Amplifiers

I go into a music store to maybe buy a guitar, I look around find a few I want to try. I check them out, play them and at some point, the salesperson asks me: “Do you want to plug it in to hear what it sounds like?” My answer is: “No, thanks. I am not buying an amp today”. This answer always brings up a very puzzled look on the person’s face and he blurts out: “I mean the guitar, not the amp”. Now I am polite and I would say something like this: “How am I going to find out how the guitar sounds like, you are suggesting I listen to how the amp sounds like”. And of course, the person thinks I am crazy, am I?

The question is, am I crazy or do I just know way too much about this stuff? Well, what I know, may not be “too much”, but I know enough and it is not the guitar that sounds, it is the amp that sounds! I do not know about what you think, but that is a fact. The guitar does not have a sound, it is a transducer that is a coordinating link between the player and the amp. The player produces the string motion and the amp acts as a vehicle for the sound that eventually exits from the speakers that also play a role in the game.

Something that the guitar produces is the output signal from the pickup(s). The strength of the signal, the amplitude if you wish, is determining how the amplifier reacts and eventually what comes out of the speakers. It is a very easy test, take your guitar and your tube amp. Set the amp to something, maybe your favorite setting with relatively high gain, if you have one. Start with your guitar volume knob low, listen to the amp, then turn the knob up a little bit and listen and do the same thing in steps up to full volume. You will hear a significant difference! Oh, why didn’t he mention the “enot” circuit in the guitar? Sorry, you probably call the circuit “tone”! I did not mention this circuit because this circuit barely survives me getting inside my front door, I rip it out of my guitars that fast. I play around with the output from the pickups in an entirely different way as I will explain in the electronics section. Do you understand my conversation with the music store person now?

Amp Design

In this section, I am attempting to clarify why I do what is described above. Now, let me break it to you gently, your sound is YOU, the way you play how you treat your instrument, so to speak. The other main factor is your amp and speaker(s). As described elsewhere, the contribution of the pickups boils down to their construction and their output, especially the output voltage that is fed to the amplifier input. In this case, we are talking about amps so the only thing that matters is the input voltage, how many volts do you feed to the input of the input is the big issue.

You do not under any circumstances feed a voltage to the amplifier that is too high! This might come as a surprise to you, but I will explain why as we go. Too low a signal is a given, it kind of defeats the purpose, does it not? I will, of course, assume that the amplifier has a frequency bandwidth large enough to accommodate our guitar signal something that should be no problem at all. That is, unless the amp designer screws up real bad, and you do not want to come anywhere near those amps! So you can rest easy there!

Amps are designed differently, but I would say as a general rule and I base this on how most amp input stages are designed, so keep the voltage below 1 Volt peak, definitely not higher than 1.5 Volt, and if that level sounds bad, dial it back to 1 Volt! I am saying this because there is nothing on the amp that can adjust that. In just about all cases, the signal goes directly into the first stage of the amp. This stage must be given special care and not only the input signal but also the noise level. That input stage needs to be the protected against injected noise from both outside and inside the amp. It is very important that the input stage is kept “clean”, any changes you want to make can be done in later stages and there is plenty that can be done, believe me. Remember, anything introduced at the input is amplified through all stages!

From the early days we have amps that were very simple designs, mostly for economical reasons but also for the simple reason that not many types of components were available: Tubes, resistors, capacitors, transformers and inductors. Tube component design has changes very little since back then, a few improvements here and there, mainly on design for reliability, otherwise there was really no need to change anything. So the tubes made nowadays are just as good, if not better that the old ones. Resistors and capacitors are much, much better now than then. Back then it was mainly carbon composition resistors, that as an electronics design engineer for many decades have come to seriously dislike, they are noisy, they are drifty, they have a very short lifetime. Absolutely no good for anything, oh wait, I have used them when you need to absorb a lot of heat quickly for a short period of time and after that nothing! Otherwise they are totally unreliable. Today, use metal film resistors. Same thing with old capacitors, way back I had to use oil-paper capacitors in a design, they are horrible, again use modern parts, film capacitors for any signal application, electrolytics for power. Transformers have not changed much in design that is, it looks like manufacturing has improved. Now, don’t get me wrong, all transformers are not equal, a transformer must be designed to fit the purpose. Power transformers are designed to handle a much narrower frequency range that the output transformer and must therefore be designed accordingly.

