Did My Fan Sabotage My Guitar Amp? A Strange Discovery and Solution

[yungman]

Bizaare experience today!

There is no question here, I just want to share a really bazaar experience that I desperately working on for like 4 hours today:

I am designing a guitar amp with high gain. I did run into sort of " motor boating" due to inadequate filtering of the B+ a day ago and I fixed it. So today I was going to finish up the amp and did final testing.When I did the final test, I can hear a tremolo sound when I hit a few notes. I went through the whole thing looking at the various HV point with a scope, play around with the ground, filter caps and all. I was stumped, I just cannot figure this out.

It's kind of warm today and I have a big fan close by blowing towards me. All of a sudden, I remember the spinning blades can give reflection of the sound. Just like if you stand in front of the fan and talk to the fan, you'll sound funny like you have mucus in your throat and has a vibrato.

I turn off the fan, everything back to normal! This is a crazy day for me. I thought I post this for a laugh!

BTW, for someone that is into tube amp, here is a little discovery. Do not put a filter cap right after the choke! It makes a good resonance circuit. It was giving problem this morning, I give a tremolo sound also, I can see the junction of the choke and the filter cap give me a slow ring when I hit the guitar hard. I was puzzled for a little before I figure out it's the resonance. You need to have a resistor at least 2K to dampen the oscillation. After I did that, then it works. That's actually part of the reason I got fooled by the fan. The symptoms are very similar. I was flat out of idea before I started to think out of the stupid box!

 

[Kholdstare]

"I remember the spinning blades can give reflection of the sound. Just like if you stand in front of the fan and talk to the fan"

Hahahaha ! That's "fan out" effect for you yung! LOL!

 

[Bobbywhy]

Yungman, very interesting discovery, that fan modulating the sound in a feedback loop!

In anti-submarine sonar systems there is a condition where the transmitted acoustic pulse strikes the submarine target in the “aft” aspect, which is where the pulse strikes the rotating propeller (screw). This gives a large and complex spectral shift to the echo, similar to the “tremolo” sound you heard. Sonar operators are trained to recognize this when they hear it. And they know what causes it: the submarine commander is trying to get away!

In airborne military fire control radar sets, an analogous phenomenon occurs when the radar is located in the “tail or aft” aspect of the target. In this case the pulse gets reflected from the jet engine turbine rotor blades and, once again, reflects a complex doppler-shifted echo. The twin-engine Phantom F4 had counter-rotating J-79 GE engines. I remember trying to explain to aircrews why the returning echo from the back side of the F4 was so bright on their radar scopes. It was called “turbine modulation” then. I expect today’s digital processors deal with it easily. Also, today’s “stealth” technology includes covering the engine turbine blades to avoid these reflecting surfaces altogether.

Cheers,
Bobbywhy

 

[skeptic2]

Yungman, here is a two part article about why tube amps and transistor amps sound different.

http://electronicdesign.com/article/analog-and-mixed-signal/-House-of-Fire-Firebottles-And-Groove-Tubes-Versus-Devices-That-Find-Their-Origins-in-Sand-Part-1-

http://electronicdesign.com/article/analog-and-mixed-signal/tubes-versus-solid-state-audio-amps-the-last-word-or-house-of-fire-part-2-

 

[yungman]

I tend to think there's a lot of hype on the different resistors and capacitors. I agree the sound is very sensitive to the value of the components. If you change the value of the coupling cap, in electronics, it only affect the BW, but when you listen to it, it sound different. Yes, human ears are very sensitive.

But I still not buy into the different resistor materials, different capacitors and opamp. I still think it's psychosomatic. I am not saying the article is wrong, voltage coef on components exists, but I don't thing that is the gating factor. There is very little money in this guitar electronics field, you don't attract high caliber people. Everyone is just copying each other. I started out working of guitar amp and I moved on to bigger and better things for the last 28 years. Now I am doing it just as a hobby and whatever come comes. I am so surprised how little the design evolve since I left the field. Everybody copy each other and myth flying around.

I just started working on the tube amp. But I have been working on effect pedals for a few months. I can assure you I don't buy the myth. I use metal film resistors, ceramic disc caps, tantalum caps, any opamp that fit my needs, none of the carbon resistor, orange or blue caps, opamp 4558 type of crap. I see to it that I don't use any of those. I actually have people interested on my designs.

