Certain methods of handling components and installing them can make the already “easy” task of PCB assembly even simpler.
With any kit or project, it is best to organize the components. If it is a kit, empty the parts out of the bag, identify them, group them, and make sure everything is present. Look at the kit notes and make sure you have a reasonable understanding of what will be involved. Kit notes are not intended to be circuit tutorials – that is what the books are for – but they will show you how to assemble the given circuit and how to tie it in to other circuits as appropriate.
Some components stay cool during operation, so they can be pushed all the way down to the PCB. That is, their leads are inserted into the holes, and the body of the component made to rest on the PCB. On the solder-side – also called the “foil side” or “copper side” – before installing the components, bend their leads outward or inward slightly. This will keep the part from falling off the board when you flip it over to solder the leads.
Install a few components at a time,starting with parts that mount fully pressed down to the PCB. This includes diodes and low-wattage resistors (1/4W, 1/2W, 600mW).
When soldering each lead,a proper connection should take less than ten seconds. Prior to soldering, add solder to the tip of the iron and then wipe the soldering iron tip on a wet sponge. This removes the solder you just applied and “wets” the tip – primes it with solder so it can more easily heat the solder you wish to apply to the connection. Touch the tip to the lead and trace. Touch the solder to the tip and then move it to the opposite side of the lead. Solder will flow around the lead and cover the pad. This should happen in just a few seconds. As soon as the solder has flowed all the way around the pad, remove the iron.
You can solder several connections in a row with each treatment of the iron tip. Never wipe the iron tip on the sponge prior to placing the iron in the holder. Leave it dirty, and even add some solder to the tip just before placing it in the holder. At the beginning of the next set of connections, wipe the tip on the sponge. Double check that the tip is shiny with solder and no dirt is present on the tip.
Trim the excess leads of the components you’ve just installed. Use flush-cutting snips to cut the leads just at the peak of the solder slope around the lead. There is no benefit in leaving long leads, and in fact, longer leads may cause electrical hazard or failure.
After all of the short components are mounted, move to taller items. If there are semiconductors other than diodes, now is the time to mount them. Push the leads through the holes but leave at least 1/4″ (6mm) above the PCB. The splaying of the leads going through the holes will provide enough friction to hold the parts in place when you flip the board over. Again, solder quickly but effectively. Do not cut across multiple leads of transistors, etc. This can reduce the effective spacing of the leads. Snip each lead individually.
Power resistors must be mounted with space between their body and the PCB. Just because a flame-proof resistor won’t burst into flames, one cannot assume that it won’t get hot enough to burn the PCB. Half-an-inch or 12mm is good here. Hold the resistor body between your thumb and finger. Use your other hand to gently bend the leads perpendicular to the body. They should curve out away from the body rather than bend at a sharp angle. Insert the leads into the holes and push the part down. You must squeeze the leads slightly at the body as you push the leads farther down. This keeps tension on the leads so that the resistor will not fall out when you turn the board over for soldering.
Next, mount taller parts, or heavier and bulkier parts. This includes small transformers, caps, tube sockets, pots, relays, fuse clips, etc. Some components are subject to external forces (tube sockets, fuse holders) while others can impart forces onto the board due to their weight (transformers, large caps). In either case, once the leads are pushed through the PCB, bend the protruding lead over for mechanical retention of the part. Then solder as usual.
Secure the PCB using metal stand-offs and/or metal bolts and nuts. Plastic mounts can be microphonic.
Note that it is important not to bend most component leads right at the body of the component. For devices with short leads that must be formed, use needle-nose pliers to bend the leads as required. Do not bend leads to precisely align with the holes in the board lest the part falls out when you flip it over to solder.
The specs for each Hammond output transformer are listed along with our own supplemental data. We have standardized the listings to make transformer selection and amp design application simpler.
This transformer information appears in print form in our book, The Ultimate Tone Vol. 5 (TUT5), along with details of output and power transformer matching procedures.
Follow these links to transformer classes and other topics:
Vintage guitar amps often produce a tone that is described as “round” or “mellow,” and they can produce a desirable overdrive sound. Some of these amps are low powered, and others have lower supply voltages than later models, but many use “cathode bias” for the output tubes instead of the more efficient “fixed bias” method. Cathode-biased output tube tone can be “round” if no bypass capacitor is used, or it can be “crisp” if a cap is used – sounding quite similar to a fixed-bias output stage.
Modern players have the advantage of an extremely wide “acceptable” tone range to experiment within. They are not limited to the relatively clean sounds that were demanded in the fifties. A greater variety of basic amp tones allows the musician to keep his system simple, but occasionally places him or her in the position of requiring more than one guitar amp for a range of music styles. Depending on individual requirements, it is often desirable to be able to have both cathode-biased and fixed-bias tones available from one amp, at the flip of a switch.
