Questions on Amp Building

[Cygnus X1]

I've been down a few roads and am always questioning and hopefully learning.

Common problem I used to have but I think I have finally conquered is grounding and 60hz hum.
I scratch build my own brews with zero grounding to chassis using a star grounding scheme I learned about here:

http://www.aikenamps.com/StarGround.html

After dealing with repairing and modding vintage Fender, Marshall, Gibson and Ampeg amps it was indeed difficult to really grasp (or accept) much of what Randall Aiken was trying to teach here.
So I took the plunge and tried it.
It works!
Problem is...you have to design it in.
Can't retrofit anything there.

But barring that, I lift the filament supply from the circuit by installing an artificial center tap to it.
Usually works, but lead dress is most critical.

Part 2: Someone asked about individual biasing to match power tubes in an amp,
self built.
I have pondered that as well.
And tried it.
Not all that great, and I haven't found a good reference why and would like a little help.
I have a theory-the bias point is just that. A base point for the power tubes to work from.
But testing tubes with a good mutual conductance tester tells a larger story.
Even when the set bias bpoint is matched, it will not necessarily tell me how that tube (or set) will operate under load.

I hate analogies, but here is a good one: A Chevy 350 at 300 hp will act completely different from a BMW engine at 300 hp. The curves of operation will be different.

So taking two tubes with different mutual
conductance behaviors will "act" different even under the same load and bias points.
(or is that inductance. Anyways, it's a Hickok 6000A tester. Thank you fett!)

That's my theory....now essplain it to me.

:D

[pel666]

My learnings over the years, is be extremely tidy and thoughtfull in the assembly. The headphone trick is where it starts with to align the trannies. I use a buzzbar on the signal section, taking into account the ground flows. A well drawn and designed schematic shows this. Seperate powersection. 60Hz or Hum is generated by the AC section. I also use a star ground connecting the buzz bar to it and the rest. The outlet has it's own seperate chassis ground. This is something I also picked up from Aiken, and for me it worked great.

Although seperate bias control is pretty cool, matched tubes are well in tolerance to have single biasing. You are right, even the Tubes are different and it's all about tolerance. Both tubes will have a slightly different operating curve anyway, the tolerance is only an indication of how much the deviation is. I was fortunate, the first build sounded great but was a bit noisy. The ones after that have been extremely quiet, even better than I could have imagined. :thumbsup

[Cygnus X1]

My learnings over the years, is be extremely tidy and thoughtfull in the assembly. The headphone trick is where it starts with to align the trannies. I use a buzzbar on the signal section, taking into account the ground flows. A well drawn and designed schematic shows this. Seperate powersection. 60Hz or Hum is generated by the AC section. I also use a star ground connecting the buzz bar to it and the rest. The outlet has it's own seperate chassis ground.

although seperate bias control is pretty cool, matched tubes are well in tolerance to have single biasing. I was fortunate, the first build sounded great but was a bit noisy. The ones after that have been extremely quiet, even better than I could have imagined. :thumbsup

I'm assuming you mean a buss bar?
Yes, I do that with my Fender builds (OK, I'm a fan of the 5C3...5E3 circuits, that's what I have built).
And have failed disastrously trying to star ground Leo's designs.
He knew what he was doing, and I don't always get it!

Lead dress drives me nuts sometimes.
I'm kind of like "Dale Earnheardt...who could "see the wind".."
When I design my own I can visualize the potential interference and noise issues by knowing what is signal and what is power.
Mine aren't pretty, they don't look like the uber high dollar bootique amps.
But BY Gawd they are quiet and deliver the signal as cleanly as possible (I hope).

60 hz hum has never been an issue in my own.

The biasing...yeah, I guess I want a techy explanation.
But I feel what I know...is probably correct.
It would be nice if it worked, but best to go with matched tubes.
Not much more expensive than buying two singles anyways.

[fett]

I have faith in you, Cyg.:thumbsup

[Cygnus X1]

Appreciate it fett.

Sometimes it takes more than faith.
It takes actual "book learning".