There is a lot of the "older is better" thing going on also. People keep going back to the older stuff in the 50s. It's so funny that you have to pay more for those junk alnico speaker that was consider to be cheap junk in those days. They want the "vintage" sound! This I called "old farts". Blame this on the baby boomer that refuse to fade away and try to prevent the world from turning in my opinion.

 

[Integral]

I thought that this was a shopping thread



A Bazaar can be bizarre

 

[yungman]

I thought that this was a shopping thread



A Bazaar can be bizarre

Can you change the word for me?

 

[jim hardy]

I have long believed the "tube sound" lies in the amplifier output stage.

A pentode tube approximates a current source, and inductance of output transformer can briefly give it the near infinite voltage of an ideal one.

A solid state amplifier generally uses feedback to make output voltage mimic input voltage.

So the two amplifiers will perform identically with a resistive load.

BUT---- A speaker has resistance, inductance and inertia. Actually it's a linear motor.

Force a square wave current through such a load and observe its response.
Voltage will have spikes because of Ldi/dt.

Force a square wave voltage across such a load and observe its response.
Current will have rounded corners because of Ldi/dt.

Force on cone is current and the two currents are not the same.


i once worked on an early Dynaco soid state amplifier for a neighborhood guitar guy. He said it was his favorite amp because it had "the warm tube sound". I noticed it derived its feedback voltage not across load but from a current sensing resistor in series.

what you think, yungman ? am i crazy?

 

[yungman]

I have long believed the "tube sound" lies in the amplifier output stage.

A pentode tube approximates a current source, and inductance of output transformer can briefly give it the near infinite voltage of an ideal one.

A solid state amplifier generally uses feedback to make output voltage mimic input voltage.

So the two amplifiers will perform identically with a resistive load.

BUT---- A speaker has resistance, inductance and inertia. Actually it's a linear motor.

Force a square wave current through such a load and observe its response.
Voltage will have spikes because of Ldi/dt.

Force a square wave voltage across such a load and observe its response.
Current will have rounded corners because of Ldi/dt.

Force on cone is current and the two currents are not the same.i once worked on an early Dynaco soid state amplifier for a neighborhood guitar guy. He said it was his favorite amp because it had "the warm tube sound". I noticed it derived its feedback voltage not across load but from a current sensing resistor in series.

what you think, yungman ? am i crazy?

That might be a good idea, sound pretty convincing to me. I have to try it when I have a chance. I am really not that experience with tube amp. I did one that sounded very good in 1978. I use variac to adjust the voltage of the power amp only to dial it into the sweet spot. I just started working on this two weeks ago only because my Marshall that I use for all the testing blew up!$%@&!( I type this, not being censored out!:tongue2:). I have no choice but to pull my old Fender Bassman 100 out and use it. It was broken too! $%@&! I had to open it up. So while I had it out, I start designing my own idea into it. It sounds good to me, but I yet to have to have people come over to judge. I mainly design the front end, the only thing I did to the power amp is to put in a switch to lower the grid bias to bring the tube into class A by lowering to -35V instead of -55V. It did improved the sound quite a bit. I yet to put in the cathode bias and power adjust.

I do talk to people on other guitar amp technical forum, consensus are again, people get use to the sound of the good old days. The solid state amps can sound very good clean. In my last days of performance in the early 80s ( after I officially quit but just kind of do it on and off for fun), I used a Yamaha cheap solid state, it sounded really good. Just because they sound slightly different don't mean it is not as good. BUT this is market driven. If people want tubes, you give them tubes!

We are in the odd ball group that we adapted to forever changing technologies. Most other people don't, a lot...a lot of the baby boomer are still rocking and rolling, they refuse to move on and they refuse to change. The guitar technology is still stuck in the 50s. They still insist to have guitar pickups wound on fiber bobbins, using cloth wires!

