OT grounding

[catalin gramada]

Hello
Did You ever feel is better to ground OT common output (except for global NFB or safety purposes) ? can we have any benefit from that (eg.stability operation, prevention oscilations etc.), and why? please. I asked cause I did not prooved any differences in my projects.(not scooped , not heared )
Thanks
Catalin

[Chuck H]

Mr. Aiken has this to say:

What about the output jacks?

The current in the secondary winding of the output transformer can be very large. For example, in a 100W amp, the secondary current into a 16 ohm load is 2.5A. It is even higher into a 4 ohm load, at 5A. This means that you need to pay special attention to the grounding of the output jacks and the output transformer. It is important not to use the chassis for this return path. The output transformer has a common wire and one or more speaker taps, usually at 4, 8, and 16 ohms. The speaker taps usually go to an impedance selector switch, and then a single wire goes to the output jack tip connection. The common wire should never be connected to the chassis right at the output transformer. It should be run all the way to the output jack and connected to the sleeve connection of the jack. This accomplishes two things. First, it maintains the continuity of the connection in the event the output jack becomes loose. Second, it keeps the heavy secondary ground currents from flowing in the chassis. Note that there still must be a return path to the rest of the circuit if the amp uses global negative feedback. This should be in the form of a wire from the sleeve connection of the output jack to the preamp ground point where the phase inverter common connections are grounded (or wherever the global feedback is returned). Note that there will be no heavy currents in this wire. The speaker output jacks can be either isolated or non-isolated if you follow this plan, but it is usually best to isolate them to maintain control of the return current path for the global negative feedback, to insure it doesn't flow through a part of the chassis that may contain power supply ground currents.

Sometimes it helps to ground the common side of the output jack to the chassis even when no global negative feedback is used. Occasionally, an amplifier will have a high-pitched oscillation noise, or other type noise that will go away if you ground the output transformer common wire at the speaker jack sleeve terminal. In addition, there may be a potential for a small AC current to flow between the floating sleeve and chassis if the secondary is not grounded. This current is due to capacitive coupling in the output transformer, and may cause a mild shock if the speaker plug sleeve and chassis (or guitar strings) are touched while running a signal through the amplifier. Even though the potential for dangerous currents is low due to the galvanic isolation of the output transformer, the shock can still be annoying. For this reason, it is best to always ground the common side of the secondary even when no global feedback is used.

[catalin gramada]

Thanks Chuck
But still not clear for me. Mr Aiken presented some situations where difernt circuits shared (intersect) the same return points. I talked about about a situation where no nfb used and output jacks are isolated from chassis. The output circuit have no potential to develop external couplings and further in reverse if talk about magnetical interference the iron is more responsable to act as antena than the coil. I'm not a tech but simply try to understand why this ground should be necessary , excepted the situations presented above. In this conditions if I tie the tip of the output jack to the ground should be the same thing, don't You think ? Why should I do it? Thanks
Catalin

[Chuck H]

I thought Mr. Aiken was pretty clear. Electronically it's best to keep output transformer current off of the amplifier chassis. But this can result in your body becoming the lowers impedance path for capacitive coupled currents so don't do it. IMHE without ever getting everything else just right, other circuit grounds, layout and lead dress, etc., it's better overall to ground the output. A few concessions or errors in ideal circuit construction can lead to unwanted noise or instability due to radiant field coupling. This is probably why every guitar amp circuit I've looked at has a grounded output. Including those with no NFB. So, yes, most of the time a grounded output is more stable and reliable for circuit construction because it's hard or impossible to make everything else exactly right also. The finer and more technical reasons, formulas and explanations go over my head. And over the head of most techs and builders without a degree in both electronics AND physics. The basics, and some very good guidelines and reasoning for the circuits we use is simple to the point we've already covered and then very complicated after that.

If I misunderstand your question I apologize.

[catalin gramada]

Thanks for You explanation.
I have to consider like a done thing. A properly grounded output never hurt, but think (supose) is more compulsory just for safety reasons.
Thank You
Catalin

[GainFreak]

Since this thread is about OT and grounding how does grounding the OT iron (laminations) affect tone (if at all)?
I've seen OTs where laminations are at ground and others which are isolated via those plastic shouldering washers?

[R.G.]

Since this thread is about OT and grounding how does grounding the OT iron (laminations) affect tone (if at all)?
I've seen OTs where laminations are at ground and others which are isolated via those plastic shouldering washers?

You're not going to like this.

The correct answer is - it depends.

The OT iron has strange effects. First, it's a nonlinear load on the primary of the transformer, as the reason it's laminated at all is that the conductivity of the iron lets eddy currents flow in the iron as the M-field changes in the iron. This is a characteristic of the iron alloy (silicon, usually, to make it more resistive) and the B-H curve. This part doesn't change much with grounding/ungrounding the OT iron. It can change with shoulder washers in the holes for the bolts that hold the laminations together. Whether this is audible or not depends on the tube output stage versus the OT design, neither of which you have any control over. Theoretically, isolated bolts are better in that it removes an eddy current path that can affect the fields in the iron. This is a big factor in some power transformers, but not usually in OTs where the flux density is lower for other reasons.

