Back around 2000 I got this stupid idea along the same lines. I took it to its Tubelab extreme though. I built a scaled down conventional tube push pull amp,.....a couple of 12AX7's feeding some 7 pin pentodes for output tubes. There are dozens of 7 pin tubes with the same pinout as a 6AU6. Some like the 6AK6 are real (low power) audio output tubes. I think I tried everything I had that fit the sockets. There are definitely different characters in each different tube in that circuit.
I tried several different "magnetic devices" for the OPT before settling on a multi tapped power transformer that came from Mouser. It's no longer available. I got all this working into a small speaker, maybe 5 or 6 inches.
I then carefully cut the paper cone out of the speaker leaving the spider intact. It still made some sound, but not a lot.
I tapped off speaker leads with a pot as an adjustable voltage divider to feed a solid state power amp that used the big National Semiconductor chip amp of the day that (LM3886 maybe).
It sounded pretty good with the tube amp cranked and the SS amp running about half power. Driving the SS amp into distortion got rather nasty sounding.
This was about the same time that I was building the Turbo Champs, and another solid state guitar amp project that used a Microchip processor and SPI programmable equalizer and tone control chips from National. I had this grand idea to build a big amp with a National chip amp and equalizer for each speaker with a selectable tube or solid state preamp, and a PIC chip to remember the settings.
This was about the time when the cell phone world went into the steep part of the volume curve, and I was a cell phone design engineer. My work week went from normal, to 70 hour work weeks, so none of these projects got finished. All of them met the trash can when Motorola sold the phone business, and I prepared to leave Florida. The multi - amp was big and heavy, mounted on a recycled steel chassis from a dead Traynor tube amp.
The low powered tube amp as a preamp was a cool idea. If I had it to do over, I'd leave out the hacked speaker.....I'm not sure it did anything, and the voice coil hit bottom so many times that it started rattling pretty bad. Some sort of reactive load on the OPT is probably the right solution, but I don't know what.
I never tried a toroidal OPT back in 2000, but I had good luck with a small Antek driving a speaker in the active part of the thread. The little Jaycar transformer was probably made for a solid state design, and may not have enough inductance, but there is only one way to find out.
The inductive load of an OPT allow the voltage swing at the plates to go to twice the B+ voltage. You can't get this effect with a resistive load and opamps. How much this contributes to the sound, I don't know.
I had a little amp wired that way. I made in it Florida when I was trying out ideas for the challenge part of this thread. I used a 6U8 tube because I have hundreds of them, and several different tubes with the same pinout. I think I got a watt or two out of it on about 200 volts using the same 70 volt OPT that I used in AMP1.4. I decided it would need a third tube for the preamp, and that part didn't get built since Amp 1.4 added up to less money. The 6U8 amp needed a real power transformer to get 6.3 volts and HV, so it didn't make the cut. We are not constrained by these rules any more so I need to find that box of real power transformers.
Really? I never would have expected it.
Around the start of the challenge I was thinking along the lines of doing a Plexi with a 6AK6 output to use as a preamp or a 2W practice/recording amp. But the challenge caught my interest and the 2W Plexi got pushed aside. I have a perf strip loaded with parts for it, will have to decide what to do. Or do both.
Eventually I did use the 6AK6's in a little 5E3 type amp, lost it to a relative though.One of the reasons I wanted to do the Fet/Plexi/switching amp is to have a pocket sized amp. And I also want to see what these little PWM amps would do. But the electronic workbench is collecting sawdust. Got to get some of the guitars out of the way first.
Thank you for confirming my hunch!
It seems as though every Internet thread or article on guitar amp speaker attenuators usually ends up discussing, in hushed tones, the importance of a reactive guitar speaker load. I've always been skeptical of this claim, firstly because the heavily reactive part of the impedance curve only affects the four or five very lowest notes the guitar can reach (below 110 Hz or so), and secondly, because there are attenuator devices sold containing an actual speaker motor - in which the voice coil has been immobilized.
Apparently the happy designers, marketers, and purchasers of said devices don't understand that if you immobilize a speaker voice coil, you also lose all the reactive characteristics of a speaker. All you have left is the tiny little voice coil inductance. Using an entire speaker motor to provide a few hundred microhenries of inductance makes for a very big, very heavy, very expensive little inductor!