In the last two weeks, I really learn a lot on tube amps. My thinking for one, is that...Tube has a plate curve very dependent on the plate voltage. You change the sound by just change the voltage! So you can dial into the sweet spot. BJT acts too much of an ideal current source, you can't play as much tricks. FET is in between. They are just different animal. On the tube amp, you change the sound by just changing the parameter a little. The imperfection of the carbon comp resistors and the caps with voltage coef like in the article provided by Skeptic2 mention might just change the operation of the tube a little. I have to put in more time before I talk more as I don't want to come out like a fool! But for solid state guitar pedals, I category reject the idea of special caps, special resistors etc. I believe people just copy each other and try to change a little here and there without really understand what's going on. People still stuck with designs of the 60s. It is like insisting of use 8085 processor, then you argue about which brand of 8085 perform better! There is no money in this field!

 

[AlephZero]

It's kind of warm today and I have a big fan close by blowing towards me. All of a sudden, I remember the spinning blades can give reflection of the sound. Just like if you stand in front of the fan and talk to the fan, you'll sound funny like you have mucus in your throat and has a vibrato.

I turn off the fan, everything back to normal! This is a crazy day for me. I thought I post this for a laugh!

I think you just reverse engineered this: http://en.wikipedia.org/wiki/Leslie_speaker

 

[jim hardy]

i have a fistfull of National LM12CLK's left over from a project

it's an 80 watt op-amp in a 4 pin to3 package
http://www.ti.com/litv/pdf/snosby8c

plan to try an audio amp with feedback selectable from output voltage or current

but it's one of those "will i ever get to it" projects.

Tube has a plate curve very dependent on the plate voltage.

triode yes pentode much less so look at 6AQ5 for example above 100 volts it's very flat for any given control grid voltage.
http://www.duncanamps.com/tdsl/link.php?target=0125A995


but you knew that
old jim

 

[yungman]

I think you just reverse engineered this: http://en.wikipedia.org/wiki/Leslie_speaker

Yes, I totally forgot about this one.

 

[yungman]

i have a fistfull of National LM12CLK's left over from a project

it's an 80 watt op-amp in a 4 pin to3 package
http://www.ti.com/litv/pdf/snosby8c

plan to try an audio amp with feedback selectable from output voltage or current

but it's one of those "will i ever get to it" projects.


triode yes pentode much less so look at 6AQ5 for example above 100 volts it's very flat for any given control grid voltage.
http://www.duncanamps.com/tdsl/link.php?target=0125A995but you knew that
old jim

Yes, I was mixing pre amp tube in and talk about it. I really have not got into the power amp yet. I am more interested in finding a way to always run the power amp at optimal. There are power attenuator that reduce the volume to the speaker. I have a THD Hot plate. That's the only way I can keep my sanity as I can't take the noise anymore. I can dial it down and crank the amp to full power and work on it in midnight. They did a pretty good job in compensating the sound in reasonable attenuation ( up to -12dB).

So far, other than biasing to class A area, the only other thing I did was changing the output transformer to a much lighter 40 watt one instead of the original huge transformer. Nobody use 100W anymore, everyone try to power down. That big transformer must be like 15lbs! I was hoping to get a little more compress sound from the smaller transformer, but so far I don't hear any noticeable improvement. I really need to get back to my pedals which is more immediate. This is really a big side track as I won't have an amp if I don't work on it. Besides the Marshall is not the best amp in the world, the one now is a mile better sounding than the Marshall.

The one I designed with variac in 1978 sounded very good, I have recording to prove it. It's just a lot of extra components. Profit margin is so low that it's not worth reproducing. I am going for other ways.

Funny one time I was only using the cheap Yamaha 50W solid state for a gig, I happened to crank it to full power because of the situation. It sounded so good my keyboard player actually turn over and said "It this for real!" There was a " Is it real or is it Memorex" moment. Solid state can sound very good.

One thing I think make it hard for people to design Guitar amp, you really need to know how to play guitar, particularly you have to be at least decent player to work the amp up to potential. Or else you really don't know what you are looking for and you cannot get the sound even if you nail the design. The sound really depend as much on the player as the equipment. It's like the saying in the audiophile, the system is only as good as the weakest link in the system and the player is part of the system.

