5e3 clone grounding question

[dobsont]

This thread goes well beyond grounding, contains new questions, documents my build. Expect frequent updates.

---------------------------------

Built as much to stock Fender spec as possible. Little/no component value deviation. New caps, NOS cc resistors, etc.

Built from the fender 5e3 schemo and layout at schematicheaven.com
Added modern power switch/standby and eliminated ground switch ala Mission 5e3a
6.3v heaters wired using two 2ndrys and a CT.
Used the Paul Ruby startup guide. All good.
I’ve got power, sound, light and no blown fuses.
She’s gonna be loud.

Wasn’t sure on the grounding, didn’t have a brass plate so I found another layout/wiring diagram – can’t remember where from at this late hour. Will look and post in the AM. Basically it was a star ground setup.

As it is it hums, bad.


This is my first amp built without a nannyguide, so bear with me.

Shields of all inputs are tied together with 22g and then wired to both the chassis (at the 12ay7 socket which was nearby – they wanted me to use a standoff bolt in more or less the same location) and to the 3rd filter cap. Seems strange to me, but this is according that layout mentioned above.

Pot grounds tied together with 22g then grounded at the 3rd filter again, as per same layout.

Mains green ground at PT bolt, as per layout.

HV center tap at 5y3 socket.

Output jack shield tied to OT ‘common’ and back to same filter cap.

I used 2 lead shielded cable for the inputs into the 68k resistors.
Red lead is ‘hot’ in, black is ‘normal’ in, repeated on the bright channel. Cable shield grounded at input jack.

That 3rd filter cap isn’t grounded to anything except back through the various ties between the input jacks to the 12ay7 socket. This doesn’t seem right to me, but what do I know!? Why they wanted me to use the 3rd filter and not the 1st I don’t know.

Tell me where to put stuff if I’ve got it wrong. Keep it simple, I’m kind of stupid.

My best guess would be to pull the ground at the 12ay7, leave everything at that 3rd cap, then tie that to the chassis and to the HV CT.

Big mess? yes. Bad Guess? Maybe
Yes, I am done for the night.

[Paul]

This is the layout for my 5E3. It is quite a lovely little amp, and sees more playing time than all my other amps combined:

http://www.trinityamps.com/ForumGallery/trinity/Doc_Files/TrinityDeluxeLay.jpg

[nonreverb]

Here's a couple of things to try...Does it hum with nothing plugged in? Can the hum be controlled by the volume?

[dobsont]

Paul - That's BEAUTIFUL! I'll be studying it. Very clean. Thank you.

nonreverb - not sure, but I'll check tonight.

As a followup to my first post, I pulled the star ground layout from dockeryamps.com:
http://www.dockeryamps.com/web/Docs/5e3-tweed-deluxe-layout.pdf

You can see everything grounded at the 3rd filter cap, etc.

[bcmatt]

Hey, I hope this helps me too. I wired a 5e3 clone for a friend but it still hums too much.
I started replacing wires with shielded wire, for very minor improvement, but still disappointed. I even took those 68k resisters and wired them right on the input jacks so that the rest of the way to the to the 1st tube could be shielded with hardly any unshielded part.
The hum increased with volume and even tone, so I replaced the that one connected to to pin 2 of the 2nd preamp tube as well. (I see on this diagram, that that one is specified as shielded). I think there was one or two more that I did, but I was looking at the schematic as I made the decisions. (I think going from the volume pots to those first two 0.1uF caps)
Anyways, no real breakthrough. He was thankful for the improvement but we both knew that there was no real breakthrough.
We've taken a break from it, but I would like to get it quieter if at all possible.

[nonreverb]

Hey...I'd look and make sure that the heater supply wires are well twisted and are away from any signal lines. This can cause a lot of headaches with respect to hum if they aren't.

[dobsont]

Here's a couple of things to try...Does it hum with nothing plugged in? Can the hum be controlled by the volume?

Yes, to both questions. With the volume all the way down the hum disappears.

Heater supplies are well twisted and tucked into the 90 deg. angle that forms the edge of the chassis as much as possible (as in Paul's layout).

The chassis I used is oriented as a 'head' (as opposed to the combo format in Paul's diagram).

I'm not totally sure where all the signal runs, but I tried to make sure the heaters were reasonably far from the shielded input cables for sure. As far as which are the most important connections to be away from, I'm just too inexperienced to know.

Heaters DO run pretty close to the jumpers between pins 4 and 8 on the 6v6s.


I've had a little more time with it, but haven't made any changes yet.

perhaps the most interesting thing (and I didn't play with it enough to notice last night) is a crazy oscillation. Sounds like a really choppy tremolo and it can be dialed out by playing with the volume knobs. I don't know what this means, but it might be a clue, or it might be a second issue.

I've been bringing it up on a variac and I've had it as high as 90% of mains. I've seen the b+ hit 300v, so I'm not all the way up yet.

