Yes I agree. Even though I work in industrial electronics, I find that audio electronics as a hobby pairs well with playing music.
Just a comment to move the topic a bit forward.. if we have a couple of jfet gain stages with soft limiting that get in the ballpark of preamp distortion (asymmetric H2+ H4 prominent) then it would be interesting to tack on a differential pair type stage and incorporate further soft limiting but symmetrical and at higher threshold for adding H3 and H5. If you look at a Boss Blues Driver gain stage you will see a jfet diff pair with a PNP follower and a feedback network. It's basically a DIY op amp, there's a good article on these types of configurations by Nelson Pass.
I have one - it might be the first guitar pedal I ever bought. I think those canny Boss/Roland engineers did indeed intend to mimic the topology of famous old tube amps (like the Fender Bassman and subsequent Marshall variants).
Then again: I was listening to Clapton and (Duane) Allman's playing in "Layla and other assorted love songs" the other day. Both guitarists
sound raw and harsh, and I think the guitars would have sounded much better with some delay and reverb added during mixing.
Clapton, probably high as a kite on cocaine, flogs his poor guitar amplifier to within an inch of its life. He screams his vocals with the same feral intensity, in a voice often as harsh as the sounds from his guitar amplifier.
But, raw guitar tone and all, one thing stood out: the boring, monotonous, unchanging buzzy sound you get from most solid-state distortion pedals was nowhere to be heard.
Clapton supposedly used a Fender Champ and no pedals for that album. Three cascaded common-cathode tube stages, including the power amp. No LTP in sight.
Even with its simple topology of three cascaded common-cathode stages, I think the Champ still produces duty-cycle modulation.
And, as each stage is increasingly overdriven, waveforms shift from nearly symmetrical for very small signals (clean), to increasingly asymmetrical for bigger signals (generating mostly even harmonics), and then back to nearly symmetrical as full output is reached, and tops and bottoms of the waveform become heavily squashed (generating mostly odd harmonics).
All of this was entirely un-intentional, of course. Just an accidental by-product of capacitor coupled stages, nonlinear device characteristics, grid current flow, and heavy overdrive never intended by the RCA engineers who suggested those gain stage designs in the back of their tube catalogue.
-Gnobuddy
After all these years it is interesting to listen to the sound of the legendary records from Santana or Hendrix. The pure sound is by far not that impressive as is seems from my recollection. But it catches me somehow even nowadays. This teaches me that there is indeed a kind of myth in it, which did not touch me alone, but millions of fans. And this secret sauce was the genius of the stoned clapton, santana, hendrix etc - and not some H2 harmonics. Another example for an ugly sounding fuzz but nonetheless legendary is Jeff Becks part in Heart Full of Soul (Yardbirds).
Last edited:
You are conflating two entirely different things.
Good amplifier tone makes even one single guitar note have a beautiful timbre. Play a single note on, say, a Steinway Grand piano, and no matter how little musical talent you have, that single note will sound beautiful. Even if you have no idea how to play piano, that single note will have beautiful tone.
Musical geniuses play combinations of notes (music) that make beautiful melodies, beautiful harmonies, or complex rhythms. They can create this kind of beauty even if they are playing a plastic bucket with rubber-bands stretched across the top for strings. Beautiful melodies, but NOT beautiful tone.
The most beautiful music results from combining both these things, combined together. That's why David Gilmour and Eric Johnson are virtually in a class by themselves as electric guitarists. Both of them produced fantastic guitar tones, both of them played fantastic combinations of notes.
Other famous classic guitarists from the 50's, 60's and 70's may not have had tone as spectacular as Gilmour, but they had at least average guitar tone for their period. None of them sounded as bad as more recent buzzy solid-state guitar amplifier do, thanks to the inherent (and accidental) musicality of period tube guitar amplifiers they played through.
Not every tube amp is great - but I have never heard a tube amp that sounds as bad, as a bad solid-state amp.
Those of us who are not born musical geniuses, will never be able to do what Clapton, Santana, Gilmour, Johnson, et al, could do as far as sheer musicality goes.
But we can still choose to use an amplifier with good tone, which will make everything we play (even if not genius) sound better.
