Pedal Mule 30 - Prototype build

[BLT]

I'm just wondering if that ground lug next to the last 22uf filter cap is too close to your new ground connection. Also wondering what the large resistor under your filter board is for.

[DocJames]

Ahh the star grounding from the filter board did the trick then, nicely done. I may have to give this build a try. I haven't built an amp in over a year now, too busy renovating my house. This might push me to make one .

[RJ Guitars]

I did a couple things as you guys have noticed... i could tell the hum was in th PS in part because it wasnt gain dependent. So i wanted to make the power supply more like the Rocket amp since they are typically quiet. FWIW - This uses a split rail configuration and i havent worked with one before but so far im not a fan. Seems like you lose the benefit of the filtering in the earlier stages when you go directly into V1 in the way this does.

I an using a Fender choke (only because i have a bunch of them) that is only about 80 ohms so i added the 350 ohm power resistor in series to make it more similar to the Vox spec choke like they use in the Rocket. That alone killed most of the audible hum... its sorta what i expected but maybe somebody can tell me why it works that way... in mind its intuitively a better filter?

I also cleaned up the grounding as some of you noticed. This might have been audible but i had to turn off everything in the shop to hear any noise from the amp at all after i applied the previous fix so the noise floor was pretty low.

Once i get it into a dead quiet room and compare ut to my Rocket i'll decide if i want to reconfigure the rail in the traditional fashion.

[didit]

I an using a Fender choke (only because i have a bunch of them) that is only about 80 ohms so i added the 350 ohm power resistor in series to make it more similar to the Vox spec choke like they use in the Rocket. That alone killed most of the audible hum... its sorta what i expected but maybe somebody can tell me why it works that way... in mind its intuitively a better filter?

So you've got a 4H & ~80 ohm. Stock Rocket 15H & ~400 ohm will provide better filtering. With the series resistor you've got something of a compromise. One could dive into the math but likely best just spending an hour or two modeling in Duncan Amp power supply designer to see how this works to scrub DC. At least for everything except the power stage, which has decent inherent cancelation due to push-pull.

Best .. Ian

[RJ Guitars]

Thanks Ian, that jives with my intuition -- not much resistance + not much inductance = not much of a PS filter.

Now this throws a bit of a discredit to something i took from one of Gerald Webers tech books that I'll paraphrase as "a choke is a choke"...

It was easier to see if my intuition was right by using the power resistor to get the ohms i needed but ultimately it seems like i need to swap out my choke for correct value.

Other thoughts?

[Zippy]

Any other thoughts??? Sure.

If you are trying to dial in voltage drop, subbing the resistor is fine. If you are looking for filtering action, you need the correct reactance. Reactance is, simply put, the resistance as a function of frequency.

X = 2 x pi x f x L

If your choke's inductance is very low, the frequency dependent resistance will be proportionally low.

Bottom line: If filtering is the question, adding a series resistor is not the answer.

The good news: Your builds seem to sound great despite all the equations.

[ToneMerc]

Thanks Ian, that jives with my intuition -- not much resistance + not much inductance = not much of a PS filter.

Now this throws a bit of a discredit to something i took from one of Gerald Webers tech books that I'll paraphrase as "a choke is a choke"...

It was easier to see if my intuition was right by using the power resistor to get the ohms i needed but ultimately it seems like i need to swap out my choke for correct value.

Other thoughts?

Looks good RJ, I'm using an AC30 choke in mine. I will get caught up with you soon I hope.

TM

[didit]

Other thoughts?

Big chokes help a lot if you are looking for clean stable power. You should hear some difference between your current compromise and Fischer's design spec. Slightly tighter low and low-mid dynamics. Also improved 120Hz filtering though you might need a 'scope to detect the difference in that. Assuming you have a two more Fender spec, then you could temporarily chain up three in series and disconnect that power resistor. You'll be close, perhaps a bit above stock 15H as inductance varies inversely with current.

Best .. Ian

[RJ Guitars]

I pulled out the Fender choke... hope to get the bigger one in there later today.

Thanks for the help in understanding a bit more about chokes...

I'm still wondering if there's anything to my intuitive thought that the further down the rail you go the cleaner the power gets? I am trying to understand the logic of the split rail rather than just copy what Komet presumably copied from Trinwreck? In the schematic I've drawn the preamp nodes both get the same starting voltage. Makes me wonder if there is an expected sag on the PI that they don't want V1 to see?

