Ground bus, it's all the rage

[centervolume]

Greetings all

Wrapping up a vibrolux reverb chassis build for an old abandoned early 70s cab I ended up with. This is the 6th amp I've built and I'm happy with the process overall. I switched from the brass plate to the grounding buss a couple amps back mostly because its easier and builders I respect have recommended it, most of whom are on this forum.

However, and here's the hilarious part, the amps I've built with the bus are exhibiting more noise (mostly 60Hz hum) than the brass plate ones. So I'm doing something wrong. I have been consulting the resources here and elsewhere and I'm not certain the switch from the plate has anything to with these results.

The 1-page diagram on this site is great for mapping a concept of how to address grounding, and I have followed that and gone back through to double check but I still have a hum.

Even though I'm not sure the ground bus is the issue, I do have a question about that which is as follows: in the bus scheme, if the pots and input jacks are being connected by a conductor that runs along the back of the pots etc., that connection would then be in addition to their natural connection with the chassis at their mounting points and then present a loop possibility, right? So are  builders that successfully use a bus also using insulated washers to prevent pots and jacks connected by the bus from contacting the chassis where they normally do (i.e. around the spot where they are mounted and protrude through the control panel)?"

Any help greatly appreciated,

Hounded by the Ground

[shooter]

the 3 tabs we solder to have no connection to the pot chassis.  so when I solder a ground to a tab to a copper wire touching chassis at 1 point only there isn't a loop

[tubeswell]

See Merlin Blencowe's chapter on grounding (downloadable from the Valve Wizard site) for the best way to implement a ground buss. IMHO


You have to pay attention to the way you cluster the ground returns according to the PS node that they source their HT supply from. If you mix it up, you're more likely to get 'low-current' grounds being jimmied around by 'high current' ground returns.

[dude]

I just ran a buss on the back of the pots with no insulation for each pot to chassis, no hum, no issues. The buss was grounded at the input jack along with the pre-amps cathode and "nowhere else". If you look at Doug's 6v6 Plexi layout here. He does the same thing, buss soldered to pots, no insulation at pots, grounds for the pots grounded at the buss and only one end is grounded to the input jack. The power supply caps, PT grounds are at the other end. Basically, two grounds one for power supply, the star near the PT, the buss at the input jack.

[d95err]

Having the ground buss connected tp the chassis at several places (e.g. via the back of pots or via non-insulated jacks) would defeat the purpose of buss grounding.

[Willabe]

....I do have a question about that which is as follows: in the bus scheme, if the pots and input jacks are being connected by a conductor that runs along the back of the pots etc., that connection would then be in addition to their natural connection with the chassis at their mounting points and then present a loop possibility, right? So are  builders that successfully use a bus also using insulated washers to prevent pots and jacks connected by the bus from contacting the chassis where they normally do (i.e. around the spot where they are mounted and protrude through the control panel)?"

Yes it is a ground loop, but it's a small/short loop. And no most builders are not using insulated washers on their pots.

I run a separate ground wire for any pot ground back to it's B+ filter cap ground star.

It seems that separating the power amp grounds from the preamp grounds is where you get the most bang for the buck in time/work. 

Link for Merlin's web site, grounding;

http://www.valvewizard.co.uk/Grounding.html

[centervolume]

Thanks all - I will review this input and go back under the hood to compare what I have going on point by point and then report back

[centervolume]

So I'm noticing that my hum is independent of volume controls (what I am taking as a 60Hz hum is present upon power up as opposed to appearing / disappearing with volume adjust., reverb adjust, trem adjust) 

This prompted me to inspect the heater wiring - the power transformer does not have a heater center tap, so I used 2 x 100 ohm resistors (actual measuring closer to 110 ohms) from pilot lamp to ground. I noticed that I initially had chosen the eyelet on the bias board that goes to ground as the grounding location for that pair of resistors so I moved them to the PT lug I have PT CT and main board ground and main filters tied to. This shift did not affect the hum.

I do have my power tube cathodes tied to a separate (closer) lug of the PT, ( I also have the power cord ground tied to a 3rd lug of the PT based on proximity) tying the lugs together doesn't seem to make an impact.

Will continue to look... suspecting heater or filtering grounding due to independence from volume controls. 

 

[sluckey]

How do you know the hum is 60Hz rather than 120Hz?

[centervolume]

that's a good question - let me see about an app on my phone to detect. Do you recommend any such download app for i-phone ?

[sluckey]

I recommend a scope or good dmm.

[centervolume]

ok, I do have a scope at the office, may have to run up there and grab it

[sluckey]

  Just connect a 1X probe to the speaker and you'll quickly know for sho.

[centervolume]

Ok I have the hum pulled up on the scope (an old leader LBO-507A). The waveform is complex showing a larger low frequency wave with a smaller one bisecting it at the midpoint. Without readily seeing how to deduce Hz I’d be inclined to say that 60 and 120 may both be present. My tuner is showing the frequency as between Bb and B which is supports this (half way between these notes is 30, 60, 120 Hz each octave up)

[sluckey]

Post a pic of that display

[centervolume]

here you go, this is taken across the speaker terminals

adding another shot of AC taken from the ground side of the heater improvised center tap and another ground location on the chassis

[centervolume]

Here is my implementation :

red grounded to PT lug(s): AC cord, power stage filter caps, improvised heater center tap, bias board, power tube cathodes, speaker grounds

green grounded to buss (except reverb/vibrato phonos): preamp cathodes (V 1-4), preamp power supply caps, pots, input jacks

notes: I have to wonder about second stage of V5 which is referenced to a ground at power trans but has grid passing through pots on buss system.

question: is it recommended to tie grounds of input jacks together then attach to buss for both channels at or near the spot where buss passes near those inputs?

