Extremely high output guitar - amp input considerations

[pjd3]

Hey everyone,

I'd really be interested in your feedback on this.

After looking at voltage output range charts for various 6 strings electrics, I thought I'd just stick my DMM on the outputs and see for myself.

Although this is just a Radio Shack 60 dollar meter the levels for my standard Strat was typical of the chart voltages for various pickup config, max strum, anywhere from 30mA to 250mA depending on pickup selection and how hard you strum (volume and tone all the way up of course). Then I believe I measured near a volt or so with the Ebow, which didn't surprise me.

THEN,, I measured the output on my fretless "Ebow only" guitar that I built. With pickups in standard series humbucker config, volume/tone on max I was seeing voltages up to 2.75 volts and I think it even reached 3.00 volts for a moment on one of the highest notes. I'm aware that the use of an Ebow generates a much higher output than just picking or strumming but, I did not expect to see anywhere near this. I know I'll need to see this on a scope to really get a good picture of what is really going on but, even considering the questionability of an older radio shack DMM, the relative values between Standard Strat and this Fretless are hugely different.

With this guitar putting out voltage values near or at line level, what consideration would you think you'd give to the input stage? Ha, Santa brought me a NOS 12AY7 and 5751 for Christmas thinking that the input gain could be "a little lower" that a standard guitar but, I have not found any Line level circuits using those tubes/gain anywhere.

This amp could end up going more or less toward "high-fi" territory but that will be seen when the hard experimentation comes.
What are your thoughts on ways to input that high level of signal?

Thank you,
PJD3

[xtian]

Ever seen Hi and Lo inputs on a Fender amp? Ever? Always?

[Stevem]

100% of the post 1945 era guitar amps I have worked on in 50 plus years of doing so only needed .150 volt ( 150 mv) to drive the amp ( be it tube or SS ) to max output and full clipping , and on many amps this could be done with that same signal level thru the number 2 of low gain input.

.150 volt is pretty much -10 db just to give you a reference.

[pjd3]

thanks guys, and sure, of course I've seen the Low and High on a Fender. My first Twin reverb in 1970 that Dad bought me (150 bucks for a 1968 Twin!!) and the single channel Fender Vibroverb I just built and use for gigging every weekend has a Low and High.

If I recall (Im' in work right now) that input difference is contingent upon the voltage division of the 68K resistors I think. And that did cross my mind that you can voltage divide many things down to nothing if you are so inclined which led me to ponder that, well, if thats the case, why are most audio device's input stage design based upon the source voltage range that exists at the output of the source device? I could also just lowered the volume knob on the guitar to compensate but, there seems to be a tendency to design input stages according to the source voltage range. I figured thats what I would try to do, assuming there was some benefit to that. I know that there is a huge difference in tonality (EQ) between having my amps on 10 and guitar on 3 verses amp on 3 and guitar on 10. Since part of this is designing the amp for this Ebow only fretless guitar, I figured I'd optimize everything for existing ranges. Also, with the much larger output of this "guitar" I would most likely have much different calls for gain staging in the amp. I'm not trying to make things simple for myself. Part of this is learning what different configurations of circuit topology yield so, I'm getting ready to try many things.

So, I am truly interested in what considerations exist as a departure from a typical guitar amp input stage for a "guitar" that has a line level output as it is. I suppose it might need to be considered that the output impedance is far higher than what you might expect from a CD player, or active line level output as in an effects rack output. But yeah, thats the thing. I'm designing this amp very specifically for this guitarish Ebow thingy with the massive output.

Thanks guys,
PJD3

[DimenBrun]

If you don't care for the std. volt. divider 68k - 68k setup, you might consider designing the input stage around a 12AU7. That way the headroom should be plenty high enough(and gain lower) and you wouldn't have to compromise input impedance. Just a thought.