[didit]

I'm still wondering if there's anything to my intuitive thought that the further down the rail you go the cleaner the power gets? I am trying to understand the logic of the split rail rather than just copy what Komet presumably copied from Trinwreck? In the schematic I've drawn the preamp nodes both get the same starting voltage. Makes me wonder if there is an expected sag on the PI that they don't want V1 to see?

Roughly speaking, yes -- power will be more stable further down along a series dropping chain. The split parallel vs series provides two decoupled supplies for PI & preamp with both at nice high voltages. Since the PI is drawing more current the voltage will actually be lower on that node. These are separated/decoupled to prevent supply-based local feedback from messing the signal across stages.

[vibratoking]

I pulled out the Fender choke... hope to get the bigger one in there later today.

Thanks for the help in understanding a bit more about chokes...

I'm still wondering if there's anything to my intuitive thought that the further down the rail you go the cleaner the power gets? I am trying to understand the logic of the split rail rather than just copy what Komet presumably copied from Trinwreck? In the schematic I've drawn the preamp nodes both get the same starting voltage. Makes me wonder if there is an expected sag on the PI that they don't want V1 to see?

Each successive RC stage adds more filtering and results in less ripple and a cleaner supply. The series choke adds another reactive element that 'chokes' off AC. If done correctly, each reactive element increases the ripple rejection and gives a cleaner supply. I think the split is done to supply a higher voltage to the preamp while adding a ground node for return currents that is separate from the PI or any other node for that matter. So, I don't think the split it is done to increase the filtering.

[RJ Guitars]

I swapped out the choke and have been playing through this amp quite a bit. I've found that the stupid hum sorta comes and goes... and seems to vary in degree.

I got it on the bench and pulled V1... it was still humming. Finally, progress. I swapped out the power tubes, still hummed. Swapped in a few different 6SL7's for the PI... some hummed more than others, but they all hummed. put my hand near the PI tubE and it got mucho worse... i have created a human detector.

Redressed leads and rerouted wires... swapped the heater wires on the PI... still hums, maybe not quite as bad. Replaced the 22uF PS filter cap going to the PI B+ with a 33uF... maybe better but still a little hum plus adds a harmony part when my hand goes near the PI tube.

I have a few more tricks but in the meantime I've put in a 6SN7 and i get a slight hiss and otherwise sweet silence...

Edit - Sadly, all the character and the gain went away with the hum. I actually didn't think the PI had that much gain but the 6SN7 kills it??

I sorted through the deep stash of old tubes and found a couple real nice JAN 6SL7's and picked the quietest one of the lot and now we are rockin' again. If this amp was an Express I'd be delighted to have this little amp noise but I was really thinking this thing would be the quietest amp I've built - it kicked my butt to get it to where it's pretty darn quiet and maybe equivalent but not better than my Rocket?? On any of my future projects with 6SL7's, I may try DC power to the heaters.

[stephenl]

A cap multiplier does a great job of reducing hum and ripple but still allows the power supply to sag. The London Power Standard that I built uses one and the amp is almost dead quiet. It's a pretty simple circuit to try. Merlin's power supply book has some good examples. Adding the current limit to it also gives you slower ramp up of B+

Maybe you've already tried, but a humdinger on the heaters might help.

Bruce Clement of BC Audio uses 6Sl7's exclusively in his amps and he uses a real basic DC heater supply. I had one of his No7 amps and it was very quiet.

Just a few ideas

[xtian]

A cap multiplier

A what now? I have Merlin's book, 2nd ed, if you want to call out a page. Thx.

[Meat&Beer]

Swapped in a few different 6SL7's for the PI... some hummed more than others, but they all hummed. put my hand near the PI tubE and it got mucho worse... i have created a human detector.

This may seem elementary, and I haven't read the entire thread here so you may have even already tried this... What about trying to shield the PI tube? A grounded piece of tin foil, or a square from an old screen window wrapped into a cylinder...

Sometimes when I get crosstalk or a hum/buzz that's getting the best of me, (which may be due to proximity) I rig up a 1"x3" piece of brass plate covered with gaffers tape with a clip lead to ground and move it around the inside of the chassis. This with the chopstick test can sometimes lead to the discovery of who's talking to who, and you can move forward to end their communication. Or a bigger plate and try it between tubes, tube/transformers, etc. Sometimes the results are surprising.

Just some thoughts!

Edit:

A cap multiplier

A what now? I have Merlin's book, 2nd ed, if you want to call out a page. Thx.

This is a neat idea, new to me too! Xtian, he's referring to the power supply book, just in case you didn't know. (I thought there was only one [discontinued] version of that, could be wrong.)

http://en.wikipedia.org/wiki/Capacitance_multiplier