Embedded Files
Few amplifiers have been played, discussed, and emulated as much as the tweed-era Fender Bassman. The 1958-59 release of the 5f6a circuit is the most popular version. That circuit is the basis of the Weber kit build described on this page.
Few amplifiers have been played, discussed, and emulated as much as the tweed-era Fender Bassman. The 1958-59 release of the 5f6a circuit is the most popular version. That circuit is the basis of the Weber kit build described on this page.
While the tweed Bassman has a signature tone and is much appreciated, it also has some limitations, and few would argue that amp designs have not improved since then. In fact, Fender made some subtle changes in its current reissue of the 59 Bassman. (More about that later.)
While the tweed Bassman has a signature tone and is much appreciated, it also has some limitations, and few would argue that amp designs have not improved since then. In fact, Fender made some subtle changes in its current reissue of the 59 Bassman. (More about that later.)
My thinking on the Bassman changed after I finished my Deluxe Plus amp. The Bassman has the same two-channel preamp structure that I modified in the Deluxe Reverb. There are two sets of jacks pairs and two level controls. While that may have made some sense in the 50's when multiple instruments were plugged into the same amp, it doesn't today. Who needs two identical preamps when you can only use one? In fact, the Bassman feeds both preamps to the same tone stack, and each preamp channel is really simple - consisting of only one triode gain stage and a volume control.
My thinking on the Bassman changed after I finished my Deluxe Plus amp. The Bassman has the same two-channel preamp structure that I modified in the Deluxe Reverb. There are two sets of jacks pairs and two level controls. While that may have made some sense in the 50's when multiple instruments were plugged into the same amp, it doesn't today. Who needs two identical preamps when you can only use one? In fact, the Bassman feeds both preamps to the same tone stack, and each preamp channel is really simple - consisting of only one triode gain stage and a volume control.
For anything but stage volume levels, the Bassman preamp and power amp stages are unlikely to clip, so modest volume levels don't produce much distortion. It's basically a clean amp. In fact, the Bassman power amp is the first stage that's likely to clip, but only once you hit 50w output levels. The amp does sound good even at lower levels because the DC-coupled cathode follower (DCCF) driving the tone stack produces nice warm "brown" tones, but it's rather one-dimensional and you have little control over that tone without changing the room volume significantly.
For anything but stage volume levels, the Bassman preamp and power amp stages are unlikely to clip, so modest volume levels don't produce much distortion. It's basically a clean amp. In fact, the Bassman power amp is the first stage that's likely to clip, but only once you hit 50w output levels. The amp does sound good even at lower levels because the DC-coupled cathode follower (DCCF) driving the tone stack produces nice warm "brown" tones, but it's rather one-dimensional and you have little control over that tone without changing the room volume significantly.
So, these two issues - a duplicate preamp channel, and little control over distortion - are points where the Bassman could be improved. The question is, what to do and how best to do it.
So, these two issues - a duplicate preamp channel, and little control over distortion - are points where the Bassman could be improved. The question is, what to do and how best to do it.
I focused on living within the chassis and tube constraints, which means the only available tube is one 12ax7 triode section of one of the two preamp channels. So, by re-purposing one of the preamp channels, I also get to use its volume control, which allows blending of the stock channel with the new channel.
I focused on living within the chassis and tube constraints, which means the only available tube is one 12ax7 triode section of one of the two preamp channels. So, by re-purposing one of the preamp channels, I also get to use its volume control, which allows blending of the stock channel with the new channel.
I opted to pursue the same strategy that worked well in the Deluxe Plus. I'd modify the spare channel to create some pre-amp distortion and mix it with the stock channel via their respective level controls. One tube stage doesn't have enough gain to generate much distortion, but the new stage - I call it the "Grit-stage" - can be cascaded with the stock preamp stage. Two common-cathode stages with gain of ~50, produce a gain of ~2500, which is plenty to generate some usable distortion. This seemed straightforward, except that there is a problem with phase.
I opted to pursue the same strategy that worked well in the Deluxe Plus. I'd modify the spare channel to create some pre-amp distortion and mix it with the stock channel via their respective level controls. One tube stage doesn't have enough gain to generate much distortion, but the new stage - I call it the "Grit-stage" - can be cascaded with the stock preamp stage. Two common-cathode stages with gain of ~50, produce a gain of ~2500, which is plenty to generate some usable distortion. This seemed straightforward, except that there is a problem with phase.
The stock clean stage feeds both the DCCF and the new grit stage. The grit stage is inverting, so it's output can't be summed into the DCCF without cancelling the clean signal. An inversion is needed, but no tubes are available.
The stock clean stage feeds both the DCCF and the new grit stage. The grit stage is inverting, so it's output can't be summed into the DCCF without cancelling the clean signal. An inversion is needed, but no tubes are available.
The solution is a passive inversion with a small interstage transformer. It's uncommon to find an interstage transformer in any guitar amp - but when they do appear, it's usually in the driver stage where they provide dual phase-inverted outputs to the power tubes. Simple inversion is needed in my preamp, and a transformer is ideal for that. It's also simpler and lower cost than adding another tube stage, and any non-ideal transformer behavior can contribute to the distortion that is the whole point of the grit stage to begin with.
The solution is a passive inversion with a small interstage transformer. It's uncommon to find an interstage transformer in any guitar amp - but when they do appear, it's usually in the driver stage where they provide dual phase-inverted outputs to the power tubes. Simple inversion is needed in my preamp, and a transformer is ideal for that. It's also simpler and lower cost than adding another tube stage, and any non-ideal transformer behavior can contribute to the distortion that is the whole point of the grit stage to begin with.
