"Like" AX84 on facebook

'Bus -> Hybrid grounding trade-offs'
Author:tristanc (registered user: 162 posts )
Date: Fri, Sep 18th, 2015 @ 03:30 ( . )

With my proposed N5X -> L2L / UHG conversion [link] it is front-of-mind to be careful with lead dress and grounding to help avoid squeal / noise / oscillations etc.

The original layout N5X included 2 buses to ground for the preamp board, and one for the poweramp (roughly). These were connected to the chassis at 3 differing locations. There were additional chassis connections at the input, speaker connection. This works fine for the original amp. I've a bit of hum when the power control is turned up, but OK for my uses.

With the extra gain of the L2L / UHG, and as I'm planning this thing to death (almost to the point of giving up...), I thought I'd review the grounding schema / layout.

I reckon it'll be possible to follow Merlin's chapter (single grounding point at the input, things grounded sequentially, grouping around filter caps etc) whilst keeping within the spatial contraints / other layout considerations.

However, this involves dividing the existing buses in places and running jumpers from one bit of bus to another.

So aiming for something that looks like Fig 15.10b from Merlin's book, my questions are:

- Would there still be a benefit from the hybrid layout even if it means more lengths of wire running all over the place?

- Would any benefit be negated if a, e.g., preamp stage grounding points aren't exactly at the filter cap ground, but the next turret along the bus?

- The alternative would be to move components around and necessitate longer signal / cathode / anode connecting wires. Which would be preferable to have shorter?

I realise this is an immensely subjective topic and there is no real answer apart from 'try it and see', but if there are any pointers I'd be most grateful.

I'll fire up Illustrator over the weekend and work on a new layout to illustrate what I'm proposing. Unless someone says not to bother and keep with the original!

-- REPLY: [With No Quote] --- [With Quoted Text]

'Bus -> Hybrid grounding trade-offs'
Author:Merlin (registered user: 5235 posts )
Date: Fri, Sep 18th, 2015 @ 05:14 ( . )

On 09/18/2015 @ 03:30, tristanc wrote :
: - Would there still be a benefit from the hybrid layout

I don't know what you mean by hybrid.


: - Would any benefit be negated if a, e.g., preamp stage grounding points aren't exactly at the filter cap ground, but the next turret along the bus?

The 'next turret along' doesn't sound very far away (an inch?), so it is not likely to make any difference. You should be fine.

-- REPLY: [With No Quote] --- [With Quoted Text]

'Bus -> Hybrid grounding trade-offs'
Author:tristanc (registered user: 162 posts )
Date: Fri, Sep 18th, 2015 @ 06:27 ( . )

Thanks Merlin.

I perhaps have used the incorrect term... I read 'hybrid' from googling the differing methods, e.g. [link] I took it to mean 'somewhere between an ideal star & bus' layout. I realise you didn't use that term in your explanation and it is a bit confusing!

Rather than having a single bus wire running the length of the board with components attached in the order they appear on the preamp board, I'll make cuts in this bus and link, say, the 1st and 2nd stage cathodes to the appropriate filter cap ground (located further down the board physically) with a jumper. I'll make sure this then runs sequentially (1st/2nd stage -> 3rd/4th -> Power) again facilitated by a jumper or two along the way.

This means the ground of the caps and cathodes won't meet in exactly the same location, and potentially not 'closer' as indicated by comparing your (a) and (b) figure. So I guess I'm trying to work out if the effort of having the single grounding point & improved order of components (wrt grounding) will be offset by the extra length of wires connecting it all.

Of course, as it stands, the design would use the chassis for these stretches, so it's likely to be an improvement on that?

-- REPLY: [With No Quote] --- [With Quoted Text]

'Bus -> Hybrid grounding trade-offs'
Author:tristanc (registered user: 162 posts )
Date: Sun, Sep 20th, 2015 @ 05:42 ( . )

I realised after an hour or so of drawing that my attempt to keep the existing layout whilst improving the grounding was going to be a complete mess to wire...

So I've re-drafted a layout based on Merlin's figures 7, 9 and 10b. I moved each stages' filter cap to be closer to that stage and run sequentially (fig 7), and wired according to figure 9. I made the adjustment to speaker grounding location as in 10b. I think it's far neater (only a few extra jumpers & will need a few tag strips to do the grid resistors) - but would appreciate another pair of eyes checking it out:

Preamp: [link]

Poweramp (only grounded parts shown): [link]

The single connection to the chassis would be on the bottom left, with that turret also taking the ground connection from the input.

The area I am most concerned about is the jumper required from C12 to C13 (on the poweramp board) and then another back to C14 (to keep things in sequence). I'm hoping these 2 stretches won't have too much of a -ve impact.

Also, I'm not sure where the tone stack / MV (VR5) should be connected. I have it currently going to the 4th stage cathode. And R14 (power tube grid leak?) goes to the power tube cathode?

I've made the decision to site the power tube's screen and grid resistors (R18 & R15) by the EL84 socket using tag strips, instead of the 6V6 which I could probably do direct from the board without tag strips. The reasoning was that I'll likely use the EL84 most so would rather avoid any extra stretches of wire from one socket pin to another following the resistors. Makes sense?

-- REPLY: [With No Quote] --- [With Quoted Text]

'Bus -> Hybrid grounding trade-offs'
Author:Merlin (registered user: 5235 posts )
Date: Mon, Sep 21st, 2015 @ 23:15 ( . )

Where's the circuit diagram to go with this layout?

-- REPLY: [With No Quote] --- [With Quoted Text]

'Bus -> Hybrid grounding trade-offs'
Author:tristanc (registered user: 162 posts )
Date: Tue, Sep 22nd, 2015 @ 00:20 ( . )

Circuit: [link]

The grounding represented on this is now out of date (and I think I had drawn the chassis connection points incorrectly anyhow).

-- REPLY: [With No Quote] --- [With Quoted Text]

'Bus -> Hybrid grounding trade-offs'
Author:Merlin (registered user: 5235 posts )
Date: Tue, Sep 22nd, 2015 @ 23:18 ( . )

On 09/20/2015 @ 05:42, tristanc wrote :
The area I am most concerned about is the jumper required from C12 to C13 (on the poweramp board) and then another back to C14 (to keep things in sequence).

That's fine, you're creating a Y shape in the ground scheme which won't be a problem.


: Also, I'm not sure where the tone stack / MV (VR5) should be connected.

They are 'part of' the PA grid leak system, so they should go to R14 or around that cluster.

And R14 (power tube grid leak?) goes to the power tube cathode?

Exactly right.

I've made the decision to site the power tube's screen and grid resistors (R18 & R15) by the EL84 socket using tag strips, instead of the 6V6 which I could probably do direct from the board without tag strips. Makes sense?

Yes, looks great.

-- REPLY: [With No Quote] --- [With Quoted Text]

'Bus -> Hybrid grounding trade-offs'
Author:tristanc (registered user: 162 posts )
Date: Wed, Sep 23rd, 2015 @ 05:42 ( . )

Thanks Merlin. Greatly appreciated.

With shielded signal runs to / from the valve stages, heater elevation and this new grounding scheme I have hopefully reduced noise / oscillations somewhat. That leaves lead dress (will do my best...), the use of metal filament resistors at the input (can swap those in post-build if there's an issue) and component/valve positioning (can't do much about).

Once I make the revision to the MV/TS grounding I should be in a position to start putting it all together. Unless I can find something else to tweak!

-- REPLY: [With No Quote] --- [With Quoted Text]


--- 43 Users Online --- 12 Recent Unique Posters

Q101=1488128912 - Threads: / 1488128912