Bias switching is a simple task, but demands the use of reliable components – if you buy surplus parts, confirm their integrity with an ohmmeter. One consideration that is not critical for every player, or on every amp, is whether a tremolo feature is present (and used). If a tremolo is used, then we must not disable the fixed-bias supply completely, as it is tied into the tremolo as a “mute” control. In this case, we simply want to disconnect the bias supply from the output stage when cathode bias is desired.
To accomplish the bias switch in the simplest way, a double-pole double-throw (DPDT) switch is required. This should be CSA or UL approved, to assure the quality of the device. It can be a miniature type, though, as the voltage and current stresses are negligible.
In Fig.1, we see a typical Fender bias-balance supply with the new cathode-bias option and switching added. To add the cathode-bias components, you must unsolder the ground straps from pin-8 of each output tube socket. Two 1-ohm 1/2 watt resistors are used to tie the cathodes together, and to the bias resistor (Rk). This resistor will be a high-power wirewound – 1W to 5W, try 500 ohms for a pair of 6L6s, or 800 ohms for a pair of 6V6s. The bypass capacitor (Ck) can be plastic or electrolytic, depending on the value you choose. Its voltage rating should be twice the value of the anticipated DC bias. For electrolytics, make sure that the positive lead is attached to the 1-ohm resistor bridge, and that the negative lead is tied to ground. This cap should be kept well away from Rk, as Rk will generate a lot of heat.
Note that Ck can be switchable also. Place a 10k half-watt resistor in series with the negative lead of the cap, and connect a second mini-switch across the resistor. With the resistor shunted by the switch, the cap allows full output power, and a loud, bright tone. With the resistor isolating the cap (switch open), the power is restricted but the tone is mellower and breaks up sooner.
To switch the cathode-biasing components in and out of circuit, contact-A of the DPDT is wired as in Fig. 1. The other pole isolates the fixed-bias supply from the grid-leak network for the output tubes, as appropriate. The second pole is series connected between the raw bias supply and the bias pot. Lift the wire that goes to the tremolo off the bias pot, and move it to the supply side of the switch contact.
For fixed-bias operation, the switch is up. Contact-A is shunting out Rk and Ck so that the 1-ohm resistors are tied to ground. Contact-B connects the bias supply to the grid leak network, for stock, fixed-bias operation.
Flipping the switch down allows Rk and Ck to come into play, where the current flowing through the tubes generates its own bias voltage across Rk. Contact-A shorts the tap of the bias pot, and thus the grid-leak network, to ground, so that the cathode-bias arrangement will be complete. Contact-B opens to isolate this newly grounded point from the bias supply. The supply maintains its usual level, so that the tremolo can still be turned on and off at will.
An alternative method using a single-pole double-throw (SPDT) switch and a transistor is given in Fig. 2. The SPDT performs the functions of contact-A in the previous circuit, while the transistor performs contact-B’s function.
In cases where there is both no tremolo and the bias supply is plate-derived (not from its own winding), then an SPDT can be implemented as in Fig. 3. Here, a very large resistance – typically 150k – protects the bias supply diode and source winding from excessive dissipation when the bias supply is shunted to ground.
If the bypass capacitor is used, its value can be determined using the capacitive reactance formula*, and the value of Rk. You must decide what the frequency response of the cathode-bypassed circuit will be. The formula tells you where the response is down by 3db. Maximally flat response occurs starting at a frequency ten times higher. For guitar amps, response to 70Hz is typical, although if response is down at 70Hz, the perceived crispness of low notes will be enhanced.
*f = 1/(2pi x Ck x Rk) and Ck = 1/(2pi x Rk x f) As an example, for a 500-ohm cathode resistance and a 70Hz 3db down point, Ck is under 5 microfarads. For flatter response, a value of 50 microfarads or even 500 microfarads should be considered.
**Note that the Rk values recommended above are conservative compared to what is typically used in cathode-biased amplifiers. This extends tube life and helps make the tone difference between fixed-bias and cathode-bias a bit wider.
IMPORTANT: Screen stops should be a minimum of 1k-1W flame-proof for all tubes and this is especially important in cathode-biased amps.
Printed circuit boards are now available in an array of colours, allowing hobbyists and manufacturers the opportunity to achieve a “signature” look.
London Power has traditionally used green PCBs, but lately we wanted to see what yellow boards would look like.
The obvious choice for the silk-screened printing on the green boards had been white, but what about the printing on yellow boards? We weren’t sure. We searched the web for examples and could find none. The manufacturer of our boards had no images, either.
So, we ordered a small quantity of yellow boards with white printing, and a small quantity with black printing. Below is an image that you can examine. Both the black and the white printing are quite visible and legible. We decided that black was somewhat easier to see in all lighting conditions.
Above: Yellow PCB with black silk-screening beside yellow PCB with white silk-screening
Myth: electronic equipment built on printed circuit boards (PCBs) is somehow “inferior” to hand-wired equipment. This belief is perpetuated by some amp builders, hobbyists and amp reviewers who do not have enough of the facts – let alone “all of the facts” – about this subject, rendering their statements too sweeping and frequently incorrect.
