Fender Bassman 5B6 amplifier

I wanted a nice clean amp to play rockabilly tunes on. I'd heard good things about the Fender Bassman line and found the 5B6 model over at the Fender Amp Field Guide. I've never heard it but it's a bit simpler than the more popular 5F6-A model and (I think) lower power. It may sound like crap, but if so I'll have to tear it apart and build it into something else.

• schematic
• layout

Below are some high-level thoughts on how this circuit works. I'm not an amplifier expert so there will probably be some mistakes.

Preamp

This is a pretty simple cathode-biased triode preamp stage. Looking at the load line the gain appears to be around 60. One different thing about the 6SC7 preamp tube is it has a single cathode that is shared between both triodes. This works for this input stage because both the inputs are similar signals. If different frequency responses were needed we would want something with separate cathodes (like a 12AX7, for instance).

The two plates are tied together for a single output to the tone/volume circuit. This prevents separate tone/volume controls for the two inputs.

Tone/volume control

The tone control here is a simple lowpass RC filter (aka a "treble rolloff"). Not much else to say here.

Phase Inverter

The signal arrives at the V2a grid (from the volume pot) and is amplified at the V2a plate. The gain here is ???. A small fraction of this output is voltage divided out between R9 and R14 and fed into the V2b grid. An amplified version of this appears at the V2b plate. Because a tube inverts its output signal the signals at C6/R9 and C7/R10 will be 180 degrees out of phase with each other.

Poweramp

I don't yet understand this well enough to say much. It's a push-pull amplifier.

Amplifiers are expensive to build. The transformers alone for this project would run a good $150. For much less than that I got an old Conn organ amplifier off of ebay with all the transformers and tube sockets I need plus a chassis that will probably work fine for my project. And I get some nice old tubes that I can reuse.

(click for full size)

(click for full size)

I checked the B+ line with all the tubes except the rectifier pulled, it's reading 480 VDC. Strangely, the filter caps seem to be holding a voltage fine, it took about 5 minutes to drain down to 20 V. Hopefully under load this will drop down to the 400 VDC that I need.

There's a few things in this amp I won't be using. There's a 3rd transformer that appears to be some sort of line-out transformer for an external amplifier. I have no use for it. There's also a small choke in the power supply circuit that I don't think I'll need.

The 5B6 uses 5881 power tubes, which are essentially identical to the 6L6's in the organ amp. So the output transformer ought to work for my amp. But there's 3 secondary taps and I didn't have a datasheet for this ancient transformer. Someone from a forum showed me a trick to measure the impedance ratios of a transformer.

I took a function generator and set it to output a 1 Vrms sine wave at various frequencies. This 1 V signal is connected to a secondary tap and the voltage at the primary is measured. The impedance ratio is (Vp/Vs)^2. If you multiply this by your speaker impedance you get the impedance seen by the output stage.

(click for full size)

The output stage needs to see an impedance of about 4 kOhm. So it looks like my output transformer will provide 8 and 16 ohm outputs (perfect for guitar amps) and a 32 ohm output that won't be useful for me.

For my reference and anyone who happens to find an amp with these transformers, here's what I figured out for the wiring.

K-48530 power transformer

Wire	Function
GRN/BLK	primary hot
BLK	primary ground
ORG	H.T. hot
ORG/YEL	H.T. center tap
YEL	rectifier heater (2x)
GRN	filament heater (2x)
GRN/YEL	filament heater center tap
ORG/BLK	n/c
YEL/BLK	n/c

48531 output transformer

Wire	Function
BLU	primary
BRN	primary
ORG	primary center tap
YEL	1:500 secondary
GRY	1:250 secondary
GRN	1:125 secondary
BLK	ground

How do you build a Bassman? You start with an amp that has everything and cut out everything which is not a Bassman. In this case that involves removing some smaller tube sockets and moving at least one of the octal plugs to the preamp section of the chassis.

(click for full size)

The original layout of this organ amp seemed reasonable. The long ground bus will make for an easy star grounding scheme. The preamp and phase inverter sockets are placed away from the transformers and as close to the input signal as possible. And the tubes are oriented at 90 degrees apart (to reduce induced noise?).

I bored out one with a 1-1/8" hole cutting drill bit and then cut an entirely new hole for the 2nd socket.

(click for full size)

Now I'm left with several extra holes. For safety reasons these need to be blocked off (you want to keep dumb people from poking their fingers into the amp and touching a 400V line). Apparently there are off the shelf hole plugs just for this but I just cut some sheet metal pieces and glued them inside. The big rectangular hole where the unused transformer was I bolted a piece of sheet metal.

