Input jack grounding

[defaced]

First, for testing, just remove the input jack and let it hang in the air and see if this solves the problem. If so, you can buy plastic phone jacks that are intended for PCB mounting that will solve your problem, or you can order some fiber washers from Mouser. These are made by Keystone. They're in the hardware section in the catalog. Don't search their website for these, you'll probably never find them. The hole required for a 3/8" phone jack with a fiber washer is 1/2 inch, I think. So you may need to open up the hole in the chassis.

Or... depending on how you're grounding it, lift the main star point from the chassis and leave the jack alone. The chassis is just a reference ground, so you have some liberty with how you handle it.

[trem]

If you do not have any success with Mouser (and I love Mouser) Doug Hoffman sells those shoulder washers....
Good Luck

[Chuck H]

I've done it in a pinch with some shrink tube on the jack barrel and nylon washers from the hardware store.

[GregS]

Do you have any suggestions for where to get plastic/rubber washers to insulate the jacks from the chassis.

Someone already mentioned Hoffman's site. That's where I get mine. There is also information on Hoffman's site on an effective grounding method when not using isolated jacks. I tried it out and found it quite effective in one of my builds. However, I tried the method of one ground point, and isolated jacks, with each circuit gounding to a wire bus in the proper order ( as in Merlin's book and KOC's books) and the amp is ridiculously quiet. I have to put my ear right up to the speaker in order to hear anything.

One issue I didn't see mentioned in this thread when using non-isolated jacks is that even if you ground everything you can to one point on the chassis, you still have another forced connection to the chassis at the speaker out jack.

Greg

[defaced]

Speaker jacks can be handled a couple of ways depending on the circuit.

First, remember that a speaker output is isolated from the main circuit by the output transformer. So it's of little consequence is it's connected to the chassis in an amp that doesn't have negative feedback in the power amp. If the amp does have NFB in the poer amp then how you handle speaker output does mater and you have some options.

Option 1: Use normal jack and omit the typical wire from the speaker ground to the circuit ground. Probably not the best idea because it uses the chassis as a signal return path for the NFB circuit, but it'll work.

Option 2: Use an isolated jack or washers and deliberately use a ground wire from the speaker jack to the signal grounding location - usually the PI ground point.

Option 3: Use a normal jack and a return wire. This will give you a ground loop, but I don't know if it'll make a difference in the circuit or not.

The other things that usually have a chassis ground which are often overlooked in these discussions are pots. Depending on how you handle the pot casings, this can be a problem or not. I make it a point to wire ground connects from the lug back to the circuit and not tie them to the back of the pot for this reason. IMO, the pot casing should be for one thing and one thing only (electrically): shielding the pot. If that connection is made via the retaining nut or isolating the pots and using a ground wire is personal preference.

[GainFreak]

Option 2: Use an isolated jack or washers and deliberately use a ground wire from the speaker jack to the signal grounding location - usually the PI ground point.

I prefer this option. For some reason I don't like UNisolated jacks. Neither for input, nor for speakers especially if your chassis is powder coated.
Concerning the RF getting through the input there are other methods as described in the EMI/RFI thread earlier.
I have the following question: If we have the amp ground separated from the chassis by a back to back diode/resistor/cap combo as found in Marshall for example (and not only) where should be the pots' casings referenced to - chassis GND or amp's GND? I've tried both and they work OK but what is the "theoretically" right place for it?

[Joey Voltage]

Thanks for the pic. I have uninsulated switchcraft jacks on my 5E3 which are grounding out to the chassis, as well as back to the star ground. Do you have any suggestions for where to get plastic/rubber washers to insulate the jacks from the chassis. I do not have TONS of hum, but I'm hoping this might clear up the little is there. Would I be better to not run grounds off the jacks back to the star ground? And if so, what would be your advise for the best way of doing this.

Well The unisulated input method only really works if everything else is insulated, and the Jack is performing the one and only connection from the circuit to the chassis (I aso put a secondary wire from the sleeve to the chassis just incase it comes loose). If you don't mind giving me a bit, I can provide a sample layout of how I would ground an amplifier like that

[deci belle]

I've never used insulated jacks, but the Mouser's PN for the Keystone 3/8" insulated shoulder washer is #534-3069.

There is a long-bushing version of the Switchcraft jack available too. It is #502-L-12A (long version of the 12A jack).

Otherwise you will have to sand down the shoulder washer to fit the shorter jack.

[R.G.]

Apparently I missed this thread when it was current.

