MODS and ODDS
Much as the Hot-Rod enthusiast is forever tinkering with his ‘wheels’ to be ‘just a little different’, so are guitar players forever tinkering with their guitars and amplifiers in an effort to be ‘just a little different’. We are always swapping pickups trying to get the sound that will put us one step ahead of the other poor soul who doesn’t yet know of the latest ‘big thing’. We look out for the latest pedal that will give us the overdrive sound of a vintage tube amplifier ‘cranked up’ (rather than just ‘crank up’ the amplifier itself). We are usually just one string gauge, or one ‘NOS’ tube change away from being the next guitar God. And amplifier ‘mod merchants’ lure us over to their benches with Svengali promises of sounding like the latest guitar hero. Since tube guitar amplifiers are the easiest thing in the world to ‘play with’, many ‘gurus’ have come out of the woodwork. Some have exceptional ideas, while others just add a 12AX7 to the input of your Marshall. And then there are also those with little or no tube training/experience, yet firmly believe that their vast solid-state experience can help you make your tube amplifier ‘better’. It seems to have helped their wallet more, and thankfully these ‘toys’ get relegated to the dust bin in due course.
There are tube gurus who have made more money selling their Do-It-Yourself modification ‘kits’ than they have by repairing or building amplifiers. Boutique amplifier builders shroud themselves in mystery with concealed circuitry (putting silicone over components so we can’t trace out a schematic) and voodoo mantra about ‘proprietary components’. When you finally do get to see a schematic, it looks suspiciously like any Fender.
Most tube amplifiers have the same phase inverter and output stage circuit. It’s thepreamplifier circuit that differs amongst them.
T he following are just a few very simple modifications that may ‘open up’ your little Fender (and Fender-type) amplifier. They are all easily reversible if you decide you don’t like the effect the modification yields. By doing any mod to the ‘Reverb’ channel, you still have the ‘Normal’ channel to compare your work up against. This applies to all of the rest of these modification ideas. Always keep in mind that the standard preamplifier circuit should yield more than enough driving voltage to achieve the full output your amplifier is capable of. These ideas are offered as suggestions for possible ways to drive your amplifier into clipping and beyond. In other words, you won’t get more watts from your little Deluxe Reverb, just more distortion. Beyond a certain point, everything will sound mushier. Know when to say ‘when’, and realize that the crazy tweak you thought of yesterday was probably thought of over at Fender years ago. If your soldering iron is ready, let’s go one stage at a time through a typical Fenderamplifier!
T his is the basic input circuit common to 99% of all tube guitar amplifiers. It is also a wonderful place to begin tweaking our amplifier to suit our tastes and sonic goals. The standard Fender values since the late 1800’s have been 68K for R1 and R2, and 1 Meg for R3. Mesa Boogie years ago had absolutely no series input resistor (R2) on their single input MkI amplifiers (Input 2 and R1 weren’t used). This ‘hits’ the first tube a lot harder, with quite a boost in comparison. You may want something a little more subtle, so I suggest replacing R2 with a 10K, and R3 with a 2.2Meg. This way, you have a modest boost at Input 1.
A n interesting arrangement is the circuit shown below, where ‘Input 1’ is the ‘Bright’ input, and ‘Input 2’ is the ‘Normal’ input. Peavey used this idea in their Classic 50amplifiers, and it works well. R1 is 47K, R2 is 100K, R3 is 5.6Meg, and C1 is .005uF or .01uF. You can play with the values a little until you find the tone you like, but be careful. The first stage cathode should not have a bypass capacitor. This is a little different than the Marshall 4-Input system, where the ‘Normal’ and ‘Bright’ inputs feed different sections of a 12AX7. There, the circuit above is used twice, and the ‘Bright’ channel cathode bypass and/or coupling capacitor are chosen to yield a brighter tone. If 12AX7’s are plentiful in your amplifier, the choice is yours. Should you not be so lucky, the circuit below is a pretty simple way of attaining added flexibility.