Today, there are just so many different component types available, but the old tube amp design is still going strong. Be careful, a lot of manufacturers like to put a lot of new functions into their amps using opamps, transistors and diodes, some for good, many for bad. I don’t know about you, but I like it simple. I prefer designs that are along the lines of the old amps, but with the economical shortcuts removed and using modern components. Don’t use those completely outdated components like carbon comps or oils, they are just being touted for hype and marketing, functionally they have a negative value, because they are bad. Any enhancement value claimed in connection with these components is dead wrong!

Amp Structure

That aside, let us take a look at the structure of the amp. We have already discussed the input stage and its importance, but the following stages serve just as an important function, I guess you can say that they shape the sound of the amp through the control of the overdrive or distortion as well as the EQ part where you can set the filtering for a specific frequency content. Finally, there is the output stage or power stage that can be designed in many ways, depending on the output power of the amp, typically class A for low power or class B (or A/B) for high power or possibly a combination. After that, there is the transformer, not to forget, because the tubes operate at a high voltage, several hundred volts, much more than the speakers can handle, so before the signal can be transmitted to the speaker(s) it needs to be transformed down to a few tens of volts. As far as the shaping of the sound, both transformer and speaker have important roles, so it is important to get the right equipment with components that complement each other.

We are not going to get deep into the design of amps, there are quite a few books about this, but please get the textbook type if you are really into it and have the background knowledge. There are a few good ones out there, there are also a lot of bad ones, even bad overpriced ones! I was dumb enough to buy one of these, it was like a spiral bound one on how to improve your “tone”, cost me about 80 bucks, not worth it!

As you can probably tell, I am itching to get deeper into the design of amps, but the purpose of this page was to discuss other issues, sort of “from the outside” issues, such as the guitar. Hence the name of this website. There is one thing that I can elaborate on here, something that I mentioned on the “Shielding” page, just to show how I would attack the signal and shielding for an amp.

Grounding

One thing to talk about is Grounding, and you know how particular I am about this. As mentioned earlier, an amp operates at high voltage so all exposed metal parts must be connected to Ground and I am talking about real Ground, the rod rammed into the ground, Earth Ground. Regulations for electrical equipment require that any equipment that has internal voltages 50 Volts or above must observe this, in order to be approved, it got to be by UL or CE or other authority dealing with safety. You may have heard about a term, “death capacitor”. Something that was used in the past when the amp power cord only had two wires and grounding was sparingly used. This capacitor was not and will never be safety ground, but it is merely noise decoupling placed in the power input to the amp. This type of approach is common in EMI engineering to mitigate noise, and only noise. As I have explained, the type of equipment I have worked on, the Ground was always completely separate from any signal wiring. And the Grounding was there primarily for person safety. So death to the “death capacitor”, it is a good thing it has vanished!

Fig 1. Amp design for signal, shielding and Earth Ground

The way I would approach the design is shown in Fig 1, the box marked “Amplifier Box” is the actual amplifier circuit itself (calling it box, may be a bit misleading). As indicated the amplifier is isolated from the so called “Shield Box” and this shield box is connected directly to Earth Ground, indicated by the well known symbol in the lower left corner by the “Power Line”. This symbol indicates the rod rammed into the ground as I have mentioned a few times before. Besides the shielding, the input jack differs from what is normally found in guitar amps, it is what is known as of the “stereo” type, not that this is a stereo amplifier, but it is a way to keep the signal and shielding separate. As indicated, the “sleeve” of the jack is connected to shield, and only to shield! Now, you do not see this arrangement in any amp, unfortunately, but even with this type of wiring you would be able to use a regular guitar cable, which means that the shield is connected to one of the signal wires back at the guitar. As indicated on the Shielding page, strings and shielding will act as safety ground connected through the cable to the amp where the “main” Earth Grounding takes place. At the amp input, the sleeve of the cable makes contact with both the ring and the prong connected to the black wire inside the amp, so the safety grounding works as well as the signal, not the best way, however!

The purpose of discussing Grounding is to make it clear that the purpose of what I have explained is to design amp and guitar in such a way that maximum personal protection is provided and a minimum of noise reaches the output of the amp. Simply, keep all interference out and connect any metal part you can touch to Earth Ground for safety.