As described in our book The Ultimate Tone Vol. 3 (TUT3), you can have good and bad hand-wired assemblies – and you can have good and bad PCB assemblies; there is nothing inherently “better” or “worse” about either process until you get past low-quantity production. At that point, PCBs become far superior.
Yes. PCBs are superior in the end. Here is why:
PCBs provide a stable platform for the small components that comprise a circuit. The card material is usually rigid fiberglass-reinforced epoxy. Copper traces on the board surface create interconnections between the components – an interconnection layout that cannot change with time, vibration, or environmental conditions. So,the circuit will be stable over time, providing consistent performance from initial use to final use, apart from component variation or aging. Also, unit-to-unit consistency is very high.
Hand-wired units can vary in the exact spacing and layout of interconnecting wires and even component mounting. This alters the parasitic elements of the circuit – the ones that will never appear on the schematic, but which are a part of every electrical assembly. Parasitic elements can cause inferior circuit stability, and low unit-to-unit consistency. Most guitar players encounter this at the music store, when they plug into side-by-side identical models, and each sounds different. A good example is old Fenders, which were were fully hand wired, so there is a lot of variability in how things are positioned internally. Old Marshalls used a PCB to support the small components, but extensive wiring to off-board parts introduced instability in the design. Our book TUT3 demonstrates how to correct this.
Even PCB assemblies can have parasitic capacitances between parts, but these will be more predictable and easier to fix or to accommodate than with hand wiring.
Hand wiring may sometimes beconsidered superior because the solder joints are physically larger. It is also said that because a mechanical joint is made first, the solder itself is less critical, so reliability is higher. The truth is that with either hand wiring or with PCBs, one can make mechanical connections prior to soldering that are fully functional electrical connections. In both cases, this will make servicing or later modifying very difficult. But… poor soldering will compromise reliability for either assembly type. And… larger solder connections are of benefit and easily achieved with either method.
Rogues Gallery
Vintage amps had PCBs and were also hand wired. Results were highly variable, depending on the execution. Early Ampegs were hand wired and are extremely difficult to service. Later Ampegs had PCBs but still a fair bit of wiring, and were relatively easy to service. They were very reliable.
Early Fenders were hand wired; later ones used PCBs. Reliability did not change significantly.
Marshall used a combination of PCB and hand wiring. Their amps failed due to poor component choice and poor wiring techniques. The card and its components was the most reliable part of the circuit.
Vox‘s hand-wired amps have a horrible service record and are extremely difficult to work on. A bad combination.
Peavey amps were always PCB construction, but later models ignore the plight of the service tech. Mesa-Boogie set a standard for new service nightmares with their PCB amps. Their choice to mount the card then attach wiring from all four sides is unbelievably ill-conceived.
Hiwatts were hand-wired and used good parts, so reliability was high. However, they are very difficult to work on.
Matchless/Badcat/Star amps are all hand-wired and very difficult to work on. Reliability problems due to overheating are the result of poor circuit value choices.
And today…
So, our own little niche of electronics gives us good evidence that hand wiring is distinctly not superior – at least in how it has been executed so far. Modern boutique amps have not been around long enough to sway the data. Rather, the large manufacturers using PCBs have demonstrated that PCB construction results in “mostly” reliable products. Not surprisingly, the detractions are related to interconnections – that is, wiring – and primarily, the use of “insulation displacement” (IDC) connectors where the wire is pushed through a knife-edge to make the connection. No wire stripping. No crimping. No soldering. Modern PCB amps of various brands have a neck-and-neck failure rate for poor solder connections and IDC failure.
But in the broader realm of consumer electronics, PCB construction is king. PCBs allow uniform assembly, leading to automated construction in some cases, and uniform quality. In the 1970s, the industry standard for acceptable equipment failure was 2.5%. Five out of every two-hundred units of a given type would fail in the field. The failure rate is much lower today, less than one-fifth of the old standard – and much of this improvement has to do with PCB construction.
The notion that the presence of a PCB means the unit is not hand wired is incomplete; it doesn’t take into account how the parts got on the PCB, how the PCB was soldered, or how the PCB was tied in to other non-card-mounted devices. Items such cell-phones, VCRs, DVD players computers may be assembled without human hands, but there are no guitar amps built that way – yet.
London Power has long since switched all of its amplifier production over to PCB assembly. These boards are hand soldered and hand assembled, just as most other low-production-quantity (custom and boutique) amp products are. Our PCBs themselves are of very high quality, with twice the industry-standard copper thickness, solder-masks on both sides of the board, plated-through holes for increased connection size, silk-screening and proper spacing for voltage and current for each circuit section. The fabrication of these PCBs is highly automated with precision to one-thousandth of an inch. This would be difficult to achieve – if possible at all – doing things by hand.
