Finished Express but have buzz

[Legin]

I finished my Express this morning and it fired up first time with the following voltages:
B+ of 399V
B+2 379V
B+3 302V
B+4 282V
B+5 266V

Sounds good if VERY LOUD however I have a quite overbearing buzz which increases intensity and loudness with the volume and is affected by the presence control. I currently have the NFB routed between the two EL34s but I am going to move that to the power board side of the first EL34 in case it is something to do with that.
http://kirkby.com.au/g2/main.php?g2_vie ... itemId=570
Has anyone any other ideas and are my voltages in the ballpark?

EDIT: Re-routed the NFB still got the buzz, I have tried a general chopstick and nothing stands out, buzz occurs with or without guitar plugged in, still looking.
Nigel

[tubedogsmith]

pull the first preamp tube with the amp turned on. if the buzz stops it's in one of the first two stages, if not pull the next preamp tube. with the noise greatly affected by the volume control I'd guess it's one of the first two stages.

[Legin]

Thanks for the reply, OK, the buzz reduces substantially with either of the first two tubes pulled although it is still audiable. Still looking.

EDIT: with none of the 12AX7 tubes in the amp is silent so it must be that pre-amp board somewhere.
Nigel

[paulruby]

Buzz is a common problem in amps with SS recto. The large current spikes that charge the filter caps every 8ms are the source of the buzz. Grounding scheme and shielding of the input lines is the key to keeping the noise out of the signal path.

First thing is grounding and keeping the large current spikes isolated from the signal ground and physically away from the rest of the circuit. Taking a glance at your pics shows you have the power supply and first filter caps off to one end of the chassis, which is half the battle. But, the grounding scheme may need to change.

First, the curent loop formed by the tranny secondary, recto diodes, stand-by switch and first filter cap needs to be a tight loop isolated from the chassis. The tranny lines should be short, twisted and go directly to the diodes. Then twisted pair wire from the recto diode to the stand-by switch and back to the top of the first filter cap (or filter cap stack). The CT of the tranny goes directly to the negative terminal of the first filter cap, again a short line. This should all be on the far end of the chassis with short leads to minimize the abiliy of this loop to radiate noise to the rest of the amp.

Next is the grounding scheme. I know the name-brand amp makers use the chassis for signal ground but it can only hurt in a DIY amp. It can be made to work, of course, but it doesn't provide any advantage and increases the potential for ground loops, RF noise and unwanted paths for cap charging current. There should be one and only one connection of the elecrical circuit ground to the chassis and that connection should be made at the input jack. The reason to make it at the input jack is to ensure a non-inductive path for RF noise direct to the chassis coming in on the shield of the guitar cord. If the guitar cord shield does not directly connect to the chassis at the input jack, RF will travel to the input tube instead. The earth ground from the mains socket should be tied to the chassis close to the mains socket.

A single, heavy wire should connect the negative of the first filter cap to the negative of the second filter cap, which is where the power tubes, bias, OT secondary and PI should all be grounded. Then a single wire from the 2nd cap negative to the 3rd cap negative, where the 3rd stage is grounded. And so one, finally arriving at the input jack shield and the chassis. This can be implemented equally well with a ground buss or multi-star scheme.

In order to further reduce the chance of buzz, use shielded line from the input jack to the first tube and from the volume pot to 2nd stage. If you've used the grounding scheme I discussed above, then you can remove the grid resistor on the input stage. That resistor is there to keep out radio but, with the chassis ground connection made at the input jack, radio is shunted away before getting to the input tube. Removing the grid resistor greatly lowers the impedance at the input tube grid and helps keep out the buzz noise that may be radiating from your power supply loop.

I've found the above to be excellent for keeping the buzz and hum out of wreck builds. The wreck is prone to buzz because of solid state recto and a large first filter cap creating very large cap charging current spikes. I'm finishing up a wreck build now and I'll eventually post pics of the power supply and grounding scheme, which has NO buzz or hum.

