|Elliott Sound Products||Project 152 - Part 1|
Bass amps are a special case for amplification. A 4-string bass has a bottom 'E' (E1) frequency of 41Hz, while most 5 and 6-string basses are tuned to bottom 'B' (B0) - 31Hz (close enough in each case). Some bass players are happy to be able to get no lower than the 2nd harmonic, which in common with most plucked string instruments is predominant. The dominant frequencies are therefore 82Hz and 62Hz. Some bass amps have deliberately limited response below ~70Hz.
Depending on the bass, the player and playing style, harmonics can extend beyond 10kHz, and many bass speaker cabinets include a compression driver and horn to cover the high frequencies. Another approach is to use smaller than normal speakers, and 410 (4 x 250mm/ 10" speakers) and similar are now common, simply because smaller speakers have better high frequency response (or at least that's the theory, which may or may not work in practice). There seems to be a consensus that you need response up to at least 7kHz if the top end is important for your sound.
Looking at the combinations that are popular, the range is very diverse, both for cabinets and amplifiers. I'm not about to even try to produce a bass speaker cabinet design, because there are so many possibilities that a single project design is simply not feasible. I do have some ideas though, and they will be discussed later. Meanwhile, the amplifier is something that can have a project design, but be warned that it has (by necessity) a great many options.
General Idea For Bass Amp Front Panel
The drawing shows one possible arrangement, incorporating most of the possibilities discussed below. To be truly useful, the amplifier should be suitable for use with electric bass (passive or active), as well as acoustic basses with piezo pickups. To get the best out of any piezo transducer, the preamp/ impedance converter should be as close to the pickup as possible, because lead capacitance will reduce the output level. However, this is independent from the amplifier, which only needs a reasonably high input impedance.
Right from the start, we need to look at some of the popular options and discuss each of them.
Most of the facilities needed in any musical instrument amp come from the preamp. It is entirely feasible to build just the preamp, and use a commercial power amp to drive the speakers. A preamp is easily made that will drive any power amp ever built, and this may be a worthwhile option given that high power amps are available at very reasonable prices.
Valve (full or partial):
There is a great deal of nostalgia for valves ('tubes'), and many people think that the mere presence of a valve in a preamp gives it some characteristic that isn't possible with transistors or opamps. Mostly, this is untrue, and some amps that boast a 'valve preamp' simply have a token valve that achieves little or nothing other than greater noise and reduced reliability. Others may use the valve to (more or less) its full capabilities, but it remains a source of noise and unreliability. It is probable that few (if any) bass players would be able to pick a valve's presence in a preamp in a double-blind test, which makes it rather pointless.
That tone controls are necessary is a foregone conclusion. The only decision that must be made as to what kind. Bass amps can have fairly rudimentary tone shaping circuits similar to those used with guitar amps, or more commonly they can have extremely sophisticated (and complex) tone controls including parametric or graphic equalisers, variable frequency bass and treble, or digital 'modelling' allowing you to set the amp to behave just like the one that your favourite bass player uses, but without having to buy the same amp (however, see below).
This control is found on a lot of bass amps (sometimes by a different name), and is basically just another tone control. Mostly, it's used to 'scoop out' the midrange and boost the highs and lows. The same effect can usually be achieved with the normal tone controls, but some players seem to like the simplicity of a single knob they can twiddle to get a fairly radical tonal change.
Also known as anything from 'growl' to 'grunge' and onwards to 'crunch' (or are the last two reversed?) with a great many variations in between, some players love it, others hate it. It can be hard to get right, but if done properly should sound like amplifier overload, but without the harsh edge that most players dislike. Switching it in and out is often a problem, because there will often be a significant level change.
The addition of a variable compressor/ limiter is worthwhile, because it allows the player to get maximum volume without distortion, and can also be used as a versatile sound effect. While some might like to be able to play around with the attack and decay times, a simple LED/LDR compressor is just about right by itself. This arrangement has the advantage of ease of use, simple and reliable circuitry, and little or nothing a user can do to make it sound horrible. However, compression and limiting should only be used in moderation. Musical instruments have dynamics, and deliberately making everything the same volume is a really bad idea (and it sounds boring!).