I was smart this time and started on the chassis finishing before I built I started building the amp. This takes a little bit of up-front planning since you can't just drill holes as you go along but it beats desoldering all the external components after you get everything working perfectly. I found some blog posts by a guy who does the markings on his amps using India ink and old cartographer lettering kits. Looks really nice when it's done.

The India ink thing looks really easy and turns out well, but I realized I can use a laser printer to print onto decal transfer sheets. I've never worked with decals but as long as I can apply them correctly this should turn out great. So the plan is:

1. Sand the chassis smooth.
2. A few coats of white primer and a little sanding.
3. Final paint finish with a glossy paint of some sort. I'm leaning towards a "celery" color.
4. Apply decals.
5. Apply some kind of lacquer coat to seal the decals.

(click for full size)

There it is after the primer coat.

(click for full size)

And there after several coats of "celery" gloss spray paint. I don't know much about painting things. It came out kind of rough. The labels are previews of the decals I intend to print with a laser printer and apply.

For labels I used "decal stock" for laser printers from Small Bear Electronics. I used GIMP to create a 300 dpi 8.5x11" image. Some text and some graphics and print (make sure you reverse the image!) and bam you've got decals. Here is the finished chassis after applying the decals and a few coats of Krylon gloss acrylic.

The Bassman logo I copied from a photo of an older Bassman and converted to black & white:

It looks a little jagged but in decal form it looks pretty good.

(click for full size)

(click for full size)

The decals worked well. It takes some practice to get right (good thing I printed many copies of each piece of text). Next time it will be easier. I would definitely recommend this for on-the-cheap labeling of DIY amps.

I had bought 4x 33 uF capacitors for the power supply but the existing stacked capacitors (the shiny cylinders on the top) seem to be fine. The capacitance meter on my multimeter reads 60 uF on the big ones and 30 uF on the smaller. This is odd because they're marked as 40 and 20 uF, respectively. It's also odd because they're probably 60 years old. Apparently USA-made caps last forever. Makes no sense but they'll do. This simplifies the internal wiring since they don't need turrets to wire them to.

(click for full size)

Well, it worked the first try. But something isn't right. The volume is way too low and the amp is distorting extremely early. The B+ is about 510 VDC - about 110 VDC too high. That can't be good. I'm lucky nothing exploded as my filter caps are only rated to 450 VDC. I've ordered a couple 50W Zener diodes (100 and 120 V) to help with this. I'm a little fuzzy on the correct way to do this but the idea is the diode is used to raise the voltage of the power transformer secondary center, thus lowering the differential voltage output.

Not much I can do until I fix that. I've played it briefly but definitely don't want to crank it up with the whole circuit biased incorrectly like it is. I was able to verify the volume and tone knobs work and aren't backwards. The tone circuit works great.

(click for full size)

(click for full size)

(click for full size)

The 120 V Zener diode dropped B+ down to 402 VDC - perfect. But it's still quiet and distorted. The layout suggests the 3 filter cap voltages should be 400, 320, and 300 VDC but I'm seeing 402, 387, and 375 VDC. The cathod of the preamp tube ought to be 2.1 V but instead it's at 1.6 V. It turns out this is because I didn't connect the 6L6 screens to the power supply!.

voltages with 120 V Zener installed but power amp tube screens floating

Node	Voltage (V)
preamp plate, V1a/V1b	113
preamp cathode	1.64
phase inverter plate, V2a	60.6
phase inverter plate, V2b	61.4
phase inverter cathode, V2a	1.09
phase inverter cathode, V2b	1.04
power amp plate, V3	416
power amp plate, V4	413
power amp cathode, V3/V4	0.502
power amp screen, V3/V4	-0.994
B+1	402
B+2	387
B+3	375

With that corrected things look and sound much more normal:

voltages with 120 V Zener installed

Node	Voltage (V)
preamp plate, V1a/V1b	81
preamp cathode	1.16
phase inverter plate, V2a	38
phase inverter plate, V2b	39.2
phase inverter cathode, V2a	0.85
phase inverter cathode, V2b	0.82
power amp plate, V3	324
power amp plate, V4	312
power amp cathode, V3/V4	23.3
power amp screen, V3/V4	283
B+1	320
B+2	283
B+3	250

Now it sounds like a real amp. It gets loud and stays clean. There's some hum and the volume never shuts off completely, but the main concern is B+ being 320 V instead of 400 V.

voltages with 100 V Zener installed

Node	Voltage (V)
preamp plate, V1a/V1b	89
preamp cathode	1.25
phase inverter plate, V2a	43.8
phase inverter plate, V2b	45.2
phase inverter cathode, V2a	0.905
phase inverter cathode, V2b	0.87
power amp plate, V3	346
power amp plate, V4	333
power amp cathode, V3/V4	25
power amp screen, V3/V4	305
B+1	354
B+2	305
B+3	296

It looks like a 50 V Zener would be ideal. These are only available rated for 10 W. Is this good enough? RC-30 says two push-pull class AB 6L6 tubes can sink 140 mA. The smaller tubes will only draw 1.5 mA between them. 50 V * 140 mA = 7 W, so a 10 W Zener should work.

