resistor(s) in ground paths to reduce hum, etc.
I am aware of a couple of different ways to break up ground loops and reduce hum: (A) the "ground loop breaker" that is applied to the earthing wire after it is connected to the chassis but before connected to the circuits within the chassis, e.g. the diode bridge + thermistor or resistor. (B) I have also seen a 4.7R (for example) resistor placed in between the incoming shield of an unbalanced signal cable and the circuit ground.
Can or should resistors be inserted both at (A) and (B), or only one or the other?
What is the upper limit for the resistance that would be useful in this regard? I have often seen 10 ohms cited on the mains earth connection, but recently found a couple of refs that said up to 100 ohms could be used. Wouldn't that start defeating the ground connection and allow the ground to float?
Please feel free to bring up any relevant issues in your replies.
If you put a resistor there I don't know what it would do for the hum. I have never had problems with ground-loop hum and my wiring techniques are not spectacular. HOWEVER, the resistor in the path of the safety ground would be just as deadly. If the line shorted to the chassis, the resistor is in the path to ground. 120vac divided by 10 ohms equals 12 amps equals 1440 watts. The resistor would violently burn (if not explode) and become open, thereby eliminating safety ground. You don't want to do either of these.
(B) seems innocuous. If the resistor acted up then nothing should happen. However, I'll share with you the technique I use with guitar amps because they are the most susceptible to hum. It should transfer over to your amplifier.
I mount all jacks insulated from the chassis. That means that no part of them is inherently connected to the chassis. You have to do that yourself. Now, all the jacks on the front panel (and all the ones on the back panel) have their ground leads tied together with some heavy bus wire. 18g or 20g is good, but if you can work larger that's okay too. At the lowest-level input, run a wire from the ground bus to the mounting screw where the safety ground connects. This point also will have connections for the wiring board, filter capacitors, and everything else (called a star ground). Now, if you want to run a shielded wire from point A to B, connect the center conductor ONLY at both ends. Now, either on the wiring board or at the input jack, connect the shield to ground. You don't want both ends connected because that makes the shield another path for current to flow.
Putting a resistor at point B won't really do much for you, just elevate the shield and let it float a little. Still, 4R7 is not a lot and probably won't pick up much. IME, it takes a lot of resistance to pick up noise. If the resistor was a WW, it might have enough inductance at radio frequencies to allow some to develop on the cable shield.
Please be careful. Just because someone posts something on the 'net doesn't mean it's safe or wise. Also, just because nothing dangerous has happened so far doesn't mean nothing dangerous could happen. When in doubt, ask. Someone will usually be able to help you because most electronics people don't want you to become a crispy critter.
Normally for grounds in audio amps I have a long, wide trace on the PCB or run some bus wire along the turret board. It's mostly overkill but if you can get resistance into the micro-ohms then it allows for faster operation of the filter capacitors. Beyond a certain point, though, adding more copper will cease to gain improvement. Usually a wide trace or an 18g wire is enough.
I am not sure I understand your description.
Bow certainly states it correctly that if one inserts the resistor in the wrong place then it can be fatal.
The third wire (PE) of the incoming mains cable must be connected permanently to chassis. Nothing should ever be added to that direct wire connection. Never ever!!!!
Now that the chassis is permanently connected directly to mains PE, other options become available.
Is that clear?
The life of any/all operators of your equipment depends on following that advice.
Wow, I thought I stated that clearly:
(A) the "ground loop breaker" that is applied to the earthing wire after it is connected to the chassis but before connected to the circuits within the chassisSo let me clearly say this again, I am NOT ADVOCATING ANYTHING BETWEEN THE CHASSIS AND THE MAINS SAFETY EARTH!!!!!!!!!!!!!
I am talking about the well-known techniques like what is described here (see "Use Of Loop Breaker Circuits" section):
Earthing (Grounding) Your Hi-Fi - Tricks and Techniques
I have also seen a low-value resistor, like 4.7R, between the shield of the interconnect and the chassis. It's a completely separate place, but also one that participates in ground loop, hum, and noise pickup.
My questions are on the values of the resistances used in these two places.
Your descriptions are not clear enough for me. Apparently not clear enough for Bow.
Here is a link to a web site talking about including a resistor (10 ohm) in the interconnect shield to reduce hum and noise pickup:
ABOUT GROUND LOOPS
See the section on the right side, under "Ground Loop causes and cures:", especially part C of the figure.
A ground lifter as you describe is a very effective way to prevent gross ground loops and provided you use a decently rated diode (100Hz repetitive surge current of at least 10x the rated circuit breaker or fuse current) you will be safe.
There is another ground loop mechanism wherein power rail decoupling capacitor charging currents (and general rail gunk) can find its way back to the 0V filter capacitor junction via two paths. One is your 0V supply return line - the normally accepted route - and the other is via the input ground connection. This happens because the input signal ground in many layouts is terminated on the PCB, and a separate signal ground run from the PCB to the star ground. To break this loop, insert a 4.7 Ohm resistor between the signal ground and the main 0V ground on your PCB.
If you follow the usual designs of the N.P, you will notice that the 0volt point of the amps P/Supply is "lifted above Chassis GND" via a 10 Ohm thermistor - it's very effective and works just as well as the cct described by Rod in the 'esp' article and you retain the protection of direct chassis earth connection - it's still a bit surprising how many people will break off the earth pin on the power plug to get rid of that insidious 'hum' problem.
Signal wire 0 volt points (sheild, etc) is far better fed directly to the circuit board 0 volt point &/or central star earth and not connected to some random chassis point, even if by the use of a 100 Ohm resistor to "float" this RCA point above 0 volt gets rid of 'hum'.
Now, it's curious that if you use the same "esp article" technique with all the other devices in your system, this avoids ground loops throughout your system - for example, on some CD players, tuners, preamps, etc the central 0 volt point is "lifted above" chassis earth by a 100 Ohm resistor, or a 100 Ohm thermistor &/or etc -this gives you the safe earthed chassis plus the "earth loop breaker" resistor but at the device central 0 volt point and not in the signal interconnect wiring.
Hope this helps reduce the confusion ...
The above is just bad information...
I never had a "ground hum" in my devices because they are all bulid corectly, connected to grounded receptacles and I always use correct cables between them. If you have "hum", your problem is in the fudged devices. "Lifting" ground is just another fudge trown in the mix. I cannot belive how many people don't have a clue how grounds should work... I guess this article is really needed:
There are quite a few ways of connecting up multiple component systems, naturally enough.
In David Davenport's article (Audio Component Grounding ...) that you have linked to, section 3.7 (Effective Interconnection Schemes) contains "Simple System of SLB Components" (3.7-3) where he has indicated that this is a viable technique of using Safety Loop Breakers in system connections.
Isn't the CL-60 thermistor that is used throughout N. Pass designs (to seperate central 0volt on the power supplies and the chassis earth) precisely that?
I don't think I would call this a "fudged" design, assuming this means poor engineering practices - but, perhaps I have mistaken your comments ....
|All times are GMT. The time now is 10:03 AM.|
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
Copyright ©1999-2016 diyAudio