also got (A-431). With 8 ohms speaker load, the FBFTO-85 will give the tubes a 3.2 kohms load and the a-431 will give about 4.5 kohms.

-------------------------------------------------------------------------------------------------------------------


I am a tube amp tech that basically repairs, make services and modifies tube amplifiers for musical instruments. I have done that for quite many years. I absolutely don’t regard myself as an amp designer. So please don’t see my words as written in stone... They are also just my own personal opinions. You could probably see the subjects here with other eyes and have other opinions.

I also hope that my English isn't too bad.

Please feel free to comment or question my thoughts and opinions.

Take care,

Sven-Johan

[Svempan]

Regarding load impedance for the output tubes. Here's one interesting article on the subject:

[briane]

firstly thanks so much Svempan, this has really helped. Also many thanks to electron avalanche and the others contributing to the related threads.

I am very tempted to build very close to your schematic, predrilling for the extra tube sockets for the dumbleator and cf, then decide once I hear it what direction to go.

Just a couple more questions if you still have the energy left to answer.

You discuss how critical your PI values are to the clean sound. I admit the 2.2k sandwiched between the 1megs is a bit odd to me. Also the 33k and 44k plate resistors off v5 seem a bit low (though probably provide less plate swing, thus less amplification, and thus a cleaner sound). Could you comment on why these values? I would have chosen an 820 ohm instead of 2.2k, and standard plexi plate resistors of 82k/100k along with the plexi snubber.

the 220 uf cathode caps on v1 and v2 seem very large for a guitar amp. Why the large values? or was this an emperical thing and it just sounded more like you wanted?

I understand the SSS was to have a cathode follwer after the PI. Was there a specific reason for omitting this? I think I am going to try and work the 12bh7 into my build, running a stage on each PI output, as per electron avalanches advice.

again thanks, this has been very helpfull.

brian

[Svempan]

...I am very tempted to build very close to your schematic, predrilling for the extra tube sockets for the dumbleator and cf, then decide once I hear it what direction to go...

Interesting!

...You discuss how critical your PI values are to the clean sound. I admit the 2.2k sandwiched between the 1megs is a bit odd to me. Also the 33k and 44k plate resistors off v5 seem a bit low (though probably provide less plate swing, thus less amplification, and thus a cleaner sound). Could you comment on why these values? I would have chosen an 820 ohm instead of 2.2k, and standard plexi plate resistors of 82k/100k along with the plexi snubber...

The values chosen are because I use a 12AU7 as the phase inverter tube. Those values fit that tube better than the values you suggest, which are better suited for a 12AX7.

I wanted a 12AU7 because it can give more “power” to the output tubes, which the 6550s or KT88s, in my opinion, requires.

If you use a 12AX7 as phase inverter tube I guess it’s OK to use the conventional Marshall values for the components. If you use a 12AT7 I guess it’s OK to use values like in a blackface Fender, or maybe like in a silverface which might give a cleaner sound. In fact I haven’t compared which one of these two, or the one in my Clean King, gives the cleanest and best sound, so I can’t comment on that...

The phase inverter plate resistors in my amp was from the beginning one 47k for each triode. Now one of the plate resistors is a 33k resistor in series with a 20k trim potentiometer. The reason for that is to be able to balance the phase inverter for a even cleaner sound. I haven’t balanced it with an oscilloscope though, but just by listening when the sound is as cleanest. That was rather easy to do as the preamp and reverb circuit distorts later than the phase inverter and output tubes.

In fact the preamp and reverb circuit doesn’t distort until the volume is at about 9. That despite that I use rather heavy strings (013-017-022-...), play rather hard and have rather high output pickups (for being vintage Strat styles).

...the 220 uf cathode caps on v1 and v2 seem very large for a guitar amp. Why the large values? or was this an emperical thing and it just sounded more like you wanted?...

Hmmmm... I guess I don’t have a good answer here... I first used 47 uF, but after reading tips in some old data sheets for the EF86 I increased the values to 220 uF and put the plastic film capacitors in parallel with the electrolytics. I think(!) the sound is better now...

If the values are “very large for a guitar amp” is a questionable statement... ... Marshall JTM45, tweed 4x10 Fender Bassman 5F6-A and some other amps has used those values

On the other hand, EF86 is rather unusual in a guitar amp... Thinking as for a 12AX7 probably doesn’t suit a EF86...

... I understand the SSS was to have a cathode follwer after the PI. Was there a specific reason for omitting this? I think I am going to try and work the 12bh7 into my build, running a stage on each PI output, as per electron avalanches advice...

Good question!

The reason is that my Clean King is built on a Fender Bandmaster Reverb chassis and I still use the original power transformer. Because the 6550's (or KTT88's) high heater current, I want as few tubes in the amp as possible. That’s also the reason for using a EL85 tube as reverb driver instead of a EL84 (the heater current for the EL85 is even lower than for a 12AT7 or 12AX7).
The amp has one empty tube socket, so it would be easy to do cathode followers between the phase inverter and power tubes, but I don’t want the power transformer to blow...
One way to go with this amp would be to put a Twin Reverb power transformer in it. That wouldn’t require making the chassis holes too much bigger...

Maybe I will make a real Clean King Amp from scratch, with a chassis and transformers designed more for my clean sound goal. Then I probably would use four 6550 or KT88 output tubes, a separate power supply for the output tube’s screen grids, cathode followers between the phase inverter and output tubes, EL84 as reverb driver and so on...

-------------------------------------------------------------------------------------------------------------------


If you build an amp with some influences from my Clean King amp, it would be really nice if you tell me what you think about it.
Again; please feel free to comment or question my thoughts and opinions.