Does this immediately grab anyone?

[dobsont]

Was initially testing with 2 12ax7s due to the lack of a 12ay7.

Swapped in a 12au7 and the aggressive trem pulsing went away.

The extremes of the volume pots seem to bring buzzing. With the volumes in the right spot I can dial up a sound with relatively low noise.

Is this a clue?

[dobsont]

Nobody?





...

[Jim Jones]

The parasitic oscillation could definitely be due to less-than-ideal lead dress...

Another cause of nasty oscillation I discovered myself when building a Champ a number of years ago is unintentional positive feedback. Try disconnecting your feedback wire. If the problem goes away you have your feedback wired backwards giving you positive rather than negative feedback.

Jim

[nonreverb]

Check that the input jacks ground blades are making contact when nothing's plugged in. You can do this by tuoching the jacks hot blade while the amp is on if you hear crazy hum then they're not grounding out properly. What do you have for cathode bypass caps on the first tube? Are they 25uF X 25V?

[dobsont]

Yeah, nonreverb -- Its a 25/25.

I'm thinking about the input jacks now. I'll test like you say. Right now I've got them loose and I'm setting things up on that end of the amp like paul's diagram. Should they have isolation washers and then the grounds collected? No isolation AND grounds collected?

Thanks!

[nonreverb]

No isolation is required and you can hardwire the ground blade if you want just to be sure...:smile:

[dobsont]

Well, it's done.

I chopped out my grounds and followed Paul's layout -- worked perfectly.

The choppy oscillation thing is all but gone
Is there any chance it might have something to do with mismatched output tubes? Maybe not.

Anyhow, it's a great sounding circuit. Pictures to follow.

Thanks guys!

[dobsont]

Turns out the 'trem' problem was a dead output tube. My understanding is rudimentary at best, but I think the 5e3 operates on one output tube until the gain reaches a certain level and then it goes into push/pull 2 tube operation. The first tube would saturate, routing signal to the second (dead) tube (rendered as silence), then the signal would return to the working tube. Repeat for effect. Anyhow, it works great now.

http://i134.photobucket.com/albums/q109/cx_deluxe/IMG_7917.jpg

http://i134.photobucket.com/albums/q109/cx_deluxe/IMG_7925.jpg
http://i134.photobucket.com/albums/q109/cx_deluxe/IMG_7932.jpg

Yep, it's a little unorthodox. Whaddaya want for $50?

[nonreverb]

Nice!:smile:

[Wild Bill]

Turns out the 'trem' problem was a dead output tube. My understanding is rudimentary at best, but I think the 5e3 operates on one output tube until the gain reaches a certain level and then it goes into push/pull 2 tube operation. The first tube would saturate, routing signal to the second (dead) tube (rendered as silence), then the signal would return to the working tube. Repeat for effect. Anyhow, it works great now.

http://i134.photobucket.com/albums/q109/cx_deluxe/IMG_7917.jpg



Beautiful old radio! Who stole the knobs?

:food-smiley-004:

[dobsont]

Beautiful old radio! Who stole the knobs?

:food-smiley-004:

Glad you like it!

I love old radios and tube junk in general...

I got that one out of my cousin's basement -- He didn't want it. It was in the house when he bought it. I've got the chassis and 'magic eye' tube and of course the cabinet you see. No knobs, no other tubes (and no tube diagram either, so it's gonna take a bit of research). Someday it'll be a radio again.:smile: Too nice to sacrifice.

[Tybone]

Hey Dobby, I was going to take a picture of my Trinity Deluxe and post it but when I saw the diagram I said "he this looks pretty close to the grounding scheme that Steve uses." hehe little did I know.

So, Forget about star grounding. You end up mixing the good with the bad in terms of ground signals. Separate them as per the diagram. Also, Are your in put jacks isolated from the chassis? You only need on path to ground for these jacks.

Also, lead dress is VERY important in terms of oscillation and hum.

Cheers
Larry

[dobsont]

The oscillation is gone. It's still got some hum, which probably does have a fair bit to do with lead dress -- the board-over-tubes layout makes things tough.

The hum is really not all that bad. It doesn't always increase with volume and some control settings are better than others. As I learn more I'll develop the skills to isolate and remedy the situation. I may also experiment with a little more power supply filtering. I'll clean up what I can.

So I'm getting mixed messages here on the jack isolation situation. I sort of intuitively 'get' that they only need one path to ground. Perhaps I'll try isolating them and see where that gets me.

At the moment the biggest issue is hiss from all those NOS cc resistors. I'm looking swapping out the ones on the preamp plates for something modern. I've also replaced the too small 22k B+ decoupler with a larger cf -- which helped a lot.

Hiss mojo is a no go...

Wish I could change the title of this thread since we're moving beyond grounding issues. I'll keep posting progress here.

[dobsont]

..............................

[dobsont]

moron?

Alright, bad measurements and misunderstandings earlier.