[tubeswell]

Here is my implementation :

red grounded to PT lug(s): AC cord, power stage filter caps, improvised heater center tap, bias board, power tube cathodes, speaker grounds

green grounded to buss (except reverb/vibrato phonos): preamp cathodes (V 1-4), preamp power supply caps, pots, input jacks

notes: I have to wonder about second stage of V5 which is referenced to a ground at power trans but has grid passing through pots on buss system.

question: is it recommended to tie grounds of input jacks together then attach to buss for both channels at or near the spot where buss passes near those inputs?

Please see the diagrams from Merlin Blencowe's grounding article above


The rest of the article is on this link:  Click here to view the article (360kb PDF)

[PRR]

This is the current in your rectifier return. Where is that routed and how is it wired?

[centervolume]

PRR

The center tap and bias board ground to the same spot, so that should be the rectifier return right?

That pic was taken of the scope set to AC with probe on ground side of the 2 x 100 Ohm resistors off lamp heater and the probe grounded to a separate ground location on the chassis.

Those resistors ground to a common ground location for PT center tap, 6L6 cathodes, filter bank (minus preamp filter), speaker ground, bias board ground and the V5 cathode 2nd stage, and V6 cathode (phase inverter).

[sluckey]

I'll bet those waveforms were not rock steady, rather they were drifting across the display and your fast camera lense just made them appear steady. I say that because you have your trigger source set to external and there's nothing plugged into the ext. trigger input jack. Push the blue button until it pops out. This is the internal trigger source mode. Now you can pull the trigger level knob out to select normal trigger mode and turn the knob until the display is rock steady on the display. This means that the timebase is now synced to the vertical input signal.

Also, looks like the variable time control (red knob is NOT in the Cal'd position. turn it max cW and you should hear/feel a switch click. Now you can accurately measure the time of one cycle. Should be 16.7mS if your timebase is calibrated.

[centervolume]

ok I have made those adjustments to the scope, but I wont be able to return to working on the amp until the morning. Thanks again to all. I'll get back to reading valve wizard's text again

[jjasilli]

Learning to use a 'scope is a plus.  Meanwhile, to simplify:  120 Hz is about a B note = 123.47 Hz, on your 5th (A) string tuned to pitch.  60 Hz is deeper than the 6th (Low E) string = about 82 Hz.  60 vs 120 Hz can be distinguished by ear using this method.



There are phone apps for tone generator, tuning fork simulation, guitar tuning, etc.  Some are free. 

[centervolume]

sluckey

My calibration is prob a bit off, but in any case I'm getting 15.6 microseconds

I used a signal generator approach (thanks jjasilli) to verify what I was seeing on the screen with the complex wave on the scope (i.e.: both 60 and 120 are present in the hum).

[sluckey]

My calibration is prob a bit off, but in any case I'm getting 15.6 microseconds

I bet that's really 15.6 milliseconds. 

[PRR]

> I bet that's really 15.6 milliseconds.

I'll take $2 on 16.6 milliseconds.

WHY are we calibrating? Hold the probe near a power cord so you see a sine.

[centervolume]

I watched a calibration video and quickly moved on to other approaches. Is getting the exact figure critical or does the fact that we are in the neighborhood of 16 microseconds enough?

I started looking for where the noise starts in the signal flow and I'm seeing the B+ return waveform at the plate of the phase inverter (pin 1). That seems to be where the noise starts.

Looking at the filter cap that feeds the PI (B+3 if im not mistaken), it is grounded along with the filter caps feeding the OT center tap and 6L6 grid 2/ pin 4 ("screen grid"). On later silverface vibrolux reverbs I've seen that cap's ground coupled with the "last" filter cap feeding plates of V1, V2.

As a guess, I'm thinking the cap feeding the PI needs to be grounded elsewhere, I'll move it from the "power amp" local grounding to the other "preamp" location. For some reason I've considered the splitter as part of the power tube circuitry. 

[shooter]

in the neighborhood of 16 microseconds enough?

frequency = 1/period time
16milli-seconds gets you real close to 60hz
16 micro-seconds gets you....?

[centervolume]

ok I have a breakthrough, besides this prob being the first time I've used a scope for a repair, I got the hum to what I consider a more respectable level. I needed to NOT consider PI as part of power amp, but as part of preamp and get that B+3 cap out of the main power filter section (1st 3 filter caps) and tied into the preamp local circuit and ground to buss along with the rest of the main board. I appear to have lost ch. 2 in the process, doh!

This is likely a no brainer for many/most of you guys... but an important step for my understanding of these cool old appliances and how they work.

Will re-dress accordingly and report back with a hopefully final note.

[centervolume]

thanks shooter, honest mistake for a first-time scope experience. thankfully I got enough of it to get the job done

peace

pl

[shooter]

aside;
once I completed the NEET program in the Navy, they put me to work as a lab tech til my next class.  when the "cocky" students got to bad I'd mess up their scope knobs 

(ya, I was a cocky student :)