- Paul

[UR12]

Paul

Welcome to the TW site! Great post also regarding grounding and buzz. I've built two Expresses now and both came out quiet as can be using Ken's original layout. Just a little hiss when the volume is cranked but that can be expected with all of the gain associated with the preamp. The only shielded cable I used was from the input to the first stage. Second stage from the vol control was plain wire run as close as I could get it to the chassis.

Picture here ---> http://maximoaudio.com/forum/viewtopic. ... 83&start=0

Ken used one ground in the corner of the chassis to ground the AC, PT CT, and bias supply 1 " from where the power cord comes in. There were three individual grounds from the filter caps to the same lug in front of the caps along with the bleeder resistor for B+1. Two more grounds at the power tube sockets for pins 1 and 8 and the 100 ohm resistors forming the pseudo CT for the heaters and then the busbar that is only grounded to the chassis through the back of the pots.

The first time I looked at Ken's grounding scheme it had me scratching my head, and it would appear, it differs a lot from the scheme you just described. I've got a couple more builds coming up in the near future and will have to give your scheme a try. Can't wait to see your pics!

Thanks!!

[paulruby]

That's a nice build!

There's no question that using the chassis for ground path can work when it's done right. But, it can accentuate mistakes made by many DIY folks. The key thing is that the charging current in the first loop (tranny, recto, 1st filter cap) needs to be short leads, off to one end of the chassis. That rule is met in your layout, which looks great looking at the pics.

I used to build all my amps with everything on one board but I would ocasionally end up with buzz because the above rule wasn't met well enough. I've had trouble when using the JJ 40x20x20x20 as the only filter cap AND with solid state recto on the same board as everything else. The result is that the tranny CT had to run under the board to the cap and the lines from the recto to the cap were also under the board. This was enough to radiate buzz into the circuits. In those builds, I've had to add series resistance to the recto to soften the charging current (essentially emulating a tube recto). All my builds with solid state recto now have the first filter cap and diodes well off to one end of the amp to be safe.

After that, I still prefer to use an explicit wire from there to the other filter caps or the ground buss. But, there's nothing wrong with using the chassis for that connection if everyhing else is done right.

Oh, and I agree the line from the volume pot to the next tube doesn't have to be shielded as long as the chassis is fully sealed electrically once in the cab. Else, that node can be suseptible to radio pickup when out on the road, playing near transmitters. I shield it to make sure...

- Paul

[Jeph]

In the original 'wrecks the artificial heater CT is grounded at the power tube. Is this the best place to ground a heater CT coming from the transformer as well?

[paulruby]

In the original 'wrecks the artificial heater CT is grounded at the power tube. Is this the best place to ground a heater CT coming from the transformer as well?

Doesn't really matter for that. It can go anywhere convenient.

[Legin]

Thank you for a most comprehensive reply Paul, I will read through a few times and check what I have done. I believe I have covered most of your points however there is bound to be something I have missed.
Nigel

[Legin]

Does anyone know if the third leg of the bias pot should be grounded? it appears so from the schematic but it is difficult to know from the pics if it goes underboard.
Nigel

[paulruby]

Does anyone know if the third leg of the bias pot should be grounded? it appears so from the schematic but it is difficult to know from the pics if it goes underboard.
Nigel

Which schematic and which pic? One end of the bias pot is almost certainly grounded.

[Legin]

I'm using the Kelly 90 A1b schematic which shows the bias pot grounded but looking at the gut shots of Francesca it is not obvious.
Nigel

[UR12]

Does anyone know if the third leg of the bias pot should be grounded? it appears so from the schematic but it is difficult to know from the pics if it goes underboard.
Nigel

Hi Nigel

I have attached three possible ways to wire up the pot. All of them will work. The bottom one is what I use. Sorry bout the rough hand drawing.
Hope this helps!

[Legin]

Thanks for that Dana, so it seems grounding the third lug of the bias pot is optional

[UR12]

Thanks for that Dana, so it seems grounding the third lug of the bias pot is optional

Yes. The nice thing about this circuit is that if the pot fails you won't burn up your tubes. I don't like leaving things hanging loose so I always use the bottom one for this reason. Just me being anal I guess.