Brown Deluxe/Constellation build...help please...

[Structo]

That's why IEC connectors are so good.
If the cord gets tripped over or pulled on, it simple breaks free from the amp.
Usually with no damage.

[dehughes]

I'd have preferred an IEC connector, but this chassis wasn't cut for one and I didn't feel like doing the work to make my own cutout...

I replaced the volume pot on the 12AX7 channel and have re-wired the ground wires for the pots. We'll find out tomorrow if that helped at all.

Another suggestion was to take the 16uf PI electrolytic off the string of electrolytics and run it to the ground point via its own wire. I may do that as well, depending on what I find tomorrow.

[paulster]

Interesting....then many professional amp builders would be in the wrong. I don't really see how it would be a bad thing...feel free to elucidate....

Just ask yourself though, how many professional builders actually know anything about UL, CE or any other certification? HAD, I believe, was aware of this because he included the note that this wasn't a production product but an experimental prototype in his 'contract'. That line doesn't seem to serve anything contractual other than to attempt to limit his liability.

Most of the boutique amps don't have any certification and I'd say that probably most would fail to meet the CE criteria for sale in Europe. A couple of major US boutique brands are distributed in Europe and don't have CE marks, which means their European distributors are selling them illegally. They could add the mark but that assumes they've done self-assessment and that they would actually pass, which is another challenge in itself.

Most builders seem to have started with classic amps, accepted that they need to get rid of the death cap and polarity switch and add a 3-conductor mains lead but don't seem to have progressed far from that point. How many, for instance, know that the ground lead should always be the longest lead? In the UK our plugs are shaped internally such that the hot wire would be pulled out first, followed by the cold and then the ground. The same mentality should apply inside an appliance.

One of my personal 'favourites' is a well-respected boutique builder who uses self-tapping screws into the chassis for tube sockets, right alongside plate wires. And a well-known attenuator that's a Class 1 device has the ground connected between the input IEC and the accessory IEC, and that's as far as it goes. And was just snipped off and earlier versions. Presumably because they got a ground loop hum otherwise. That's being sold in Europe too, at 240V!

[Phil_S]

All is good Phil.

As for people worrying about shocks if a wire came off, this is why you use a fuse. All should know that the black power wire is hot, the white wire is neurtal and the green we call ground. The fact is white is grounded at the grid in all USA electrical power household distributions. The extra added green wire came about as an extra precaution to provide and extra means of grounding the recepticle boxes and lighting boxes. If you opened a 200 amp service panel the white (neutral) and bare wires of your 14/2 romex wire goes to the same buss bar. The black (hot) goes to the breaker.

Building guitar amps has a theory and application all of its own and my electrician brother does understand one thing I am doing inside that little tiny box I call an amp.Mark

Yes, Mark, my British ex-patriot friend who is an engineer refers to things like the green wire as "redundant" and I always get a laugh at his use of the vernacular. I was aware that the green and white go to the same buss bar in the service panel box. I'm not sure that everyone realizes this. I've done a bit of electrical work and was actually surprised the first time I saw this.

[M Fowler]

Phil,

I think paul has hit on something that you are saying, we as builders need to make sure we are paying close attention to our builds and make them safe not only for ourselves but for future owners. From what I have seen I think we are good at doing that and by posting pictures of our builds allows other builders to find mistakes and give feedback keeping us on our toes.

Thanks guys. Hopefully dehughes will have good news about his amp today.


Mark

[dehughes]

Thanks Paulster. Makes sense to me. Perhaps I just assumed that people who are selling their amps online and at storefronts did whatever due diligence was required in the design and building of the amp. Good points for sure.

As for my amp....we'll go see....I just got up so it's breakfast first...

[Structo]

All is good Phil.

As for people worrying about shocks if a wire came off, this is why you use a fuse. All should know that the black power wire is hot, the white wire is neutral and the green we call ground. The fact is white is grounded at the grid in all USA electrical power household distributions. The extra added green wire came about as an extra precaution to provide and extra means of grounding the receptacle boxes and lighting boxes. If you opened a 200 amp service panel the white (neutral) and bare wires of your 14/2 romex wire goes to the same buss bar. The black (hot) goes to the breaker.

Building guitar amps has a theory and application all of its own and my electrician brother does understand one thing I am doing inside that little tiny box I call an amp.Mark

Yes, Mark, my British ex-patriot friend who is an engineer refers to things like the green wire as "redundant" and I always get a laugh at his use of the vernacular. I was aware that the green and white go to the same buss bar in the service panel box. I'm not sure that everyone realizes this. I've done a bit of electrical work and was actually surprised the first time I saw this.