Or we can settle for a crappy-sounding amplifier that sounds like a kazoo. This will take our limited musical talent, and make it sound much worse. Now we have combined our mediocre musical melodies and harmonies and rhythms with even more mediocre timbre. Not surprisingly, the result will be even more mediocre.
In particular, a lot of solid-state guitar amplifiers, and guitar distortion pedals, fall into the "sounds like a kazoo" category.
In the past, you have commented that you cannot separate playing technique from guitar tone in a video or audio clip. Perhaps an analogy will help: great food comes from excellent ingredients, as well as a great chef.
If you give the great chef (musician) sufficiently bad ingredients (buzzy solid-state amplifier), she cannot make good food (music).
Several years ago, I heard John McLaughlin play on a DVD of one of Eric Clapton's Crossroads Festivals. McLaughlin had fantastic guitar technique, as always. But his guitar tone was terrible. His guitar sounded like a rat being tortured and shrieking in agony. Absolutely awful.
I have literally never heard a professional musician, particularly a top-notch one, sound that bad.
What the heck had happened? McLaughlin had no visible amplifier on stage, and in a subsequent interview with a guitar magazine, he said that he had abandoned playing through a guitar amplifier, and was instead plugged into a small solid-state multiFX pedal, running straight to the P.A.
Apparently McLaughlin has stayed with his horrible new sound. In late 2021, I watched a You Tube interview with Brett Kingman. The interviewer mentioned going to a McLaughlin concert, and being incredibly disappointed by McLaughlin's horrible guitar tone.
When you take a gifted musical genius with incredible technique, and give him a tortured-rat guitar amp, the sound he makes is horrible. All the scales and all the speed in the world don't change that.
Bad amp = bad ingredients. No chef can make good food from those.
I have no explanation for why McLaughlin chose to sound like a kazoo. Maybe he stopped caring? 😕
-Gnobuddy
The debate between us two is pointless. Neither you nor me will change their point of view for one millimeter. That's it.
Maybe he was more deaf than me. Or he was being paid a lot of $$$$ to promote it.
A good chef can do something with bad ingredients. But is usually involves compromise and/or lots of smoke and mirrors: Multi tracking, reverb, echo, chorus. See certain Gilmour set ups, Gilmour can make sawing perspex sound good. All though some of the bootlegs sound a bit screechy - mine and others
Last edited:
Yes I agree a single ended low watt tube amp still gets duty cycle modulation. and there's a difference between the types of distortion you get between that and a push pull amp (cranked). But to me the two don't sound that different for similar settings of distortion. Though I must admit that to get LTP/Power tube distortion in a tube amp of even modest power, it's getting (very) loud. It sounds great but usually at levels that bother my hearing these days. A couple of years ago I was in a music shop where this guy was trying out a new Marshall head, it was a tube version, but with some extra overdrive trickery going on I think. The distortion sounded harsh, and not just a little bit either. I don't know if its the manufacturers trying to appease new young players who want to just "chug-chug" on their guitars and sound like jackhammers and bandsaws..
When I did the three stage jfet (common source) circuit, I measured DC source voltages while I hit a few loud chords on the guitar. From what I recall it would shift around 80 to 100 mV on the second stage, and similar on the third stage. Once the notes started to decay the DC source voltages would settle down to their normal "idle" settings. Just to say that there does seem to be some bias shifting that goes on similar to what can happen with common cathode vacuum tubes.
That man is absolutely a sonic magician.
Bjorn Riis from Gilmourish.com can make himself sonically indistinguishable from David Gilmour:
Some years ago my wife and I were watching a documentary film about something or the other, and there was an appearance by a musician from one of the African countries. He plunked out a repeating melodic motif on his sansa (thumb piano) and started to sing, and it was so beautiful that it moved both of us nearly to tears in an instant.
I have to add that the sansa had much better tone than McLaughlin's multiFX pedal. 😀
-Gnobuddy
Too loud - the bane of all tube amps!
Circa 2010 I found my first ever tube amp (a 15 watt hybrid Super Champ XD) far too loud to play at its "sweet spot", except outdoors in a bad part of town, where one of my friends lived. His neighbours had learned to live with frequent gang shootings, and nobody there would call the cops over something as minor as loud music.