Some misconceptions

Like I mention in other parts of this website, I watch a lot of YouTube videos, not so much to learn about guitars, pickups or electronics. I already have a bunch of knowledge in those areas, I watch to find out exactly what these people are saying and more importantly what they get wrong. Sad to say, they get a lot wrong, which is why I decided to put together this website. They keep getting stuff wrong and I keep watching and updating this site.

Here is a good one from a pickup maker and YouTuber whom I have quoted in other places on this site, he said: “Once you lose treble in your guitar pickup or other circuitry, these frequencies are gone forever”. You may have noticed why he is wrong about that, specifically if you know about amps and what happens when they run at higher gain, the start to “break up”. The sound gets more and more distorted, and what gets added? You are correct, higher frequencies! What happens when a signal gets distorted in the amp, it gets “squared up”, it gets flat at the top, so to speak. When a signal, and we have used sinusoidal signals because they represent a clean note which makes it easier to determine the change in the signal, gets distorted more and more higher harmonics get created. I said squared up, which alludes to the signal get rounded off at the top getting closer to looking like a square wave, which per definition can be dissolved into containing an infinite number of frequencies.

That is why I was talking about limiting the input signal to the amp earlier, if the gain setting is constant the higher the signal amplitude the more the waveform is distorted! You might have seen elsewhere what I say about judging any improvement that they just made to the guitar, simple truth is, the sound will be different if the pickup output voltage is different! Maybe that is the same thing when people are switching magnets, for example. As you can see, and I have mentioned it elsewhere, that it bothers me when people do this as “proof”, I don’t know. Some are guitar teachers, maybe they should find another way of advertising their business. Hey, maybe they love to play, that’s fine as long as I don’t have to listen to it and to hear them talk about mellow highs and mid-scoops and all that BS.

There is plenty other BS, especially when it comes to so-called “boutique amps”, here the BS is really flying! Please duck! Claims like hand built is better that PCBs, that is nonsense. Using a PCB is actually much better, if you know how to design these for the purpose. I can say that I have quite a bit of experience with PCB design after having supervised the design of a large number of circuit boards some for very tough applications, high voltage and power much higher than 100W. You see amp design books operate with a thing they call “star” grounding where you run all wires to a single point on the chassis. Designating one specific point to being the “ground” point is not necessarily bad, just not the way it is done there. The good thing about PCBs is that it is easy to create a Bus structure, mainly consisting of the DC power traces on the board. As mentioned above, it is not really necessary to connect any part of the signal amplifier to Earth Ground, but way too often you see that Earth Ground and Common of the circuit is confused to be one and the same thing! No need for that! Now you say , what is Common? Common is the amp’s negative DC Bus for example, sad to say this Bus is very often and wrongly referred to a ground. Hence the confusion, which is why this text is dedicated to righting that wrong by explaining what Earth Ground really is.

There is probably a lot more to add and go into further detail with, so there is more to come!

More Amp Design (with some repetition)

It is usually my approach to look at electronic designs from the most modern view, doing my best at incorporating the latest available components which has always been the engineering way to go, the latest components are the products of the longest development and based on prior experience with similar type parts. Now, as it is supposedly common knowledge, I have designed electronic circuits for products for very demanding applications so a lot of care has to be put into the design, but since these were industrial products, product cost was also an important factor. I go so far back that I did this when we were just about the only company designing these types of product, but at the end of my career there were so many companies manufacturing these that the competition was so fierce that you would pick the right design approach. In that long period, a lot of new components and a lot of new techniques had been developed and everything had to kept up to date. The important part is that the performance and quality of components underwent a vast improvement during this period. Bottom line, this is what I experienced as “the way of life” so that is why I believe in developing and improving products. In many instances, I have been lucky enough to be involved in improvement and development of components that were used in the products we developed.

Starting with vintage amps

Based on the above, it should not come as a surprise that the following views on designs and components is what will be expressed in the following even though it is 40 years of industry and research experience that is talking, so why would I start with vintage amps, amps that are almost as old as I am, sometimes older. First, however, let me say something about electronic circuit design in general. Typically, a design starts with a specification that in some detail describes what this circuit is supposed to perform. More often than not, two engineers would design circuits that would meet the same requirements and display the same performance, but you examine the circuits and find that they look very different. Morale is, none of the two are more correct than the other, basically there are more than one angle to a certain performance. Something to remember before judging a circuit design.