Things I've tried so far to eliminate the background hum:

• Wired the power tube heaters in phase.
• Rewired all the heaters with a really tight twist and careful layout.
• DC biased the heater center tap at 25 V (tied to power tube cathode).

And... no change so far.

With some help from a couple guys on some forums I found out I had the cathode connections swapped on the phase inverter. How did this work at all?

voltages with 100 V Zener and correct phase inverter cathode wiring

Node	Voltage (V)
preamp plate, V1a/V1b	84.6
preamp cathode	1.25
phase inverter plate, V2a	41.7
phase inverter plate, V2b	43
phase inverter cathode, V2a	0.87
phase inverter cathode, V2b	0.82
power amp plate, V3	339
power amp plate, V4	339
power amp cathode, V3/V4	21.9
power amp screen, V3/V4	305
B+1	348
B+2	305
B+3	295

More help from a forum guy and I found out I had wired R4 in series with R7 and R8 instead of in parallel. This caused really low PI plate voltages.

voltages with 100 V Zener and correct phase inverter plate wiring

Node	Voltage (V)
preamp plate, V1a/V1b	119
preamp cathode	1.67
phase inverter plate, V2a	164
phase inverter plate, V2b	164
phase inverter cathode, V2a	2.44
phase inverter cathode, V2b	2.44
power amp plate, V3	347
power amp plate, V4	347
power amp cathode, V3/V4	22.5
power amp screen, V3/V4	305
B+1	354
B+2	305
B+3	287

This will sound unsurprising but there's way too much bottom end in this Bassman. It sounds fine with a telecaster bridge pickup but with the neck pickup (which as a lot more bass response) all you hear is the bass strings shaking the windows. According to the Valve Wizard, the half-boosted frequency is:

f_1/2 ~= 1/(2*pi*Rk*Ck)

where Rk and Ck are the cathode resistor and capacitor. For the 2500 ohm/250 uF combination this comes out to about 0.25 Hz. Great for a bass amp (as far as I know). Replacing the 250 uF with 1 uF the cutoff is around 63 Hz. Since the low E on a guitar is 82 Hz this is just about right.

Initial testing showed it sounds much better. But there was still way too much bass. I put another 1 uF cap in series to bump the cutoff up to 127 Hz. This did the trick. It's still got a lot of bass response but it sounds bad-ass instead of just boomy.

Armed with a cheap oscilloscope I now have more visibility into the signal chain. Here is what I see with inputs grounded.

At the output of the preamp what looks like a 100 Hz 10 mVpp signal with some jaggedness to it. Grounding the preamp plate doesn't make the audible noise change.

(click for full size)

At the V2b phase inverter plate a very complicated signal that measures 5 mVpp and I think repeated at 200 Hz. The image was taken with the preamp plate grounded.

(click for full size)

At the power amp plates there is this 1.5 Vpp, 200 Hz sawtooth signal. Is that my hum? I'm not sure.

(click for full size)

V2b plate (upper) and V2a plate (lower). V2a plate shows 20 mV @ 120 Hz signal that looks really bad.

(click for full size)

An attempt at plotting the load line for V2a based on voltages measured above. Looks like we're in cutoff?

(click for full size)

I'm not sure what to make of any of this. The noise level at the phase inverter seems too small make a big difference. I think the sawtooth at the power tube plates just indicates the power tubes aren't perfectly matched.

No phase inverter is perfect, especially not the "paraphase" type used in this circuit. The below oscilloscope shots illustrate the mismatch of the PI outputs. The first shows the two outputs and the 2nd shows the two outputs summed.

(click for full size)

(click for full size)

You can see in some cases one side of the wave is a little larger than the other. This generates some distortion since the positive half of one output and the negative half of the other are amplified by the power stage to construct the entire waveform. Ideally the two PI outputs would sum to zero. The two levels can be fine-tuned by adjusting R14 (or so I'm told) but I may just leave it be, it sounds good.