Later I built an attenuator that let me play at home while pushing the amplifier into its sweet spot, but the attenuator itself caused some "tone suck". Or maybe it was Fletcher-Munson curves, my ear hearing things differently at lower loudness levels.
Since 15 watts was far too loud, around 2015 I designed and built a two-watt push-pull tube amp around a pair of 6AK6 pentodes from the era of tube radios. Surely ten times less power would do the trick? Nope, I found out two watts was far too loud for an apartment.
A couple of years ago, I designed and prototyped a single-ended tube amp based around a 6AQ6 preamp, and a little 6JW8 triode-pentode used as both driver and output. Running on 135 volts DC, with a 25k output transformer primary, power output was a thumping quarter-watt or so, roughly 200 -250 milliwatts. One more order of magnitude reduction in power output. Plugged into the stock speaker in my '65 Princeton reissue, it was still very loud in my apartment. But now I was at least in the ballpark. Clean tones were at about the right loudness, though it got too loud for my tastes when overdriven.
I've never owned a tube guitar amp with an LTP, so I've contemplated using the same tubes, and making a tiny push-pull amplifier with two 6JW8s. The triode sections of the 6JW8s could be used for an LTP, and the pentode sections as outputs.
The tricky bit is finding an output transformer with a high enough primary impedance. You'd need a 50k anode-to-anode impedance to use the same load-line as the SE version, and that's not exactly an off-the-shelf item. Mebbe resistors in series...after all, we don't care if we lose some power!
I don't understand the appeal of guitars so distorted till they sound like power tools, either. I guess subtlety is almost a lost art in the era of social networking, "influencers", round-the-clock shock and awe, and almost complete separation from nature.
No doubt! If you put a sinewave through a device that conducts better one way than the other (like a square-law JFET transfer function), you automatically get partial rectification, and therefore, and average DC level that depends on the peak amplitude of the signal.
I wonder if there was also enough signal strength for the gate-source diode junction to become forward biased on positive peaks, flowing one-way gate current and adding to the bias shift.
I tinkered with the idea of adding a diode-resistor network to either the emitter or collector side of a BJT common-emitter amplifier. LTSpice thinks this can produce some very interesting effects. Not only duty cycle modulation, but even dramatic changes to the envelope of the waveform, particularly the transient at the start. (Output waveform in red, input waveform in blue. Input is just a repeating damped sine wave at 200 Hz.)
-Gnobuddy
Attachments
Indeed interesting, again an example of a simpler (non op amp) gain stage behaving somewhat badly which in turn could produce good "lively" sounds.
And in this case, we have considerable control over the amount and type of badness, by tweaking R9/R10/C3 values.
I have no idea what values would actually provide the right sort of musically attractive badness, so some experimentation is called for.
If I zoom into the simulation, I can see the output waveform looks almost half-wave rectified at the start of the note, but looks like an almost undistorted sine near the end. Hopefully that will translate into interesting variations in timbre throughout the note.
I forgot to attach the LTSpice simulation file to my last post, so I'll attach it here, if anyone wants to fiddle with it.
As an aside "Behaving Somewhat Badly" sounds like a pretty good name for a punk band. 🙂
-Gnobuddy
That's the thing right there. Most of the commercial solid state guitar amps absolutely can't do the transitional from overdriven to clean. There's usually two channels, one clinically clean and one overly distorted. The distortion channel doesn't overdrive naturally or very progressively, it's an all or nothing situation. Not to say that they were all built that way, but a lot of them.
I couldn't agree more.
I was just posting on another thread, about a solid-state guitar preamp evaluation board designed and sold by Texas Instruments, and intended to pair up with one of their class D power amplifier eval boards.
That preamp has exactly the failings you're describing. And if that wasn't bad enough, there are other even clumsier mistakes in the design.
Here's a link to my post in that thread: https://www.diyaudio.com/community/...xas-instruments-thoughts.339994/#post-6973018
In 1970, over five decades ago, Electro Harmonix put four capacitor-coupled transistor gain stages into their Big Muff pedal, two of the stages with diode clippers in them. There would have been enough imperfect circuit elements to cause the circuit to "behave somewhat badly", giving some life to the distorted output. David Gilmour made full use of that.