So, with a transformer inversion, the grit stage is feasible and it address both goals - better use of the spare channel and a means for controlling the amount of distortion. However, the physical panel changes in the amplifier also have to be considered. One set of the input jacks is now unused, since the grit channel is fed by the clean channel instead. While there is nothing wrong with unused jacks, (one pair is normally unused anyway) I considered how to use that panel space for something more worthwhile.
So, with a transformer inversion, the grit stage is feasible and it address both goals - better use of the spare channel and a means for controlling the amount of distortion. However, the physical panel changes in the amplifier also have to be considered. One set of the input jacks is now unused, since the grit channel is fed by the clean channel instead. While there is nothing wrong with unused jacks, (one pair is normally unused anyway) I considered how to use that panel space for something more worthwhile.
As mentioned above, the DCCF is the main source of the Bassman tone-quality. If the amp had a Master Volume control after the DCCF stage, it would be possible to drive the DCCF stage harder or softer without impacting room volume. A master combined with the grit stage would provide two independent controls over two different distortion sources. That would open up a huge new range of tones for the Bassman.
As mentioned above, the DCCF is the main source of the Bassman tone-quality. If the amp had a Master Volume control after the DCCF stage, it would be possible to drive the DCCF stage harder or softer without impacting room volume. A master combined with the grit stage would provide two independent controls over two different distortion sources. That would open up a huge new range of tones for the Bassman.
Versions and Demo Clips:
Versions and Demo Clips:
The Bassman Redux design now has three versions that have evolved over several years. The first is described is detail immediately below. The second version (BR3) follows that and there are demo clips of that version. More recently, a third version (BR4) was added at the end of this page.
The Bassman Redux design now has three versions that have evolved over several years. The first is described is detail immediately below. The second version (BR3) follows that and there are demo clips of that version. More recently, a third version (BR4) was added at the end of this page.
If you get that far, you'll get a sense of how a design starts and evolves over time. It's really hard to get everything right on the first try. After several years of playing and working on a design, it does improve. I'm quite pleased and proud of BR4.
If you get that far, you'll get a sense of how a design starts and evolves over time. It's really hard to get everything right on the first try. After several years of playing and working on a design, it does improve. I'm quite pleased and proud of BR4.
Initial Design (BR2)
Initial Design (BR2)
With all the above goals in mind, I measured the panel spaces and bless Leo, the space used by the jacks was just right for adding a new control. I moved the Norm Vol and Grit Vol toward the remaining input jacks and added the Master Vol next to the tone controls. See the complete full scale panel layout here.
With all the above goals in mind, I measured the panel spaces and bless Leo, the space used by the jacks was just right for adding a new control. I moved the Norm Vol and Grit Vol toward the remaining input jacks and added the Master Vol next to the tone controls. See the complete full scale panel layout here.
Note that simply getting a new panel is only part of the problem. You also have to drill some holes in the chassis for the two little switches (1/4" holes are pretty easy) and the added volume pot (3/8" hole).
Note that simply getting a new panel is only part of the problem. You also have to drill some holes in the chassis for the two little switches (1/4" holes are pretty easy) and the added volume pot (3/8" hole).
Since I needed a new panel anyway, I also decided to flip the labels so they're readable from the front of the amp (I rarely play behind an amp). I also never liked the white text on chrome. It's hard to read in good lighting, never mind dark rooms, and the letters wear off too easily.
Since I needed a new panel anyway, I also decided to flip the labels so they're readable from the front of the amp (I rarely play behind an amp). I also never liked the white text on chrome. It's hard to read in good lighting, never mind dark rooms, and the letters wear off too easily.
So, I settled on a brass panel with black letters, which looks great with the tweed. I had the panel made by PlaqueMaker.com. I sent them the PDF file and they sent back exactly what I asked for.
So, I settled on a brass panel with black letters, which looks great with the tweed. I had the panel made by PlaqueMaker.com. I sent them the PDF file and they sent back exactly what I asked for.
In summary, the grit channel, master volume, and new panel are the key goals of the design. The actual circuit schematic and related discussion are below.
In summary, the grit channel, master volume, and new panel are the key goals of the design. The actual circuit schematic and related discussion are below.
Power Supply
Power Supply
The high-voltage supply chain is unmodified. The filament supply is now elevated. It's common practice to elevate the filament supply to a positive DC level to reduce the stress on the DCCF stage (V2b). Its cathode is idling at ~200v and that's beyond the spec (~180v) for cathode to heater voltage on a 12ax7. Actual tubes seem to tolerate this pretty well, but I've blown at least one over the years and Blencowe describes the start up stresses that can damage these stages.
The high-voltage supply chain is unmodified. The filament supply is now elevated. It's common practice to elevate the filament supply to a positive DC level to reduce the stress on the DCCF stage (V2b). Its cathode is idling at ~200v and that's beyond the spec (~180v) for cathode to heater voltage on a 12ax7. Actual tubes seem to tolerate this pretty well, but I've blown at least one over the years and Blencowe describes the start up stresses that can damage these stages.
I tried something that seems like a simple low cost solution that reduces the stress on this tube and reduces heater-induced hum at the same time. Elevation of the heaters is done through R47 and R48. The reference is a +60v supply created from the bias supply tap and D2 and C19.