Take care,

Sven-Johan

[benoit]

I saw yesterday the website of a guy who has done a lot of tube pro audio/hi-fi/guitar amp design, and one of his projects was some champy type thing. Before he built it into a chassis, he just had it all mounted to a board: trannies, tubes, everything. He had some kind of special tube sockets that were much bigger than usual and had contacts on either side with screw terminals. Components were mounted on something similar to those boards in old ac30s (I forget what they're called). Everything was connected with alligator clips. What I want to know is, is this...

a. safe?

b. useful?

c. likely to get you in trouble with interference from other signals?

[briane]

If you build an amp with some influences from my Clean King amp, it would be really nice if you tell me what you think about it.
Again; please feel free to comment or question my thoughts and opinions.

of course!

Thanks for the methods behind the designs, it helps a lot...Did not mean to say anything is too unusual, just different, and its obvious this circuit is a bit different, even if its just by running 2 ef86's.

I thank you, this should be enough info for me to go forward.

this is definetly in the works....it might take a little while, as I am adjusting some things a bit....and still deciding...its definetly going to be an experiment, within some closly defined component choices...but I do want to do some more ef86 builds.

Well that and a true SVT, and I need to get a couple of cabs built for the latest builds.

I do have parts, I will use a weber 22756 6A/400ma PT and 2k/100watt bassmaster iron since I have that around.

And the SVT cathode follower after the PI might be simplified, and made switchable.

and to benoit: no, yes, yes. its for prototypeing, but due to those considerations I consider a full build the best prototype.

[Svempan]

Because of this thread, today’s rehearsal (and recording) with my band, made me more focused than usual on my sound. Especially the sound from my Clean King Amp (Bandmaster Reverb).

Here are some of my thoughts:

I must admit that the sound with the Genalex GEC KT88 is even better than with the GE 6550 I’ve used for a while. It’s tough, but if I will continue to think that, it’s an expensive experience...

I really like having the opportunities that the two rotary switches give. The difference in sound between each position isn’t very, very big, but is enough to fine adjust the amp’s original sound. Just as I’ve thought on a couple of gigs when the sound in the particular rooms has been different/strange.

For my purposes, the combination of my pedal board, the Clean King Amp and the blackface Vibroverb Amp (with LS-mod on the reverb, as described in another thread) is very good.

Another thing I also noticed again today, is that different tubes/labels of EF86s give quite a different sound. More than I would say that different tubes/labels of, for example, 12AX7/7025/ECC83 gives.

To simulate a sealed cabinet speaker I also tested to put the amp against the wall, with a soft…thing… (sorry I don’t know the English word for the soft, thick thing you have in a bed to sleep on) between the open back amp/speaker and the wall (just for 1-2 minutes not to overheat the amp). The sound then made me rather convinced that I shall go for a sealed 4x10 or 2x12 speaker. (Hmmm... and I think I'm an open back fanatic...)

I most certanly do love my Clean King Amp!

Thank you,

Sven-Johan

[briane]

Well Svempan, I did end up building a very similar clone to your design, aside from a number of changes and tweaks, so of which are not fully explored as yet. The amp has only been running fully healthy for 3 days, but I did have an entire day of tweaking and listening last sunday.

Like I say, very similar circuit, 2 x ef86, 1st triode, second pentode, basic dumbelator 12ax7 with mods, into 2x 6550, no reverb. I admit I might like the kt88 better at this point, as it is a truly clean amp, I kept wanting more loud. Kt88 also had a bit more high end. Funny the psychological effect of no distortion, I think my brain felt it was less loud, even though items were rattling off the table near me. I think it just takes more clean to sound the same loudness as a distortion amp.

Very nice harmonics, but I really liked what the dumbelator added, but initially it had to much distorttion. I modded in a dlite style OD ground drop (220k to a 100k pot), and that works pretty well. The dumbelator adds a certain bite and bark to the initial attack that I felt was lacking without. By adjusting the drive resitors in my little mode there, I will be able to get much more volume, as I have seen in previous tests...I just took it a little to low for my tastes, and now need to adjust the OD drive on the dumbleator to get a hair bit more. As is very loud, and will never distort, even with all controls maxed.

Nice amp, I will need to modify even more as I have time, definetly something I think many will enjoy. Its really a different beast, must be the first time I really have heard a truly clean amp.

Did not have as much of that classic ef86 pentode bite I find familar to my (not quite) ac-15 lite clone. Its aint a 12ax7 tone either though, and thats why the dumbleator was nice, to be able to mix the ef86 and 12ax7 harmonics. I think if the pentode was run as v1 more of that classic bite would be preserved, though of course thats a different amp.

Very much the sound I believe those who ask 'Is there any more clean?' will like. As long as they want exactly zero distortion max (though the dumleator can do that!).

some mods: either cause thats what I had on the shelf, or what I was used to, or just what I did:

82k input to V2 grid (little more drive from v2)
500 k bass pot
presence 2k, with .68 (instead of typical 1 uf) and 560 ohm resitor
220p off v1, .1, .05
v2 2.2k/50uf cathode - liked better
v2 plate 220k/ 750k - liked 750k better than 1m, as in other designs
v1 560->1k/220uf cathode
20k/1k/1m/1m PI
pab switch
dumbleator, cf + gain with 220k/100k pot drop pre-gain

and my feed voltage was 440vdc, with 407vdc on the ef86 plates.

Some pics attached, Note not in a cab as yet, and the pics may not reflect the current state of the the build. Still working out a few minor issues. Theres also a post PI cf soldered in, but currently not hooked up (just need time to test), as well as an extra 9 pin socket.

again, very cool, very unique, thanks a lot..