:jump

[pel666]

Yup, we are on the same page! I am a step by step builder, I kinda look at it like a chemical plant with flows (must be my Petroleum background ;) ). For the tubes, I think it's not that hard. If a Tube has a different 'idle current', it will have an impact on actual operating voltage opposed to the other. This again has a direct impact on the behaviour and output of the tube. Taking the car example, have two engines, one that powers the left and one that powers the right wheels. Both the same engines, but one operates at an exponential increasing RPM and drive a straight line. At 10mph it's not that bad, pretty good actually...at 40 you get a sensation...and at 100 it's outta control. I had a customer once and his tubes were so out of whack it just sounded...well sharp/out of tune is not really the word but it was interesting. Did a swap, rebias and voila back in business...

If you're interested, I had to learn the actual tube operating theory by reading 50's theory books. Forgot a lot, but what sticked are the tube charts and finding the best operating voltages etc. They are pretty heavy to read, and pretty big too, but they go down to tube component explanations and theoretical backgrounds. I am pretty sure it should have in there what you are looking for!

[Cygnus X1]

Thank you pel.
I'm a fan of Jack Darr's book.
And the RCA tube bibles!

:thumbsup

[RacerR]

Hey Cyg - I'm guessing you've seen this site, but I've found lots of great info there, as well as a pretty helpful community of amp builders:

http://www.ax84.com/

If you haven't seen it, there might be some good stuff for ya.

[pel666]

That's actually the site that started it for me...I build the P1 and the Hi-Octane. Wanted to do the November, but instead found out about Ceriatone Kits (they also did AX84 kits in the past) and did a 18W version :thumbsup

[RacerR]

Right - I have one of the old kits (bought it from Chris Hurley about 8 years ago... a low wattage Moonlight with the high gain option).

[KramerBassFan]

My older brother built a ax84 P1. He's now almost an EE, and my dad is by profession.

Interesting... wonder where i got the inspiration to build a Superbass? ;)

[Cygnus X1]

My not-so-famous (yet) Cygnus-7 amp
is based on the AX84 SEL.

In fact, I used to call it SEL, but took the warnings and disclaimers to heart.

I made enough changes to make it my design though.

:D

[KramerBassFan]

http://i892.photobucket.com/albums/ac123/KramerBassFan/db.png

Helocene suggested this in my other thread.... thoughts?

[Cygnus X1]

It's worth a shot Ben.
But it won't cure all mismatches.

Read back a little about the "curves".
Pel spelled it out quite well.

:toast

[Baretta57]

I do understand the purpose of the schematic and demonstrating bias controls, but my first thought looking at the schematic is that you can go WAAAAAY higher with the primary filter caps. This is a SS rectified circuit and doesn't have the restriction that tube rectified circuits have. Why wouldn't you want to go much higher than the 50uf primary caps? Also there are specific RC filter combinations that virtually eliminate 60Hz, don't recall the calculation but I believe this is in Torres book. Great thread by the way:headbang I have found that layout is primary as well as star grounding and like the neat and tidy approach. It is always easier to troubleshoot.

[helocene]

Silicon rectifiers can handle a larger ripple current than tube rectifiers(Ripple is the capacitor drawing a large pulse of current from the rectifier and transformer in its charging cycle. Larger capacitor, larger ripple.) but I still use the following formula to calculate cap size:

10% ripple. The HT is 345V, so 10% of this is:
(345 / 100) * 10 = 34.5Vp-p.

C = (t * I) / V
Where:
I = average load current drawn
V = ripple voltage peak-to-peak
t = duration between charging cycles

So for the example above the reservoir capacitor required will be:
C = (0.008 * 0.188) / 34.5
= 44uF.
The nearest standard is 47uF which gives just over a 10% ripple. Very large values require a PT that can handle a lot more current than you need otherwise you run the risk of overheating your transformer.

//me puts 2 pennies on the table.