Some players still seem to like the sound you can get from a stereo bass rig, but it seems that few of the current commercial offerings include this option. It's not hard to include it if you build your own system, but it usually does mean that a great deal of the preamp circuitry will be duplicated. This makes it a fairly expensive inclusion, and also means that the front panel will be very crowded. It can be implemented with a less complex tone circuit for one of the inputs
Several commercial bass amps include an electronic crossover and separate amp for a high-frequency horn. Some also include the option for running two main power amps in bridge or separate, with an electronic crossover (usually variable) to split the full range signal into high and low ranges to be amplified and connected to separate speaker boxes. The LF might go to a pair of 380mm (15") speakers, and the HF to 4 x 250mm (10") speakers (with perhaps that box having a horn driver as well). This allows the system to be operated as a triamped bass rig, which is likely to sound louder than an equivalent single amplifier of the same total power.
DSP (digital signal processing) is now exploited by many systems, providing amp/ speaker emulations, special effects, and most of the functionality of the preamp. Unfortunately, this comes at a price - not necessarily in hard currency, but in long term reliability. With many of these systems, a fault in the DSP circuitry may render the entire system inoperable, and it may or may not be possible to get it repaired. 'Repair' generally means replacing the entire board, and while there shouldn't be too many issues in the first couple of years, the chances of getting such a system repaired in 10 years are far from certain.
This is rarely included, which is a shame. It's surprisingly easy to generate subsonic frequencies with a bass guitar, especially with slap techniques or if the strings are muted by 'palming' - laying the palm of your hand (or even just a finger) across the strings. This can use truly vast amounts of amp power at subsonic frequencies, and should you be using a vented cabinet it can cause excessive cone excursion and possibly speaker damage. A high-pass filter should be set up so that any frequency below the lowest open string is rolled off. This /will/ change the sound if you actively mute strings or use a lot of slap bass, but the frequencies you are getting rid of are below the cabinet tuning frequency and not reproduced properly anyway. If fitted, it should be switchable. Rolloff slope needs to be at least 12dB/ octave, and the filter shown below is 24dB/ octave, with a design frequency of 27Hz.
It's quite common for bass amps to have an output that is intended for use with an external tuner. It's usually taken from one of the early preamp stages, so Volume can be turned down and the bass tuned without any noise reaching the audience. At the cost of a jack socket and a resistor, it's an easy addition.
Providing a balanced send to the front-of-house PA system or recording console is common. It's useful to be able to switch it for pre or post EQ, because what comes out of your speakers can be highly equalised, and that's not always useful for the PA or recording mixers. The level should be adjustable.
That's a lot of functionality for the preamp, and by no means all bass amps provide all the facilities listed. Some do manage to include most of them though, but usually only for fairly expensive systems. After the preamp, then we have to decide on what kind of power amp is needed, the power level, and make sure that it won't destroy every speaker connected to it, so power levels must be sensible.
However, bass generally needs a lot more power than guitar for a variety of reasons. Unlike many guitarists, few bass players push their power amps into hard clipping, and will often try to avoid any clipping because it just doesn't sound nice. The big question about power is "how much?". This actually depends on a vast number of variables, but ultimately is limited by the power each speaker can accept without melt-down or severe power compression. Loudspeakers are often less efficient because they must have a lower resonant frequency, and therefore heavier cones.
The connections to the speaker box(es) should only ever be via Speakon connectors. 1/4" jacks have been standard for many years, but the risk of a short circuit is too great, and they are totally unsuited for high current applications. The only other connector that can be considered is XLR, but Speakons are still the preferred option. This is especially true because many project power amps do not include short circuit protection, and a short will cause the amp to fail. Short circuit protection is not as easy as it may appear, and it's common for the circuits used to react badly to reactive (speaker) loads, generating spikes and gross distortion.
Valve (full or partial):
While still popular, high power valve amps are expensive, heavy, and comparatively unreliable. Hybrids (using valves and transistors) are also common, but if the valve stage is just at the front end (as a first gain stage) it's mostly a marketing exercise. Valve output stages need large output transformers and at least 4 (preferably more) output valves. These are only available from China or Eastern Europe, and quality is variable. Failures are common, and expecting more than 120W or so is generally unrealistic. This is rarely enough for bass.