The way to wire ground to chassis is
(1) Incoming third-wire safety ground must be tied to chassis. This is a separate issue than signal grounding for either hum or RF.
(2) The chassis should be though of as a mechanical holder/assembly to keep everything from rattling around, and as a metallic external shield. It should carry **zero** current, other than any unavoidable RF currents as part of its RF shielding duties. The way to make this happen is to have the chassis and any metal connected to it isolated from signal ground completely, then add 1.000000 wires from signal ground to chassis. With no return wires, the chassis cannot carry currents, and also cannot cause hum.
(3) There are a couple of ways to add that 1.000000 wires.
(3a) One is to make the input bushing be the wire, by using a metallic bushing. When you do that, there is a wire from the input bushing back to the amplifier power ground which carries only signal currents. No hum or RF can run on that wire because the RF is held on the chassis by Maxwell's Laws, and the hum is held in the power supply because there is no return wire. This works really, really well as long as there is only one input. When there are two or more inputs, there are two or more connections to the chassis, and it works less well. That still may be OK for most situations. It's not guaranteed trouble, but the door is open a crack.
(3b) Another is to insulate the input jack bushing from ground with either insulating washers or non-metallic jacks, and then to explicitly wire the input jack ground to wherever you want it to go, and separately wire the chassis to signal ground either there or somewhere else. With insulated jacks, you have the option to wire ground wherever you want it to go, even with multiple input jacks. The chassis can now be a shield and the input jack grounds can carry signal currents only, which is the best situation. This may require a small ceramic cap from input jack ground to chassis right at the jack, because RF is one form that Maxwell's Demon takes, and it will leap to the lowest inductance path. The ceramic cap forces it to the shielding chassis and away from the input circuits. A resistor or ferrite bead on the input signal wire from the jack is another sledge hammer on the Demon's head, and it sulks off to the chassis. (I can hear the next question. No, it won't affect your tone to put a small resistor or ferrite bead on the input wires. Really, it won't.) It makes sense to put the small ceramic and resistor or ferrite on each input, including the reverb return, which is also a high gain input from a magnetic transducer.

Pot bushings and housings do not carry signal current. They are shields, and should be firmly grounded to chassis. They should (for hum and RF) not be used for signal grounding, no matter what you've seen in old amps. Well, OK, unless you like RF and hum, or just feel lucky.

For lowest hum, the power supply needs to be wired correctly to keep rectifier pulse currents out of the signal ground node, silicon rectifiers need to be either soft turn off types or snubbed to prevent them from emitting RF at twice line frequency, and either star grounding use or some other grounding scheme that keeps power ground returns (i.e. "sewer ground") out of the signal ground reference wires.

[GainFreak]

Although it's a SS amp with one preamp tube I think this schematic illustrates well what R.G. said about inputs, caps, beads etc. Maybe it's a little bit more than necessary because of the opamps and of course all those regulations.

AVT50.pdf

[diagrammatiks]

seems like it's just easier to use insulating washers then worry about completely changing the grounding scheme because of what input jacks are available that day.

[Steve Conner]

Yep, if you allow for an insulating washer on a Switchcraft-style jack, you can always wire round it straight to the chassis. But if you didn't allow for a washer, and discover that you need one at the troubleshooting stage, it's a whole lot more hassle to take everything apart and drill the hole bigger.

[catalin gramada]

Hello All
supposed to use input jack to chassis grounding. how seems to be the right way to tie the jack to the ground , relative with the drawings below ? directly to the first ground stage point or directly to the bus at common local star point -as time as I share the same cap for both stages - ? Thank You.
Best Regards
Catalin

[Chuck H]

Thw star point is usually a grounding lug that is bolted to an aluminum chassis or soldered to a steel chassis. In which case you would use an isolated input jack and ground it at the star only. Nothing should ever have two paths to ground.

[soundguruman]

R. Aiken writes, "If you are using non-isolated input jacks, you can instead directly connect their shield to chassis directly with a short wire (don't depend on the nut to make contact, because they corrode over time). If you do this, you cannot also ground the main power supply star node to the chassis, or you will likely develop massive hum."

I personally tend to prefer isolated jacks anyway and implement a star system, which is usually foolproof for dummies like me, but if one is forced to use non-isolated jacks, then my question is where DO you ground the main power supply star node if you can't ground it to the chassis? I'm having a dense moment as I can't see where else you could ground it; it has to end up at the chassis eventually, somewhere, doesn't it?

The power supply ground should be on the P transformer side of the chassis.
The preamp ground should be on the far opposite side of the chassis.
The signal audio ground and power supply filter ground should never be connected to the same place.
The resistance of the chassis should be between them.
Although "star" grounding may seem like a good idea, it's not.
Connecting the PS and signal ground to the same place always results in more hum.