My hunch is that the contribution is a small one. If the output valves are overdriven to, say, 50% total harmonic distortion, and the transformer core's hysteresis curve is contributing 10% THD, how much does it matter? Maybe not very much, in practical terms?
The little guitar amp I built has an 8 ohm, 5 watt power resistor wired across the (switched) output jack; if I forget to plug in a speaker, the amp still has a load. Maybe I should just wire a couple of resistors across that load, feed the line-level signal to my EQ pedal, then into the P.A., and see how it sounds. It would be nice not to have to lug the speaker enclosure to all the jams I attend!
-Gnobuddy
You forgot the tongue-in-cheek smiley.
From the clip you posted, I think you got something really special out of that little 5E3. Make him give it back!
-Gnobuddy
You forgot the tongue-in-cheek smiley.
From the clip you posted, I think you got something really special out of that little 5E3. Make him give it back!
-Gnobuddy
Didn't forget, thought my character would carry through my posts.
Want to build something for him with a little more headroom and may get it back then. Maybe a 8W amp with a 60's 10" home speaker. At the moment building him an acoustic to use on stage. He really like the family he married into.
I'm convinced that I killed the fun when I cut away the cone, and the speaker was too small and a POS at that. I think it was a cheap car speaker.
I spent some time trying to measure the dynamic effects of a moving speaker cone on the output of an audio amp. My measurements were rather crude with a scope on the voltage fed into the speaker, and the current drawn by the speaker as the speaker is driven through and around resonance by a tube amp with low damping factor. They are almost never in phase, and often change wildly in the presence of transients, say slapping the strings on a bass.....
The speakers were my Yamaha NS-10M Studio monitors and the amp was a SSE (single ended HiFi amp). With some music I became convinced that the speaker's impedance can go negative for brief instances. I also measured unclipped peaks of nearly 60 volts on an "8 ohm" speaker. The track that demonstrated this was Duende by Bozzio - Levin - Stevens on the Black Light Syndrome CD. Also on the Gypsy Soul Flamenco sampler CD. Maybe this was just the right combination of music, amp and speakers to provoke this, but it's another thing to explore further down the road.
I saved the 6AK6's for a battery operated tube amp. They have the lowest heater power to output power of any small tubes I have tested. The heater is 150 mA on 6.3 volts. I can get 3 watts from a P-P pair on 160 volts.
I believe that. Negative impedance equals power generation; that's perfectly plausible, since once the voice coil is already in motion, you can remove the driving voltage, and the voice coil will continue to generate voltage/current/power until the moving mass (coil/former/spider/cone/surround) comes to a complete halt. In a poorly damped system, that could take a few bass cycles, which is a relatively long time, maybe as much as some tens of milliseconds.
On the question of how much the reactance around Fs matters in a guitar amp, most guitar speakers I've looked at have Fs in the range of 80 Hz to 100 Hz. Qts tends to be really high, so the impedance peak around resonance is very high Q, and quite narrow in bandwidth.
So, looking at it in the frequency domain (rather than the time domain), the heavily reactive part of the impedance curve only extends over just a handful of the lowest notes on the guitar (maybe the five lowest frets on the 6th string, E2 through A2 or so).
If you're playing in a band, those are the same low notes that create sonic mud if you let them clash with the bass guitar, keyboards, or kick drum. So, IMO, any reasonably aware guitarist would probably avoid those very lowest notes. In a mix, the recording engineer would usually high-pass the guitar parts, filtering out these low notes, for the same reason. For that matter, the guitar speaker itself also frequently rolls off those lowest few notes to some degree.
(Unfortunately, the bit about avoiding the lowest few notes on the guitar does not apply to self-indulgent, long-haired, heavily tattooed fellows playing flat-black 7-string guitars very loudly through Peavey 5150 heads. But I did say "reasonably aware guitarist"...)
So it's not that I don't think the speakers fundamental resonance can play a role in a valve amps output. I know it certainly does, in a Hi-Fi amp, particularly one with a high ouput impedance. In the case of a guitar amp, though, I'm not so sure this effect matters much.
(Maybe this only applies to my own playing - I certainly don't lean on those lowest few notes on the guitar when playing with other musicians.)
Those 6BK5s I've been thinking about are on the other end of that spectrum. Heater current is 1.2 amps at 6.3 volts, for only 9 watts max. plate dissipation!