Some very early/crude amplifiers have a volume control immediately following the input jacks, but typically the following ubiquitous circuit is seen;
Standard values are R1 is 100K, R2 is 1500-ohms, C1 is .022uF (or higher), and C2 is 25uF (or higher). Remember, C2 is not there at all if you have the Normal/Brite input jack option seen above. This is because you are close to having grid-leak bias on the tube, courtesy of the ‘Brite’ input capacitor. R3 is either a 1Meg resistor, or a 1Meg volume control. There are many postings on newsgroup bulletin boards for suggested mods for smaller Fender amplifiers such as the new Blues Junior. Many aspiring technicians immediately go for the jugular, and increase R1 to 220K. While this does work quite well, there are usually other places further on down the line that have ample gain not being used. Keep in mind that the combination of R1, C1, and R3 form a Thevenin equivalent circuit to the input of the following stage. Increasing R1 would require a decrease in the value of C1 in order to maintain exactly the same frequency response. That said, a popular mod idea is increasing the plate load (R1) and reducing C2 just enough for a more ‘biting’ lead guitar sound (R1 provides the overdrive, C2 provides the ‘bite’). You can also switch a large value C2 in/out of the circuit, as the ‘Fat’ switch on the Fender Tone Master amplifiers do. The following is a brief summary of the effects of altering each component value from the circuit above;
|Increase gain, to a point. Risk of ‘current starved’ amplifier.||Decrease gain. |
Too low a plate load will cause the tube to draw much more current.
|Lower gain/increase bias.||Decrease bias.|
|In conjunction with C1, increase bass response.||In conjunction with C1, decrease bass response.|
|Increase bass response, risk of ‘motorboating’.||Decrease bass response, risk of a ‘thin’ sound.|
|Not much you can actually hear.||Decrease low frequency gain.|
Result can be a ‘treble peaking’ circuit.
In some high gain ‘boogie’ type designs, the danger of high frequency oscillation is met and dealt with by adding what old-timers refer to as ‘snubbing’ capacitors. They work, but at the risk of removing high end ‘sparkle’ or ‘sheen’ to your tone. They can either be across the plate load resistor (as seen below) or from plate to grid. My preference is to either have a lead layout that avoids any of these oscillations, or reduce the preamplifier gain. Values for snubbing capacitors range from 250pF up to 1000pF. This idea is also seen in many Fender phase inverters. Wait and see.
‘Snubber’ capacitor (seen in red) removes high frequency oscillations, but also some ‘sparkle’ from the tone.
I n multi-stage amplifiers, and where aspiring technicians were hot for adding 12AX7’s, a voltage divider is often seen between stages. In older ‘pawn shop’ type amplifiers, this was done to prevent preamplifier distortion. In high gain amplifiers, this is usually necessary to prevent your amplifier from ‘farting’ when the preamp controls are ‘dimed’. This is usually caused by what is called Blocking Distortion.
If C1 is present, it is referred to as a treble peaking circuit. Marshall used this idea in their amplifiers because the standard Bassman circuit (that Marshall ‘borrowed’) featured too much preamplifier drive for EL34’s. This circuit cut down on the drive voltage, and gave a nice treble boost for a guitar amplifier copying the circuit from an intended bass amplifier. The standard values are 470K for R1 and R2, and 470pF for C1, although anything that tames the preamplifier enough can be ‘standard’ values. Experiment. In your ‘no name’ amplifier, removing the voltage divider may be just what the Doctor ordered! Just make sure on the following stage that you have some kind of grid resistance to ground, OK? Don’t go higher than 2.2Meg, unless you want to get into ‘contact bias’.
Hi-Fi tube amplifiers often have no tone controls, and carefully design the stages to have a flat frequency response.