Nine years later, Ibanez launched their Tube Screamer, and settled for a single op-amp stage with symmetrical clipping diodes. The op-amps perfect behaviour ensured the duty cycle modulation was gone, and so was any subtlety in the sound that resulted.
But if you followed the Tube Screamer with a tube guitar amp, the circuitry in the amplifier would add duty cycle modulation, assymmetrical waveform distortion, et cetera to the sound. Eager young guitarists missed out on the second part of that equation, and bought the same pedal their guitar heroes were using, not realizing that you needed to follow the pedal with a tube amp to get the same sounds as those famous guitarists.
In the last couple of decades, we seem to be willing to settle for one op-amp stage with symmetrical diode clipping, feeding a sterile-clean solid-state power amp. The worst of the worst, combined together into one truly atrocious combination. Exactly as you described: either clinically clean, or overly distorted, with nothing in between. The Texas Instruments Sidegig eval board is one more example.
Incidentally, according to Premier Guitar magazine, it seems Boss/Roland already had a patent on assymmetrical solid-state clipping, which forced Ibanez to use symmetrical clipping in their copy of the Boss OD-1, which they sold as the Tube Screamer ( https://www.premierguitar.com/gear/tube-screamer-history ):
-Gnobuddy
I've been working on improving the accuracy of my power tube emulation (see attached anode characteristics – green is an EL34 tube model and blue is my circuit) and I started to wonder about output transformer saturation and whether it actually happens to any significant degree in a cranked guitar amp. Does anyone have any data on the DC saturation current for guitar amp output transformers?
When a typical LTPI is overdriven, it exhibits significant bias excursion, causing duty cycle modulation of the output waveform. This means mismatched average currents in the output tubes and a DC component in the output transformer. Simulations suggest somewhere around 150mA of DC offset for a Marshall-style 50W power amp with the phase inverter driven hard.
When a typical LTPI is overdriven, it exhibits significant bias excursion, causing duty cycle modulation of the output waveform. This means mismatched average currents in the output tubes and a DC component in the output transformer. Simulations suggest somewhere around 150mA of DC offset for a Marshall-style 50W power amp with the phase inverter driven hard.
Attachments
Last edited:
This thread suddenly has more reason to exist. There is a serious shortage of vacuum tubes now, caused by a combination of world events: the closing of the Chinese Shuguang tube factory in 2019, current politics surrounding export of tubes from the New Sensor tube factory in Russia, and the supply chain disruptions caused by the never-ending COVID-19 pandemic.
There were only three working tube factories in the world in 2019, cut to only two after the Shuguang factory stopped making tubes. A couple of weeks ago, Russia banned exports of all tubes made in the New Sensor factory, though that situation might have eased slightly a couple of days ago. That situation remains as murky as the rest of the politics surrounding Russian's invasion of Ukraine.
What is clear, is that there is now only one single tube factory in the world that is still cranking out tubes, and able to sell them: the JJ factory in Slovakia.
The fact that there's currently only one single functioning tube factory in the entire world is bad enough. But if your memory of European geography is a bit fuzzy, take a look at a map, and you'll find Slovakia is slap-bang next to Ukraine, immediately west of it, with a shared border. Putin's invasion of Ukraine puts Slovakia in the line of fire, too.
Not only is there a serious tube shortage, prices have gone up for the small remaining supplies, and the JJ factory is backlogged for months. There is also a shortage of integrated circuits used in solid-state guitar amps. This has forced guitar amp manufacturers to raise prices and reduce production of both tube and SS guitar amps.
Fender says that 16 million people in America alone have bought their first guitar during the last two years, as a reaction to enforced lockdowns and social isolation caused by the pandemic. Demand for guitars has virtually doubled during that time, and guitar manufacturers are selling every guitar they are able to crank out, as fast as they can.
All of which means there may be much more interest in DIY builds of relatively simple analogue solid-state guitar amplifiers over the next year or two, as production shortages of commercially manufactured guitar amplifiers are expected to last at least that long.