Servicing Considerations
Servicing a printed-circuit board amplifier does not have to be difficult. However, when designers look at a layout on a computer monitor, it is easy to forget about both the finished product and the tech who might have to service it. This is why you see so many amps with one big PCB inside, supporting all the pots and jacks along the front and rear edges. Servicing for something like this is a nuisance: every knob and every nut has to come off every control and jack just to release the PCB. Although this provides the lowest-cost assembly for the manufacturer – which contributes to the affordability – it also contributes to the extra cost to repair such units.
London Power remembers what it is like to service amplifiers. We try to avoid large PCBs (although in low-profile chassis their use may be unavoidable). Our PCB assemblies are more complex and thus cost more to assemble. This added up-front cost and effort makes servicing much simpler and less expensive. These products can last for generations – something will inevitably break or wear out. It is the fate of all things made by human hands, so why not make that inevitable condition easier to deal with? Routine maintenance, such as spraying potentiometers with lube, is accommodated and does not require removal of the PCB. If an actual control fails and needs replacement, only one or two controls need be released to remove the failed component.
Given that human-made parts eventually fail, servicing a properly designed PCB assembly is no more trouble than servicing a hand-wired amp. In many cases, the hand-wired amp can be more difficult to make look right. Wires might have to be unsoldered from a pot or jack, and this can be difficult to redress and make to look how it did before. The icon of “beautiful” for hand-wired amps is Hiwatt. But – their string-tied wire looms cannot be retied properly without special skill or experience. So, a minor alteration casts a pall over the entire assembly.
The design of the earliest guitar amps was not so much “beginning with a clean slate” as it was an application of hifi circuits for use with pickup-equipped guitars and basses. Portable hifi gear was not new -just the intended use. Clean sound reinforcement was still the mandate.
Tube manufacturers “of old” provided application circuits for their products, and included advice for designing within the safe limits of the tube. They generally advised that the cathode of the tube be at its operating temperature prior to application of the plate voltage, to prevent “cathode stripping”. Cathode stripping shortens the life of the tube and reduces the amount of power that the tube can control. For very expensive and high-powered tubes operating with extremely high voltages >1kV, separate switching for the heater and the plate supply is recommended. The plate supply switch is called a “standby” switch, as the tubes are standing by with warmed cathodes, waiting for B+ to appear.
Many tubes have controlled warm-up times for their heaters, minimizing heater surge currents and allowing plate voltage to be present before the cathode emits electrons. This appears to contradict the concern about cathode stripping – but here is where the application makes a difference. Plate voltages for tubes equipped with controlled heater warm-up are always <1kV, and most often <500V. The danger of cathode stripping is insignificant, so the need for delayed turn-on of the plate supply is unnecessary.
Guitar amp power tubes are inexpensive compared to the cost of, say, a transmitting tube for a radio station. Guitar amps operate at 500V or less – most at under 100W. Low cost, low voltage, and low power: three reasons that guitar amps should not require standby switches.
So, why do many guitar amps have standby switches?
Answer: Fender essentially misinterpreted the requirements, and everyone else copied Fender.
Leo tended not to put anything into the circuit that he felt was unnecessary – but he came from a repair background where a standby switch is a service convenience. There was no other logical reason for its presence: proper use of the standby is lost on most players. Many modern safety agencies even insist that the standby be removed, since the standby’s removal of the plate voltage reduces the idle power dissipation and quiets the amp, and in such a state, it is easy for the user to forget that the amp is still on, which produces a fire hazard. Fender’s rationale about the service convenience is also flawed: a B+ fuse works just as well, as does any easily unsoldered connection in the plate wiring. Although cathode life is theoretically extended – if so, by an infinitesmal amount – heater life is not extended.
Apart from misunderstanding when to use a standby switch, Fender also chose a poor implementation. Fender’s switch is in the DC path, between the main filter caps and the feed to the output transformer and the rest of the plate supply. This leaves the main filters charged during standby, producing no useful advantage, especially for servicing, as this leaves high DC voltage present in the circuit.
Mechanical switches arc when DC is present, so switch life is shortened in this application. Despite the fixed bias voltage also being present at all times, there is often a surge of current through the tube upon transition to the operating mode. The situation is improved substantially if the standby is used only to switch the screen circuit of the output tubes. Tube dissipation can still go to zero except for the heater being powered, but the current through the switch is reduced by 10-100 times. Other manufacturers switch the plate winding AC, which does lengthen the switch life, as AC switching is what most switches are designed for.
A few older amp models had standbys that simply muted the audio signal. There are several ways to go about this, and it does not alter the idle heat of the amp. Such circuits allow the use of a physically small switch that is easier to fit into the front panel. However, the quieted amp still presents the same fire hazard described above.
How to Use a Standby If You Have One
The best way to use a standby switch is to leave it on all the time – that is, leave it in the operating position. If you feel compelled to manipulate it since it’s there, flip it when you flip the power switch.
Never leave the amp in standby between sets. Just turn down the volume control.
You can rewire the standby to control just the screen circuit. Switch stress will be reduced by 10-100 times, and the output tubes will run at zero plate current. Or, rewire the switch as an audio mute.
