5F6A-M Bassman Micro

[robrob]

This is my latest micro amp project, the Bassman Micro tube guitar amplifier. It's a compact, moderately easy to build, three tube, 3 to 5 watt output practice amp (around 3 watts with a 12AU7 power tube and around 5 watts with a 12BH7). The Bassman Micro's single channel preamp circuit was inspired by the 5F6A Bassman's Bright Hi channel. The Marshall JTM45 preamp is an exact copy except for the use of a 12AX7 in V1 so this amp could also be called the JTM45 Micro. The Bassman Micro features two stages of gain, a tone stack buffer, a cathodyne phase inverter and a true push-pull power amp.

More inof on the Bassman Micro: http://robrobinette.com/Bassman_Micro.htm

[img738]http://robrobinette.com/images/Guitar/B ... _small.png[/img]

[img600]http://robrobinette.com/images/Guitar/B ... _small.png[/img]

I'm a big fan of the 1950's Fender tweed amps and the 5F6A Bassman is a legendary amp that was copied by Marshall and became the foundation of that company's amp lineup so I decided design a Micro version of the Bassman. I used an almost exact copy of the Bassman's preamp and paired it with a cathodyne phase inverter and small bottle push-pull power amp. Why a cathodyne PI? Because I didn't need the extra gain of the Bassman's long tail pair phase inverter and the it would have required another tube. The greatest weakness of the cathodyne PI is if driven too hard it can develop nasty double-frequency distortion. I use a pre-phase inverter master volume paired with a 270k phase inverter grid stopper resistor to eliminate double-frequency distortion.

I liked the output and flexibility of the 12AU7/12BH7 tubes for the true push-pull power amp. The 12AU7 with both triodes in push-pull will develop around 3 to 4 watts of output power. Swap in a 12BH7 and the output increases to 4 to 5 watts. My H&K Tubemeister 5 uses a 12BH7 in push-pull and it sounds very good. The 12AU7 and 12BH7 tubes are very similar in output tone but the volume boost from the 12BH7 is noticeable. Both tube types are readily available with the 12BH7 going for about $5 more than a 12AU7.

One of the reasons I chose to use the Fender 5F6A tweed Bassman preamp in this project is the Marshall JTM45 uses an exact copy of the circuit except Marshall calls for a 12AX7 in V1 instead of the Bassman's lower gain 12AY7. All that's required to convert the Bassman Micro into a JTM45 Micro is to put a 12AX7 in V1. A 12AX7 in V1 gives you much more overdrive and pushes the Bassman Micro's tone into the 1960's.

I installed the Bright/Normal Channel Switch because the only difference between a 5F6A's channels is the Bright Cap. By putting it on a switch you can choose between the two Bright and Normal channels.

The Negative Feedback Switch allows you to disconnect the feedback loop and get a grittier, more 5E3 Deluxe type tone with earlier breakup, or turn on the feedback for the standard Bassman tone. I went with a 82k feedback resistor because the 5F6A used a 27k resistor with a 2 ohm speaker tap. To get the equivalent feedback from my 8 ohm speaker tap the resistor must be bumped up to 56k. Use a 39k resistor for a 4 ohm speaker tap and 82k for a 16 ohm tap. Feel free to adjust the feedback resistor value to your liking or even install a pot to vary the feedback but I'd use about 12k as the minimum feedback resistance (12k resistor in series with a 100k pot).

The Tone Stack Bypass Switch is an easy mod that removes the signal sucking tone stack from the circuit. The signal will be 'boosted' when the tone stack is bypassed.

The EZ81 rectifier tube uses a standard 9-pin socket and 6.3V filament heat so no 5V power transformer output is needed. If your power transformer does have 5V available and you'd rather use a standard, full size 8 pin rectifier tube like a 5Y3GT then just wire the 8 pin socket with the HT to pins 4 and 6, 5V heater lines to 2 and 8 and the B+ power line to pin 8.

[chordball]

You're really killin' it with these! Thanks again for sharing your work. Do you have a preference between this and the 5e3 micro? Or is it just like asking if someone prefers a bassman to a tweed deluxe?

[dorrisant]

Nice work. Gonna try this very soon.

Tony

[robrob]

You're really killin' it with these! Thanks again for sharing your work. Do you have a preference between this and the 5e3 micro? Or is it just like asking if someone prefers a bassman to a tweed deluxe?

I haven't built the Bassman yet but the Deluxe Micro exceeded my expectations--it's a very cool little practice amp. I'm ordering parts and working on the turret board now for the Bassman. It'll probably be a while before I get it built, tested and tweaked but I'll report my progress here.

Rob

[pdf64]

It looks a great project, but I'm not sure that having the global feedback loop encompass the 2nd gain stage / tone stack / master vol will give best result.
Rather I would look to put the loop return at the cathodyne grid somehow, or perhaps just leave it open loop, being as the triodes have their own inherent local nfb.

I have a hunch that Marshall used a 12AX7 in V1 because 12AY7 and 30% audio taper pots may have been (and for the latter, still are) hard to source in the UK.
The 10% audio taper volume control gives a better 'user interface' / resolution with the 12AX7.
Also they may have needed the extra gain at the top end to make up for higher degree of negative feedback implemented in the JTM45, by the use of the 16 ohm output for the 27k global feedback return (rather than the 2 ohm output used on the 5F6A).
I wonder if that was accidental or intentional?

