Tube amplifier layout guidelines and tips

[arnoldc]

guys, as i come near completion in having all the parts for my 2A3 project, i would like to seek your assistance for useful information on the guidelines and tips in laying out your components for a tube amp.

my basic knowledge are:

1) PSU should be as far as possible to input stage
2) the importance of the grounding
3) the importance of twisting the filament wires
4) keeping the power supply wires as close to the chassis as possible
5) the orientation of the power xformer in relation to the OPTs
6) keeping the ac, dc, input and output lines separate

I'm pretty sure there are a lot more to consider. What are those? Furthermore, which grounding scheme - bus or star and why?

I made three layouts. Comments are highly appreciated.

#1 - input and output stages left and center and PSU on the right. going from left to right, front to back - 6SL7, 2A2, OPT; 6SL7, 2A3, OPT; 5U4G; filament, choke and power transformaers


#2 - PSU on the center. filament and choke mounted underneath.


#3 - PSU on the center, V-layout. filament and choke mounted underneath.


#4 - PSU on right side, variation of #1


does these layouts make any sense at all?

[johnmarc0]

Arnold,

The first layout is the most logical as it will be easier to align and true during the plate hole drilling and cutting process.

The second layout is a bit complex as more centering lines

The third and fourth layout is almost the same in complexity as the first one.

Remember the more lines you have will mean the more probability the error in cutting will drift ( I make my errors below 1mm, my mill machinist makes around .5mm errors totally acceptable).

Also, your layout is your 1st line of presentation of the amp (you see it before you hear it), it should show character, the questions is what character do you want to project with this amp?
Take your pick whether its a kickass? bad? open arms(like layout # 3)? I could care less? Look at me? am cool your not? am working so buzz off? I do not see anybody? type of layout.

Then that's when the art and complexity exponentially increases.

JM

[arnoldc]

prof. JM thank you very much for your thoughts and reco on layout #1. i'm going to use hammond's 125ese and these OPTs are in no way "pogi" so i'm not after that na. i'd go for cleanliness na lang and aim for 1mV-2mV of hum.


I found this nice article about star ground by Randall Aiken

[arnoldc]

nakakabaliw pala 'to! just designing the grounding system, avoiding ground loops, minimizing hum, isolation, ek ek... buti na lang may internet!

ok, it looks an "H" style ground bus is appropriate for this stereo amp.

[johnmarc0]

Pare that's the fun part your chest will sink kung napatunog mo na yan

[arnoldc]

i'll bet my pwet pare

[yeo]

hi arnold,

i like layout #2 and it's the way i would do my amp. i don't like #1 as the rectifier tube is sandwiched between 2 huge iron. poor heat dissipation.

yeo

[vintage_dog]

I like #2 as well but I would align the 2 driver tubes with the 2A3's like this:


OPT.........PT........OPT

2A3-------5U4-----2A3

_X.........................X_

X= driver tubes

[ambel]

Grounding is very important .
I tried buss ground and it didn't work for me power and pre-amp.
But the one that I use is the power ground and signal is tied together
on power amp and star grounding on pre-amp.
I twist all the wires that carry AC except the HV for the rectifier that's
a tip that was given by Jack Eliano of Electra Print.

[ambel]

Grounding is very important .
I tried buss ground and it didn't work for me power and pre-amp.
But the one that I use is the power ground and signal is tied together
on power amp and star grounding on pre-amp.
I twist all the wires that carry AC except the HV for the rectifier that's
a tip that was given by Jack Eliano of Electra Print.

[johnmarc0]

Arnold,

You better get your grounding sorted if you choose #2 (considering eveything is from the front panel perspective), if your input is let say from the left side and your output is in the right side and power is also in the right side. This will what happen...

1. your left channel input is ok just one short run, our left channel output on the other hand will cross the PT magnetic field because your output ports in the right side, using shielded cables for this run will usually create coloration, your opt should be really close to where your output port is.

2. For your right channel output is ok since it is near the output ports but the input on the other will span from left to right. again it crosses the PT magnetic field and may possibly cross the left channel OPT.

3. If you are star grounding your best star point is the negative terminal of the filter cap. Star grounding is done best if you sink the star points to the cathode of the highest power bearing tube, then you tie the two tube cathode groundpoints together with a high mass conductor or bus but made as short as possible, this layout gives you that advantage to do star grounding effectively as left and right star points can meet in the filter cap.

4. I will not worry about thermal issues, if you have done your amp correctly, there should be no motorboating or standing waves to create thermal runaway (as if the rectifier acted like a capacitor but instead it stores thermal energy) on the rectifier tubes (and your just running below 20 watts on the tubes anyways) so it does not really matter where you place the rectifier tube, the whole unit generates a consistent amount of heat even if the 2A3 goes runaway which is rare for DHT tubes as the wire filaments also acts as fuses (i learned that the hard way). The only instances that thermal runaway occurs is when the rectifier oscillates building up more energy and goes to thermal oscillation or when you have found a way to violate the laws of thermodynamics and conservation of energy.