In a Power Scaled amp, the standby should be rewired into the screen circuit after the Power Scale regulator. This is especially important in amps with a ground-lift standby in the centertap of the plate winding.
There are alternate preferred wirings for standby switches, as shown in The Ultimate Tone (TUT1) series: all are safely implemented using a subminiature switch such as those in the C&K 7000-series.
The Safest Standby Switch
Bypass the standby switch internally so that it does nothing.
Electronics often allows for the accomplishment of a particular goal in many different ways. For guitar amp users, controlling how loud or quiet their amp is while achieving a “cranked” sound has always been difficult. Attenuators can be placed between the speaker and amplifier to provide some means of control, but they tend to alter the sound and shorten tube life. The overdriven amp will clip its output continuously at its full rating.
Controlling Power
Power control methods have been around longer than electronics itself, but electronic control of power provides the greatest versatility of control attributes combined with low cost and fine resolution of control. Electronic methods had their beginnings with had active gain elements, such as vacuum tubes. With semi-conductor technology, power control is much easier to implement, using very low-cost devices and minimal space.
Back to our guitar player with cranked tone. His goal is to get that same sound in the same way, at reduced volume, BUT without having to get a whole new smaller-quieter amp. The goal, then, is to “Power Scale the amp”, which will achieve “Power Scaling the sound”. The mission, then, is Power Scaling.
To top it off, how we get there (electronically speaking) is alsoPower Scaling.
If we maintain the “sound” of the amp while reducing its power output, we have by definition Power Scaled it. So, Power Scaling is the “methodology” we use to attain the sonic goal of Power Scaling. It is not overblown semantics to say that both the goal and how we get there are the same thing – Power Scaling. Both are intertwined: the goal defines the method, and the method achieves the goal.
Foundations: Transfer Curves
The amplifier has a certain “response” to the input signal. It produces an output with characteristics specific to the amplifier design and component choice. If we swap tube types, it sounds a little different, because now it is effectively a different amp with a new response. That response is referred to as a “transfer curve”. This is just a way of relating the output to the input. If we maintain the shape of the transfer curve, then we will maintain the sound, even if we make that curve “look” smaller. This is what Power Scaling achieves.
Transfer curves are generally considered on a single-stage basis in traditional tube electronic analysis, but I have extended the application here. An entire block of circuitry has a net transfer curve, as does the entire amplifier. For the most part, when we implement Power Scaling, it is not necessary to alter or control the entire amplifier circuit, but just a portion of it. Usually, this control is confined to the output stage, or possibly the whole power amp. This is the circuit area in which the greatest signal dynamics occur, and where signal processing from preceding stages is “swamped” by, say, hard clipping of the power stage.
Reducing the power output of the power stage *without altering* how it handles signals is quite simple, and any approach used to achieve this IS Power Scaling. There are also simple ways to control power output which result in *alternate tones* as we dial down; these are simply “Variable Power”.
There are yet other things that achieve variable output power and are labeled as such, but have nothing to do with either Power Scaling or variable-power methods.
Examples of Power Scaling
The following are examples of Power Scaled products and approaches: the techniques presented in our The Ultimate Tonevolumes – particularly volumes 4, 5 and 6; London Power’s amplifiers; London Power’s former PSK-, DCPSK-, SB- and SF- kits, and current SV- Power Scaling kits representing preferred methods. There are countless circuit variations that will achieve this goal. The licensed amp products using the Power Scale trade names, of course, are Power Scaled amps and achieve the desired sonic goal. Unlicensed implementations are also plentiful, and those companies choose to call the control something else, like “Voltage”, “Power”, “Variac”, and others. Whether a company uses the Power Scale name or not, if they use our circuitry or variations of it they must pay licencing. Claret’s and Trentino’s approaches can be implemented well or poorly, achieving either Power Scaling or just variable power. Zimmerman’s approach achieves Power Scaling down to a still too-loud level.
Examples of Variable Power
Examples of “variable power” include: any implementations in which the sound changes as one dials down, but reduced loudness is nonetheless achieved; Mesa’s D-180 Limit control using a variable current-source for the splitter (as does Mojave’s copy of it); Carvin’s power reduction switches; Marshall’s “virtual” power reduction; Moore Amplification’s Power control; and many others.
Examples of “Something Else”
Things that are neither Power Scaling nor “variable power” include: all master volumes. Seymour-Duncan’s “Juice” control is an electronic post-splitter MV; most controls labeled “variac”, which are simply variable current sources for the splitter; and, the use of light bulbs to restrict power gives limited effect while “browning” the sound. One of the benefits of the preferred Power Scale approach is extended tube life. One can still achieve the sonic goal of Power Scaling without this benefit – but why would you want to?
Note that a “trade mark” and a “trade name” do not have to be registered to be recognized and protected. “Power Scale” and “Power Scaling” are trade names and trade marks of London Power and Kevin O’Connor and are legally protected, but are not registered with the US Patent Office. That organization would like you to believe that unregistered marks are without merit and protection. The only mark that is not legal to use is the circle-R for “registered”. The “TM” symbol can be used for both registered and nonregistered marks alike.