This type of amplifier is by far the most common, and it's quite easy to get around 350W into 4 ohms from a reasonably simple design (See Project 68 as an example). There are many bass amps with a lot more, but excessive power comes at a price - reduced reliability, blown speakers and serious loudspeaker power compression being the most common problems. Unfortunately, this general class of amp has fairly low efficiency, so substantial heatsinks are essential (preferably with fan assistance). However, they are generally easy to fix if a problem develops, and most should be able to be repaired easily - even 10 years from now.
While one of the most common designs for dedicated power amps, Class-G (or H if you prefer) seems to be fairly uncommon for bass amps. I'm sure that some manufacturers do use Class-G amps, but I didn't find any circuits on the Net. Class-G amps are more efficient than Class-B, but also use more output devices and filter caps in the power supply. While there's no doubt that such an amp will run a little cooler than Class-B, it's doubtful if there is much to be gained.
Switching power amps (Class-D does not mean digital) are common now, and many can deliver truly scary amounts of power. It's very common to include a switchmode power supply as well, which reduces weight considerably. Just like DSP based preamps, many Class-D amps are (or will quickly become) impossible to service, and again, 'repair' means replacing the entire circuit board. Once the manufacturer runs out of spare modules or replacement circuit boards, the amp is a write-off (for the power amps and power supply, and both may be on the same circuit board).
Regardless of the type of transistor amplifier, it's potentially worthwhile to include a precision diode network just before the amp to create a 'soft clipping' effect. This will provide an alternative to the normal 'hard' clipping you get from these amps, similar to the way a valve amp clips. Distortion will start to increase as the peaks approach clipping, rather than appearing suddenly as is normally the case. If done properly, the maximum power output isn't restricted. You can still get the full rated amp power, but with gradual onset distortion becoming apparent from around 3/4 power. Because this also provides significant compression, the amp will sound as if it has more power than it has, but the distortion will be audible in some cases.
If included, the soft clip function should be switchable so it can be disabled. This isn't something that you'd do often, so the switch can be on the back panel. If you have a triamp system (2 main amps plus a horn amp) all three should have the soft clip function. This also means that the amps must be operated in voltage mode, because the variable gain of a current output amplifier makes it impossible to get predictable performance.
Having gone through the options, the design I suggest will use a combination of the following features, and in order ...
- Input Gain - Can be switched between high and low gain from the front panel (or a footswitch, not shown in this design)
- Tuner - Output for electronic tuning meter
- Variable-Frequency Tone Controls - More-or-less conventional tone controls, but with variable turnover frequencies for both bass and treble
- 2-Band Parametric Equaliser - Variable frequency boost and cut controls that can be varied over the range 70Hz to 3kHz in 2 bands
- High-Pass Filter - Set for 27Hz, it removes high-level very low frequency signals to improve clarity (switchable)
- Effects Send/ Return - Dual phone jack sockets for external effects
- Inbuilt DI - A balanced feed via XLR connector for a send to the FOH (front-of-house) PA system or recording console, variable
- Compressor/ Limiter - A LED/LDR based adjustable limiter to maintain consistent output levels or prevent power amp clipping (Optional but recommended)
- Variable Crossover - An electronic crossover network (with defeat switch) so the signal can be split and sent to two separate power amps (Optional)
- Fixed Crossover - Another high pass electronic crossover set for 2kHz to drive a separate horn amplifier, no low-pass filter is needed (Optional)
- Power amp drivers, incorporating 'soft-clip' circuits
- 3 Power Amps - Two 300W amps (P68), plus a 60W amp (P27A is ideal) for the compression driver. (Multiple amps optional)
The plan is that you can include or ignore any of the options described, so if you only ever intend to use a single cabinet with no horn, then the electronic crossovers can both be omitted. You might not need the facility for a direct feed, so the DI can be left out. If you wanted to run a full stereo rig, the second channel might only use the variable frequency tone controls but not the parametric EQ sections, or you might not want or need an 'overdrive' capability.
Figure 1 - Block Diagram Of Bass Amp
The block diagram shows where each of the modules is located within the amplifier. This is the overall structure, so you can see how everything fits together. It can be difficult to imagine how all the modules are interconnected without a simplified diagram such as this. The nominal working level throughout the amp is intended to be around 1-2V RMS, and the Overload LED will be triggered by any instantaneous peak above 8V. Playing is always very dynamic, so expect the LED to flash every so often during normal playing, especially if you use slap bass techniques.