I tried something that seems like a simple low cost solution that reduces the stress on this tube and reduces heater-induced hum at the same time. Elevation of the heaters is done through R47 and R48. The reference is a +60v supply created from the bias supply tap and D2 and C19.
Two single wave rectifiers are used to produce both the negative bias and a positive heater reference supply from the same transformer tap. It's a textbook concept to produce both positive and negative voltages from a single tap transformer, but I've never seen it implemented or used in a guitar amp. In fact, even textbooks only mention the idea in passing. (An online description of this generic circuit is found here.) (Update: I just found the same dual diode idea used in a Dynaco Series II Stereo 70 tube amp - although it's not used for heater elevation.)
Two single wave rectifiers are used to produce both the negative bias and a positive heater reference supply from the same transformer tap. It's a textbook concept to produce both positive and negative voltages from a single tap transformer, but I've never seen it implemented or used in a guitar amp. In fact, even textbooks only mention the idea in passing. (An online description of this generic circuit is found here.) (Update: I just found the same dual diode idea used in a Dynaco Series II Stereo 70 tube amp - although it's not used for heater elevation.)
Most often, elevated references are created with a voltage divider off the high-voltage supply chain. While that also works, an independent reference switched on/off with the heaters (before the standby switch) and drawing no current from the high voltage supply is a good, if not better option.
Most often, elevated references are created with a voltage divider off the high-voltage supply chain. While that also works, an independent reference switched on/off with the heaters (before the standby switch) and drawing no current from the high voltage supply is a good, if not better option.
This could and probably should be done in any amp with a bias tap. Heater elevation is also helpful for reducing preamp hum, so for the cost of a diode and small cap - why not? The Bassman, a low-gain amp, is not known to be particularly noisy or have a hum problem, but with the elevated heaters, what little hum I recall with the volume pots at zero is now completely gone. V2a has an unbypassed cathode, so it's susceptible to heater-induced hum, which was otherwise audible even with preamp volumes set to zero.
This could and probably should be done in any amp with a bias tap. Heater elevation is also helpful for reducing preamp hum, so for the cost of a diode and small cap - why not? The Bassman, a low-gain amp, is not known to be particularly noisy or have a hum problem, but with the elevated heaters, what little hum I recall with the volume pots at zero is now completely gone. V2a has an unbypassed cathode, so it's susceptible to heater-induced hum, which was otherwise audible even with preamp volumes set to zero.
The bias supply has independent control for each output tube, as described in the initial amp build pages. Two resistors R41 and R42 are added for safety. They force the power tube grids to high bias (lower plate current) should the P8 or P9 wipers become intermittent or open. This is a good idea I picked up from Enzo in his blog postings (thanks).
The bias supply has independent control for each output tube, as described in the initial amp build pages. Two resistors R41 and R42 are added for safety. They force the power tube grids to high bias (lower plate current) should the P8 or P9 wipers become intermittent or open. This is a good idea I picked up from Enzo in his blog postings (thanks).
Preamp
Preamp
The Bassman normal channel is preserved with V1a. Note the two series 820 ohm bias resistors (R5 and R6) bypassed by C1. These set this stage to operate exactly like the original. The bright switch (SW1) is a three position (center off) switch so it allows for two degrees of brightness or none at all. The bright switch only impacts the clean channel whose output is controlled by P1 and passed to the DCCF (V2) exactly as in the original circuit. The clean channel preserves all the characteristics of the original Bassman. None of the original Bassman qualities are lost in this modification.
The Bassman normal channel is preserved with V1a. Note the two series 820 ohm bias resistors (R5 and R6) bypassed by C1. These set this stage to operate exactly like the original. The bright switch (SW1) is a three position (center off) switch so it allows for two degrees of brightness or none at all. The bright switch only impacts the clean channel whose output is controlled by P1 and passed to the DCCF (V2) exactly as in the original circuit. The clean channel preserves all the characteristics of the original Bassman. None of the original Bassman qualities are lost in this modification.
The grit channel is created with V1b. Its input is the output of the normal channel. Note there is no volume control in the signal path to this point. The guitar signal passes through both V1a and V1b without any level control in the amp. The guitar volume is used to control the signal, as needed. V1b amplifies the signal from V1a and drives the transformer, T1. R8, R9, and C5 produce a 228 vdc supply created just for this stage. The bias point is shown in the curve set image below.
The grit channel is created with V1b. Its input is the output of the normal channel. Note there is no volume control in the signal path to this point. The guitar signal passes through both V1a and V1b without any level control in the amp. The guitar volume is used to control the signal, as needed. V1b amplifies the signal from V1a and drives the transformer, T1. R8, R9, and C5 produce a 228 vdc supply created just for this stage. The bias point is shown in the curve set image below.
The AC signal load to V1b depends on the load on the secondary. A switch system (SW2) with three positions varies the load impedance, equalization, and levels to create three Voice options. The reflected primary impedances for the three SW2 settings are shown as three AC load lines through the bias point. Notice the bias point is fairly centered in the curve set. I opted for center-bias since the DCCF already creates a lot of warm-bias tone for the amp. The whole point of this mod is adding more variations of tone to the Bassman.
The AC signal load to V1b depends on the load on the secondary. A switch system (SW2) with three positions varies the load impedance, equalization, and levels to create three Voice options. The reflected primary impedances for the three SW2 settings are shown as three AC load lines through the bias point. Notice the bias point is fairly centered in the curve set. I opted for center-bias since the DCCF already creates a lot of warm-bias tone for the amp. The whole point of this mod is adding more variations of tone to the Bassman.