You will learn new habits in your first amp, I suggest you keep the good and safe design/building habits.Throwing your questions in the forum is a sure good way to understand building techniques and good, habits and practices.

JM

[arnoldc]

francis, yeo, ambel, jm, thank you very much for your support.

i guess this should work:

1. two point star ground
1.a the first star ground point will be the negative terminal of the first filter cap (before the choke) tied with the PT ground. this point will be screwed/soldered to the chassis
1.b the second star ground point will be the second filter cap negative terminal and connects to the first star ground point

2. the input RCA ground will be isolated from the chassis, and will be connected together with the driver component ground point. this will in turn be connected to the second star ground point

3. the output component ground point, will be connected to the second star ground point

4. the output binding posts ground will be isolated

5. the IEC input ground is not connected to first star ground but rather to the chassis in the immediate vicinity

i hope i'm in the right track...

[johnmarc0]

Arnold,

Good to hear about your progress one comment though.

star point 1.a and start point 1.b should be in the same point. Logic behind star grounding is that different ground points may induce a few hundred microvolts of ripple potential that can be amplified as hum.

My verdict on that is 50/50, I will rather ground both points in the filtercap 2 (post choke) just to assure me that in the back of my head it does not come from that point when hum does show up.

Second, your chassis, signal and power ground meets in the chassis? You better be properly grounded on the power line side (check continuity between the two points better if you have a MEGGER Tester) as the cathode voltage bias of the 2A3 may be present in the chassis plate depending on how you did your circuit. This is not advisable if you are doing a loftin white direct coupled 2A3 as it is sensitive to unstable local ground feedback and chassis voltages can be several hundred volts (almost equal to the plate voltage of the driver tube). in such case explore other methods like isolated grounds or floating ground schemes.

Again, always remember, unplanned grounding plan is a tricky back stabbing bastard.

JM

[yeo]

1. two point star ground
1.a the first star ground point will be the negative terminal of the first filter cap (before the choke) tied with the PT ground. this point will be screwed/soldered to the chassis
1.b the second star ground point will be the second filter cap negative terminal and connects to the first star ground point

i like sakuma's grounding system. one power ground and one signal ground. then one fat wire connecting both grounds.

2. the input RCA ground will be isolated from the chassis, and will be connected together with the driver component ground point. this will in turn be connected to the second star ground point

i prefer one connection to chassis, and that is usually thru input rca jack. this is part of signal ground.

5. the IEC input ground is not connected to first star ground but rather to the chassis in the immediate vicinity

yup!

as you can see, there are many ways to do this. this is what i have done and tried and it works well for me but please feel free to try out anything you like and post your experience on this forum.

there's more than one way to skin a cat.

yeo

[johnmarc0]

Yeo,

I would have recommended Sakuma's two point earth scheme, but it may be too much for at this stage, I was thinking he evolve layout and grounding skills so he can comparatively transcend and appreciate what we do now.

Also, two point earth is very good if you use oil caps in the 10uF to 20uF range. I think he is just using 100uF cerafine electrolytics.

John

PS.

BTW, I tested the transformers they are in the 10K to 20K range good for 6SN7 and 6SL7s. They can be both used as single ended or push pull interstage.

Ciao

[arnoldc]

yeo, thanks much

jm, i'm using 5uF pre choke and 200uF post choke

more comments welcome!

[arnoldc]

yeo, thanks much

jm, i'm using 5uF pre choke and 200uF post choke

more comments welcome!

[arnoldc]

so here's what's gonna happen...



primary star gound point is the negative terminal lead of the first capacitor, where the PT center tap is also connected. this ground point is screwed to the chassis, via locking screws.

secondary star ground point is the negative terminal lead of the post choke capacitor, that supplies the B+. the voltage divider (for 80V) and capacitor ground is also connected to this ground point. this ground point is connected to the primary via a wire, not to the chassis.

input section: the input RCA ground is isolated and connected to the 6SL7 cathode resistor and capacitor ground side in a local ground point. the volume pot ground is also connected to this ground point. this ground point is then connected to the secondary star ground point.

output section: the output banana plugs are totally isolated. the ground only goes to the speaker ground. the hum balance resistor and capacitor is tied to the grid resistor in a local ground point, which is then connected to the primary star ground point.

[arnoldc]

This is how it will look like now. The upside-down transformers are to be mounted under the chassis, like most of the components.

[johnmarc0]

Ano choke values mo?