[amerturk]

I have a little amp knowledge, less than most and a little more than some...I leave the big jobs for my amp builder in Germany. I have learned that when working with amps, it takes knowledge of electronics and priniciples. This isn't something you can learn online or in a book. It takes years of training.

I do wish I had the skill to make my own, but I would just probably kill myself in the process.

Hat's off to those that can build and tinker.

[paul525715]

A little trick I learnt on my last build to remove a tiny amount of hum...

1) Separate IEC outlet ground (as mentioned by Mies/Pel)

2) Star ground all your other earth wires to one point, but isolate this point from the chassis using a rubber grommet

3) Connect your grounding system to the chassis at the opposite corner of the chassis (usually near to the input jack).

Silence is golden :thumbsup

[pel666]

Am a bit puzzled by that Paul...highest ground current come from the power section, and the ultimate endpoint is actually at the AC power connector. The way you show insulated grommet connection (the point where basically all ground currents meet and has the highest 'potential') is fed back to the input :scratchhe I might see it wrong, but it is exactly opposite to the 'Aiken' method. Personally I seperate the AC section ground from the signal ground section, coming together at that starground since from there the ground current flows back thru the AC ground to the net. That's why the star ground is next to the mains...but not directly the exact same point but the short path ensures ground current doesn't go 'upstream'.

[Cygnus X1]

Nor much to add to that pel.
I keep power grounds as far as possible away from the input ground.

[paul525715]

That's what I thought too, but an amp tech did exactly what's on my diagram and it cured what little hum I had. Oh well...

[Cygnus X1]

That's what I thought too, but an amp tech did exactly what's on my diagram and it cured what little hum I had. Oh well...

That's cool.
Could be any number of things that was causing the hum.

Most of what I was referring to is theoretical and needs to be built in, not modded on to anything that already is built.

[helocene]

I've been doing mine pretty much like Paul described(at least how I interpreted his description) Never had a problem with noise or hum, dead silent.

I have a single buss bar across the entire chain, the ground points on it follow the schematic. It's chassis grounded without a grommet right by the input jack.
IEC connector has it's own ground.
Here's a shot of the guts, I had it apart to experiment with a push pull setup instead of single ended and some cap changes to mica(Thanks Cygnus :thumbsup). Ignore the mess on the right, transformer leads will be trimmed properly once I settle on SE or PP.

http://i599.photobucket.com/albums/tt80/helocene/PlexFleaguts.jpg

The red wire in the middle of the turret board should be underneath. Don't you hate it when you are buttoning everything up and realize you forgot to solder on the lead for the cathode follower plate. :banghead

[Cygnus X1]

Bumping this up for a tip o' the hat to fett.

And the tech knowledge ought to be up front.

fett and I have traded labor for stuff back and forth for a while now-if it wasn't for him I would not have gotten back into working on amps.
He was incredibly generous and sent a Hickok 6000A mutual conductance tube tester for...I guess my work but he went over the top.
Any builder that does not have a tube tester ought to snatch one up!

This one is a dynamic, tests the basics plus gas leakage, tube life expectancy...really everything except full power matching.
I just bought a load of tubes and it is a great tool to have!

Thank you again fett.
:toast

[KramerBassFan]

Star grounding sure would of helped me when i grabbed the speaker cable of my Ampeg V4 on saturday... not to mention, i had two of my 100w'ers go out on that gig alone... but the 16 gauge extension cord was the culprit there :whistle

[Cygnus X1]

Star grounding sure would of helped me when i grabbed the speaker cable of my Ampeg V4 on saturday... not to mention, i had two of my 100w'ers go out on that gig alone... but the 16 gauge extension cord was the culprit there :whistle

Whuh?

Your speaker outputs ought to always be isolated from the chassis.

'Least that's the way I always do it.

What happened?

[KramerBassFan]

Its a 1970 Ampeg V4, so not star grounded, and doesn't even have a ground plug on it. :whistle

I grabbed it, and you know those shocker pens? The joke ones? Imagine holding maybe, five of them. :D
My hand was tingly for 15 minutes.

Not terrible, or painful, but i'd rather not do it again. :thumbsup