I don't know if it is a USA electrical code or just locally by state but, electrical main boxes on residences as well as commercial main wiring have required a separate buss for grounding which actually has a heavy gauge copper wire that goes outside to a deep ground stake for years.

It used to be that the neutral and ground were combined in the main breaker box and shared a buss.
My house was built in 1975 and it has the neutral and ground combined but also has an earth ground.

Not sure when they changed the code here but it has been years since it went into affect. It may have happened when they started polarizing the two prongs in the recepticles, making the hot smaller than the neutral prong.

I know ultimately that the neutral is earthed at the sub-stations and probably at the transformers in the outside wiring but the have separated the ground and neutral inside the house for years here.

[dehughes]

Well, the hum is still there after replacing the (faulty?) pot. The new pot doesn't jump in volume like the old one did, and the hiss is slightly less now though it wasn't too prominent before.

As well, I took that tone cap off the pot and put it on the buss bar.

Still though, that hum is still there, and the amp sounds more strident and midrangey now after moving that cap. The tonal response of the 12AX7 channel is much more harsh...maybe due to moving that cap?

Still though, I can't shake the suspicion that I need a ground reference before the PI input cap...or that I've done something incorrect in how I've mixed the channels before the PI. I have a voltage divider between the two channels, but this provides no real ground reference. I know the 6G3 schematic doesn't have this leading to the PI input, but then again, it blends both channels BEFORE the last triode, whereas I'm blending the channels right before the PI input.

Thoughts?

[dehughes]

Soooo, I tried the following:

Running a 220k to ground before the PI input cap....still humming.

Reversed the tone pot wiring....still humming (though I do like the tone control better this way).

Re-wired the PI cathode resistors to 1k2 and then a 47k to ground, and replaced the 1k5 to ground off the .1uf cap with buss wire...still humming (though I like this PI setup better...a bit warmer and smoother).

So, I'm nearly at a loss. The only other thing I'm going to try (at this moment, unless someone has a better idea) is to separate the PI electrolytic from the other ones and run it to ground via its own wire, in the hopes of avoiding ground issues...

The other thing that's odd is that even when the volume controls are both "off", I can still hear some guitar signal coming out of the speaker. Maybe a ground issue?

[MBD115]

First I want to say you have a very nice looking build!!!!

The other thing that's odd is that even when the volume controls are both "off", I can still hear some guitar signal coming out of the speaker. Maybe a ground issue?

I am probably of little to no help but I would like to try to anyway.

If both your volume pots are indeed turning down to 0k then the signal should be killed at those points, so no guitar should be present after them. The only other way around those pots looks like down the 220k resistor to ground on the first stage of the 12AX7. The 6sn7 side looks like it would be closed up by the pot. My thinking is if that 220k resistor isn't making it to ground that this might explain 2 of the things you have said. The tonal response of the 12AX7 channel is much more harsh maybe from the extra gain you would have without this resistor making it to ground, and this volume issue. Down the 220k and back up the 15k to the signal path after the second stage. (the drawing)

I'm a noob so look over my stupidity. But I tried!!!

Best of luck finding the fix. You'll have a nice amp when you're done.

[dehughes]

Thanks for your help, man. It's good advice to check all the solder joints again, so I'll do that and make sure that 220k is getting to ground. Perhaps the volume issue has something to do with the two channels run in parallel? I'll have to unplug one of the tubes and see if the other channel still carries on when the volumes are down.

[Structo]

I didn't go back and read every word so forgive me if this has been suggested.

I do know that the typical 12ax7 tubes and their offspring (12at7,12ay7,etc) have hum bucking filaments.

Not sure about the 6sn7.

That being said, did you make sure to keep the heater wires in phase with each other from tube to tube?

The filter caps should be grounded as near to the negative side of them as is possible to ground lugs on chassis floor.

Power transformer center taps should ground to transformer bolt.
Make sure that there is good contact to the bare metal chassis there using a star washer.

All front control pots should ground to a common buss wire that is grounded near the input jack.

The input jack should run straight down to where the buss wire grounds.

I remember reading about a guy that had his input jack grounded to the buss wire vs running a ground wire down to where the buss wired was bolted to the chassis.
He changed it so the sleeve was grounded through a wire to the buss wire ground at the chassis and solved the hum. Doesn't make much sense but he says that made all the difference.

All preamp cathodes should ground to the buss wire.
All shielded signal cable should ground to buss wire.