 

[Mike_In_Plano]

In my early days, I loved designing, building, and testing amplifiers. Tube types, Bipolar, Mosfet, and a couple of class-D. One was even a class-D made from bipolars (1983).
I noted that there were differences in the devices with extreme open loop gain, versus those with lower gain or nested gain (like Harmon Kardon). But most of all, the tube amps varied. Speakers, tubes, operating points, and output transformers all had interactions.

Discussing this with a friend, I speculated that this may be due to the presented output impedance. After all, the ultra high damping ratio (low output impedance) amps of the 80s definately sounded different from their lower gain brothers.

So, I performed an experiment. I built two identical output stages, with identical drive, biasing, and nested feedback. Each used complementry MOSFETs and a the typical long tail pair and split pole voltage amp.

One, I treated normally, by feeding it back to the inverting input. But with the other, I place a low value resistor in series with the speaker return and feed this back into inverting summing node. This I did to spoil the output impedance of the amp.

I met with my friend, and we swapped the two amps with a common speaker, and there was a drastic difference in sound. At the time, I attributed this to the somewhat convoluted impedance of the speaker and how that impedance reacted differently when driven from different source impedances.

Mike

 

[jim hardy]

But with the other, I place a low value resistor in series with the speaker return and feed this back into inverting summing node.

what you describe is not unlike my friend's old Dynaco .


edit do i understand it, that'd make Zout larger---> more like a current source?

 

[yungman]

In my early days, I loved designing, building, and testing amplifiers. Tube types, Bipolar, Mosfet, and a couple of class-D. One was even a class-D made from bipolars (1983).
I noted that there were differences in the devices with extreme open loop gain, versus those with lower gain or nested gain (like Harmon Kardon). But most of all, the tube amps varied. Speakers, tubes, operating points, and output transformers all had interactions.
What is nested gain?
Discussing this with a friend, I speculated that this may be due to the presented output impedance. After all, the ultra high damping ratio (low output impedance) amps of the 80s definately sounded different from their lower gain brothers.

So, I performed an experiment. I built two identical output stages, with identical drive, biasing, and nested feedback. Each used complementry MOSFETs and a the typical long tail pair and split pole voltage amp.
What is that again, I am not familiar with these terms.
One, I treated normally, by feeding it back to the inverting input. But with the other, I place a low value resistor in series with the speaker return and feed this back into inverting summing node. This I did to spoil the output impedance of the amp.

I met with my friend, and we swapped the two amps with a common speaker, and there was a drastic difference in sound. At the time, I attributed this to the somewhat convoluted impedance of the speaker and how that impedance reacted differently when driven from different source impedances.

Mike

Do you mean two identical amp, one have simple closed loop feedback to lower the impedance first, then drive the speaker. The second amp have the speaker in the feedback path? If so, that is a completely different thing. The voltage across the speaker depends on the impedance of the speak at that particular frequency. It's neither constant voltage nor constant current. Well bottom line, which one sounded better?

 

[Mike_In_Plano]

Yungman,

Yes, I had essentially two bulding block amps. Each resembling an op-amp from the black box perspective, The first was constructed as your typical non-inverting amp, while second used some degree of the voltage feedback and some of the sampled current.

Each was tested on a cheap stereo speaker. Neither exhibited any instability, and neith exhibited fuzz or clipping.

The normal one sounded normal, while the one with impedance run up seemed to emphasize certain bands in the upper mid range and lower mid range. Bass and treble seemed suppressed.

We were both thinking the speaker's characteristics were coming to dominate the response.

Mike

 

[carlgrace]

I think you guys (and the article linked) are making this too complicated.

Tubes have more even-order distortion that transistors (for a given level of odd-order distortion). For whatever reason, humans find even-order distortion more "warm" and "musical" than odd-order distortion. Simple as that.

Also, many VOX guitar amps have a tube in the output stage but are solid-state everywhere else. They sound great (I can't tell the difference) and weigh a TON less than a Marshall amp of equivalent power.

 

[Septivalian]

Does the resistance impact the oscillating factor?

 

[Studiot]

Old hands may remember a fault with valves (tubes) and some early transistors called microphony.

That is not what was happening here, but it is due to the electrodes vibrating in sympathy to some mechanical excitation and translating this into an (additional) electrical signal.

Sorry, just realized how old this thread is.