Guitar amps are built to give us fun, and we should have fun building them, too!
Often, when implementing mods or improvements, you’ll need a raw bias supply that is higher in magnitude or capable of more current delivery than your existing bias supply.
Low Impedance and High Impedance
“Low-impedance” supplies are either a tap or dedicated winding, with a rectifier and filter cap. Minimal or zero resistance is present between the winding and the cap. This type of supply can usually provide many tens of milliamps.
“High-impedance” supplies are known for high resistances between the winding and the filter cap. An example is the use of the plate supply with grounded centre-tap, with a 100-220k resistor in series with a diode feeding the bias filter cap. To keep the filter cap voltage from rising to the same magnitude as the B+, but negative, a place a resistor or zener across the cap. The zener is preferred as it gives a constant voltage-clamping action at a specified value.
A second form of high-impedance supply is the type that is capacitively coupled from the plate winding. In this case, the winding uses a full bridge to generate B+. A cap, resistor and diode feed the bias filter cap. The coupling cap “scoops” charge into the bias filter cap on a half-cycle basis. The cap impedance is very high, so maximum current from this supply is limited, as is maximum output voltage.
Raw Bias Voltage and Applied Bias Voltage
We should also make a distinction between the “raw bias voltage” and the “applied bias voltage”.
“Raw” bias is the value of the supply at the first filter cap and prior to regulation. This voltage should be fairly constant regardless of the loading, or not, of a bias-set network.
“Applied” bias voltage is the voltage actually applied to the tube control grid to set its plate current. In most resistive and capacitively coupled bias supplies, the “raw” and “applied” voltages are the same. This is an economic choice made by the manufacturer. Marshall, Fender, Peavey, Traynor, Hiwatt, Mesa, Trace Elliott, and all the other major builders have done this and still do this. But, for our own builds, we do not have to pinch pennies in such an important part of the amp.
Solutions
If your amp has a capacitively coupled bias supply and you are thinking about adding Power Scaling – or merely wish to add a tracking bias regulator for other purposes – add an auxiliary transformer to generate the bias supply. This will provide a low impedance that can support multiple bias pots, etc., and have sufficiently high voltage overall to let the regulator function properly.
The resistively coupled high-impedance supply can be modified to provide more voltage and more current. We simply parallel a few resistors to decrease the impedance between the winding and the final filter cap. Typically, three 100k-1Ws are paralleled to make a 33k-3W. We add 80-100V worth of zeners across the cap to clamp the voltage and protect the cap. Now, we have a -80V to -100V “moderate-impedance” supply.
The ultimate source of accurate information regarding Super Scaling is found in our book, The Ultimate Tone Vol. 4 (TUT4) (and the two Super Scaler projects in TUT5).
• Super Scaling is a technique that allows the power output of a small amplifier to be boosted to a higher level to drive a speaker without changing the sound.
Doesn’t a regular power amplifier do this?
• No. Conventional power amplifiers monitor the input voltage only and scale this voltage to drive a load to a higher power level. The final load is fully isolated from the signal source.
What about a Guytron amp? It has a small amp driving a large amp.
• The large amp in the Guytron is like any other conventional power amplifier: it monitors the voltage across the load for the small amp, then boosts the voltage only. This produces an increased output, but the characteristics of the large amp dominate the final tone and interaction with the speaker. Besides, the player has no access to the output of the small amp, and thus has no way to know that the boosted sound is true to the small amp’s tone.
So, how does a Super Scaling amp monitor input power?
• Super Scalers are essentially just the final power stage of a type of power amplifier in which a driver stage would provide some amount of drive power to achieve output power. The difference is that the driver stage is missing, so this drive power must be obtained from an external source. Common tube output stages require no power to drive them, and so cannot be used as Super Scalers.
How does this retain the sound of the driving amp?
• In a Super Scaling amp, no output power can be produced unless input power is provided. The two quantities are intimately related. As you drive the Super Scaler harder, the speaker is driven harder. The speaker’s back-EMF tries to buck the signal from the Super Scaler, which then bucks back at the source. The driving amplifier can then ‘feel’ the speaker and interacts with it.
The Super Scaler is still a big amp, so doesn’t it provide better damping to the speaker than the small driving amp would?
• The Super Scaler is transparent. Its inherent damping is low – in the same range as a typical tube amp or less. However, the driving amp is able to ‘see’ and ‘feel’ the speaker through the Super Scaler, so that the sound is the same as if the small amp itself was driving the speaker directly, but to the new, higher power level.
How much boost does a Super Scaler provide?
• The boost for a typical tube Super Scaler is 4-100, with most in the 4-20 range. Nested Super Scalers can provide further boosts up to 1,000 times. The boost for a solid-state Super Scaler is anywhere from 4-1,000.
Can the boosts be varied?