No matter how you look at it, this will be an expensive undertaking. However, it will also be extremely versatile, and will have the features you want, tailored if necessary to suit your style. Many of the facilities described are available in commercial amps, but you'll only get all of them in up-market models. Don't expect to find all the features described here in a $300 bass amp.
Please note that all drawings that involve opamps omit the ±15V supplies for clarity. Naturally, all opamps require power supplies and local supply bypass capacitors placed as close to the IC package as possible. The capacitors should only ever be 100nF. 50V monolithic ceramic types, and I recommend that every opamp package (normally with two opamps in an 8-pin dual inline plastic package) have its own bypass capacitor. If one opamp in a dual package is unused, join the output to the inverting input, and connect the non-inverting input to earth/ ground.
The limiter circuit is optional (but recommended), and if you don't need the high/ low bass outputs, a single output can be taken from the output of the 'master' volume control following the limiter. If the limiter isn't used, the output to the power amp comes directly from the 'FX Ret' (Effects Return) input. The output amps (shown in Part II, Figures 16, 17 & 18) are also optional, but I would recommend that the Figure 18 circuit should be used as the minimum. If you don't need the 'soft clipping' provided, the LEDs/ transistors (etc.) can be left out and the circuit shown in Figure 18 should be used. Only one trimpot is needed, so you can set the gain structure properly (the gain needed depends on the amplifier used).
If you really want to include a valve stage, it's not too difficult. The hardest part is the high voltage supply, which can be produced easily by a small switchmode DC-DC converter but will more likely come from a separate transformer. Alternatively, it can be derived from the main power transformer and a voltage multiplier. The DC needs to be at least 70V, and I ran tests at this voltage and got quite good results using a 12AU7 valve. A 12AX7 is far less forgiving, and needs a higher supply voltage or the distortion will be excessive even with quite low input voltages. Despite what you might read elsewhere, there is no difference in 'tone' between a 12AU7 and a 12AX7 provided they are biased correctly.
It's better to use a higher voltage than 70V if possible, and we should aim for between 100 and 150V, which doesn't require too much fancy multiplication. If at all possible, a specialised transformer should be avoided as it would have to be made to order - usually a very expensive option. It's also possible to use a small transformer wired backwards, powered from one of the AC windings of the main transformer.
The following circuit was tested both with each half of the valve run separately, and with the two in parallel. There isn't a huge difference, but parallel operation has the edge, with slightly more allowable input voltage and marginally lower distortion. Tests were done with a 70V DC supply. It's noteworthy that some bass amps use the input valve as a cathode follower, which doesn't achieve anything even remotely useful. All that does is raise the overall noise level, but it contributes nothing in terms of 'sound' - unless you like a noisy amplifier of course.
The heater is connected to 12.6V from the power supply shown below, and you'll use pins 4 and 5 (series connection). A 12AU7 draws 150mA at 12.6V, making it easy to filter and regulate.
Figure 2 - Valve Preamp Stage
If the cathode is bypassed the gain is higher (as expected), but the input voltage is limited to no more than 500mV (3% THD). Above that, the distortion increases dramatically. Most commercial amps that use a valve stage deliberately avoid operation at high gain and high input level, because the distortion becomes highly intrusive. From the tabled results below, you can deduce that the stage has a gain of 4.2 (12.5dB) if the cathode resistor is not bypassed, rising to 9.5 (19.5dB) with a 47µF capacitor.
|With C1 Bypass||Without C1 Bypass|
|Input (RMS)||% THD||Volts (RMS)||% THD||Volts (RMS)|
For a single channel amp, it's best to run the two halves of the 12AU7 in parallel as shown in Figure 2, and normally without the bypass cap if you use a bass with high-output pickups. The bypass capacitor can be switched in and out to get different input sensitivities, rather than have separate high and low level inputs. Typical basses can deliver anything from about 50mV up to 1V or so (RMS), depending on pickups, playing style, etc. The two zener diodes at the output protect the following circuitry from high voltage transients. Even though the supply is only 140V, it's more than capable of damaging the input stage of the opamp that follows.
