Output transformer mounting onboard

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Deepdark

Well-known member
Joined
May 19, 2013
Messages
1,321
Location
Quebec, Canada
Hi

I would like to know one thing. When we install an output transformer, ie, an Cinemag CMOQ, directly on a ocb and screw it there, I guess one bolt should be grounded, so any RFI or whatever the noise will have a path to ground, instead of sticking on the transformer, right? It's like, the same as installing the transformer with brackets on the chassis, we must have a path to ground. Is my understanding right?
 
As you said, when you bolt it to the chassis it is connected to the chassis. When you bolt it to a PCB you should also connect it to the chassis. The chassis and the signal 0V should be connected together at one point only.

Cheers

Ian
 
Thanks Ian.

In fact, I saw a lot of people simply connecting one mounting hole to the ground plane. I know it's not directly to the chassis as it would ultimatly be, but it seems to work (here we speak about keeping a good grounding scheme, so the signal ground ground will met the chassis at the main bolt only). I thought it would be a good solution, since we do the same with input transformer shield and can.
 
abbey road d enfer said:
This wouldn't be acceptable for an RF product, but for audio there shouldn't be any issue.

Thanks.

I guess given the relative low impedance of auido, it's acceptable and isn't a problem, compared to hi-z circuit? Or there is another reason why it's not that of an issue for audio?
 
Deepdark said:
Thanks.

I guess given the relative low impedance of auido, it's acceptable and isn't a problem, compared to hi-z circuit? Or there is another reason why it's not that of an issue for audio?
The problems related to leaving a transformer core floating are much more sensitive when dealing with high frequency signals. There are some cases when audio equipment is used if RF-heavy environment, though. In that case, teh more stringent constraints of RF design are to be used.
 
ok. But in general, I'm ok to have one mounting hole connected to the ground plane for teh output transformer, as well as having the body of the input transformer connected to the groundplane, too?
 
Deepdark said:
ok. But in general, I'm ok to have one mounting hole connected to the ground plane for the output transformer, as well as having the body of the input transformer connected to the groundplane, too?

The body of the input transformer may be different. As well as reducing magnetic interference, the body of an input transformer also acts as a screen from electric fields. The secondary, which is usually unbalanced, would benefit if the body was connected to a local 0V in the same way that an unbalanced screened cable would have the screen connected to 0V. For example, Sowter mic transformers do not generally have a wire connection to the body. So, if you use a PCB mounted version, the body is unconnected. If you touch the body with your finger, audible hum is induced in the mic input circuit. This is especially so when the input stage is a high impedance tube. In the identical circuit, using a PCB mounted Jensen or Cinemag transformer ,which do bring out a connection to the body, if the body connection is tracked to 0V then the hum is not induced when you touch the body

In an output transformer there generally is no overall electrostatic screen so this does not apply.

Cheers

Ian
 
ruffrecords said:
The body of the input transformer may be different. As well as reducing magnetic interference, the body of an input transformer also acts as a screen from electric fields. The secondary, which is usually unbalanced, would benefit if the body was connected to a local 0V in the same way that an unbalanced screened cable would have the screen connected to 0V. For example, Sowter mic transformers do not generally have a wire connection to the body. So, if you use a PCB mounted version, the body is unconnected. If you touch the body with your finger, audible hum is induced in the mic input circuit. This is especially so when the input stage is a high impedance tube. In the identical circuit, using a PCB mounted Jensen or Cinemag transformer ,which do bring out a connection to the body, if the body connection is tracked to 0V then the hum is not induced when you touch the body

In an output transformer there generally is no overall electrostatic screen so this does not apply.

Cheers

Ian

Thanks Ian. Pretty good explanation. And for the output one, there is no can around the transformer, but is static noise tends to stick to the core? I guess yes. So, having a path to the ground plane on one of the transformer mounting hole would be just fine, right? Or, better, a trace from its mounting hole to the nearest mounting pad of the pcb, which will touch the chassis with metal standoff. This gives a direct path to the chassis. But I guess the first option will work just fine, given the pletora of people who just did it.
 
Deepdark said:
, but is static noise tends to stick to the core?
"static noise" is a wrong word. It dates from early days of wireless, and does not relate to audio, except when used as "similar to".
  I guess yes.
In fact, NO. Noise  does not stick to anything. Outside interference is capacitively coupled from secondary to primary. Grounding the core makes it act as a partial electrostatic screen. 

Sorry for being punctilious, but I believe some day you'll have fond memories of my lectures... ;)
 
abbey road d enfer said:
"static noise" is a wrong word. It dates from early days of wireless, and does not relate to audio, except when used as "similar to".  In fact, NO. Noise  does not stick to anything. Outside interference is capacitively coupled from secondary to primary. Grounding the core makes it act as a partial electrostatic screen. 

Sorry for being punctilious, but I believe some day you'll have fond memories of my lectures... ;)

haha thanks. It's ok to be punctilious, I prefer to learn the good termes :) Thanks a lot. So I was ok to ground the core :)
 
Deepdark said:
haha thanks. It's ok to be punctilious, I prefer to learn the good termes :) Thanks a lot. So I was ok to ground the core :)

I just learned the term punctilious in this thread!

One thing that may be of note is that in more complex multi-layer type boards, there may be things grounded to the PCB that are specifically a chassis ground net.  Most boards I've worked with professionally will have chassis ground present and typically connected to any PCB mounted hardware, used for ESD filtering, etc.  We may pick up chassis ground from exposed pads on mounting holes, or even bring it over on a connector from some other part of a system that's supposed to pick it up from a chassis stud.

Something like a typical two layer audio board may not go so far.  If you're using a ground plane it'd likely be signal ground.  There's potential to do something like run a chassis ground net and keep traces around the edge of the board, providing a point to get to a connector,  shunt RF/ESD currents to chassis at some of the inputs/outputs, or provide a chassis connection for a shielding can or something.  Could pick this up by mounting holes or feed over a power connector pin from a power supply board or something like that.



 

Latest posts

Back
Top