• Not easily. Solid-state circuit are more easily adapted to variable boost ratios, but the circuits become much more complex. Nested tube Super Scalers can be made variable more readily than simpler tube Super Scalers.
Won’t a solid-state Super Scaler change the sound of a tube amp driving it?
• Not if it is properly designed. Our solid-state Super Scalers use special mosfets with characteristics approximating tube performance, in a circuit optimized for drive by a tube amp. The bulk of mosfets on the market are not suitable in this application. Solid-state Super Scalers can be made with multiple selectable boost ratios for different performance situations, while still retaining tone.
What is a Nested Super Scaler?
• This is a unit with two or more cascaded Super Scaling circuits for very high boost ratios. A half-watt input might produce 700W of output, but with the tone of the half-watt source amplifier.
What is tube life like in a Super Scaler?
• The simplest Super Scalers use conventional receiving-type power tubes in pure class-B circuits. Tube life is longer than in a conventional amplifier even though the internal voltages are somewhat higher.
Do I need matched tubes in a Super Scaler? Can I mix tubes like I can in London Power’s Power Scaling amps?
• Matched tubes are not required in the simplest Super Scalers, but they should be the same type. The circuit itself assures ‘matched’ performance by taking advantage of some of the tube characteristics that cannot be fully utilized in conventional amplifiers. See the Super Scaler project in TUT5, which uses standard common pin-out octal tubes.
In Super Scalers using filamentary triodes such as the 811A, SV572-160 or 572B, the tubes should be similar, and there is a benefit if matched pairs are installed.
Can I drive a Super Scaling amp with a Power Scaling amp and still have power amplifier distortion at any level?
• Yes. The Super Scaler will shift all the power levels up by six times typically, for a simple tube Super Scaler. Most players find this to be an ideal situation because they can use the same “quiet” home settings on their Power Scaling amp when they are on stage or playing with a full band. See the SV572 chapter in TUT5 for a Super Scaler project using these tubes.
Using a Super Scaler with a Power Scaling amp gives a modular system where you can bring along only as much power as you need.
Power Scaling’s goal is to achieve the same tone as one’s preferred “loud sound,” but at a much lower volume. The method can involve one of over sixty distinct approaches, each with many variations.
Power Scaling™ is a methodology developed by Kevin O’Connor of London Power.
Is Power Scaling simply power reduction?
• No. Earlier designers made attempts to achieve effective scaling of power, but never quite got there. London Power refined and fully developed the technology to allow the maximum power of an amplifier to be dialed down to whatever level a player needs. It was first used in the amazing London PowerSTUDIO amp.
Isn’t this just like a speaker load box?
• Not at all. Speaker load boxes, speaker emulators, and speaker attenuators are all forms of attenuation interposed between the power amp’s output and the speaker. They work for some people, but are notorious for sounding “buzzy” at high attenuations.
A speaker attenuator forces your amp to be run flat out, producing its full power all the time. The power that is not needed is thrown away as heat, with only the required power going to the speaker. It is quieter than full-tilt, but now the speaker is isolated from the amp and cannot interact with it, so some tone is lost.
Power Scaling is none of these.
The key to Power Scaling is that it is applied to the power output tube stage itself, and so comes before the output transformer. Power Scaling allows a dynamic power range over 40dB. Most speaker attenuators alter the tone before they reach 8dB reduction. Minus 8dB is just a little bit quieter than full blast; minus 40dB is literally a whisper.
Wouldn’t a master volume do the same thing?
• Only in specific situations. If you only play clean or you only use preamp overdrive or distortion tones, then a master volume will satisfy you.
Power Scaling is the best solution for players who incorporate some output stage “effect” in their sound.
This effect can be light or heavy clipping, or just that cusp of compression you get in a tube power amp approaching clipping. Power Scaling allows you to live at that cusp or beyond, but at ANY loudness you need.
So, Power Scaling will help my overdrive sounds. How clean will a Power Scaled amp play?
• All London Power Power Scaling amps are designed to provide smooth, sweet, clean sounds up to their limit of power. Set to full power, Power Scaling will not alter the clean sound of your amp. If you set Power Scale to a lower setting, the amp now behaves as a lower-power amplifier with a lower maximum loudness for clean sounds.
How does Power Scaling affect tube life?
• With the POWER SCALE dial set to any setting less than maximum, tube life will actually be extended. In accelerated tests, power tube life is as long as that of a preamp tube … up to 100,000 hours if the tube is not mechanically upset. This is a great attribute for players using NOS tubes (new-old-stock).
Output transformer life is also extended, since it is subject to much lower voltage stress even with fully squared output signals and unexpected load disconnection.
Can I run a Power Scaled amp without a load, right into a mixer?
• Yes. It is perfectly safe to do this, although you lose the benefit of frequency shaping provided by the speaker and the interaction of the output stage with the speaker.
The POWER SCALE control reduces voltage and current stress on the output tube, so even at a fully saturated distortion output, the tube is under less stress than it would be in, say, a 3W amp.