-Gnobuddy
As a guitarist who gigs regular I cant really see why we need to go this way...I made a small push pull amp using 6aq5's and was able to fit it into a kustom cab with an 8" speaker which I replaced with a bose speaker ....the cab was roughly 12" x 12" x 10" deep... and with a sm57 in front worked just fine....absolutely no heavy lifting here...
But each time I do something like this I find myself going back to the real deal...18 to 50 watts 1 to 4 x 12" speakers
But each time I do something like this I find myself going back to the real deal...18 to 50 watts 1 to 4 x 12" speakers
I have a cheap 8" speaker in my project amp, bass is definitely on the weak side at present. I can understand why you go back to bigger speakers.
Maybe it's not possible, but the dream of the "real valve amp in a pedal" is a difficult one to drive out from one's brain!
-Gnobuddy
nothing wrong with valve pedals....its that stage you don't have in your unit that you wish you had......but when you sit down and listen to everything hifi these days you can still pick that a guitar is being played through tubes...that in itself is a statement that valve sound can make it through transistor sound...
To explore strange new tubes, to seek out new circuits and topologies, to boldly go where no tube has gone before......
To explore strange new tubes, to seek out new circuits and topologies, to boldly go where no tube has gone before......
I started out playing electric guitar in a (cultural) vacuum - no one to get advice from. I didn't know that valves were the magic ingredient in good electric guitar sound, so I kept building - and later, buying - solid state guitar amps. After all, I knew was that valves had gone extinct before I was even born, that they were obsolete, completely replaced by solid-state devices. Why would I ever think twice about them?
So, for 25 years (!) I had horrible electric guitar tone. For a good ten-year chunk of that period, I didn't touch a guitar at all, I was so fed up with the whole thing. I thought I must be a terrible guitarist, or all my guitars were awful, or something.
I was this close to giving up for good. But by this time, the Internet and guitar forums existed. After lots of online research, I bought my first amp that had valves in it. I plugged in, waited 30 seconds for the heaters to warm up, and voila, for the first time in my life, an electric guitar in my hands sounded good instead of nasty. You could have knocked me over with a wet noodle. It wasn't me, it wasn't my guitars, it was my amps that had let me down all these years. Wow.
And that was just a budget-priced hybrid amp, less "valvey" sounding than an all-valve design.
I'm perfectly happy with solid-state devices in all my electronics except my electric guitar amps. There, there is no substitute - nothing else sounds good to my ears, at least, not yet.
Which is why I'm on this thread. Even a $100 valve amp has a good chance of sounding better than a $1000 solid state amp...
-Gnobuddy
Back in post #1927 I shared a photo of my amp, which had just grown a working second channel.
Testing at the next couple of jams revealed a rather abrupt onset of distortion on the drive channel. When the distortion did arrive, it sounded a bit harsh for my tastes, though not too dissimilar to a Vox AC15 or similar. It was also too thin-sounding.
I knew from previous measurements that the pentode (final stage in the preamp drive channel) was slightly cold-biased, and clipping began on the positive peaks. So, in the last hour before my next jam, I replaced the 270k screen grid resistor with a 180k one, for a warmer bias. I also made some changes to capacitor values to let more bass through.
This is the part I thought some of you might be interested in: the amp was noticeably more touch-sensitive on the drive channel after I'd made these changes. As you turn up the gain on the drive channel, there is a quite noticeable "squish" applied to the attack of each guitar note, followed by a slight swell or rebound. Very much like a compressor.
The overdriven tone is also a bit subtler and more progressive than before.
I think warming up the pentode bias has caused grid current to flow more readily during overdrive, and that in turn is charging up the coupling and bypass caps, shifting the stage towards cooler bias. In short, warming up the quiescent bias has increased the amount of "sliding bias" in response to overdrive, as well as softening the initial onset of clipping.
Whether those are good things or not depend on your tastes and the music you play, but this shift in pentode playing qualities with warm-biasing seems a useful thing to know.
This is the first guitar amp I've ever played through that has pentode(s) in the preamp, and I really like what they bring to the party. To my ears, both better clean tones, and better overdriven ones.
-Gnobuddy
Testing at the next couple of jams revealed a rather abrupt onset of distortion on the drive channel. When the distortion did arrive, it sounded a bit harsh for my tastes, though not too dissimilar to a Vox AC15 or similar. It was also too thin-sounding.