Regardless of everything that people might claim, a valve in the audio path is not magic. If it's working linearly, there is no difference between a valve and any other amplifying device - transistor, JFET, MOSFET or opamp. When it's operating non-linearly (but not clipping), there's still almost no difference, except distortion is higher and it's harder to provide the necessary power supplies. Even though it adds a considerable extra cost, including the valve will make some players much happier.
It's up to the constructor to work out a way to mount the valve socket to protect the valve from vibration. If the internal mica supports are damaged by constant vibration, the valve may become noisy, microphonic or may even fail completely. Make sure that you always carry a (well protected) spare valve and that it's easy to replace it if necessary.
The Gain control is 100k (it must be linear) with the valve circuit because it has a relatively high output impedance. We also need to protect the following opamp from excess voltage swing because that can damage the input circuits. Although the valve stage can be greatly improved by providing feedback from the following stage, I suspect that this would rather defeat the purpose, so it's shown warts and all. Note that the valve stage is inverting, so if you think that absolute polarity is important you may want to include an inverting buffer somewhere.
The valve's B+ power supply is most easily obtained from the power amp's mains transformer via a simple voltage multiplier, but this may not be feasible for a variety reasons. The use of a simple switchmode boost circuit can be tempting, but noise might become an issue because switching supplies always create noise, some of which may be within the audio band. It's easier to use another small transformer. Let's assume that the transformer for all the low voltage circuits has a 30V centre-tapped winding, and is rated for not less than 30VA.
Note that the '15AC1' and '15AC2' connections are used for a Project 05 (or similar) power supply for the rest of the circuitry. If a smaller (but also 15-0-15V) transformer is connected with its full 30V secondary winding connected across 15V AC, then the voltage on the 'new' secondary will be around 100V RMS (assuming a 230V transformer). This is perfect for the valve preamp's supply, and will give a DC voltage of ~140V easily after filtering, as shown below.
Figure 3 - Valve Preamp Power Supply
For those who use 120V mains, if possible use a small transformer with dual 120V primaries that can be wired in series. You can connect a 30V secondary of one transformer directly to the 30V secondary (but now used as the primary) of another, but it may draw excessive current, and must be tested before you commit to doing so. Expect the driven transformer to draw around half its allowable current - for example, a 30V, 150mA transformer may draw up to 75mA (no load current) when connected in reverse with the full voltage applied to the secondary.
As shown, I have assumed that a dual primary winding is not available for those using 120V, so the 5VA transformer is connected with its secondary (now used as the primary) between 'C' (Common) and 'A' for a 230V tranny, and between 'C' and 'B' for a 120V unit. The 1 ohm resistor allows you to measure the current easily - it must be less than the transformer's rated secondary current (if you measure 75mV RMS across 1 ohm, the current is 75mA).
I tested a fairly typical 12V, 150mA transformer (I didn't have a 15-0-15V transformer immediately to hand), and it draws about 60mA when driven in reverse with 12V RMS across the 12V winding and with the 230V winding unloaded. Output voltage was about 200V RMS. This doesn't leave much capacity, but a single valve stage doesn't draw very much current (about 1-2mA) so the transformer will not be overloaded. Should you think about it logically, there's rather a lot of extra work and cost just to include the valve, all for a rather intangible 'benefit'.
The power supply will not be inexpensive, because you need the extra transformer plus high-value, high-voltage capacitors to keep ripple to the minimum. The filtering shown will keep the ripple to less than 0.1mV peak-peak (about 35µV RMS) with a current of around 1.5mA, but you may choose to add another 100µF cap in parallel with C2 to reduce noise even further. The transformer only needs to be 5VA or so because of the low current, and the two 100µF caps need to be rated for a minimum of 200V DC. Then you need a bigger transformer for the main supply because you have to drive the HV transformer and another regulator for the valve heaters, as well as the normal ±15V supplies.
|Please note that power supplies are not shown for any of the opamp circuits. All opamps need power, and if you use dual types they use Pin-8 for positive and Pin-4 for negative. Supplies are normally ±15V is provided by Project 05 or similar. Supply rails aren't shown for clarity, as they make the schematic more difficult to read clearly.|
It's more logical, simpler, cheaper, lower noise and far more reliable to use an opamp as the first stage. A suitable design is shown below. The first stage is the one that would otherwise be replaced by a valve if you go that way, and everything after the Gain control will be the same from here on in.