The transformer is a Hammond 148Q. It's priced rather high at ~$20, but it's widely available and fits the need. (Most any small transformer would also work.) The secondaries are connected in series to obtain a 10k:2.5k (4:1) impedance ratio, which is a 2:1 turns ratio.
The transformer is a Hammond 148Q. It's priced rather high at ~$20, but it's widely available and fits the need. (Most any small transformer would also work.) The secondaries are connected in series to obtain a 10k:2.5k (4:1) impedance ratio, which is a 2:1 turns ratio.
All three voice options feed similar levels to the grit volume control, P2, that also match the level of the V1b input signal. In other words, the grit stage has an overall gain of ~1.0. That means mixing grit signals from V1b and clean signals from V1a through R18 and R19 is straightforward since they're at similar levels.
All three voice options feed similar levels to the grit volume control, P2, that also match the level of the V1b input signal. In other words, the grit stage has an overall gain of ~1.0. That means mixing grit signals from V1b and clean signals from V1a through R18 and R19 is straightforward since they're at similar levels.
An overall stage gain of ~1 does not mean that there is no gain in V1b. There is sufficient input signal on the grid to easily drive V1b into cutoff and grid current, resulting in clipping, which is really all that's needed from the stage. Any actual voltage gain in V1b is reduced by the divider network set by SW2 to maintain an overall unity stage gain.
An overall stage gain of ~1 does not mean that there is no gain in V1b. There is sufficient input signal on the grid to easily drive V1b into cutoff and grid current, resulting in clipping, which is really all that's needed from the stage. Any actual voltage gain in V1b is reduced by the divider network set by SW2 to maintain an overall unity stage gain.
The schematic shows signal voltages in green text. There is a measured ~190vpp on the plate of V1b for a ~200 mvpp input with SW2 in the mid voice position. Many pickups easily produce that signal level so clipping is easily reached and controlled by your playing technique and guitar volume setting.
The schematic shows signal voltages in green text. There is a measured ~190vpp on the plate of V1b for a ~200 mvpp input with SW2 in the mid voice position. Many pickups easily produce that signal level so clipping is easily reached and controlled by your playing technique and guitar volume setting.
The Voice switch (SW2) produces the following settings, which are tuned to my taste by ear. There are many degrees of freedom to this stage, allowing a wide variation of tone.
The Voice switch (SW2) produces the following settings, which are tuned to my taste by ear. There are many degrees of freedom to this stage, allowing a wide variation of tone.
This chart below lists the secondary load impedance, stage gain, and RC corner freq's for each switch setting. These calculations are not exact models of how the circuit behaves due to interactions between circuit elements, however these numbers convey an idea of what's happening. In the end, your ears matter most with such parameter tunings - not the calculations.
This chart below lists the secondary load impedance, stage gain, and RC corner freq's for each switch setting. These calculations are not exact models of how the circuit behaves due to interactions between circuit elements, however these numbers convey an idea of what's happening. In the end, your ears matter most with such parameter tunings - not the calculations.
Mid-voice 15K//.022uf = 480Hz LP stage gain = ~10x
Mid-voice 15K//.022uf = 480Hz LP stage gain = ~10x
Produces symmetric spiky waveforms on plate of V1b.
Produces symmetric spiky waveforms on plate of V1b.
Up voice 15K//2.7k //.022uf = 3200Hz LP stage gain = ~4x
Up voice 15K//2.7k //.022uf = 3200Hz LP stage gain = ~4x
Produces hard negative clipping waveforms on plate of V1b.
Produces hard negative clipping waveforms on plate of V1b.
Down voice 15K//6.8k // .032uf = 967Hz LP and 1M//4.7M//240pf = 755Hz HP stage gain = ~8x
Down voice 15K//6.8k // .032uf = 967Hz LP and 1M//4.7M//240pf = 755Hz HP stage gain = ~8x
Produces negative clipping at high signal levels, decaying to symmetric spiky waveforms on plate of V1b.
Produces negative clipping at high signal levels, decaying to symmetric spiky waveforms on plate of V1b.
The HP filter gives this voice it's most audible quality.
The HP filter gives this voice it's most audible quality.
The point is just to produce some variation. I know words don't mean much, but I created options I think of as treble-bite, medium crunch, and low-snarl. The bias-point, EQ, and load parameters can easily be tweaked for personal taste. The grit and clean signals are merged (in phase!) to drive the DCCF, and the rest of the amp. The tone stack follows the combined signals so further EQ shaping opportunities lie there.
The point is just to produce some variation. I know words don't mean much, but I created options I think of as treble-bite, medium crunch, and low-snarl. The bias-point, EQ, and load parameters can easily be tweaked for personal taste. The grit and clean signals are merged (in phase!) to drive the DCCF, and the rest of the amp. The tone stack follows the combined signals so further EQ shaping opportunities lie there.
Power Amp
Power Amp
The power amp is slightly modified from the tweed 5f6a version. These changes are made by Fender in its current 59 Bassman Reissue. I marked the changed components in red on the schematic.
The power amp is slightly modified from the tweed 5f6a version. These changes are made by Fender in its current 59 Bassman Reissue. I marked the changed components in red on the schematic.
Basically Fender added small grid stoppers on the output tubes - always a good idea for stability, with little impact on tone. The negative feedback components are changed, and the phase inverter tail resistor, R24 is lowered. These last two impact the tone to a slight degree, but I made the changes and I'm quite happy with the results. In fact without them my initial build had an occasional low buzz at the very end of a decaying note. I didn't trace it's source (maybe a spurious oscillation?) but with these changes, the output stage is rock solid (verified on a scope at low and full output) and I'm much happier with the behavior of the new presence control.
