Tube Pre-amp Project

[v8pete]

Got this one back on the workbench at the weekend! It's a design I've been messing with for the last year or so. Vari-Mu triode with adjustable bias into Pentode stage which can be morphed between pentode and triode modes. All in stereo, with a 3-band EQ at the output (2 bass bands and treble)

 

http://i1278.photobucket.com/albums/y507/v8pete1/IMG_00000179_zpsbaa4a542.jpg

 

 

[Joe Muscara]

Recordings or it didn't happen.

[v8pete]

Just messing about with it at the moment! I ran the Electro 3 through it for a few hours to see what I could achieve on the B3 and the Rhodes pianos. It's certainly capable of adding in quite a lot of extra grit, and treading that fine line between adding "enough" extra drive but not decending into clipping where things rapidly start sounding very harsh. In the workshop I only have a pair of JBL Control 1 monitors, so it's really crying out to be checked out a a decent gig volume to see how it really works in practice. During the limited tests that I did, it certainly seemed capable of taking the everso slightly clinical nature of the Nord away, making things feel quite a lot more analog - but I'd like to run it up at a gig to see how much difference I notice in a real situation. It's an undeniably good feeling though to know that your signals going through a NOS RCA 6BC8 and a couple of EF86's !! When I get a chance I fully intend to record some audio, which I'll post up. Everyone always talks about the "tube sound" in relation to preamps and clonewheels, but there are virtually no examples of this online to downloand and make a judgement. Might be a while though as I'm also chopping the Leslie 760 down at the moment - decided to make a 145 sized tube driven cab that I can take on the road! Will post some pics too when I get a bit further

[Kawai James]

Wow, great project!

 

Nowhere near as cool, but I've recently been experimenting valve pre-amps to warm-up my Electro 3 too.

 

http://imageshack.us/a/img541/2935/preampmess.png

 

Cheers,

James

x

[Griffinator]

Just a word of warning, James... Those ART preamps are what are known as "starved plate" tube preamps. They don't give the tube anywhere near enough voltage to be driven properly, because they can't generate that kind of voltage with a 9v, 1000mA power supply. As a result, you wind up with Eddie Van Halen's "brown sound" - on a massive scale. The tubes sound "hairy" because they're being grossly underdriven.

 

Translation: You're not getting a real "tube" sound out of those things.

[JMcS]

A well designed circuit is a well designed circuit. High voltage doesn't guarantee good sound.

 

Here is some information about starved plate tube theory and circuits:

 

Tube sound

 

There is more information on pages 22 & 23 here:

 

Tubehead manual

[Griffinator]

A well designed circuit is a well designed circuit. High voltage doesn't guarantee good sound.

 

Here is some information about starved plate tube theory and circuits:

 

Tube sound

 

There is more information on pages 22 & 23 here:

 

Tubehead manual

 

From your link:

 

Operating tubes at low plate voltages exaggerates natural warming distortions ...

 

Precisely the point I was making. It sounds "hairy" because it's grossly under-powered.

[Kawai James]

Thanks for the info guys.

 

Yeah, I expect I may not be getting the 'full tube' sound, but for $50 a pop (actually a little less as one of the units I purchased was second hand), I'm pretty happy with the results.

 

James

x

[JMcS]

The ART preamps aren't designed or intended to be distortion boxes. They are designed to add a little warmth to the sound. A starved plate device designed to do that can do it just as well as a high voltage design. The tube in a starved plate circuit does exactly what it would do in a high voltage circuit. However it can do it sooner and a little more when pushed, thus the "exaggeration". To some degree, a starved plate circuit designed for distortion may be considered as being able to go past 10 to 11 or 12. A high voltage circuit pushed to the same level of distortion may not sound any better.

[J. Dan]

One of the most disappointing aspects of my Kronos is any kind of overdrive or amp modelling. At the Midwest Keyboard Hang, the gritty sound of RedKey's real Wurli was apparent. Among the clones, the SV1 had the upper hand on the dirty sounds, although I preferred the Kronos for clean. I ran into similar assessments when it came to B3. IMO conclusion: a tube pre could solve the problem. My idea: use an aux out and assign any sound I want with teeth to that out, I could even bring it back in and mix to the main out. Haven't tried it yet, but this thread has me thinking.

[v8pete]

Yep, I don't know whether the ART MP is true "starved plate" or just runs at a reduced anode voltage, but if it's the former then you're operating down in the bottom corner of the plate cahacteristics plot - and you're going to see a lot of variability between tubes and some funny grid current stuff with the straved plate designs.