Figure 4 - Opamp Preamp Stage
We'll use half of an OPA2134 for the front end, because it's a very high performance opamp with JFET inputs, so high impedance isn't a problem for it. The input gain is switched, and is designed to have very similar gain in both 'Low' and 'High' gain settings as the valve input stage. As noted, the Gain control needs to be 100k for the valve stage, but can be reduced to 10k (linear) with the opamp because of its much lower output impedance. Note the diode feeding the 'O/L' bus - if any section of the preamp exceeds the threshold voltage the O/L LED will come on to warn you that the signal level is too high.
This stage is not inverting, and the remainder of the circuit is arranged so that the overall preamp stage provides 'normal' polarity. That means that a positive-going signal at the input provides a positive-going signal at the output. However, be aware that all the EQ stages can introduce significant phase shift anyway, so it really makes little or no difference in real terms. If you use the valve input, the signal is inverted. You can add an inverting stage to restore 'normal' polarity if you wish, but it's not necessary.
The gain pot is linear (either 100k or 10k) because this is a Gain control, and is not intended to give the normal logarithmic characteristic of a Volume control. The idea is to be able to control the gain through the preamp in a nice linear manner, aiming for just enough level to cause the O/L (overload) LED to come on every so often but no more. In both versions of the preamp stage, VR1 is a dual-gang pot. The second stage has a gain of 6.56, so maximum total gain is near enough to ×67 (36dB, high gain) and ×30 (30dB, low gain). This gives a maximum input level of 15-30mV RMS on high gain and 33-66mV on low gain for a nominal preamp output level of 1-2V. Much higher input levels can be handled on the low gain setting, and the input level can typically be up to 1.4V RMS without clipping the input stage.
The output marked 'Tuner' goes to a jack on the rear panel, and is intended for a guitar/ bass electronic tuner. The other output (PreEQ) is used for the balanced send which goes to the front-of-house or recording mixer. This send can be switched between pre and post EQ. In case you were wondering, the terminal marked 'B&T' connects to the next stage - bass and treble controls.
EQ is the heart of a bass amp. Many commercial offerings have very comprehensive EQ, but the budget versions usually have the bare minimum. The nice thing about DIY is that you can do your own tests and determine what's right for you, but there's no reason to skimp on good tone controls because the cost isn't that high and the results far better than you'll ever get with a typical 'tone stack' as used for guitar.
Bass & Treble
The primary controls are (as always) bass and treble, but normal Baxandall type fixed controls are next to useless for any instrument, and this is especially true for bass guitar. There are countless different ways that variable-frequency controls can be implemented, but this method is fairly straightforward and works well.
The tone controls are shown below, and both bass and treble have a variable turnover frequency. Bass 3dB frequency (at maximum boost or cut) needs to be adjustable from around 100Hz, up to about 500Hz, with boost and cut of 12dB. The treble control has a frequency that can be varied from around 335Hz up to 1.7kHz. In each case, this allows a range of a little over two octaves, and while a wider range is possible, it's unlikely to be useful.
Figure 5 - Bass & Treble Controls
There are simpler and more complex ways to achieve the same result, but the version shown is a good overall compromise. The bass control makes use of a variable (gyrator based) inductor, and the frequency where the pot works is determined by the inductance. It's a shame that the pot for bass frequency control needs to be 100k (almost all others are 10k), but the gyrator won't work properly in this circuit if the resistance is too low. The treble control uses a variable capacitance multiplier. The output of the bass & treble controls goes directly to the parametric stage. C0 (100nF) will likely be needed if you use an NE5532, because they always have some input DC offset that can cause the pot (VR1, Figure 2 or 4) to be noisy. It's not needed if you use a TL072 or OPA2134.
The ability to switch between bass shelving and peaking is optional. The switch changes between the two, and when closed (C1 shorted) the bass control is shelving, giving 'traditional' tone control operation. The 4.7µF cap will be something of a nuisance, because it's a large value for a polyester cap, and it may be easier to use lower value caps in parallel. The value isn't overly critical, so you could use 5 x µF caps in parallel with only a small frequency change (the lowest frequency will fall by less than 1Hz). Don't use an electrolytic cap here, and definitely not a tantalum cap!