Basically Fender added small grid stoppers on the output tubes - always a good idea for stability, with little impact on tone. The negative feedback components are changed, and the phase inverter tail resistor, R24 is lowered. These last two impact the tone to a slight degree, but I made the changes and I'm quite happy with the results. In fact without them my initial build had an occasional low buzz at the very end of a decaying note. I didn't trace it's source (maybe a spurious oscillation?) but with these changes, the output stage is rock solid (verified on a scope at low and full output) and I'm much happier with the behavior of the new presence control.
I was looking for some use of the ground switch (SW5) and I decided to use it for controlling negative feedback. I know this causes a pop when its flipped, and I solved that problem in the Deluxe Plus, but I keep it simple here and I can live with the small pop.
I was looking for some use of the ground switch (SW5) and I decided to use it for controlling negative feedback. I know this causes a pop when its flipped, and I solved that problem in the Deluxe Plus, but I keep it simple here and I can live with the small pop.
The Master Volume control, P3 is added between the tone stack and the phase inverter, so I think of it as part of the power amp. Nothing special about it - but it allows me to freely control the amplitude of signals fed from the preamp stages into the DCCF stage (V2) and thereby the degree of its distortion produced for any room volume.
The Master Volume control, P3 is added between the tone stack and the phase inverter, so I think of it as part of the power amp. Nothing special about it - but it allows me to freely control the amplitude of signals fed from the preamp stages into the DCCF stage (V2) and thereby the degree of its distortion produced for any room volume.
That's it for now. It's a simple set of mods that really increases the utility and range of the Bassman. I highly recommend these to anyone sympathetic to my views about the stock amp limitations.
That's it for now. It's a simple set of mods that really increases the utility and range of the Bassman. I highly recommend these to anyone sympathetic to my views about the stock amp limitations.
These changes are simplest to implement if you're building a new amp from scratch or from a kit. The hardest part of the mod process was drilling the three holes in the chassis. The circuit changes are simple and the parts are cheap and easily found.
These changes are simplest to implement if you're building a new amp from scratch or from a kit. The hardest part of the mod process was drilling the three holes in the chassis. The circuit changes are simple and the parts are cheap and easily found.
The gut shot below shows most of the changes clearly. I don't follow the strict classic parts layout. I mount numerous components on the controls and switches or tube sockets to keep wires short. That leaves some eyelets unused for additions. I did add the small perf board for bias controls, but otherwise I had no problem fitting the remaining parts on the standard eyelet board.
The gut shot below shows most of the changes clearly. I don't follow the strict classic parts layout. I mount numerous components on the controls and switches or tube sockets to keep wires short. That leaves some eyelets unused for additions. I did add the small perf board for bias controls, but otherwise I had no problem fitting the remaining parts on the standard eyelet board.
The transformer (T1) and the filter cap (C5) are clearly visible near the right-most tubes. All the preamp components are on the right end of the chassis.
The transformer (T1) and the filter cap (C5) are clearly visible near the right-most tubes. All the preamp components are on the right end of the chassis.
Update 4/2/14 - BR3 hum reduction
Update 4/2/14 - BR3 hum reduction
The amp described above is relatively quiet, but with full volume, there is some noticeable hum and I just could not leave it be. I tracked it down and an updated schematic (v3) is below.
The amp described above is relatively quiet, but with full volume, there is some noticeable hum and I just could not leave it be. I tracked it down and an updated schematic (v3) is below.
The hum was not due to any component or circuit changes - but rather simply due to layout of those components. The problem is that the input jacks are on the front panel, the eyelet board spans about 4-5 inches of chassis and the V1 tube socket lies on the opposite chassis side, away from the jacks. Initially (in v2 described above), I shielded the wire from the jacks to the V2 grid (pin 2). This is common practice and helps reduce hum. However the bias network (R5, R6, C1) was on the eyelet board with about 4 inches of unshielded wire running to the cathode (V1 pin3). It turns out that it was enough to pickup an annoying and persistent hum.
The hum was not due to any component or circuit changes - but rather simply due to layout of those components. The problem is that the input jacks are on the front panel, the eyelet board spans about 4-5 inches of chassis and the V1 tube socket lies on the opposite chassis side, away from the jacks. Initially (in v2 described above), I shielded the wire from the jacks to the V2 grid (pin 2). This is common practice and helps reduce hum. However the bias network (R5, R6, C1) was on the eyelet board with about 4 inches of unshielded wire running to the cathode (V1 pin3). It turns out that it was enough to pickup an annoying and persistent hum.
I must admit, this wiring issue was the last in a series of hum reduction measures I pursued. I never considered the cathode circuit a hum risk and I thought it was heater hum. I put in a DC heater (regulated) using a Canon power module I had laying about, and the heaters already were elevated.
I must admit, this wiring issue was the last in a series of hum reduction measures I pursued. I never considered the cathode circuit a hum risk and I thought it was heater hum. I put in a DC heater (regulated) using a Canon power module I had laying about, and the heaters already were elevated.
The rest of the amp from volume control to speaker was and still is dead quiet - you literally can not tell it is on. Yet the hum from V1 (pre-volume stage) was still there. I could see about 3-4 mvpp on the plate of V1 with the input shorted. (You can see the note I made about this on the v2 schematic.)