 

For my design I wanted to be able to easily get a high level of even harmonic distortion from the fist stage, so hence going for a remote cutoff tube (as used in famous UA 175A) and being able to swing the plate voltage over a reasnoably wide area. Tjis way you can get as much or as little distortion as you desire, depending on how you set the bias and the level of input signal that you apply. This stage then drives a pendode (and can over-drive the pentode if you so wish), and by fooling around with the dc conditions and feedback applied to the screen grid you can create all sorts of possibilties - I was particularly intersted in the compression effects that would be possible by allowing the screen grid to "sag". The original Hammond preamps (AO28 etc.) contained a number of pentode stages, so I've always though that they probably have an input into the elusive Hammond sound.

Something else that I'm keen to explore further is the distortion effects produced by inductor saturation in the Leslie 145/147 crossover - I think that the true nature of the Leslie grind/growl sound is at least in part due to someting happening here, and certainly not just a facet of tube amp distortion!

 

[Theo Verelst]

Always interesting to go back to effects/mix/mastering basics ! Of course the advantage of a hardware tube distortion/overdrive effect is there's no sampling to mess things up. I mean any signal you give extra overtones is either not rich in high frequencies when you begin, or becomes a problem when sampled at CD rates (because of the higher than ~20khz components causing aliasing distortion).

 

Using a sampled source isn't really great, unless it is prepared, at least somewhat, for that. So if you want the "warmer" additions to the sound, you may want to filter out the higher frequencies before driving the tube, like with an eq, or by using a secondary output of the synth, where you kill all frequencies above 6 kHz or something, and then mix the output of the tube circuit in with an analog mixer. Also, you may want to kill quite a bit of very-to-sub low from the input of your tube side-chain, to prevent harmonics coming from the not-so-great transient componenta out of your digital synth.

 

I had a tube amp when I was really young, too, and I know that the tubes and the schematics they were used in had to be tuned for reverberation purposes, which is an interesting subject, but difficult: those electron cloud characteristics can be quite challenging when put in a feedback loop, and give very special effects (involving long averages), hard to obtain otherwise.

 

In my Kurzweil tube effect (which is a big word at the moment) experiment, I make use of mastering paradigms from the good mixes, where essentially FFT-type transforms on the samples reveal tube information, which is then used as a harmonizing+waveform enhancing effect, most prevalent when listened to in a reverberant space. Like a "dimension" coupled with deeply hidden sample preparations (as also present in good CDs). I'd try some natural instruments like voice and guitar with your hardware, that's probably easier to follow than a digital synth which has no preparations for using tubes, because the sampled outputs might not give you the signals you want much.

 

Theo.

[Griffinator]

The ART preamps aren't designed or intended to be distortion boxes. They are designed to add a little warmth to the sound. A starved plate device designed to do that can do it just as well as a high voltage design. The tube in a starved plate circuit does exactly what it would do in a high voltage circuit. However it can do it sooner and a little more when pushed, thus the "exaggeration". To some degree, a starved plate circuit designed for distortion may be considered as being able to go past 10 to 11 or 12. A high voltage circuit pushed to the same level of distortion may not sound any better.

 

What I'm saying is that the real "warmth" of a tube amp isn't all hairy like an ART. It's much cleaner. I'm not talking about overdriven tubes. Quite the contrary, I'm saying the ART and other starved-plate designs deliberately introduce distortion into the signal to make it sound "tubey", when a well-designed full-voltage tube preamp circuit only does that when you choose to overdrive it.

[JMcS]

The ART preamps aren't designed or intended to be distortion boxes. They are designed to add a little warmth to the sound. A starved plate device designed to do that can do it just as well as a high voltage design. The tube in a starved plate circuit does exactly what it would do in a high voltage circuit. However it can do it sooner and a little more when pushed, thus the "exaggeration". To some degree, a starved plate circuit designed for distortion may be considered as being able to go past 10 to 11 or 12. A high voltage circuit pushed to the same level of distortion may not sound any better.

 

What I'm saying is that the real "warmth" of a tube amp isn't all hairy like an ART. It's much cleaner. I'm not talking about overdriven tubes. Quite the contrary, I'm saying the ART and other starved-plate designs deliberately introduce distortion into the signal to make it sound "tubey", when a well-designed full-voltage tube preamp circuit only does that when you choose to overdrive it.