Output transformer black wire should ground to sleeve of output jack(s).

Hope I'm not being too redundant.....

[M Fowler]

Structo,

Sounds right to me.

Mark

[dehughes]

Thanks Structo....that humbucking filament thing might be the ticket.....hadn't considered that. However, IIRC the amp does hum even with the 6SN7 out of the socket.

I'll go through that list line by line when I get a chance to work on the amp again. Thanks.

[Phil_S]

With all proper respect to Structo, he an I don't agree on a few things. I am not saying the advice is bad advice, and Structo does follow much of the common wisdom I see given elsewhere. So, let's be clear, I'm just going to present a bit of a different viewpoint.

Hum can be an elusive thing to conquer and since you have been tinkering with it for some time unsuccessfully, I am going to suggest that you rework the whole ground scheme. Normally, I wouldn't suggest this, as the act of changing more than one thing at a time means you will not learn where the error exists. I am suspecting you have multiple errors, maybe even ones that are compounding the problem. Here's how I do it, and I know this works because I've got a half dozen amps, all homebrew, that are done this way.

First, no grounds on transformer bolts. While the ground connection made there is OK, the bolts might loosen over time and that will be a problem. Drill a dedicated hole for a #10 screw near the power transformer. Put a star washer (with the pointy things on the inside preferred), then several ground lugs you will later solder to, another star washer and the hex nut. Use NoAlOx if you are bolting to an aluminum chassis, but not needed for steel. NoAlOx prevents the oxidation caused by dissimilar metals comming into contact. Also, while I'm on it, the mains green ground wire gets its own bolt, as close as possible to where the a/c power supply enters the chassis and nothing else on that bolt; same treatment with the star washers and hex nut; better yet, solder the ring lug to the chassis if you can, then drill a hole through the ring lug insert the bolt, and finish with loc-tite.

Good, now to the circuit grounds. Work from lowest ground potential to the higher ground potential. Each stage is one section or tube in the preamp. Gather the grounds for each section and bring them together on the ground buss.

First stage: input jack, V1 cathodes, all the pots, the 220K ground ref resistor for V1B grid, AND the filter cap ground for the 8mf that supplies V1. Draw a vertical line on your schematic after the 100K plate load resistor for V1B, between the following 15K and the .01 coupling cap, so the 01 cap and the 22K divider on the power rail are to the right of the line. That is stage one.

Second stage is the 6SN7. This is everything to the left of the 10K 2W divider resistor, including the 16mf filter cap ground. These go on the ground buss very close the the first stage ground. Theoretically, these are at the same ground potential as V1, as V2 is parallel to V1.

Third stage is the PI 12AX7. This is everything to the right of that vertical line I described in stage one up to the edge of that "row" of the schematic, the last item being the junction of the 220K grid leak resistors. Yes, include the ground for the 16mf filter cap. Your A B and C indications to connect to the power section are the dividing line. The PI is the last thing you put on the ground buss.

Now, from the high potential end of the ground buss (the PI end), run a wire to that ground bolt near the PT. The other end of the ground buss, near the input jack should not be grounded.

Next, you ground directly to the one chassis ground bolt everything else: power tube cathode R/K, 16mf filter cap, output transformer, output jack sleeve, those other two 16mf filter caps shown on the schematic near the standby switch, and anything else in that region I may have failed to mention.

Do this. If you are not otherwise introducing hum from some sort of input or RF source, your amp will be quiet. My work is not as neat and tidy as yours, but my amps don't hum, so, I think I must be doing something right. Your neatness works in your favor here, as you will be able to find everything you need to find, gather it, and ground it where I suggest.

You are actually pretty close on what I say. One big difference is that you have the preamp buss grounded at the input jack end. I see this style of work and there are many who swear by it. I think it runs counter to the theory of starting with low potential and work towards high potential and can inject hum. Also, it may be too inconvenient to make a dedicated bolt and your use of the transformer bolt, while not recommended, will do the job OK. You can even cheat a bit and take advantage of the bit of ground buss between the three power section filter caps and the bolt by soldering to that bit of buss instead of using ring lugs; your choice there and a convenience of your build style. BTW, I can't really see, but assume that short bit of buss for the power section filter caps is only grounded at the bolt and not the other end, as it should be.

This is a link to my 18W build. I think you can see here how I made the grounds. There is a gaff or two shown, and since corrected, but not related to the grounds. This one is quiet.
http://home.comcast.net/~psymonds/TMB18W.htm

You've come this far, why don't you give this a try?