While we're not supposed to discuss politics on this forum, I have to say that I am quite aware that shortages of tubes and guitar amps (and guitars) are trivially inconsequential problems compared to the life and death issues the people of Ukraine are facing right now.
-Gnobuddy
There were only three working tube factories in the world in 2019, cut to only two after the Shuguang factory stopped making tubes. A couple of weeks ago, Russia banned exports of all tubes made in the New Sensor factory, though that situation might have eased slightly a couple of days ago. That situation remains as murky as the rest of the politics surrounding Russian's invasion of Ukraine.
What is clear, is that there is now only one single tube factory in the world that is still cranking out tubes, and able to sell them: the JJ factory in Slovakia.
The fact that there's currently only one single functioning tube factory in the entire world is bad enough. But if your memory of European geography is a bit fuzzy, take a look at a map, and you'll find Slovakia is slap-bang next to Ukraine, immediately west of it, with a shared border. Putin's invasion of Ukraine puts Slovakia in the line of fire, too.
Not only is there a serious tube shortage, prices have gone up for the small remaining supplies, and the JJ factory is backlogged for months. There is also a shortage of integrated circuits used in solid-state guitar amps. This has forced guitar amp manufacturers to raise prices and reduce production of both tube and SS guitar amps.
Fender says that 16 million people in America alone have bought their first guitar during the last two years, as a reaction to enforced lockdowns and social isolation caused by the pandemic. Demand for guitars has virtually doubled during that time, and guitar manufacturers are selling every guitar they are able to crank out, as fast as they can.
All of which means there may be much more interest in DIY builds of relatively simple analogue solid-state guitar amplifiers over the next year or two, as production shortages of commercially manufactured guitar amplifiers are expected to last at least that long.
While we're not supposed to discuss politics on this forum, I have to say that I am quite aware that shortages of tubes and guitar amps (and guitars) are trivially inconsequential problems compared to the life and death issues the people of Ukraine are facing right now.
-Gnobuddy
Thanks! You can see that it's missing the kink at low plate voltages, but I'm not sure that it's worth bothering with especially considering the extra complexity required.
As I'm sure you know, that kink is the last remaining trace of the negative-resistance region, which was a massive flaw in the predecessor to the pentode, the tetrode valve.
Philips engineers invented the pentode by adding the supressor grid to the tetrodes, and then patented it. Other valve manufacturers were not particularly happy about being either shut out of the lucrative pentode market, or being forced to pay Philips a pentode licensing fee. The beam tetrode was invented, and by careful shaping and positioning of the electrodes and grids, eliminated the troublesome negative resistance region of the original tetrode.
All that remains of the massive original flaw, is those little kinks at low anode voltage and low anode current...
While I see those kinks in 6L6 and 6V6 and EL34 anode curves, datasheets for some later-generation beam tetrodes have no kinks at all. Evidently the engineers managed to completely eradicate the last vestiges of tetrode misbehaviour before the valve era came to a complete end.
All of this is to say, those little kinks have always been a flaw, and IMO, unlikely to play much part in the sound of a tube guitar amp. I doubt you'd need to reproduce them before you could convincingly mimic output tube characteristics.
IMO, the actual spacing of the curves is much more important. If you read Schade's white-paper on the design of the 6L6, he talks about one of the failings of the beam power tube / beam tetrode: compared to proper pentodes, beam power tubes have high second harmonic distortion.
While that was a flaw for Hi-Fi and other linear applications, that very failing is probably why "6V6's sing the blues, while EL84s rock", as Tubelab George puts it. More singing 2nd harmonic from the beam tetrode, more growling 3rd harmonic from the pentode.
-Gnobuddy
I decided to try to fit the anode characteristics of my solid state pentode emulator to data I extracted from the Philips EL34 datasheet rather than a SPICE model. The black solid lines are my circuit and gray dashed lines are from the EL34 datasheet. No fundamental change from last time, just changing some component values. I suppose the question now is how the circuit performs in practice. I've got a fair amount more work to do on the rest of the circuit, but I'll build an updated version of the circuit I mentioned in post #134 at some point incorporating these improvements.