I knew from previous measurements that the pentode (final stage in the preamp drive channel) was slightly cold-biased, and clipping began on the positive peaks. So, in the last hour before my next jam, I replaced the 270k screen grid resistor with a 180k one, for a warmer bias. I also made some changes to capacitor values to let more bass through.
This is the part I thought some of you might be interested in: the amp was noticeably more touch-sensitive on the drive channel after I'd made these changes. As you turn up the gain on the drive channel, there is a quite noticeable "squish" applied to the attack of each guitar note, followed by a slight swell or rebound. Very much like a compressor.
The overdriven tone is also a bit subtler and more progressive than before.
I think warming up the pentode bias has caused grid current to flow more readily during overdrive, and that in turn is charging up the coupling and bypass caps, shifting the stage towards cooler bias. In short, warming up the quiescent bias has increased the amount of "sliding bias" in response to overdrive, as well as softening the initial onset of clipping.
Whether those are good things or not depend on your tastes and the music you play, but this shift in pentode playing qualities with warm-biasing seems a useful thing to know.
This is the first guitar amp I've ever played through that has pentode(s) in the preamp, and I really like what they bring to the party. To my ears, both better clean tones, and better overdriven ones.
-Gnobuddy
The first guitar amp that I completely made from scratch had a pentode (6SJ7) in the first stage. I tried several tubes in the output stage. It was based loosely on a Fender Champ 5C1. I was 12 or 13 years old. I tried several different output tubes, and speakers. Most were whatever I could pull out of junk TV's, radios, or HiFi sets. It had a horrible overload / bias shift / blocking distortion when driven by a fuzz pedal. A ham radio guy figured out that it was in the input stage and caused by the grid leak bias. Cathode bias fixed that.
I would discover the 12AX7 and 6SL7 soon after that, but that old pentode - pentode amp was something I kept coming back to. Don't have a clue how many I made as a kid. Some I kept for a while, some I gave away, and some I blew up or took apart.
Somewhere before high school I discovered the transistor. First it was the big fat germanium power transistors that I took off the back of car radios. They kinda rocked, and had a sound that worked rather well with the surf music of the day. Then silicon happened......
At age 20, I got a job at Motorola. UH, they MADE silicon, and to employees, it was FREE (within reason) just by filling out a sample request. So, we made amps. HiFi amps, guitar amps, car audio thunder boxes, home computers (MC6800, 6809)......all powered by silicon......
Then one day a guy asked me if I knew how to fix his Bandmaster.....UH, this thing rocked. Time to start making tube stuff again. This was somewhere around 1980.
Full circle......Amp 1.5 was reborn about 6 months ago. It's still the only working guitar amp I have, and its pentode - triode - push pull pentodes. Maybe next time I'll try the triode first.......no I'll probably make something way off the deep end, but I have been tinkering with this synthesizer thing. I got this out there idea......like tubes in the synth, pentodes at that.....Just the ticket to warm up those digital VCO's.
I would discover the 12AX7 and 6SL7 soon after that, but that old pentode - pentode amp was something I kept coming back to. Don't have a clue how many I made as a kid. Some I kept for a while, some I gave away, and some I blew up or took apart.
Somewhere before high school I discovered the transistor. First it was the big fat germanium power transistors that I took off the back of car radios. They kinda rocked, and had a sound that worked rather well with the surf music of the day. Then silicon happened......
At age 20, I got a job at Motorola. UH, they MADE silicon, and to employees, it was FREE (within reason) just by filling out a sample request. So, we made amps. HiFi amps, guitar amps, car audio thunder boxes, home computers (MC6800, 6809)......all powered by silicon......
Then one day a guy asked me if I knew how to fix his Bandmaster.....UH, this thing rocked. Time to start making tube stuff again. This was somewhere around 1980.
Full circle......Amp 1.5 was reborn about 6 months ago. It's still the only working guitar amp I have, and its pentode - triode - push pull pentodes. Maybe next time I'll try the triode first.......no I'll probably make something way off the deep end, but I have been tinkering with this synthesizer thing. I got this out there idea......like tubes in the synth, pentodes at that.....Just the ticket to warm up those digital VCO's.
I'm a little stunned by those Metasonix designs featuring the tips of glass valves sticking out of the front panel. Hot, fragile glass, containing high voltage, and placed right next to knobs you're supposed to turn with your easily-burned fingers.