Here are some the videos thatt these were taken from:
 

 

At the first filter cap (B+1) I was measuring 16 V of ripple. That's about 4.6% of 350 V, pretty decent especially for a push-pull amplifier. In my layout I used a single 33 uF cap instead of two 16 uF caps. Since I've been having problems with 120 Hz hum I tried adding another 47 uF (80 uF total) cap here. This brought the ripple down to 2.3%. There was a noticeable reduction in hum but it's still there. My theory is there is that ripple is somehow leaking into the preamp or phase inverter stages.

Mucking with the preamp cathode capacitor helped quite a bit but was never quite enough. I decided to try changing the treble rolloff to a bass rolloff control. Someone suggested using the bass control from a Fender 5C8 "Twin" model amp. This worked very well. If I hadn't finished the chassis already I might have both treble and bass controls but I don't want to redo the chassis and usually don't find much use for treble knobs anyways (most guitars have them).

I tried a couple master volume circuits to see if I could get a little grit out of this amp at reasonable volumes. The first was a "cross-line" master volume (CLMV) where part of each side of the phase inverter is bled to the other side. The bled signal is 180 degrees out of phase with the main signal so it cancels out a portion of the signal going to the power tubes, resulting in a lower output volume.

The second was a post phase inverter (PPIMV) design by Bruce Collins. This uses a dual-ganged pot to simply attenuate the signal going to the power tubes.

Both of these methods were functional but I wasn't able to detect any distortion even with the (preamp) volume cranked all the way. I'm not sure yet why there isn't some distortion at max volume.

I had been playing around with smaller preamp plate resistors to bias the preamp tube hotter. Someone suggested decreasing R15 to give a higher voltage at the preamp plate. Earlier I had left the B+ voltage at 350 VDC, much lower than the spec'd 400 VDC. Swapping in a different Zener diode brought this up:

voltages with 50 V / 10 W Zener and correct phase inverter plate wiring

Node	Voltage (V)
preamp plate, V1a/V1b	132
preamp cathode	1.77
phase inverter plate, V2a	198
phase inverter plate, V2b	200
phase inverter cathode, V2a	2.43
phase inverter cathode, V2b	2.48
power amp plate, V3	392
power amp plate, V4	392
power amp cathode, V3/V4	26.7
power amp screen, V3/V4	305
B+1	323
B+2	340
B+3	399

The higher voltage resulted in a slightly clearer sound. The preamp tube may still sound better if biased a hotter, but this was an improvement.

One odd thing I noticed was the bass control will bring the preamp plate voltage from 132 V to around 40 V. It still sounds fine but I thought this was suspicious. I added a .1 uF coupling cap between the preamp tube and the tone circuit and now it's steady.

6V6 tubes have the same pinout and similar characteristics as 6L6 tubes but have less output and tend to distort earlier. I had some old 6V6's from an amp I retubed so I plugged them in. They were signficantly quieter. I think there's two reasons for this. First, 6V6's expect a plate load of about 8 kOhm vs the 4 kOhm for 6L6's, so the impedance wasn't matched very well. Second, I don't think the biasing is correct. I drew the loadlines for each:

(click for full size)

I think the 6V6's would be operating entirely in class A based on the load lines, but I don't have a good understanding of that yet so may be wrong. I need to look at this under the scope to be sure.

In any case, the output transformer isn't matched well for 6V6's unless I switch to a 16 ohm speaker. I didn't care for the sound of the 6V6's as-is, but maybe properly biased and the right output impedance would make them sound good.

A friend helped me get some boards cut with some hand tools (hand planer, hand saw, etc). He turned out some pretty bad-ass, flat, square boards. The wood is 3/4" Peruvian walnut.

(click for full size)

There's the box glued/nailed together. It still needs sanding and finishing.

(click for full size)

There's the grill I'm making for it. I found some 1/8"x1/2" aluminum stock and a sheet of flattened aluminum mesh on ebay. Ebay has a surprising assortment of metal in hobby quantities. All the aluminum together was about $25. That may seem expensive but most metal vendors don't do small quantities.

(click for full size)

The grill turned out really well. Aluminum is surprisingly easy to work with and paints well. I used the same primer/paint/clearcoat as for the chassis.

This clip was recorded with an SM57 clone about a foot from the cabinet. The cabinet is a DIY 1x12 open back cab with an Eminence Swamp Thang (a bass-heavy speaker). The guitar is a MIM Telecaster with treble maxed out. The first bit is the neck pickup and the rest is the bridge pickup. Note: the occasional clicks are an issue with my recording setup, not the amp.