Tone controls are typically next in the circuit, and probably yield the most options for tweaking. The early Fender amplifiers used a simple, crude treble ‘cut’ control (still labeled ‘Tone’), like so;
The Treble Cut control may not necessarily be in the preamplifier stages. Vox put a similar control across the output tubes. Typically, the ‘Treble Cut’ control is a 500K/1Meg potentiometer, with a .005uF capacitor. The capacitor value is wide open for experimentation. It can be used for a very subtle treble cut, or a ‘Deep Rhythm’ tone, as was emphasized on the Telecaster guitar tone controls (another example of a crude treble cut). Later, the ‘Tweed’ era circuitry gave a whole new advanced set of tone options, with fancy features like a ‘Bass’ control! Actually, these circuits were borrowed from the home audio industry, which had been using them for years, and they looked something like this;
You can obtain the values from any similar Tweed-era schematic from Ampeg, Gibson, etc. While Fender didn’t use this circuit verbatim, they did use similar ideas, but eventually went to the ‘Blackface’ era controls. I prefer this circuit hands down to the ‘Blackface’ tone controls (see below). The ‘Tweed’ circuit has much less insertion loss, and by playing with the value of the resistor between the ‘Bass’ and ‘Treble’ control (if it’s not the usual 100K) you can minimize the interaction of the tone controls. I also like this circuit because you can use all 1Meg controls (Treble, Bass, Volume). This simplifies your stocking arrangement. A typical ‘Middle’ resistor is 47K or 56K, so a control pot is possible. You can easily play with the value of a fixed resistor to get the midrange response you like. The mid 60’s Fender ‘Blackface’ controls looked like this;
This setup had huge insertion losses, to the point that an early mod favorite (that eventually showed up in the Mesa Boogie MkI) was to send the ground end of the middle resistor/control pot to a footswitch. This could ‘lift’ the ground from the tone circuit, effectively removing the entire tone control section from the signal path, resulting in a huge boost. Of course your tone controls aren’t functioning in this ‘mode’, but for many the effect was well worth it. As a side note, if you use a switch to lift the ground from the middle resistor in the Tweed tone circuit, the Treble control will still function, and you’ll receive a nice boost suitable for lead work. The Tweed Bassman (andMarshall) circuit fed the tone controls from a cathode follower, or the output was taken from the cathode as opposed to the plate loaded circuit shown above. This results in a little less insertion loss, but that’s about it. The slope resistor will determine where the ‘dip’ in the response curve will occur, and to what attenuation level. I usually leave this resistor alone, since I don’t like the Blackface tone circuit. If you want to play with the capacitor values or slope resistor value, you won’t hurt anything. Typical Fender values are 250pF for the Treble capacitor, a .1uF for the Bass and Middle capacitors, and 100K for the slope resistor. Lowering the slope resistor tends to make the tone appear more ‘midrangey’. Try 47K or 56K and find out if you can hear the difference and prefer it to 100K. Marshall used .022uF capacitors and a 56K slope resistor for their characteristic midrange ‘honk’. Finally, the ‘Blackface’ era tone control is the beginning of actual Fender patented circuits, and for my tastes when thing started going wrong. Of course, a few years later Fender proved that there was something even worse than a ‘Blackface’ design… a ‘Silverface’ design. (This is one more of my personal opinions and another story altogether.)
All tone control circuits suffer from insertion loss. That’s the sacrifice you make for being able to adjust the frequency response.
Immediately following the tone controls on most ‘Blackface’ Fender amplifiers was another simple voltage amplifier stage. This was done to make up for the losses in the tone control circuit. Succeeding this stage was a ‘mixer’ or ‘summing’ resistor that along with a similar resistor from the ‘Reverb’ channel fed the phase inverter. It is usually 220K, and is there to prevent the two channels from ‘interacting’ with one another.
A few uneducated people have a marvelous ‘mod’ whereby removing the first 12AX7 (belonging to the ‘Normal’ channel) gives a very modest boost to the ‘Reverb’ channel. They erroneously state that this works because the voltage on the preamplifier stages in the ‘Reverb’ channel will increase without the ‘Normal’ channel 12AX7 in place. The typical current draw on any preamplifier stage is never more than about 1 or 2mA. You won’t ever know about the piddly little voltage increase. This ‘mod’ works because without the ‘Normal’ channel 12AX7 in place, the signal from the ‘Reverb’ channel can’t bleed back through the ‘Normal’ channel as much and get loaded down. The summing resistors don’t work perfectly, and this is one more reason I much prefer the ‘Tweed’ circuitry. But let’s move along, shall we?
