DIY Audio Projects Forum
http://diyaudioprojects.com/Forum/

Grounding and Sheilding for the KT120 Oddblocks
http://diyaudioprojects.com/Forum/viewtopic.php?f=9&t=5941
Page 1 of 4

Author:  ozapaydin [ 08 Jun 2017, 07:49 ]
Post subject:  Grounding and Sheilding for the KT120 Oddblocks

Yes Poty, I'm referring to "2 wires in the shield" cable.

So, here is the recipe:
- Using a shielded cable between poweramp and preamp, where its shield is NOT connected from power amp end
- Using a shielded cable inside preamp, where its shield is connected in both ends,
- Using a shielded cable between preamp and source where its shield is connected in both ends.
- No chassis grounding of signal ground or no earth grounding of chassis ground, which means no need to use a metal case.

This way, ground loop is avoided and signal is mostly shielded except on the connectors and the pot. Please correct me if I'm wrong.

poty wrote:
ozapaydin wrote:
There is no earth grounding in the circuit you posted, only chassis grounding if the case is metal, right?
Not exactly. You've correctly noted that
ozapaydin wrote:
If we use a metal case, shouldn't we earth-ground it?
the case is not grounded in any point. So
ozapaydin wrote:
to complete grounding against RF and EMI, we should use both side terminated shielded phono cable to give unwanted RF and EMI to the source side.
is correct. (I assume that "shielded phono cable" is one wire in the shield.) In the case you always have to "terminate" the cable to both sides or the return path would be broken. There is no ground loop in this case (beside the ground loop between channels).
If you think about "2 wires in the shield" cable or "double shielded" cable (one wire and two shields) then one of the shield (external) may not be connected from one end and that way breaks the single channel ground loop.
Personally I would not ground the case and more than that - not used any connection from a signal ground to the case in such passive design.

Author:  poty [ 08 Jun 2017, 08:38 ]
Post subject:  Re: KT120 Oddblocks

ozapaydin wrote:
...a shielded cable between poweramp and preamp, where its shield is NOT connected from power amp end
Well... It depends on your active systems ground scheme. I'd connect the shield to the active part, rather than to the passive.
ozapaydin wrote:
...a shielded cable between preamp and source where its shield is connected in both ends.
Why is the cable different from the cable between poweramp and preamp?
ozapaydin wrote:
No chassis grounding of signal ground or no earth grounding of chassis ground, which means no need to use a metal case.
Metal case is a parasitic fields isolator itself. While in many cases the connection to the mains ground helps with beating hum, the connection is needed mostly as a protective measure IMHO.

And... I agree with Juancho about using a passive preamp here. You certainly may have other experience.

Author:  ozapaydin [ 08 Jun 2017, 09:07 ]
Post subject:  Re: KT120 Oddblocks

When we use a shielded cable inside a PASSIVE preamp do we need another shielding like a metal case??

Author:  gofar99 [ 08 Jun 2017, 10:32 ]
Post subject:  Re: KT120 Oddblocks

To Moderator: Suggest moving this portion of the thread to a new one on grounding and shielding.

Some rules/guidelines on shielding are in the tech section.

Generally I follow these...
1. If the cable between pieces of equipment has only a single center connector then I ground both ends to the signal ground of the gear (not chassis).
2. If the cable has two inner conductors I use one as signal ground and the other as signal hot. If the equipment it it attached to has only RCA or similar connectors I connect the shield to only one end and the one that seems best for me is the source end.
3. Inside a device I use single center connector shielded wire for any high impedance low signal level connections longer than about 6 cm. I will typically have a signal ground buss at the inputs and will attach the shield at that end only. The circuitry should have its own ground that goes to the same place. The chassis is isolated as noted in next paragraph.
4. Anything that has high impedance and low signal level (below about 10 volts) should be in a metal (preferably steel) case that is connected to the signal ground inside it via either a capacitor and resistor in parallel or a pair of reversed diodes in parallel between the signal ground and chassis. In really low level signal areas I will run a separate ground between the devices cases.

Other things may work...but I'm fanatical about hum and noise and these work for me. YMMV

Good listening
Bruce

Author:  poty [ 08 Jun 2017, 11:55 ]
Post subject:  Re: KT120 Oddblocks

In many cases - not, but in my case the metal case helped me to eliminate EMI impact on pots, especially when I used metallic knobs and touching them produced rather loud hum.

Author:  ozapaydin [ 08 Jun 2017, 13:36 ]
Post subject:  Re: KT120 Oddblocks

Bruce, can you explain the function of capacitor and resistor in parallel?

Author:  ozapaydin [ 08 Jun 2017, 13:46 ]
Post subject:  Re: KT120 Oddblocks

Poty, thanks for sharing your experience. I just try to bring it to life in my head. So, your experience is valuable for me.


poty wrote:
In many cases - not, but in my case the metal case helped me to eliminate EMI impact on pots, especially when I used metallic knobs and touching them produced rather loud hum.

Author:  gofar99 [ 08 Jun 2017, 16:13 ]
Post subject:  Re: KT120 Oddblocks

Hi, They provide just enough separation between the signal and chassis grounds to prevent ground loops there. The chassis still works as an EMI shield and in systems that have three wire AC mains with the earth ground connected to the chassis it is a safety feature. BTW most countries now require either earth grounding of all exposed metal parts or double insulation for them. The latter is difficult for DIY builds.

Good listening
Bruce

Author:  ozapaydin [ 09 Jun 2017, 08:06 ]
Post subject:  Re: KT120 Oddblocks

Hi Bruce,
I took following part from grounding and shielding section. What do you mean when you say active portion of the circuit and power output stages?

An obvious question now is that if you provide power to something in the active portion of the circuit, won't the current end up in the signal ground. The short answer is yes. It is also unavoidable and generally rather small in magnitude and normally does not cause problems. An exception is in power output stages. The large amount of current involved can cause noise in the signal ground so I ground them separately to the input jack ground.

Author:  gofar99 [ 09 Jun 2017, 09:53 ]
Post subject:  Re: KT120 Oddblocks

Hi, Not exactly. Take a preamp for example. The active circuitry (tubes) ground including the small amount of current going through the cathodes should be separate from the main power supply ground. Then each section will have a single connection to the main ground. The main ground I place at the input jacks. It will typically be a buss wire between all inputs and outputs grounds. This is a sort of modified star arrangement where all the grounds go to one place. The purpose is to keep at a very minimum the amount of power flowing through any signal ground path. Sure some as noted above will, but the bulk of the stuff flowing around in the power supply will not. You sort of have to think in terms of relative voltage sizes. Say the preamp is handling 100 millivolt signals. Then its input side ground will be doing the same. For the moment ignore the output side. The power supply depending on what type and number of filters and such will have between some components (like two filter caps separated by a resistor) many volts of ripple there. The other half of the ripple needs to go through the ground path between them. Even if it is very low impedance it will develop a voltage there. If you tie this point to the the active circuit ground in such a way that the path to the signal ground is part of that circuit you will introduce that ripple on its ground. The active device doesn't know what is signal and what is not. It just amplifies whatever is between its grid and ground. If the ground has ripple it amplifies it as well.

A simplified description for sure....the key part is that you don't want any power flowing through the signal ground if you can avoid it.

Good listening
Bruce

Page 1 of 4 All times are UTC - 6 hours [ DST ]
Powered by phpBB® Forum Software © phpBB Group
http://www.phpbb.com/