An interesting interconnect problem

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Jump to Latest
DECKY999

DECKY999

Member
Joined 2007
OK - I built a BrianGT gainclone (non-inverted 3886) and I built a pair of short interconnects to link it with a pre-amp. The initial interconnect configuration was two runs of silver wire insulated with cotton in a twisted pair. I use this interconnect throughout my system without any noise problems.

However, when connected with the amp, it produced a mid-audible buzzing. Mid-audible means it was present on normal listening levels - just enough to be annoying. If the interconnect was replaced by any normal double shielded coaxial cable the noise was gone. I tried yesterday adding a shield to the cables - very heavy copper braid. Tried with a single end connection to the ground and both end connection - the results were the same - the buzz was transformed into slight hum but this time barely audible and almost not influenced by the volume level. Although the results are better I still do not understand what is happening here. Is the amp that sensitive to EMI/RFI pickup? Anything that I can to the amp to stop this? Poly-caps at the input RCAs maybe?
 
DECKY999

DECKY999

Member
Joined 2007
I do not think it is a grounding issue. I have no noise with normal double shielded coaxial cable. It is only my home made interconnect that is creating a problem. I only wonder if there is some way of attenuating the RFI input to the amp or I sould change my interconnect.

... BTW my grounding is done as per your schematics including a ground breaker form eliott sound projects.
 
A

AndrewT

R.I.P.
Joined 2004
I suspect that the chipamp does not have any passive band limiting filters on the input to the power amp nor at the RCA input terminals.

I too suspect that the grounding can be improved.

A buzz is more usually associated with misusing the PSU zero volt connection as the audio star ground
 
A

AndrewT

R.I.P.
Joined 2004
DcibeL said:
Sounds to me like your twisted pair is picking up noise from the AC mains. Is the cable running in any proximity to AC mains? Does the amplitude of the hum change when you move the cable around?

My advice would be to stick with the shielded cable, connecting the shield to ground at one end only.
Click to expand...
is that shielded two core or using the shield as the return?
 
DECKY999

DECKY999

Member
Joined 2007
Well - first about the cable(s). The power cable goes into the box in the centre - the input cables connect at the far ends of the box. The power cable is shielded.

Inside - the signal star ground is connected to the input of the ground breaker (diode bridge, cap and 10ohm resistor) - exit from the ground breaker connects to the chassis and the mains ground wire. Signal star ground point is formed by connecting to ground wires from the PCBs. I do not have a direct wire connection of the signal grounds to the star point. I was planning to do this at some stage. PSU zero volt is not connected to the star ground.

Signal input cables are twisted pair and shielded (pseudo balanced)- no difference between single or both end shield connection. At least not in the hum presence - probably yes in overall sound quality (have not had any time to check). Shield is only copper braid (very heavy - at least 96% coverage) but without foil shield. Standard dual shield coaxial had no hum issue but the sound was dreadful.

Just to make things even more interesting. Yesterday hum disappeared. Simply nothing - changed levels, inputs - simply not there. On a shorted input of the preamp - simply dead silent.

Any suggestions?
 
DECKY999

DECKY999

Member
Joined 2007
Well, honestly I do not see the difference apart from the fact that one chanell (I have dual mono configuration) will have longer ground wire path then the other. This in my opinion ruins the symmetry - apart from slight increase in resistance which is probably not an issue. My amp boards are less then 5cm from the input RCAs. So the wires from RCAs to PCBs are very short. On the PCB they are on the common ground plane and then two wires of maybe 15cm in lenght go to the ground breaker.

I will try as you said and see if it makes any difference. Thanks
 
D

decky

Member
Joined 2006
Just to report

I rewired my interconnects - the shield is now only linked to one end (pseudo balanced configuration). I added another ground line - from the input ground directly to the star ground connection.

No hum - well - the amp is dead silent to be fair.

Something fishy was happening there, I will follow up if I notice anything strange.

Thanks for all help.
 
A

AndrewT

R.I.P.
Joined 2004
decky said:
I rewired my interconnects - the shield is now only linked to one end (pseudo balanced configuration). I added another ground line - from the input ground directly to the star ground connection.
Click to expand...
I think you have described the Leach method of connecting the input RCAs to audio ground. I have found this to be effective in the majority of two channel builds.
 
A

AndrewT

R.I.P.
Joined 2004
artfuldodger said:
Ive seen other builds where they run the rca input grounds to the star ground and others say to keep them isolated from the chassis. I guess it is best to try it both ways to see what works better?
Click to expand...
these two options describe the same thing.
The RCAs are isolated from the chassis AND the RCA grounds are run to the audio ground, also called the star ground .
 
digi01

digi01

Member
Joined 2003
assemblycraft.com
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Top Bottom