On Fender amplifiers equipped with Reverb, the Reverb is fed from the tone controls in a parallel circuit similar to that shown below. The combination 10pF/3.3Meg network is necessary, otherwise the Reverb circuit would be ‘swamped out’. This is an interesting area to attempt at increasing the gain through the amplifier. By loweringthe 3.3 Meg resistor (and increasing the 10pF capacitor) we increase the gain of the preamplifier, at the expense of reducing the Reverb driving signal. Unless you play with your Reverb control on ’10’, and still need more Reverb, you can live with a little less, right? Start with 2.2Meg/20pF and work from there. It has been said that with this modification in place, the average guitar player who set his Reverb control on ‘3’ now needs to set his Reverb control on ‘4’.
Reverb circuit has definite potential for modification yielding more gain, but less Reverb. So what?
Phase inverters have had a very interesting evolution. We began with transformer coupled output stages, and then we had simple split load phase inverters, that looked like this;
This simple setup uses only 1/2 of a 12AX7, and frees up the second half for other ideas we may have. It is still used today, in the Peavey Classic 50. A few magazine articles and books written by hyper-active cheerleaders who moonlight as wannabe tube engineers suggest that R2=R3+R4 for ‘improved’ performance. Don’t worry about it. Every phase inverter is not balanced quite perfectly. The split-load phase inverter actually has awful balancing properties, but sounds great for guitar amplifiers. Using typical values, R1 is 470K/1Meg, R3 is 1K/1500 ohms, and R2 and R4 are anything from 47K to 100K. Using 100K results in the least losses, but your gain from this stage will always be less than ‘1’. Output 2 can be taken from either the top of R4 (as shown) or the top of R3. The old-time theory was that you chose the output spot that yielded the best balance between the two outputs, and that had the least distortion. Try it out at home, and scope the results. Can you see a difference? Play the amplifier with both setups. Can you hear a difference? The biggest concern from wannabe engineers was that the frequency response from both outputs would never be the same, because the effective plate impedance and the effective cathode impedance are not equal. This would mean that the Thevenin equivalent time constants, and hence the frequency responses, are not equal. To quote one of my tube ‘mentors’…. “It worked for Fender and everybody else for 900 years, so why fuck with it?” Good enough for me. My theory is that since I like the old-style tube circuits, I’m better off to stick with the old-style tube circuits. I’ll let the wannabe engineers reinvent the wheel.
The standard Fender cathode coupled phase inverter (aka ‘long tail’) is shown above. Standard plate load resistors are 82K for R1 and 100K for R2. The ‘mismatch’ of plate loads is done in an attempt to balance the two outputs. Since the plate load of 82K is not quite the same as the cathode load of about 22K, the plate load of the cathode input side is raised to 100K to try and have the same output level as the grid input side. Some Bassmans used 47K for both R1 and R2. This works a little better, but gain is sacrificed. The resistor labeled ‘Presence?’ is typically 100 ohms. In the Tweed Bassman circuit (and Marshall) this resistor was raised to 4700 ohms. This did a few things all at the same time;
- It reduced the driving voltage (good in an EL34 Marshall).
- It balanced the two outputs slightly better.
- It made the way for an easy ‘Presence’ control to be added.
No phase inverter is balanced perfectly, even in Hi-Fi equipment.
The balancing of the two outputs is not a matter of life and death, seeing as the output transformer won’t be perfectly balanced. The adding of a Presence control gives a little more flexibility, but even with a 100-ohm resistor in this spot a ‘Presence’ control is still possible (it would need to use about a 1uF capacitor). Of course you know about trying different phase inverter tubes. The standard Fender ‘Blackface’ phase inverter is a 12AT7. A 12AX7 will give more drive, and a 12AU7 will give less drive. Sometimes across the phase inverter plates will be seen either a 47pF or a 100pF ‘snubbing’ capacitor, that we were introduced to earlier. This was done over at Fender to facilitate speedy production methods. Rather than learn neat and tidy lead layout, the capacitor was added as an insurance measure to guarantee a stable amplifier. You have a 50% chance of cutting it out and not having an oscillating amplifier, but instead getting a little high end ‘sparkle’ or ‘sheen’ back. Most people don’t notice the difference, though. As a final thought, I have seen snubbing capacitors through every amplifier stage and even across the output tubes as well. What ever worked for the prototype amplifier was used in the production run.