In our specific case here we are discussing amps, and by that we mean guitar amps. Circuitry wise, the god part is that these amps are pretty simple layouts and component choice even though some functions are harder to understand than others. Even better, the design of tube amps has not changed much since the beginning. Unfortunately, as I have observed, some of the design mistakes have prevailed to a point that they have survived up to present day. At least, some design practices are still alive and well in present day. As far as I can tell, for purely commercial reasons, not functionality. For instance, using circuit boards, PCBs, is something I have used in design for almost my entire career, mostly for low power designs, which is less than a few kW (kilowatts). For higher power circuits, a different approach is necessary, of course, but these products still include circuit boards. But properly designed, a circuit board design will by far surpass designs as seen in amps that are so-called “hand wired”.

In an amp design, it is important to realize what contributes to the “sound” of the amp and which components do not have any influence. Here, of course, we do run smack into various myths and fairy tales that have developed over the years that from a technical point of view have no merit. One example is carbon composition resistors, these resistors were used basically because that technology was one of only a few resistor types available back then. Since electronics has developed significantly since the early days and lot of different resistor types that are far superior in both size, performance and quality, basically leaving carbon comp resistors at the bottom of the barrel. Why this type of resistors is still in use is a mystery to me, they are noisy, they drift in value, they deteriorate in other words, they are bad and should be considered undesirable for modern designs where so many different resistors are much better, cheaper and smaller than those old carbon comps. I have used a carbon comp resistor in a design once. A good side of carbon comps is that they can tolerate high peak power if limited to “once in awhile” application of power. Because it is a “bulk” component, it can absorb high power on a single pulse basis. Unfortunately, as we came to experience, this carbon comp would catch on fire, so we nick named it the “fire log”, needless to say it was subsequently eliminated from the design, because if it did not burn up it could split in half, invisibly, so it would quit working.

My recommendation would be to use metal film, or metal oxide film resistors for by far the most applications, for higher wattage resistors Lo-L wire wound resistors, typically 10 W and up.

Now that we are at it, a few words about capacitors. Forget about these oil-paper capacitors that are out there at astronomic prices. They are not worth the money, better spent it on good quality film capacitors for the “signal” applications, such as the coupling from stage to stage and if possible, AC decoupling capacitors for cathode biasing. Electrolytic capacitors have improved tremendously since I started my professional career in electronic design. The parts available today are much better quality and performance than back then, when manufacturers actually recommended changing them after maybe 5 years because these parts would actually dry out and lose their capacitance.. As I have mentioned, the products I have been involved in designing were actually somewhat “hard” on electrolytics. Besides the voltage rating, they also have a ripple current rating. It is obvious to stay within the voltage limitations, ripple current is a whole other story. Luckily, in tube amps, we do not have to worry too much about ripple current if we pick new capacitors of a reasonable capacitance value. I emphasize value, because a capacitor is NOT a rating, it is the value. I think I mentioned this before, but it cannot be repeated enough. You do want to sound like you know what you are doing, right? In case you work on old “vintage” amps, the electrolytics definitely need to be changed out and please use brand new ones.

Transformers (and chokes) technology wise have not changed much, maybe except that they have become more efficient through use of better materials and manufacturing methods. Also, design tools have become much more accurate, well, in the beginning they only had paper and pencil plus trial and error. Now you can design a transformer from ground using sophisticated software. Bottom line, if old transformers are still good, there is no need to change them and please don’t save them because of the mythical belief that they have a magic contribution to the “tone” of the amp. They Don’t!

Tubes, oh, here we go! You can pretty much say the same as said above about transformers, except from an extremely important difference, tubes DO have an influence on the sound of your amp. There are books written about this, so I won’t mention it here in much detail.

Sometimes you see ceramic capacitors use in amp circuits in the signal path. I have some advice, if possible, use film capacitors, the come in similar voltage ratings and capacitance values.

A practical example, Gibson GA-77 (1954-version)

Recently, I viewed a video on YouTube that was about the repair of the indicated amp. I have watched quite a few amp repair videos online, not so much to learn about amps, but more to see what people say, how they explain things and most importantly to get an insight in how amps were designed an built almost 70 years ago. What does an amp look like when it is almost as old as I am!