 

I have several ART preamps and I guess we will have to disagree about how they sound.

[B3Nut]

The original Hammond preamps (AO28 etc.) contained a number of pentode stages, so I've always though that they probably have an input into the elusive Hammond sound.

 

The AO-28 only has a 6AU6 pentode in the first gain stage of each channel (vibrato and non-vibrato.) The rest of the preamp circuit itself is all triodes.

TP

[Griffinator]

The ART preamps aren't designed or intended to be distortion boxes. They are designed to add a little warmth to the sound. A starved plate device designed to do that can do it just as well as a high voltage design. The tube in a starved plate circuit does exactly what it would do in a high voltage circuit. However it can do it sooner and a little more when pushed, thus the "exaggeration". To some degree, a starved plate circuit designed for distortion may be considered as being able to go past 10 to 11 or 12. A high voltage circuit pushed to the same level of distortion may not sound any better.

 

What I'm saying is that the real "warmth" of a tube amp isn't all hairy like an ART. It's much cleaner. I'm not talking about overdriven tubes. Quite the contrary, I'm saying the ART and other starved-plate designs deliberately introduce distortion into the signal to make it sound "tubey", when a well-designed full-voltage tube preamp circuit only does that when you choose to overdrive it.

 

I have several ART preamps and I guess we will have to disagree about how they sound.

 

Have you compared them to a full-voltage tube preamp? If you have, then yes, we'll just have to agree to disagree.

 

You're talking to someone who has employed tube preamps in every application from instruments and microphones to hifi turntables. An ART doesn't sound like a full-voltage preamp at nominal volumes. The bogging-down of the circuit from lack of juice is very audible to my ears - it sounds hairy, like a slightly overdriven Marshall head.

[JMcS]

Yes I have. Probably not in as wide a range of applications as you list. One thing I have noticed is that with the ART preamps I've used I haven't had to use EQ to boost the bass and treble to get the overall tone to sound balanced. One thing often spoken of here by many of those who use a popular high voltage preamp is that they have the bass and treble turned up somewhere in the signal chain. Maybe that has something to do with the circuit design.

 

I have replaced the tubes in my ART preamps with old 50's/60's era tubes (either 12AX7s or 12AT7s) as well. They don't seem to get fizzy when pushed as the new manufacture tubes do. Again, I don't use the preamps to produce an overdriven tone so they don't get pushed to the point where the sound gets "hairy".

[v8pete]

Always interesting to go back to effects/mix/mastering basics ! Of course the advantage of a hardware tube distortion/overdrive effect is there's no sampling to mess things up. I mean any signal you give extra overtones is either not rich in high frequencies when you begin, or becomes a problem when sampled at CD rates (because of the higher than ~20khz components causing aliasing distortion).

 

Using a sampled source isn't really great, unless it is prepared, at least somewhat, for that. So if you want the "warmer" additions to the sound, you may want to filter out the higher frequencies before driving the tube, like with an eq, or by using a secondary output of the synth, where you kill all frequencies above 6 kHz or something, and then mix the output of the tube circuit in with an analog mixer. Also, you may want to kill quite a bit of very-to-sub low from the input of your tube side-chain, to prevent harmonics coming from the not-so-great transient componenta out of your digital synth.

 

I had a tube amp when I was really young, too, and I know that the tubes and the schematics they were used in had to be tuned for reverberation purposes, which is an interesting subject, but difficult: those electron cloud characteristics can be quite challenging when put in a feedback loop, and give very special effects (involving long averages), hard to obtain otherwise.

 

In my Kurzweil tube effect (which is a big word at the moment) experiment, I make use of mastering paradigms from the good mixes, where essentially FFT-type transforms on the samples reveal tube information, which is then used as a harmonizing+waveform enhancing effect, most prevalent when listened to in a reverberant space. Like a "dimension" coupled with deeply hidden sample preparations (as also present in good CDs). I'd try some natural instruments like voice and guitar with your hardware, that's probably easier to follow than a digital synth which has no preparations for using tubes, because the sampled outputs might not give you the signals you want much.

 

Theo.

 

Yep, agree with all of your points there Theo. I'd like to try a band-limited / multi-band approach sometime to see what kind of difference that makes. Certainly HP filtering the signal which is applied to the tube stages and adding tha bass information back in externally would be easy to try. Bandwidth limiting at each tube stage also seems very important (just like in well designed tube guitar amps).

 

Pete