You do need to be wary of the variable capacitance multiplier. While it works exactly as described, on occasion it may be found in an unstable state after power is applied. I used this circuit is Project 199, and (predictably) it refused to misbehave when test equipment was anywhere near the circuit. In use, and very occasionally, it was found to be unstable. I've not been able to create a guaranteed fix, but the 10 Megohm resistor shown (R9) seems to work. If you prefer, just use a switch to select different capacitors, and leave out U2B and associated circuitry. The nominal minimum capacitor value is around 22nF, and the maximum value is 120nF.
|Bass (Hz) - Peaking||30||34||40||50||72|
|Bass (Hz) ±3dB||66||85||120||205||720|
|Treble (Hz) ±3dB||310||400||565||970||3.4k|
There is considerable overlap between the ranges of the two controls, and this is intentional. When combined with the parametric sections, the overlap provides for a vast range of tone adjustment. It's certainly possible to add even more variety by including a co-called 'contour' control, but that's something that can be achieved easily (and with far better control) using the parametric sections.
Figure 6 - Bass & Treble Control Response
In the above, you can see the response of the controls. Frequency is stepped through by 25% increments, from zero to full rotation, boost and cut are shown at maximum cut, flat and maximum boost. The maximum range has been limited to ±12dB, and although more is possible it's unlikely to be useful. The response of the bass control in peaking mode is not shown. A peaking bass control has a similar response to the parametric midrange controls shown next, but at the frequencies shown in the above table.
Next we have the two parametric sections. These are configured to have a maximum boost and cut of 12dB, and the Q is 1.26 at maximum boost or cut. The frequency of the 'Low Mid' section can be varied from 66Hz to 720Hz. The 'High Mid' section has the same boost, cut and Q, and is variable from 310Hz to 3.4kHz. You can change the frequency range of both by varying the values of C1 & C2 (Low Mid) or C3 & C4 (High Mid). Keep each pair of caps the same value. Calculate the frequency using the standard RC frequency formula ...
f = 1 / ( 2π × R × C ) (where R is resistance in ohms and C is capacitance in Farads). For example ...
f = 1 / ( 2π × 110k × 22nF ) = 65.76 = 66Hz
This is the frequency for the Low Mid section when VR3 (A&B) is at maximum resistance, and is exactly as expected.
Figure 7 - Parametric EQ Stages
There are countless options for parametric equalisers, but the circuit shown is quite straightforward and doesn't need 4 opamps for each frequency (that's standard with state variable filters) [ 1 ]. You don't have the ability to modify the filter Q, but that's not usually an option on bass amps anyway. Provided you can change the frequency, two midrange parametric sections will usually be sufficient, especially when combined with variable frequency bass and treble controls. Obviously, if you want more control, another section can be added, but the tone control arrangement described provides four variable frequency controls.
Each parametric section's frequency is tuned with a dual-gang pot (VR3 and VR4). The Q remains constant as frequency is varied, and as shown each section has a ±12dB range. Because of the number of stages, low impedances are used to minimise noise. The circuit shown is not quite as well behaved as the alternate version with buffers as described in Project 150, but is simpler and it's highly doubtful that you will hear any difference at all. The down-side is that the frequency pots are 100k, so there may be a small increase in audible noise if you use maximum boost. The gain (and Q) are increased by the addition of R6 and R11, and while these can be reduced for more boost and cut, it's not recommended. 12dB is an increase (or decrease) of 4 times, and increasing the amount of boost makes it far too easy to clip the opamps with a high level signal.
Figure 8 - Low Mid & High Mid Control Response
The response of the two sections is shown in Figure 8, at maximum boost and cut, and maximum and minimum frequencies. Intermediate settings aren't included because the graph would be a complete mess of traces and you'd be unable to see anything even remotely useful. Be warned that if the Low Mid and High Mid are set for the same frequency (for example 500Hz) the maximum possible boost is 24dB (x17)! This is grossly excessive and will definitely cause the opamps to clip, because just 300mV input will cause the O/L LED to come on.