[ER]

Hi low inputs are good, you see the same thing on bass preamps with different hi/lo inputs for passive or active pickups. But is it something you really need? Build whatever sounds best on your other guitars, I wouldn't do anything that much different. The chosen input resistor is mostly part of the grid stopper low pass to block high frequency oscillation coming into the first gain stage. Work for a good impedance match. The typical 68k only drops the voltage around 6%. If it sounds fine, don't worry about it, think about all the high gain pedals people stick in front of amps, even how far you can push a clean boost in that position.

I worked at EMG as a kid back when e-bows were popular, and never saw any issues with output levels. One of the other guys that worked there was crazy about those things and most of us had the PA2 20dB boosters on board back then too. I think there was some specific testing as the e-bow was becoming a thing to see if that was going to be a problem. I think the best solution to overdoing it was technique, to play the e-bow close to the bridge and use distance from the strings. Are the highest voltage readings with the e-bow directly over the pickup? The amplitude can really get away from you playing out away from the bridge and that will translate directly to volts. You might be getting a bit of a Paul Gilbert Makita effect from the e-bow being over the pickup. If anything maybe think of building an "un-boost" pedal for that particular guitar, so lowering the volume doesn't change the tone as much as with the pots on the guitar. Also try the e-bow on your other guitars and see what the increase in output voltage is from the device alone. Don't worry about what the meter says to much, if it sounds good it is good.

Are you wanting to play the amp clean or overdriven with that particular guitar? Maybe more gain isn't a bad thing? I'd worry more about the overall amp design and power requirements for the tone you're going for and the way youll be using it than get hung up on the input itself.

Another option besides a un-boost pedal is to put a Dumble style FET input on the amp and trim it back. On my Dumble style amps, I add a relay for the FET and put the gain trim pot on the back of the amp. If you did something like that you could have it set perfect for the e-bow guitar, and then switch easily to a regular guitar as well. You might note that most of the guys that play the old high output lap steels with 1meg pots on Dumbles all use the FET input. Lap steel and fretless e-bow are going to share a lot of the same vibe and tonal centers.
I find it's better to work toward flexibility than limitations on a amp within reason and if it can be done elegantly.

[pjd3]

Great feedback, thank you.

These are the kind of ideas I"d be happy to run with once its time to start putting together the actual stages. And I have every intention of trying as many configurations as I reasonably can. I imagine my
requirements will be very close to that of a pedal steel player or lap steel player, and possibly an electric violin player. The reason I'm initially focused on the input is that's where it all begins, and can
have a big bearing on what will, could, or should come following, especially in capitalizing on the inherent signal amplitude coming straight from the Ebow guitar.

And sure, I"m happy to have a good handful of flexible features built into it. It doesn't have to be a total one trick pony although, it is going to be made very specifically for the Ebow guitar.
I am fairly confident that I will be looking for something very clean with alot of dynamic range. The volume range on the Ebow guitar is very wide and I want it to fit cleanly in there. The Ebow guitar does have a big tendency to get harsh and to produce a huge amount of frequencies in the 7K to 9K region at its higher dynamic, when the Ebow is the closest to the pickup magnet. The pickup by the way is a Seymour Duncan Dimebucker, the last pickup I would have thought to use but, many claimed it had very extended highs and a very high resonant frequency for the dc resistance that the windings give. For these reasons, I believe this will benefit from a good choice or selection of high frequency EQ control. When I record this instrument direct through DI-Neve preamp-Kurzweil Rumor-Scarlett in interface-Reaper I find it tends to benefit from a good dip or roll off at 8K-9K or else you are sonically attacked by a hive of angry hornets.

The adjustable mosfet input sounds like an interesting idea, as well as the 12au7. I've seen a bunch of Line input tube stages using that tube. I've also stumbled up some Jensen input transformer tube circuits that were interesting.

So, there seems to be a number of options available. Guess I'll just have to try some out to see what works best for this instrument. Maybe I'll even consider something like a switchable cold clipper just to have some funny distortion as a departure from what will likely be a very clean chain with good dynamic range.