The Metasonix D-2000 drum machine gave me flashbacks to the late 1980s, when I made a DIY drum machine. It used a 4017 decade counter to provide 3, 4, 6, or 8 beats per bar. The pulses from the 4017 drove various "ringing oscillators" built around a 4049 hex inverter, with the inverters biased into linear amplifier mode. DIP switches let me program which oscillator would sound on each beat.
Later I built that drum machine into a solid-state guitar preamp that I also built. The preamp was based on a design from Elektor magazine.
The drum machine worked, but (predictably) sounded horribly monotonous. The preamp also worked, with the typical unlovable solid-state guitar preamp sound.
-Gnobuddy
Last edited:
Sometimes I wonder if Fender switched to triodes simply because they were cheaper, or because there were actually problems with the pentode designs.
To my ears, the now-classic "Fender cleans" are super-easy to get from a pentode, and much harder to extract from a triode.
But I'm guessing Fender was simply interested in getting enough voltage gain, at the lowest possible cost. And a single 12AX7 did the job.
I also got started with germanium PNP transistors (right after the crystal radio with the germanium detector diode). The transistors came out of a little radio, and my big brother helped me get them out, because I was 8 or 9 years old.
I don't remember part numbers after all these years, but they were little metal-can types, with the "power transistors" only good for a few hundred milliwatts, and requiring an output transformer to drive an 8 ohm speaker.
I remember the shock of seeing my first epoxy-encapsulated silicon transistors. They didn't look right to me, but I quickly found out that they worked much better than the germanium ones!
I have two small Fender amps, but neither has been turned on in months. The little amp I built is the one that gets used at the jams, because, (at least to me), it sounds better, and is more versatile. Score a point for $1 valves that (almost) nobody wants!
My goal of good tubey/valvey sound at apartment-friendly volume, though has not been met. When I turn it down that quiet, I can no longer hear those subtle harmonics. It just sounds as uninspiring as any solid-state amp. Perhaps headphones really are the only way.
That arrangement is working well for me. Triode -> pentode in the clean channel, triode -> triode -> pentode in the drive channel.
It would be cool if that works!
I'm still baffled at the sonic magic that is so easily worked by simply sticking just about any kind of valve into a guitar amp. What a curious coincidence that old, primitive, triodes and pentodes just happen to have transfer functions that sound good, and none of our more modern amplifying devices have that property!
-Gnobuddy
I have known about Eric Barbour ever since he was the US sales engineer for Svetlana vacuum tubes before they went poof, and the name was acquired by Mike Mathews.
There have been lots of weird projects from him over the years, but the Butt Probe and the Fu^%ing Fu^%er guitar amp were over the top. The amp sold for $5,000 and was introduced at the 2008 NAMM show to mixed reviews. It's been gone for some time, but still can be seen on Youtube.....warning, language and graphics not suitable for some children....and this was the mildest video I could find.
https://www.youtube.com/watch?v=YuAaTEn8fcw
I'm not sure how practical a vacuum tube synthesizer is, or how practical vacuum tube powered modules are for a modular synth. Adding a tube module to an existing modular containing modules that are clones of old Moog or ARP designs is just asking for drift. Adding a heat source inside the box would make for guaranteed warm up drift, unless it's never turned off.
Amp 2.whatever (don't remember) used a pair of UL84's (a 45 volt 6CW5) on 330 volts driven by some UCC85's (26AQ8's) and a circuit typical of the "18 watt Marshall" it just fit the $100 price point, with an attenuator between the OPT and speaker. You could dial up a full metal racket, then leave the amp cranked, and dial back the loud with the attenuator. You will still lose some of the sound created by a cheap 8 inch guitar speaker trying to eat about 25 watts of overdriven "84's, but it beats a chip amp!
I used this:
Speaker L-Pad Attenuator 15W Mono 1" Shaft 8 Ohm
I have no idea if it will work, but who else is going to try a vacuum tube ladder filter........
There is another vacuum tube synthesizer, the Knifonium, with a "4th order ladder filter." It's made in Sweden, I believe. All 25 of the tubes are inside a cage at the top of the unit.
https://www.youtube.com/watch?v=4zf_d8tbhb8
https://www.youtube.com/watch?v=c-gGuo17lYA
http://www.synthtopia.com/content/2013/01/10/the-knifonium-vacuum-tube-synthesizer/
Found them when I was looking to see if anyone had a guitar amp using the 6/12BK5. Disappointed that the only vids were of noise.