The rest of the amp from volume control to speaker was and still is dead quiet - you literally can not tell it is on. Yet the hum from V1 (pre-volume stage) was still there. I could see about 3-4 mvpp on the plate of V1 with the input shorted. (You can see the note I made about this on the v2 schematic.)
The lesson learned is that the cathode connection to the bias RC is also important to either keep short or shield, and the grid and cathode connections should be routed together to minimize their loop area which leads to induction by magnetic fields. It's common to shield the grid line, but it seems that the same care is needed on the cathode line as well.
The lesson learned is that the cathode connection to the bias RC is also important to either keep short or shield, and the grid and cathode connections should be routed together to minimize their loop area which leads to induction by magnetic fields. It's common to shield the grid line, but it seems that the same care is needed on the cathode line as well.
My fix removed the bias RC from the eyelet board and attached it directly to the jacks (this means there are very short wires to the jack ground). I then used 2-conductor shielded wire (about 5-6 inches) between the jack circuitry and the V1 grid and cathode. The loop area of the grid and cathode connections is minimized by running the lines together in the same cable. Now the hum is completely gone. No component changes - no cost - just layout changes.
My fix removed the bias RC from the eyelet board and attached it directly to the jacks (this means there are very short wires to the jack ground). I then used 2-conductor shielded wire (about 5-6 inches) between the jack circuitry and the V1 grid and cathode. The loop area of the grid and cathode connections is minimized by running the lines together in the same cable. Now the hum is completely gone. No component changes - no cost - just layout changes.
I am going to leave the DC heater, but I suspect it is not really needed. Adding the DC heater for V1 made no noticeable change in the level of hum. I suspect removing will make no difference, however, it can't hurt. I think this layout approach is something worth remembering and sharing since it's simple, costs nothing, and makes a big difference.
I am going to leave the DC heater, but I suspect it is not really needed. Adding the DC heater for V1 made no noticeable change in the level of hum. I suspect removing will make no difference, however, it can't hurt. I think this layout approach is something worth remembering and sharing since it's simple, costs nothing, and makes a big difference.
Demo Clips
Demo Clips
These clips were recorded with a NANDY CM-100 pencil condenser mic. The amp is on the ground and the mic is on an eight-inch stand about three-feet away and directly facing the amp. When played on a good hifi system these clips sound pretty close to what I'm hearing as I play. The recording was a single 20 min session, and edited later using REAPER, which is a great DAW at a great price (free).
These clips were recorded with a NANDY CM-100 pencil condenser mic. The amp is on the ground and the mic is on an eight-inch stand about three-feet away and directly facing the amp. When played on a good hifi system these clips sound pretty close to what I'm hearing as I play. The recording was a single 20 min session, and edited later using REAPER, which is a great DAW at a great price (free).
To minimize the number of variables between clips, these settings are constant for all clips:
To minimize the number of variables between clips, these settings are constant for all clips:
Bass = Treble = Mid = 4, Presence = 0. NFB is on and the bright switch is off.
Bass = Treble = Mid = 4, Presence = 0. NFB is on and the bright switch is off.
The guitar is my American Standard Telecaster with its volume and tone at max using only the neck pickup.
The guitar is my American Standard Telecaster with its volume and tone at max using only the neck pickup.
The room and mic levels stay roughly constant. I didn't touch the mic gain during the entire session and room volume was a comfortable practice level (for me), even with others in the house. You'll hear me talking between segments to announce when I switch the grit channel voices.
The room and mic levels stay roughly constant. I didn't touch the mic gain during the entire session and room volume was a comfortable practice level (for me), even with others in the house. You'll hear me talking between segments to announce when I switch the grit channel voices.
This first clip has two segments and uses only the Norm channel. This shows the effect of the Master volume control. The first segment has the MV =10 (out or 12) and Norm = 3. A high MV setting makes the Norm channel work the same way it does in a standard Bassman, so the tone is clean vintage Bassman in the first segment. The second segment has the MV = 3 and Norm = 12. This setting pushes extremely large signals through the DCCF stage (V2) and creates a nice cranked Bassman tone, but at room volume. This illustrates the kind of distortion tone produced by the DCCF stage.
This first clip has two segments and uses only the Norm channel. This shows the effect of the Master volume control. The first segment has the MV =10 (out or 12) and Norm = 3. A high MV setting makes the Norm channel work the same way it does in a standard Bassman, so the tone is clean vintage Bassman in the first segment. The second segment has the MV = 3 and Norm = 12. This setting pushes extremely large signals through the DCCF stage (V2) and creates a nice cranked Bassman tone, but at room volume. This illustrates the kind of distortion tone produced by the DCCF stage.
This clip has three segments and only uses the Grit channel, so Norm = 0. The MV = 12, so there is low signal in the DCCF stage, and all the distortion is coming from the Grit stage. The three segments correspond to the Voice switch settings of Up, Mid, and Down. You can hear me announce the switch changes between segments. The clip starts in the Up position. Notice that the bass is reduced in the Grit channel (due to R10 and C6), relative to the bass levels in the Norm channel. The voice variations are pretty apparent and I would describe the Up voice as a balanced tone, the Mid voice as a dark/woody tone, and the Down voice as a bright fuzzy tone.
