Best Place For Heater CT?

[rp]

I know this has been beaten to death and I googled much before tiring you all with it but where do you guys usually ground the heater CT when not elevating it? Apparently there is just no clear answer out there and till now I've always connected it to the chassis AC ground lug figuring since it's near the PT and fairly close to where the heater run starts at the power tubes, it'd be fine. If I had to far too go to reach the chassis ground I just grounded it to the chassis at the same terminal strip I usually start the heater run at. Haven't had trouble with either yet.

But now it occurs to me that as I usually use a bus bar (grounded only at the signal input end) for the system ground maybe the heater CT would be better on the bus at the power supply end along with the HT CT, C1, power tube cathodes, and OT output ground?

[Bob S]

I have seen some builds with the heater ct tied to the input/v1 ground point.
Never done it that way myself.
I nail it down with the power grounds.
The only hum probs I've had have been tube or wire dress related.
If it ain't broke I don't fix it.
My last few builds I've split up the pre-amp/pi node grounds.
Interesting that the quietest amp I've ever built didn't have the heater wires twisted. Ran parallel in heat shrink the shortest distance between the tubes.
Less wire ...
Bob

[Phil_S]

I'd ground the heater CT at the same place as the HT CT. Placing it with the input ground seems wrong to me.

[gktamps]

I'd ground the heater CT at the same place as the HT CT. Placing it with the input ground seems wrong to me.

For my recent JMP50 type build I did a lot of reading up on the Larry grounding scheme, which implements this approach. Couldn't find an explanation at the time from Larry about the logic behind making the long wire run to do this, but thought I'd give it a try, any my amp is really, really quiet.

[rp]

I actually have the Larry info on my computer. I didn't think to check where he put the heater CT. Sure enough, at the input. Seems weird but I don't have any real knowledge to go on. For me as for gktamps it sure would require an annoyingly long run under the whole board to reach the input ground. Maybe when it's done I can use a jumper and try it and see what happens, or build it with a jumper and then remove it as the test.

Anyone know if it really would be better on the circuit bus rather than on the chassis? I'm starting to think it should. I always put it on the chassis to get it out of the way thinking the heaters weren't really part of the sonics.

[roberto]

This is how I do the grounding in my amps (except bias caps, of course).

[IMG1275]http://i52.photobucket.com/albums/g39/joeyvoltage/groundscheme.jpg[/img]

[Structo]

If I recall, Larry's theory was that the V1 input ground is the quietest point on the chassis.

A lot of Metro Amp members have used his grounding scheme with good results.

Seems to work on Marshall style amps.

[Phil_S]

Roberto: That's a nice diagram. I do something very similar and it has produced good results. The big difference is that I make the connection to the chassis from your cap #7, not at the input. I think that is why I feel the filament CT at the input makes no sense. The way I do it, it doesn't. The way you do it makes perfect sense.

Hey, let's debate whether the chassis connection goes at the input or at the first filter cap! That will take someone with more technical expertise than I can offer but I'll put my beak in it anyway and see if someone will tweak it.

What I was told is that the grounds should be collected in order of low potential to high potential, with the input and first gain stage at lowest and the first B+ filter cap at the highest and to make the chassis connection where the highest potential exists.

I'll observe that I've seen many discussions on the topic of ground schemes and what I've gathered is that there are several ways of doing it and they all work. They only work well when the builder understands the "rules" and that things get more critical in higher gain amps.

[roberto]

I have to say that sometimes I connect cap 7 to ground instead of the input.

The input has more sense as it's the most sensible point of the amp to noises, and it needs to be well grounded. On another point of view, grounding cap 7 is more safe as is internal and without any possibility of mechanical failure.

[jhaas]

I could be wrong, but I seem to remember that Kevin O'Connor recommends the single connection to ground be from the cap supplying the screens (#6 in Roberto's diagram). I don't remember the rationale behind it.

FWIW I do it at the input jack and can't remember the last time I didn't elevate the heaters.

[Jana]

In my experience the grounding of the heater CT isn't that critical. Think about it for a bit--the purpose of star grounding and other grounding schemes is to avoid ground loops and differing voltage potentials from one ground point to the next. The heaters are a closed system and are not supplying voltage to any of the other parts of the amp (besides the heaters).

I just ground the heater CT to a lug near the PT and be done with it (unless I'm elevating the heaters).

[roberto]

I've not read KOC books, but the reason is that screens are more sensible than plates. The same reason why the power tubes are grounded to the screen supply cap.

[cbass]

I like to hook it to the cathodes on cathode biased amps

[gingertube]

I could be wrong, but I seem to remember that Kevin O'Connor recommends the single connection to ground be from the cap supplying the screens (#6 in Roberto's diagram). I don't remember the rationale behind it.

FWIW I do it at the input jack and can't remember the last time I didn't elevate the heaters.

Kevin recommends an elevated heater with the elevation voltage sourced from a divider off the screen supply node. The rational is that in any good design the screen node will already have decent filtering and hence give a clean elevation voltage.

If NOT elevating the heaters but simply wiring the heater centre tap or "Pseudo" centre tap (via a pair of resistors) then I've always found that the same 0V point as the B+ centre tap works for me. The caveat here is that I always use bussed 0V, 0V buss connections follow schematic from input at one end thru to the B+ filter cap at the other end with a single chassis connection at the Rg1/Rk 0v end of the input tube. I use an isolated input jack, screened cable with the screen tied to that same Rg1/Rk 0V point PLUS I add a Radio Frequency chassis connection on the input jack using a 10nF ceramic cap from teh jack 0V side to chassis alongside the jack. This helps extend the screening fuction of the chassis along your guitar cable screen as well.


Cheers,
Ian

[katopan]

We all know as has been said that they all work, so take this as ideal theory but not what you need to do.

Heater hum and any other noise gets into the audio path via the proximity to the cathode that the heater is heating (assuming heater wiring lead dress isn't super terrible). Some valves hum more than others because the level of coupling between the heater and cathode is different. Subminis are more susceptible because they are smaller and the cathode heater clearance is much less. The cathode bypass cap provides additional filtering to this hum, but unless the valve is stuffed everything is more than acceptable. If you ever have the opportunity to stuff up a build (like I did on one amp) and induce rectification spikes into the AC heater supply, you'll see and hear the coupling I'm talking about come to life in a big way.

Like everything the most sensitive part of the circuit is V1 because every bit of signal there is amplified to the output more than anywhere else. So any difference between the heater CT ground and the V1 ground will show up as noise induced on to the cathode.

In practice with any decent grounding system the noise between ground points is insignificant, the amount that gets coupled from the heater into preamp cathodes is worlds lower than that, and none of it matters. But that's my takeon why theorectially the V1 ground might be the best spot for the heater CT.