Some recent Fender designs attempt to balance the phase inverter with a circuit that looks like the standard circuit, but adds one resistor like this;
The added resistor (Rx) forms a voltage divider for the Output 1, or grid input side. Here, R2 is still 100K, but now R1 is 91K and Rx is 9.1K. Many technicians miss this little resistor, and in the phase inverter this is no big deal. However, many Bassman heads had the same idea applied to a voltage amplifier stage, and this really neuters the gain/tone. You can really open up things when you locate and banish this offending resistor.
I have read many articles pertaining to various output stage coupling methods, other than plate loaded, and wondered why none were in use today. I suppose the best answer is that the plate-loaded setup was the most efficient. In a 50’s era Audio Anthology Magazine, a circuit was suggested for a single ended 6V6 amplifier that was cathode loaded. The output transformer recommended was a 5K primary unit, a standard value. However, when I actually built the circuit, the output was very clean and about 21/2 watts. The article mentioned that the ‘benefits’ of this cathode coupled output stage were;
- Improved low frequency response.
- Improved high frequency response.
- Improved damping.
- 100% degenerative feedback.
This all ‘sounds’ wonderful (no pun intended), but the efficiency of the circuit is still terrible. Stick with plate loaded output stages. With high powered amplifiers (Marshall, Twin Reverb, Showman, etc.) everyone is looking for a way to tame the wattage without sacrificing the tone. The idea of ‘pulling’ a pair of output tubes is fine enough, but there are better ideas out there. First of all I would switch a pair of tubes ‘off’ rather than ‘pull’ them. Secondly, a DPDT Center ‘Off’ switch yields possibilities that are far more imaginative and interesting. Make sure to mount the power resistors on the chassis with a little heat sink grease, and keep the capacitor(s) away from the resistor(s). You now have switching between three power output levels. Use the other 1/2 of the switch for the other pair of output tubes (make sure they are ‘paired’ properly), and don’t worry about the bias.
Fender Twin Reverb/Showman owners trying to reduce the power output levels can also use the following circuit variation. This works only on non-bridge rectifier circuits, and is more clever than removing two output tubes. You could also use this circuit where you want ‘Tweed’ B+ levels, with reduced output wattage and ‘browner’ tones in any Fender amplifier, especially the reissue Bassman.
Feedback circuits can get complicated, and enough articles have been written that you should have an idea what you want to accomplish here. My advice is to scope the output waveform and decide where on the Volume control you want the amplifier to ‘break up’. Then using a decade box, determine how much feedback you need. If you have a little too much preamplifier gain, and the amplifier ‘compresses’ at high volumes, either reduce the drive voltage or add a little more feedback until the compression doesn’t occur. I will say that some Bassman heads used the feedback circuit below to increase the low end frequency response. If you plan on using the amplifier for guitar, replace it with a conventional feedback circuit. Trust me, you’ll thank me later.
One of my favorite mods for the feedback circuit was designed around a few tweed Fender amplifiers. The Champ (5E1, 5F1), Harvard (6G10), and Princeton (5E23, 5F2, 5F2A) amplifiers of this era were all similar electronically, and contained a feed back circuit shown below, complete with my modification suggestion ‘installed’.
The circuit works very nicely and requires an extremely modest parts list. Build a ‘stomp-box’ type footswitch, using a simple SPST footswitch and a 1/4″ jack. With the footswitch ‘open’, the amplifier is stock. By closing the footswitch, we get a great boost for lead work from two sources:
- Negative feedback is effectively removed from the circuit, giving approximately a 3dB voltage gain.
- A bypass capacitor is added to the cathode of the tube, giving another voltage gain of 3dB.
The circuit can also work with the 5E7 Bandmaster, the 5E6 Bassman, the 5E5A Pro, the 5E4A and 5F4 Super, and the 5D8 Twin amplifiers, just remember that the ‘Presence’ control doesn’t work in the ‘Lead’ mode. You could also play with the value of the capacitor, thus voicing the ‘Lead’ mode to taste. The footswitch is detailed below. Use any shielded guitar cord you have laying around the house between the amplifier and the footswitch. Finally, keep in mind this voltage gain does not equal more watts from your amplifier (as mentioned earlier, none of these mods do that). This modification yields way more driving voltage to the output tubes, thus driving the output stage into distortion, something earlier manufacturers tried to avoid. We’re just putting some in!