I'm really getting a sense of how dirty the power is in my house, how much things sag and draw, etc.

Last night I went to confirm the plate voltages and found things 7 to 10 volts low. I went and turned off a space heater and my readings were back to normal. Yikes.

So here's where its running now
I get 17v across the cathode resistor (seems way low, but bear with me)
its a 220 ohm resistor
so 220/17 -- .077 and change, so 77ma -- sound about right?

figure about 4 less for the screens, so 73ma (I didn't actually measure this)

Ive got 333 and 336 on the plates (depending on space heaters, lamps, etc)

so lets just go with 334 as a conservative average.

334v*.073ma = something like 24 watts dissipation between the two tubes
so assuming the tubes I've got in there now are reasonably matched (I don't really know how to test this, I don't have a tube tester -- is there any way to figure it from measurements in the amp?) I'm getting something like 12 watts dissipation -- which is good.

I'm not sure why that cathode voltage is so low (17v when I'd expected 20 to 23), but the current draw and dissipation numbers seem to be pretty much perfect, unless of course I'm misunderstanding how to figure this stuff.

Feel free to chime in here.

[dobsont]

I'd like to know what you guys think of those voltages.

Is the cathode bias voltage simply a means to an end (and thus, since the currents and dissipation are in line its fine)?

[Wild Bill]

I'd like to know what you guys think of those voltages.

Is the cathode bias voltage simply a means to an end (and thus, since the currents and dissipation are in line its fine)?

Things seem reasonable to me!:smile:

You should understand that your line voltage from the wall sets all the voltages coming from the power transformer. I do think that the plate voltage should be more like 375-400 volts. What's the line voltage? Is it close to 120 vac?

When you calculate plate dissipation you should subtract the voltage at the cathode from the plate voltage. This is because a tube only cares about the voltage difference from cathode to plate. Most of the time the cathode is grounded, as in a fixed bias amp. So we may not realise that it's not the same thing! With cathode bias we've raised the cathode up from ground by the amount of voltage drop across the cathode resistor, which of course is the bias! Again, bias is determined by how negative the grid is with respect to the cathode, NOT to ground! A tube doesn't know or care what ground is! It only cares about the voltages at its various electrodes.

When a tube(s) is being run at or close to its plate dissipation rating it's likely operating Class A. Class A normally will have lower plate voltages and higher idling current. In fact, the idle current doesn't really appear to change when driving the amp harder. The tube is always working full out, which also is why the plate voltage must be lower to keep the dissipation from getting excessive. So Class A gives less power, but a very different (and nice!) tone!

Anyhow, I'd check that line voltage. If it's low it also will make the heater voltage low. That's hard on the tubes.

Are you sure about the HV winding connections? Both red wires going to pins 4 & 6 on the 5Y3? The red/striped wire CT to ground? Maybe a dumb question but something wrong here can cut your voltage down.

If you have low line voltage check it around the house. Often the wiring in a basement workshop can be screwy or excessively loaded, dropping the level.

Are we having fun yet?:smile:

:food-smiley-004:

[dobsont]

Hey Wild Bill - Thanks for your post. I'm learning, and I appreciate it.:smile:

When you calculate plate dissipation you should subtract the voltage at the cathode from the plate voltage.

I did know that (and I'm a bit embarrassed about missing it).

This is because a tube only cares about the voltage difference from cathode to plate. Most of the time the cathode is grounded, as in a fixed bias amp. So we may not realise that it's not the same thing! With cathode bias we've raised the cathode up from ground by the amount of voltage drop across the cathode resistor, which of course is the bias! Again, bias is determined by how negative the grid is with respect to the cathode, NOT to ground! A tube doesn't know or care what ground is! It only cares about the voltages at its various electrodes.
And now I understand why! Thank you - this makes perfect sense.

When a tube(s) is being run at or close to its plate dissipation rating it's likely operating Class A. Class A normally will have lower plate voltages and higher idling current. In fact, the idle current doesn't really appear to change when driving the amp harder. The tube is always working full out, which also is why the plate voltage must be lower to keep the dissipation from getting excessive. So Class A gives less power, but a very different (and nice!) tone!
Again, all super interesting. So for a 5e3 how hard should I be pushing the output tubes? Is 220 ohm/17v at the cathode too hot? I wish I could nail down some ideal voltages for a 5e3 - you'd think it'd be easy to find but I haven't had much luck.

Anyhow, I'd check that line voltage. If it's low it also will make the heater voltage low. That's hard on the tubes.
Today line was 123.1vac and I've got exactly 6.3vac on the heaters. That's without the space heater on...

Are you sure about the HV winding connections? Both red wires going to pins 4 & 6 on the 5Y3? The red/striped wire CT to ground? Maybe a dumb question but something wrong here can cut your voltage down.
Yep, that's how I have it wired alright!
I don't think I've got anything miswired.

Anyhow, thanks again for your input. Much appreciated!