Couldn’t a low-power tube amp do the same thing?
• No. A low-watt amp only has one compression point, one cusp of distortion and one maximum loudness level through a speaker.
Power Scaling amps can play both louder and quieter than amps of less nominal power.
And, the compression point stays in the same position relative to the cusp point for all settings, allowing the touch responsiveness to remain consistent.
How is this possible? Is the circuitry complex or expensive?
• In technical terms, all that must be accomplished is to keep the “transfer curve” of the amplifier the same. The transfer curve is simply the relationship between the input and output signals, but as we know, tube amps respond differently to different-size signals. This is because the transfer is not straight and not uniformly curved.
Think of the transfer curve as a mirror.
A flat mirror, parallel to you, will reflect your image perfectly and full size. If you move the mirror away, the image is smaller but still perfect. A tube’s transfer curve is like a slightly curved or rippled mirror. In this case the image is slightly distorted, but this is exactly what we want – it is why you chose a tube amp in the first place. So, moving the mirror farther away reduces the size of the image, but it is still perfectly imperfect.
Electronically, it is very simple and inexpensive to achieve this goal. The diversity of electronic circuits allows countless approaches by different designers, with greater expense added or bulkier components used. In the end, it is all Power Scaling.
Does it affect the output impedance of the amplifier? People on the web say Power Scaling changes the tube plate resistance.
• No. Because the shape of the transfer curve is maintained, the plate resistance of the output tubes is also maintained. So, the output impedance of the amp does not change even though much less power is available once you dial down.
Can Power Scaling be added to any tube amp? Is the circuitry large?
• The circuits are simple and small, and are easily retrofitted into existing tube amps. But… the tech performing the installation must be very good at mods and not just a good repair tech – as these are two different skills.
Depending on the amp, some heatsinking or a fan might be required to cool the Power Scale circuit.
Is the Power Scale circuit just converting the unwanted audio power into heat?
• No. If you only need 3W of power, only 3W is produced; if you need 29W, then you get 29W. The simplest Power Scale circuits are soft regulators, and as such, they divide the voltage available from the raw power supply between the amplifier and itself. In that voltage sharing, there will be some waste heat.
At POWER SCALE control settings between fully clockwise and about 12-o’clock, there will be waste heat from the Power Scale regulator. At settings between minimum (counter-clockwise) and 12-o’clock, the regulator runs cool. Meanwhile, as audio power is reduced, waste power in the power tubes goes down in direct proportion, which increases their reliability.
How does Power Scaling compare to Maven Peal’s Sag circuit?
• Power Scaling allows the player to dial the sound down to whisper levels – actually unusably quiet. If you have a detuned speaker cabinet or any other design that extends dynamic linearity, you can play below a normal speaking level. With London Power’s Power Scaling design approach, the player can choose how quiet to play, with typically 44dB of dynamic range – that is 100W down to less than 0.01W.
Maven Peal’s designer chose the half-watt lower power limit for their WATTAGE control, so that the player would not encounter the point where the matching open-back combo’s speaker loses its tone at lower volumes. To put this into perspective, most guitar speakers produce 90-100dB of sound, with just one watt of input. Half as much input power only reduces loudness by 3dB, so you are at 87-97dB of sound, which is a loud party level.
So, even though Maven Peal has amps that go from 100W down to 1W (20dB range), and 20W down to 0.5W (16dB range), those lower power levels are still fairly loud. To some players, they are “quieter” or “quiet enough”. But it is obvious that a power range that is smaller than these – say, 100W down to 5W – is even less useful at just 13dB dynamic range. The quietest level is still too loud for most players.
“Sag” is an inherently signal-dependent effect exhibited by all amplifiers with conventional power supplies. A regulated supply with a very beefy transformer will exhibit much less sag. Sag is merely a voltage drop under loading, which affects the attack of a note. Once the supply sags, its stays sagged and then power-limits the entire signal.
Maven Peal products use stiffly regulated power supplies to reduce noise, and thus their amps have no inherent sag effect. Sag must then be added by letting the signal modulate the supply reference. Despite the potential of the system, the limits imposed by the designer restrict the range of sounds, both in the sag effect and the possible power reduction.
Power Scaling supplies, in contrast, have a soft-rectifier sound and do not tighten up as you dial down. Diode noise is inherently filtered, and you can play as quietly as you want. “Sustain” is the sonic quality of sag, and London Power uses various Sustain circuits to increase this tonal characteristic.
Must I buy a London Power amp to get Power Scaling?
• No. There are many amp brands licensed to use our technology and trademarks. Each builder begins with traditional or original audio circuit ideas enhanced by Power Scaling. You can also have your own an amp fitted with Power Scaling by a recommended installer.
London Power offers a range of Power Scaling Kits suitable for all tube guitar amps on the market, including vintage amps and new designs.
Where can I get help with Power Scaling?
If you purchase Power Scaling technology licensing or a Power Scaling kit from London Power, technical assistance is included.
