Amp Build Report - Micro-Bassman tube amp in Marshall Lead 12 SS

[klooon]

I thought it would be nice to share my last amp build. A Bassman Micro, designed by Rob Robinette [link] with one ECC99, according to this schematic: [Bassman Micro Schematic]. *)

I used an 80's Marshall Lead 12 solid state amp with a 10" speaker as a donor amp.

This is my layout:

it includes some voltage measurements/calculations.

*) EDIT: I see that the schematic has been updated. The version I used is this one (V2.6 10/2015):

[klooon]

Some pictures of the build process - part one:

[klooon]

Part two:

[klooon]

Then the sound... With the initial design, I wasn't so much happy. The clean sound was excellent, but the overdriven sound was very fizzy and a bit harsh/splatty. Using the master volume is even worse.

Then I made a few changes trying to reduce the horrible fizz. I removed the bright cap and added an extra 10uF filter cap for the preamp which not only reduced the hum quite a bit (now it's really silent) but also lowered the B+ in the preamp stage. This helped quite a bit, but still I am not happy. Again the clean tones are fine and also lead tones are nice. Especially with strumming chords it is still quite fizzy.

Any ideas to improve the design?

[loudthud]

I think the root of the problem is in the phase inverter. Opinions vary on this but I think the large grid stopper cases symmetrical clipping of the power amp. This suppresses even harmonics. There are several variations I have seen with components right on the grid of the PI. Experiment there, I'm not sure what will work. IMHO you need grid current to skew away from symmetrical clipping.

[The Dude]

Thanks for sharing! Nice work.

[klooon]

I think the root of the problem is in the phase inverter. Opinions vary on this but I think the large grid stopper cases symmetrical clipping of the power amp. This suppresses even harmonics. There are several variations I have seen with components right on the grid of the PI. Experiment there, I'm not sure what will work. IMHO you need grid current to skew away from symmetrical clipping.

Thanks louthud!
I will try jumpering the 270k grid stopper here (or replacing it with a lower value resistor??).
Further, if I have a look at the Princeton Reverb schematic [link], it has a cathodyne PI which is very similar except the 0.1uF coupling cap is put right at the cathode instead of the current situation in which the coupling cap is placed at the junction of the 1k5 cathode resistor and the 56k tail load resistor. I could also experiment with this.
Also, in this project [HotLucas Tube-Town], they have a slightly different PI with a voltage divider, somewhat lowering the grid input??

[mikepukmel]

Nice job, nice build. thanks for he photos!! Any tips for newbies on wiring and solder joints?

[mikepukmel]

Then the sound... With the initial design, I wasn't so much happy. The clean sound was excellent, but the overdriven sound was very fizzy and a bit harsh/splatty. Using the master volume is even worse.

Then I made a few changes trying to reduce the horrible fizz. I removed the bright cap and added an extra 10uF filter cap for the preamp which not only reduced the hum quite a bit (now it's really silent) but also lowered the B+ in the preamp stage. This helped quite a bit, but still I am not happy. Again the clean tones are fine and also lead tones are nice. Especially with strumming chords it is still quite fizzy.

Any ideas to improve the design?

I'm the newbiest of the newbies, regarding electronics (although Im reading every day!!). I had two Fender Princeton Reverb's over the years, and they both had that problem. Nice cleans, nasty, ugly clipping sounds. The only thing I could verify on various amp build and repair sites is that it is due to the Cathodyne phase inv, as other more experienced people have mentioned. I have read that they are very sensitive to "tuning", but I don't have the engineering chops yet to know how to fix it. Check out this article,

www.valvewizard.co.uk/cathodyne.pdf

He also references some other older articles from the era. Most of its over my head yet, but maybe it will help. Please let us know how you resolve the issues! thanks. MP

[klooon]

I'm the newbiest of the newbies, regarding electronics (although Im reading every day!!). I had two Fender Princeton Reverb's over the years, and they both had that problem. Nice cleans, nasty, ugly clipping sounds. The only thing I could verify on various amp build and repair sites is that it is due to the Cathodyne phase inv, as other more experienced people have mentioned. I have read that they are very sensitive to "tuning", but I don't have the engineering chops yet to know how to fix it. Check out this article,

www.valvewizard.co.uk/cathodyne.pdf

He also references some other older articles from the era. Most of its over my head yet, but maybe it will help. Please let us know how you resolve the issues! thanks. MP

Thanks! Interesting stuff. I have checked the website, but not the article yet.