The high pass filter is tuned for a -3dB frequency of 27Hz, the frequency being selected so as to not attenuate the lowest notes, but still using standard value resistors and capacitors. It has a small (1dB) peak before rolloff, but this will not cause any problems. When the bypass switch is operated, response is completely flat. Rolloff is 24dB/ octave, and will remove subsonic frequencies very effectively. The response at 10Hz is down by 35dB.
You could also build just one of the filters shown, but performance is nowhere near as good and for the sake of a couple of dollars in parts its worth the extra effort.
Figure 9 - High Pass Filter
The subsonic filter can be switched in or out without affecting the gain. All that's needed is to bypass the capacitors, and the circuit is flat to DC. Naturally, this isn't really the case because capacitive coupling is used in several other places in the preamp so there is a natural bass rolloff, but it's below 20Hz and not well defined. That's the reason for the filter in the first place - to keep frequencies out of the amps and speakers that will just eat power but not produce useful sound.
With the values shown, the -3dB frequency is 27Hz. You can use higher or lower value caps to change this is you wish. For example, 150nF gives a -3dB frequency of 22Hz, and 100nF raises the frequency to 33Hz. You could make this switchable, but I doubt that there's any need - the values shown should be fine for most players.
There's nothing remarkable about the filter, other than the slightly unusual bypass which simply shorts the two capacitors. Most of the time the filter should be in-circuit to prevent any subsonic signals from causing excess cone travel and possible speaker damage. It also helps to conserve power, since less of the amp's output power will be used to create sound that can't be heard anyway.
The point shown as 'PostEQ' is used for the balanced send. This is the post-EQ output. The main output goes to the effects insert jacks.
Because of the amount of gain in the preamp section and especially the huge tone control variations available, it may be quite easy to cause various stages to clip. Using a LED indicator to show that the preamp is clipping lets you turn down the input Gain control and increase the level with the Volume control. I have only shown one 'O/L' bus, but you can add as many as you like. That can make it easier to determine which section is overloaded. Note that the GND connection should be direct to the power supply, and should not be shared with any of the signal stages. Figure 10 shows 2 separate detectors. The level can be varied by changing the value of R2/ R6. As shown, the detection threshold is 4.7V + 0.65V (for the diodes), so any signal above 5.3V peak will activate the LED.
Figure 10 - Clipping Indicator
This is a greatly simplified version of the clipping detector described in Project 146, but the constructor can use a 'better' detector if preferred. The main change suggested would be to detect both positive and negative peaks, but the circuit shown will work well enough for most users. As shown it will turn on the 'O/L' LED for any voltage above 5V peak.
The signal is sampled at the input preamp (twice), after the tone controls, the effects return and optionally before the crossovers. If any of these approaches clipping, the LED will come on. In most cases, the signal level should be high enough to cause the LED to flash briefly and very occasionally, as this indicates that the amp is running at a fairly high level throughout and is well above the noise floor.
Part 2 covers the compressor/ limiter, the crossovers, including a variable network to work with two cabinets using different speakers and one for the high frequency horn loaded compression driver. It will also cover the power amplifier drivers including distortion/ soft-clip circuits.
The tuner output and effects send and return circuits are also described, along with stereo operation. There's also the balanced output intended for the front-of-house PA mixer, or for recording. This can be switched to be pre or post EQ, because in many cases the sound you want on stage is not the same as that needed by the mixer, and your 'stage EQ' might be a bit radical for the PA system or studio mixer.
Many of the controls discussed will be on the rear panel, and these include the effects insert jacks, balanced send, pre/ post EQ switching, tuner output and the main outputs (assuming the unit you build is a preamp). Otherwise, the speaker outputs will also be on the back panel. We'll also look at power requirements and loudspeaker efficiency, amongst other things.
|Copyright Notice. This article, including but not limited to all text and diagrams, is the intellectual property of Rod Elliott, and is © 2015. Reproduction or re-publication by any means whatsoever, whether electronic, mechanical or electro- mechanical, is strictly prohibited under International Copyright laws. The author (Rod Elliott) grants the reader the right to use this information for personal use only, and further allows that one (1) copy may be made for reference while constructing the project. Commercial use is prohibited without express written authorisation from Rod Elliott.|