The the laminations are theoretically "insulated" by that blackish/blue layer of iron oxide. That's there for a reason - prevent the laminations from shorting together to cut eddy current losses. In real life it's more of a resistance than an insulation the way we think of insulation, but it cuts eddy current losses a lot. The resistance from lam to lam is - variable, but not a short circuit. And it varies with how tight the bolts are. So it's not really possible to ground all the laminations; they have a variable resistance to one another and to chassis. In some power transformers, all the lams are welded by a single bead of weld on the side right at the joint between Es and Is. This means the laminations can actually be grounded. But it's done for safety and ground continuity, not for "tone". I've never seen an OT treated that way.

One thing that would be affected by grounding/ungrounding the laminations - the capacitance from windings to laminations. That is, if you could really ground the laminations, which you really can't. There will be some distributed capacitance from place to place on the windings through the laminations, but it's both small and terrifically difficult to get a handle on what it does.

I think that (1) the effect is likely to be small as you've presented it and (2) I've never seen an OT that is welded so the laminations can actually be grounded.

[trobbins]

If you were the speaker load, then it doesn't matter whether one 'side' or the other is connected to amp chassis. The caveat being in a combo where the speaker metalwork may capacitively couple back to parts within the amp.

If you were the output stage circuit connecting to OT primary windings, then yes there are extra parasitic capacitances coupling to chassis then back to circuit 0V. The coupling can be thought of as capacitances from each anode all the way along the windings to the CT, and over to the 'chassis'. The 'chassis' referring to any path back to chassis, such as via core laminations, or bell-ends, as well as via the secondary windings. This parasitic influence could become noticeable at higher frequencies (eg. above 10kHz) as the capacitive impedance levels decrease, but a guitar amp is usually not set up for much higher frequency response, so the subtlety on tone of that capacitive loading should typically be negligible.

The core may be a stack of resistively connected laminations, which may be resistively connected by bell-ends or other mounting method to chassis, but in essence each of those resistances is bypassed by local capacitance between laminations, and that impedance contribution is likely to be low compared to the capacitive coupling from primary winding over to the laminations- bell ends - secondary winding.

If you ground or float the secondary windings, then yes that would change some of the parasitic capacitive coupling paths to chassis (and hence the total impedance loading on the primary side), but its likely that would be benign for guitar amp operation. Capacitive coupling between windings depends a lot on the interleaving used in the OT.

Old PA amps would float the speaker line. The floating speaker line could then suffer one 'fault to ground' without causing loss of function. Some amps included a capacitor bypass to chassis on one side - probably to alleviate capacitive 'shock', but also to stabilise higher frequency impedance loading which may become an issue for badly mismatched speaker loading (eg. light or open speaker load)

[catalin gramada]

Thanks for answer
I work curently with Hammond. I found no diferences scooped with signal up to 20k and no instability tendencies even with square at max power for very short period with or without output ground. practical no changes. I will do further tests to see how far can I go before output grounding become compulsory to be used. Thanks
Cheers
Catalin

[trobbins]

Every amp would be different, for many and varied reasons - so there is no compulsory or absolute outcome. Just connecting measurement probes to a floating output may influence what you believe you are trying to observe.

[Chuck H]

Hmmmm... I had a thread recently where there was an oscillation on the reverb circuit. Not the typical acoustic feedback, but the more insidious 15kHz whine that would spontaneously pop to life when the reverb gain knob reached 8-ish. The output of this amp is grounded at a selected place and the jacks themselves are isolated from the chassis. The only thing I tried that stopped the oscillation was grounding the speaker frames. Most of this discussion goes over my head, but I get the feeling my problem was covered in here somewhere

[Malcolm Irving]

Since this thread is about OT and grounding how does grounding the OT iron (laminations) affect tone (if at all)?
I've seen OTs where laminations are at ground and others which are isolated via those plastic shouldering washers?

If you have a steel chassis, bolting the transformer to the chassis allows the magnetic field of the transformer to leak into the chassis. For a PT this can cause hum, not sure if it would be a big problem for an OT. On the other hand, connecting the bell-ends to ground will screen the transformer windings and prevent the transformer signal coupling (electrically) into any nearby tubes. The bell-ends won't screen the leakage magnetic field though. Perhaps the best of both worlds might be for the transformer to be decoupled (magnetically) from the chassis via plastic washers, but the bell-ends to be grounded through a separate wire connection.

[catalin gramada]

Hi
what do you think about doing an artificial center tap over OT secondary to keep the output balanced, ? I wish to use the speaker output for a balanced line out. Do you think it works, please ? (no nfb will be used) Thanks
Catalin

[Enzo]

A balanced line has no ground connection, assuming neither side of your output is grounded, why have the artificial center tap at all?

[Leo_Gnardo]

If you insist on having a grounded center tap then let your 4 ohm winding be the ground connection, 0 (com) and 16 ohm will be your balanced signal connections. Nothing artificial there.

It's a good safety practice to ground one of the OT secondaries. If there was a short between primary and secondary, that should pop the fuse and shunt fault current to ground rather than have those stray electrons find a way to ground through you and your guitar. Other than that, as Enzo says, there's no reason to ground the secondary at all.