This clip has three segments and only uses the Grit channel, so Norm = 0. The MV = 12, so there is low signal in the DCCF stage, and all the distortion is coming from the Grit stage. The three segments correspond to the Voice switch settings of Up, Mid, and Down. You can hear me announce the switch changes between segments. The clip starts in the Up position. Notice that the bass is reduced in the Grit channel (due to R10 and C6), relative to the bass levels in the Norm channel. The voice variations are pretty apparent and I would describe the Up voice as a balanced tone, the Mid voice as a dark/woody tone, and the Down voice as a bright fuzzy tone.
This first mixed clip starts with a clean segment where MV = 5, Norm = 5, and Grit = 0. This is a balanced setting where some distortion comes from the DCCF, but it's still pretty clean. The second and longer segment blends in some grit with Grit = 5, so the Norm and Grit channels contribute equally to the tone. Note that by blending (rather than switching channels) you get both the clear full bass of the clean channel and dirt from the Grit channel This longer segment uses first the Up voice, then the Mid and Down voices. You'll note that switching voices doesn't make that much difference since the grit voice only contributes to half the signal.
This first mixed clip starts with a clean segment where MV = 5, Norm = 5, and Grit = 0. This is a balanced setting where some distortion comes from the DCCF, but it's still pretty clean. The second and longer segment blends in some grit with Grit = 5, so the Norm and Grit channels contribute equally to the tone. Note that by blending (rather than switching channels) you get both the clear full bass of the clean channel and dirt from the Grit channel This longer segment uses first the Up voice, then the Mid and Down voices. You'll note that switching voices doesn't make that much difference since the grit voice only contributes to half the signal.
This second mixed clip has three segments, one for each Grit voice, but now the balance between Norm and Grit favors Grit. I keep MV = 5, and Grit = 5, but now Norm = 3. The three segments use the Up, Mid, and Down voices, in sequence. With this blend balance favoring the grit channel, the voice changes make more difference.
This second mixed clip has three segments, one for each Grit voice, but now the balance between Norm and Grit favors Grit. I keep MV = 5, and Grit = 5, but now Norm = 3. The three segments use the Up, Mid, and Down voices, in sequence. With this blend balance favoring the grit channel, the voice changes make more difference.
In summary, it's pretty clear that this amp has a wide range of tones that extend the range of a Bassman. Of course, it's also nice that the amp still is a Bassman and can do everything a normal Bassman does - only now it has some surprises too.
In summary, it's pretty clear that this amp has a wide range of tones that extend the range of a Bassman. Of course, it's also nice that the amp still is a Bassman and can do everything a normal Bassman does - only now it has some surprises too.
Update (5/5/15) - BR4 improvements in distortion and crosstalk
Update (5/5/15) - BR4 improvements in distortion and crosstalk
After playing v3 for a while, and hearing from some people that tried the mods, I wanted to change two things. First, there is a tone quality to the BR3 grit channel that I found appealing at first, but it's become too fizzy and harsh for my taste over time. (I think I took the "grit" concept too literally.) Second, there is a slight distortion in the clean channel, even with the grit level at zero. It's a form of crosstalk and that should not be there. The BR4 circuit fixes both of these issues.
After playing v3 for a while, and hearing from some people that tried the mods, I wanted to change two things. First, there is a tone quality to the BR3 grit channel that I found appealing at first, but it's become too fizzy and harsh for my taste over time. (I think I took the "grit" concept too literally.) Second, there is a slight distortion in the clean channel, even with the grit level at zero. It's a form of crosstalk and that should not be there. The BR4 circuit fixes both of these issues.
The BR4 distortion is softer in the sense that it's less abrupt and less harsh. It's also more responsive to the guitar volume setting. Backing off on the guitar volume gives a smooth reduction in the degree of distortion. Even at full guitar volume levels, the distortion is less harsh. It a subjective thing, but I find it more responsive and pleasant. A worthwhile change, IMO.
The BR4 distortion is softer in the sense that it's less abrupt and less harsh. It's also more responsive to the guitar volume setting. Backing off on the guitar volume gives a smooth reduction in the degree of distortion. Even at full guitar volume levels, the distortion is less harsh. It a subjective thing, but I find it more responsive and pleasant. A worthwhile change, IMO.
A 12at7 as V1 is a lower gain tube so it will give a cleaner tone in both the BR3 and BR4. In the BR4, it's almost too clean (IMO), but let your ears be your guide.
A 12at7 as V1 is a lower gain tube so it will give a cleaner tone in both the BR3 and BR4. In the BR4, it's almost too clean (IMO), but let your ears be your guide.
As for the distortion crosstalk, the BR4 changes have eliminated it. I can't hear any trace of it anymore.
As for the distortion crosstalk, the BR4 changes have eliminated it. I can't hear any trace of it anymore.
The BR4 schematic is below.
The BR4 schematic is below.
The BR4 changes from BR3 are all in the grit channel. First, the distortion leakage turned out to be a high-frequency coupling of wires around V1 that induced small spurious pickup in both channels.
The BR4 changes from BR3 are all in the grit channel. First, the distortion leakage turned out to be a high-frequency coupling of wires around V1 that induced small spurious pickup in both channels.
V1 is used for two cascaded triode gain stages. They have a combined gain of over 3000 and there is no level control between them. That means the V1b anode driving the transformer is almost always clipping at about 400v of peak to peak signal. That's happening right next to the input stage triode V1a. In my layout, the V1b anode signal has only about 2-3 inches of wire from the tube socket to the transformer pin. However, that was radiating enough signal throughout the chassis and causing what was essentially crosstalk into the clean channel.