[klooon]

Nice job, nice build. thanks for he photos!! Any tips for newbies on wiring and solder joints?

I especially try to take care of the grounding scheme. I follow the guidelines in this article: ValveWizard Grounding article. Especially figure 15.14 is very interesting. I try to use only one chassis ground point (close to the input jacks). I use jacks that are not ground connected to the chassis for instance.

For the rest with the wiring, I like to use a vintage push back, solid wire. Works very neatly. If a wire is moved too much, I remove it and replace it with a new one as I want to avoid 'broken' wires. Also here I try to run all ground return with a separate wire back to its 'ground node'. Have a look at grounding colours of this schematic of the Princeton Reverb. I learned a lot from the more experienced guys from this forum. Further, the wiring of the tube sockets, I try to use the 90 degree wire crossing rule as much as possible.

A tip for soldering. I try to hit (at least) all the wires at the same time with the soldering iron (sometimes also the eyelet if there are not too many wires), hold it for a second or two), push hard and supply the solder for another second or two (/three).

A few two cents...

[klooon]

Thanks louthud!
I will try jumpering the 270k grid stopper here (or replacing it with a lower value resistor??).
Further, if I have a look at the Princeton Reverb schematic [link], it has a cathodyne PI which is very similar except the 0.1uF coupling cap is put right at the cathode instead of the current situation in which the coupling cap is placed at the junction of the 1k5 cathode resistor and the 56k tail load resistor. I could also experiment with this.
Also, in this project [HotLucas Tube-Town], they have a slightly different PI with a voltage divider, somewhat lowering the grid input??

I will try to experiment and try different things I found after some reading on web. My "action plan":
1) Zobel network - I will get a 5uF (>50V) and a 8.2Ohm/5W-10W resistor. (I followed this calculator: for an 8 ohm speaker, the proper DC resistance of the speaker is about Re = 6.56Ohm ==> Rz = 8.2Ohm and with a Le of 0.35mH --> C ~ 5 uF).
2) Conjunctive filter - 2.2nF (3kV) and a 20k to 30k 10W resistor (1.2 to 1.3* the primary OT impedance (~21.5kOhm)).
3) or only a 2.2nF (3kV) cap across the primary leads of the OT (just like the FireFly, see this link)
4) Play around with the Phase Inverter design. Get rid/change values of the grid stopper, maybe Paul C mod (fixed bias mod) or something like that.
5) Maybe implement a Vox cut control...
6)....

[klooon]

still waiting for the components to arrive...

[J M Fahey]

I´d build a real Tweed Bassman, period.
No master , a real PI , meaning a long tailed one, not a cheesy cathodyne, and, most important, a couple power pentodes.
If you want low power, fine, forget 6L6, 6V6 and even EL84, but use a couple 6AQ5 or to simplify things, a couple ECL82, where you have 2 pentodes and 2 triodes (LTP wink wink) in just 2 small bottles.

Even if you forgo push pull out and use a single ended pentode stage, á la Champ/AX84/etc. you will still be closer to original Bassman sound than if you try to drive speakers with soggy noodle triodes.
And small signal triodes at that !!!!

[Dave H]

I had two Fender Princeton Reverb's over the years, and they both had that problem. Nice cleans, nasty, ugly clipping sounds.

I think the problem with the PR is the low (240V) B+ voltage on the cathodyne which causes the cathodyne to clip before the output tubes reach full power so it's ugly cathodyne clipping you hear not power tube clipping. The cathodyne is powered from the 240V preamp B+ node. I think the solution is to move its 56k plate resistor B+ connection over to the unused 320V B+ node.