When the person explained the circuitry, he got to the output tubes which were 6L6GA, and there were some peculiar and not common design and parts around these tubes. There was an attempt to built in a “stand by” switch that would disconnect the cathodes of the two output tubes (the two cathodes were connected) from their common bias circuit which was a resistor – capacitor to B- which is the negative DC bus of the high voltage power supply (I do despise the “ground” used in this connection, as explained elsewhere). This was unusual to begin with, but here is something even weirder, see Fig 2.

Fig 2. Gibson GA-77 (1954-special) Screen shot from video presentation

If you look to the left and below the 6L6 tubes, you see the stand by switch to the left and how it connects to cathodes. This switch has 3 points of contact, ON, STAND BY and OFF. The “other end” is connected to the bias circuit 150 Ω and 50 μF capacitor that again is connected to B- (wrongly indicated by an Earth Ground symbol). In OFF and STAND BY positions, this point would be “floating” un-connected to the cathodes, such that no bias nor screen current could flow. In the OFF position, the power transformer is disconnected from the power line, so this situation is not interesting. On the other hand, if the switch is in STAND BY position, the power is on and the bias network is not connected to the cathodes, but here is the weird part. If you look to the right of the tubes and below, there is a 4k resistor connected to the screen with a 10 μF capacitor to B-, so far it is normal. There is, however, a 25kΩ resistor from the screen point to the switch point of the bias circuit. With the switch in STAND BY, we have a series connection from B+ to screen through the 4kΩ resistor, through the 25kΩ resistor to the RC bias circuit to B-. The capacitors in this case do not have an active role in the circuit other than adhering to the voltage determined by the resistor chain between B+ and B-. With a total of 29,150 Ω and assuming the supply voltage is 300V, there would be a voltage of about 1.55V on the bias circuit. So the bottom 50 μF capacitor is charged to 1.55 V when switched to ON position, where after it would be charged to the bias voltage of around 20V, (I could not find a 6L6GA data sheet). Now we also have screen current flowing, but that is not the important part, even with the tubes now being biased and the current through the 25kΩ resistor is added, it does not affect the operation of the amplifier output tubes. Why is the 25kΩ resistor there? First thought is that it helps the transition from STAND BY to ON, by the capacitor being charged. Could it prevent the so called current inrush phenomena, even though the first few volts are typically where the surge happens, it is hard to see that with only less than a couple of volts across the capacitor prior to switching (and not 300V like it was stated in the video, I believe). Under normal operation, it does not do any harm (the 25kΩ ), but it might increase the bias voltage a bit, but that is somewhat irrelevant. So why is the 25kΩ resistor there? Hard to tell, since there is apparently no practical explanation because it being there makes no difference. As I have mentioned, sometimes designs or the mind of the designer is sometimes unexplainable, so the original intent is not clear. Also, why the cathode bias? Don’t know, but here is a hint, I was able to find numerous schematics of GA-77 (and other GAs) and NONE of them had that 25kΩ resistor there and all of them had fixed bias! One of these small mysteries that you sometimes are dealt. A fluke?

One more thing, please notice the capacitor in the bias circuit, the value is 50 μF ! Why such an insanely high value? Is it just mindless copying, like “two tube to bias, half the resistor value and what about the capacitor, oh we’ll just double it”. That is not the right way to do it, a much lower capacitance could do the trick, that capacitor would give a corner frequency of 22 Hz. Even if we cut the value in half, it would give us 44 Hz.

Finally, it was mentioned in the video that they had attempted a bus structure design for the power distribution. I love that, I have used that approach my entire career in electronic circuit design. Not something you see much at all in guitar amps, even modern ones, except maybe the PCB based amps. Too bad.

Also this guitar amp is not using any grid stopper resistors in the instrument inputs. It is always a good practice to use some resistance there and for all in the world, do not use carbon comp resistors, they are very noisy, the more the higher the resistance value.

Next Step

Now, it has been some time since I wrote the first part about guitar amplifiers and it has occurred to me that I cannot get away with just that, I would have to write some additional and more specific about how a signal from the guitar enters the amp and how this signal develops to the point where it is presented to the amp’s speaker. So please stay tuned, I need to create some figures to make the explanation easier to understand. This is of course based all the misinformation that is being pumped out on YouTube.