[robrob]

I'm going to evaluate the feedback loop closely since like you say it may not be necessary at all with non-bypassed triode power tubes.

I've always heard the almost 3x increase in feedback on the Marshall JTM45 over the 5F6A was an accident.

[pdf64]

Regarding the 270k grid stopper on the 2nd stage, to be faithful to the original it may be more accurate to use 120k-150k, as the 2 x 270k mixers are effectively in parallel for most control settings.
Regarding the unbypassed shared output cathodes, bear in mind this will introduce a degree of compression (assuming class AB operation); for lower signal levels ie neither enters cut off, the cathode impedance of either tube tends to bypass the other. For higher signal levels, that will only occur for part of the cycle, maybe not at all.
How about independent unbypassed cathodes, as that may act to increase output impedance, so better mimicking pentode behaviour?

[robrob]

pdf, I don't see how the bright and normal channels are parallel electrically. They appear to be separate until they hit the 270k mixing resistors, the controls are only connected through ground.

Regarding the power tube shared cathode resistor, during my research I never found a 12AU7 or 12BH7 power amp with separate cathodes. Actually, I don't recall ever seeing any pair of cathode biased power tubes without a shared cathode resistor but it's worth a try.

[robrob]

pdf, I see on the 5F6A schematic when the non-use channel's volume is full down the 270k mixing resistors actually form a voltage divider to cut the active channel's input signal in half but I don't see the mixing resistors in parallel.

[pdf64]

As I see it, the key issue is the total effective grid circuit impedance, as this then interacts with the tube's Miller capacitance to set the HF roll off.
Given that, probably the best way to find the grid circuit impedance is to do a Norton analysis of the V1 plates - 2nd stage V2 grid.
It's beyond the scope of the area to go through that in detail, but it should be clear that with both volumes at zero, the grid circuit impedance is 135k?
Both at max will be ~150k, and both electrically halfway (ie ~80% setting) ~260k but all these are ignoring the effect of the bright cap.
So for a single 'best fit' value, it may be seen that something rather lower than 270k is probably appropriate.
Hope the above makes sense?

[robrob]

Speaking of the 5F6A Bassman here's an annotated schematic I did with signal flow and component function:



[img814]http://robrobinette.com/images/Guitar/B ... ematic.gif[/img]

[johnnyreece]

That would be very handy for new folks coming in! It has a lot of frequently used terminology with its associated parts. Thanks for sharing!

[electrochimp]

the bassman is close to the JTM45 and this one has the 12au7 for a power tube .. sounds pretty decent
https://www.youtube.com/watch?v=w71tBeuLn2Y

[robrob]

I believe the bassman in the video uses the 12au7 in self-split push-pull mode which gives 1/2 half the power of a 12au7 in true push-pull mode like in the Bassman Micro.

[robrob]

Using the Deluxe, Bassman or Champ Micro As a Tube Tester

The Deluxe, Champ and Bassman Micro's can use the same power and preamp tubes so all three tweed Micro amps can function as a preamp tube tester. If you do plan to use the amp as a tester then try to match V1's 100k load resistors and 1.5k cathode resistors so you can take bias readings of the two halves of the tube to check for balance. Balanced triodes are important for long tail pair phase inverters or push-pull power micro amps. You also want to measure and record the exact values of the two cathode resistors for precise bias measurements.

Insert the tube to be tested into V1 and power up the amp. To read the bias of V1A and V1B make sure the input jack does not have a guitar plugged into it, turn the volume up to 11 (we're trying to match V1A's 33k grid stopper) and turn the Tone to 12. Set your multimeter to DC Volts and place its probes on each leg of the cathode resistor. The voltage drop across resistor will be shown so record it. If the voltage is negative just ignore the sign. Then place the black probe on a ground and touch the red probe to V1's pin 1 to measure V1A plate voltage and record it. Place the red probe on V1's pin 6 to measure V1B's plate voltage and record it.

With these numbers we can now calculate the tubes bias. We need the Plate-to-Cathode voltage so subtract the cathode voltage from the plate voltage and record the numbers. Then use my https://robrobinette.com/Tube_Bias_Calculator.htm webpage to do all the calculations. As an example we'll say the measured cathode voltage drop was 2.3 volts and the plate voltage was 150. We subtract the 2.3v from 150v and get 147.7 plate-to-cathode voltage. We go to the calculator and select our tube type of '12AY7' and you'll see it's rated for 1.5 watts per triode. We then plug in our plate-to-cathode voltage of '147.7'. Move down to the next section, Tube Dissipation Using Cathode Resistor Voltage Drop, and enter the number of tubes that share a cathode as '1' because V1A and V1B have separate cathode resistors. If you were testing using V2 then you would enter '2' because V2A and V2B share a cathode resistor. Next enter the cathode resistor voltage drop of '2.3' and then the measured resistance of the cathode resistor of 1522 ohms. Press any Calculate button and our answer is shown in the triode line that states, If your tube is a triode then there is no screen current so your plate current is 1.5 milliamps, your plate dissipation is 0.2 watts and 13.3%. You would do the same for the other triode and compare their numbers--the closer the better for phase inverters and micro power tubes.

Below I offer up layout diagrams for a traditional eyelet board, point-to-point wiring and ValveWizard B9A development board PCB's.[/url]