Tube rectifiers receive much credit for the vintage ‘sound’ from those who ascribe to the tonefulness of a GZ34 over a 1N4007. ‘Reissue’ amplifiers often come with a ‘plug-in’ rectifier option to appease everybody. Unfortunately, this doesn’t really work. Older amplifier designers, under the delusion that we would never turn our amplifiers up to ’10’, cut their final costs by utilizing a light-duty power supply. Turning to our RCA tube manual, a pair of 6V6’s, at full output, draw about 92mA of plate current. Screen grid current will add another 13.5mA. Since 12AX7’s draw a negligible amount of current, we can use 105mA as ‘the number’. Adding on a 50% safety factor, we would normally require a power supply capable of furbishing about 155mA of DC current. Of course, seeing as we won’t be playing our amplifier at full output, we can use a transformer rated at about the original 90/100mA, and save quite a few dollars. This is exactly what early tube amplifiers did. Also, in days of yore, large value filter capacitors were very expensive. Engineers chose the smallest value filter that would just get rid of the 60/120Hz hum. Early Fenders can been found with preamplifier filters as low as 8uF! Modern amplifiers will use much higher values of filtering, as capacitors today are much smaller physically, and cost far less than their ancestors. Later Fenders (‘Blackface’) would use the following typical rectifier circuit;
Part of the real ‘secret’ to great vintage guitar amplifier tone is the ‘underdesigned’ power supply. When you crank up the smaller powered amplifier to play in the modern gigging environment, the power supply ‘sags’, and you get the wonderful power amp distortion thrown in for good measure. There are sonic trade-offs, however. The audible result of having a high power supply impedance is a ‘mushy’ bass response at high volume levels and a peak compression. However, if you grew up with classic Rock ‘n’ Roll albums, you heard underdesigned power supplies with a high impedance. If you’re the type that wants to build a better mouse-trap, power supply impedance can be reduced in any of a number of ways.
- Decreasing the rectifier resistance (using a 5AR4 instead of a 5U4, or using FRED rectifiers).
- Decreasing the winding resistance of the power transformer (using a transformer rated for more current).
- Increasing the filter capacitor values.
Of course reducing the winding resistance yourself is highly impractical, but a point to be taken anyway. You can get a transformer rated for the same voltages, but more current, to achieve the same results. Some people like a tight power supply that doesn’t sag, using their amplifier more as tone reproducer than a producer of its own tones. Those folks can use FRED rectifiers and increase the filter capacitance in their amplifiers. These steps aren’t quite necessary today with most modern and ‘reissue’ amplifiers. These amplifiers usually have a low-impedance ‘rock-solid’ power supply that can’t ‘sag’, even by plugging in a tube rectifier on dual-rectifier models. Many posting on the Internet Newsgroups ‘alt.guitar.amps’ and ‘AMPAGE’ enquire about putting a tube rectifier in their silver-faced Fenders to ‘soften things up’. It can sometimes be done without adding an auxiliary heater transformer (regardless of what the other ‘guru’ types may think), but below is a far simpler (and much better) solution.
Mesa Boogie used similar principles, calling their switch something like ‘BOLD’/’SPONGY’. Value of the resistor depends on how solid the rest of the power supply is, and the threshold of when you want the sag to ‘kick in’. A good starting point is 100-ohm/10 watt. I sometimes reduce the filter capacitor(s) value while I’m at it. Be careful not to go too low, or ‘hum’ will result. You then have two threshold ‘levels’ of sag, as well as the sloppy bottom end common to vintage amplifiers. Reissue Bassmans with reduced filtering and a 200-ohm/10 watt resistor can have more sag than Grandma’s triceps!
There are ways to induce sag into the amplifier circuit other than going through the bother of adding a tube rectifier.
There are many other modifications possible to your little Fender. You can easily find magazine articles giving ideas that will keep your soldering iron going the whole weekend. I have given you just a small sampling of how I like to go through an amplifier and tweak it up for my tastes. I will add to/remove from this sampling according to time/space allowances and feedback (no pun intended) from you.
Let me know what worked/didn’t work for you.