Thanks very much guys, some good ideas and food for thought here. By the way, guess I'll mention that I gutted my Bogner Alchemist and used a couple of the circuit boards as the bed for a tube amp design station. One board has 5 9-pin sockets and the other has 2 Octal sockets - both with common heater traces. The boards even have some preamp tube plate capacitors and theres some high power screen and bias resistors on the Octal socket board so. Got a nice little design station ready to go, all mounted on a nice Oak board.

Thanks again!
PJD3

[ER]

I totally get it, I have a pointy guitar with a Dimarzio X2N, a 15k+ bladed style pickup. It's basically my heavy metal guitar, but it has the most amazing clean tone, bold and with tons of fundamental and clarity that my single coils or PAF style pickups just can't touch, they sound so colored in comparison. When ever I plug that thing in I end up play a bunch of Jazz stuff because it plays and sounds so good.

Even starting from scratch it would be hard to design something better than a Dumbleland or Steel String Singer type amp for what it sounds like you're going for. That would already have the FET input, and the tone filters could be really handy for what you're doing too. Read the description and check out some of the schematics in the sticky in the Dumble files. Every stage in those things you would be hard pressed to improve on for clean high head room, tonal flexibility, and ability to be happy with whatever you plug into it.

Just a weird out of left field thing, it would be cool if you could run that E-bow guitar through some Hammond drawbars and be able to shape the harmonic signature. I think Larry Cragg does that sometimes on his pedal steel, he cops an organ sound with leslie and everything, but with just the drawbars you could really shape the sound.

Just some food for thought, good luck with your project, sounds like a fun one!

[M Fowler]

For the impedance side of things I use the standard 1m resister on the #1 or high gain input jack and 150K resister on #2 or low gain input jack, this works very well for my Les Pauls or humbucker pups.
You could also employ a rotary switch to add a selection of resistors to match you guitar.

Mark

[pjd3]

Thanks guys, again, excellent and useful feedback on this topic,

And Yes, switchable input resistor impedance or gain is a considerable idea. My Seventh Circle DI has 3 positions for 1M, 100K and 10K ohm input impedance selector and they are good for more than just matching to source impedances. They also shift the frequency response in very useful ways. I'm pretty sure I use the 10K ohm selection for the Ebow guitar as it provides a very distinct high frequency roll off that helps with those intense high freq's inherent in the upper dynamics of the sound. I would normally use a 1M or 100K ohm input impedance for guitar pickup coils.

It looks like a good close look at these Dumble circuits would be a smart thing to do. YOu can bet I will be doing just that. What remains a mystery, and that's Ok, is whether more of a guitar design or high-fi design will work with this weird but engaging "guitar" that I've cast upon myself. I'll see what I can do about posting a clip so you can hear this thing. I'll crop a clip from an ambient thing I've been working on. Its doused with Valhalla reverb but, that's the way I like it for that project.

Thanks guys, great ideas once again.

Best,
PJD3

[pdf64]

For very large input signals, a 12AU7 / ECC82 seems to be the most appropriate input stage tube, as it will centre bias to quite high cathode voltages and so accommodate large input signals, see https://tubedata.altanatubes.com.br/she ... /ECC82.pdf for some suitable operating conditions / characteristics.

[pjd3]

I had initially imagined that a 12au7 could be a contender for an input tube seeing that its a mu of 20 compared to a 12AX7 that's 100, not because I've experimented yet, or cranked out the math, but because the output of the Ebow guitar seemed so much greater that it would simply make sense to use a tube who's gain was relatively lower to compensate. I've been trying to better verse myself looking at these tubes in terms of gain and dB, plate resistances and of course being able to read load line charts. Getting there but, still mystified on a couple of things. We'll get there. I have too!

Santa brought me a couple of very nice and expensive NOS 12AY7 and 5751 but, I'm getting that the gain is a bit too much for an input tube for this. Maybe they will come in handy for a 2nd or 3rd gain stage, eq recovery, something like that. Or maybe even an optional input for something else.

Thanks,

PJD3

[pdf64]

If you want to avoid attenuating the input signal (which isn't ideal for good signal to noise performance) then as well as low stage gain, tube characteristics that enable centre bias at a highish cathode voltage are beneficial.