V1 is used for two cascaded triode gain stages. They have a combined gain of over 3000 and there is no level control between them. That means the V1b anode driving the transformer is almost always clipping at about 400v of peak to peak signal. That's happening right next to the input stage triode V1a. In my layout, the V1b anode signal has only about 2-3 inches of wire from the tube socket to the transformer pin. However, that was radiating enough signal throughout the chassis and causing what was essentially crosstalk into the clean channel.
The worst coupling was to the wiring at the input jack. Although the input jack signal is run through a shielded cable to the V1a grid, the 68k resistors and their connections are exposed. Only a few millivolts of signal on the V1a grid is enough to create noise and oscillations given the cascaded gain.
The worst coupling was to the wiring at the input jack. Although the input jack signal is run through a shielded cable to the V1a grid, the 68k resistors and their connections are exposed. Only a few millivolts of signal on the V1a grid is enough to create noise and oscillations given the cascaded gain.
Crosstalk into other wires was also happening, so the best solution is to stop the source. The solution is simple (fortunately) - a shielded cable from the V1b anode to the transformer input pin. The shield is grounded at one end and the exposed unshielded connections are kept to less than 1/4" in length.
Crosstalk into other wires was also happening, so the best solution is to stop the source. The solution is simple (fortunately) - a shielded cable from the V1b anode to the transformer input pin. The shield is grounded at one end and the exposed unshielded connections are kept to less than 1/4" in length.
Problem solved. I can't find or hear any trace of crosstalk anymore. It's common to use shielded cable for sensitive inputs (like grid wires), but it's also important to keep the noise emission down by shielding the "loud" high-signal amplitude connections. In fact, shielding the signal sources is arguably even more important since all of the chassis circuits benefit from a quieter environment.
Problem solved. I can't find or hear any trace of crosstalk anymore. It's common to use shielded cable for sensitive inputs (like grid wires), but it's also important to keep the noise emission down by shielding the "loud" high-signal amplitude connections. In fact, shielding the signal sources is arguably even more important since all of the chassis circuits benefit from a quieter environment.
To change the distortion character, the drive to V1b was changed. Standard practice is to use a large grid stopper to isolate the source from an over-driven grid. This is almost universal in over-driven stages and BR3 used a 47k resistor for that purpose.
To change the distortion character, the drive to V1b was changed. Standard practice is to use a large grid stopper to isolate the source from an over-driven grid. This is almost universal in over-driven stages and BR3 used a 47k resistor for that purpose.
The problem is that the isolating resistor (or grid stopper) also strongly impacts the distortion of the driven stage, making it harsher with more abrupt distortion onset. A compromise is often struck to select a resistor that isolates enough, yet minimizes undesirable distortion impacts.
The problem is that the isolating resistor (or grid stopper) also strongly impacts the distortion of the driven stage, making it harsher with more abrupt distortion onset. A compromise is often struck to select a resistor that isolates enough, yet minimizes undesirable distortion impacts.
BR4 does away with that resistor and instead uses a transistor buffer (Q1) as a dc-coupled emitter-follower. This has two benefits. First, it eliminates any loading issues on the V1a anode so the clean channel is truly isolated without any compromises. Second, it drives the V1b grid from a low impedance source, driving the grid more smoothly into overdrive. (More about low impedance overdrive or LIO is discussed in this book.)
BR4 does away with that resistor and instead uses a transistor buffer (Q1) as a dc-coupled emitter-follower. This has two benefits. First, it eliminates any loading issues on the V1a anode so the clean channel is truly isolated without any compromises. Second, it drives the V1b grid from a low impedance source, driving the grid more smoothly into overdrive. (More about low impedance overdrive or LIO is discussed in this book.)
By now you're probably thinking - won't the transistor ruin the tube tone? Well, for several reasons - no. This transistor is benign. It can't be over-driven and doesn't add anything of its own to the tone. It's simply a "better" wire since it acts as a unity gain buffer to isolate the two stages and drive the second stage. (BTW, any NPN transistor or N-channel MOSFET rated for 300v and ~1watt dissipation is usable in this position.)
By now you're probably thinking - won't the transistor ruin the tube tone? Well, for several reasons - no. This transistor is benign. It can't be over-driven and doesn't add anything of its own to the tone. It's simply a "better" wire since it acts as a unity gain buffer to isolate the two stages and drive the second stage. (BTW, any NPN transistor or N-channel MOSFET rated for 300v and ~1watt dissipation is usable in this position.)
The results are very pleasing. The distortion character is much improved. If you've built a Redux, or added a grit-stage to a similar amp, you'll want to consider this idea.
The results are very pleasing. The distortion character is much improved. If you've built a Redux, or added a grit-stage to a similar amp, you'll want to consider this idea.
Update (7/12/16) - Adding tremolo to the Bassman chassis
Update (7/12/16) - Adding tremolo to the Bassman chassis
The stock Bassman has a useless ground switch on the front panel. The Redux mods use that switch to turn NFB on/off. Since I rarely find myself using that switch, I hardwired NFB "on" and replaced the switch with the Amp-Aid tremolo module. Now I have tremolo depth and speed on the front panel. It's a great addition to the amp. Highly recommended.
The stock Bassman has a useless ground switch on the front panel. The Redux mods use that switch to turn NFB on/off. Since I rarely find myself using that switch, I hardwired NFB "on" and replaced the switch with the Amp-Aid tremolo module. Now I have tremolo depth and speed on the front panel. It's a great addition to the amp. Highly recommended.
Page updated
Google Sites
Report abuse