Transformer diagram question

Hi all,
Pardon my ignorance, but could someone please explain to me what the black lead on the secondary in this cinemag input transformer diagram is for? I've never encountered such a thing in a transformer diagram.


I'm trying to figure out the how the colors correspond to this sowter model (for use in the drip 'four seven' rev1 pcb, which only provides the sowter color codes on the pcb and not a schematic):

Thanks in advance for any hints!

Electrostatic shield.  Tie it to chassis ground.

Best,

Bri

Thanks Bri!

if the Sowter has an e shield, it is probably connected to the the core or metal can if it has one, so it gets grounded when and where the transformer get mounted,

the extra black wire is probably for fussy people who may not want to ground the shield at the same place as the transformer, thinking that the black wire should go somewhere else, like directly to a chassis ground where the green wire goes, or maybe to an audio star  ground,

what is on the shield?

noise?

maybe, but the main function is to reduce capacitance between windings,
it breaks up the natural capacitor formed by the pri-insl-sec, no e field can exist between the pri and sec if there is a zero potential conductor between the windings,

won't this prohibit signal from passing from pri to sec?

no, because the tranny works on the mag field, which is not disturbed by the e shield which works on the electric field, so the mag field gets to the core which means it is coupled to the sec winding, even with the e shield, 

now the e shield on the pwr trans, that is for noise, or longitudal (sp) currents, which is another fancy name for noise, or excess baggage that you do not want to claim,


well, it is

It's a Faraday (or Electrostatic) shield.  As CJ said, it's for eliminating most of the interwinding capacitances.
The generally accepted connection for it is Audio ground so in the case of your Cinemag, the hookup would be to tie Black (9) to Green (11).
White (10) would go to chassis.

Jean Clochet said:

The generally accepted connection for it is Audio ground so in the case of your Cinemag, the hookup would be to tie Black (9) to Green (11).
White (10) would go to chassis.

Click to expand...

I don't know if it's "accepted", but it should really go to chassis. No advantage in discharging EMI/RFI into audio gnd.

abbey road d enfer said:

Jean Clochet said:

The generally accepted connection for it is Audio ground so in the case of your Cinemag, the hookup would be to tie Black (9) to Green (11).
White (10) would go to chassis.

Click to expand...

I don't know if it's "accepted", but it should really go to chassis. No advantage in discharging EMI/RFI into audio gnd.

Click to expand...

Just to clarify, we're not talking about the outer can which sheilds EMI.
In quite a few tome's on this stuff, it's suggested that Electrostatic shields for reducing inter-winding capacitance in audio transformers should go to audio ground.  To name one that you may have a copy of;  the chapter in Audio 'Cyclodedia on grounding and wiring has a relevant section where it discusses what they call "Transmission Grounding". 

Of course, YMMV. 



Edit:  Here's a Jensen application note for a 990 op-amp mic pre showing what I'm talking about:

http://www.jensen-transformers.com/as/as083.pdf



 

Hey CJ, Jean and Abbey,
thanks for the sage advice. I'll experiment to see which ground point is quieter in the Drip board design.
I should also re-read the transformer chapter in the Audio Cyclopedia and hope that the knowledge doesn't deflect off my brain into outer space this time.

Jean Clochet said:

abbey road d enfer said:

Jean Clochet said:

The generally accepted connection for it is Audio ground so in the case of your Cinemag, the hookup would be to tie Black (9) to Green (11).
White (10) would go to chassis.

Click to expand...

I don't know if it's "accepted", but it should really go to chassis. No advantage in discharging EMI/RFI into audio gnd.

Click to expand...

Just to clarify, we're not talking about the outer can which sheilds EMI.

Click to expand...

That is absolutely clear to me. The can shields radiated EMI, but does nothing against conducted EMI that appears as common mode voltage on the primary of an input xfmr.

In quite a few tome's on this stuff, it's suggested that Electrostatic shields for reducing inter-winding capacitance in audio transformers should go to audio ground.  To name one that you may have a copy of;  the chapter in Audio 'Cyclodedia on grounding and wiring has a relevant section where it discusses what they call "Transmission Grounding". 

Click to expand...

I'm not familiar with this book; how old is it? For many years, not much distinction has been made between chassis ground and audio ground, although RF engineers certainly knew the difference.
I agree that inter-stage xfmrs can have their e-shield connected to audio gnd because they are not submitted to incoming EMI/RFI, but input and output xfmrs are.

Edit:  Here's a Jensen application note for a 990 op-amp mic pre showing what I'm talking about:

http://www.jensen-transformers.com/as/as083.pdf

Click to expand...

Well, to me it just doesn't make sense.

now if you leave that black wire ungrounded, what happens to leakage inductance?

abbey road d enfer said:

Jean Clochet said:

Audio 'Cyclodedia on grounding and wiring has a relevant section

Click to expand...

I'm not familiar with this book; how old is it? For many years, not much distinction has been made between chassis ground and audio ground, although RF engineers certainly knew the difference. 

Click to expand...

The book is old but does make a distinction -  the main point of the chapter is to highlight and show the difference and then to explain the application of a correct scheme. 

I do find it interesting that some of the Jensen schematics aren't consistent,  some showing both Faraday and can tied to the audio ground symbol.  Generally, it looks like the older schematics and app. notes (drawn by Jensen himself or Whitlock?) show the E.S.S. connected to audio but that the later ones are over the map.   

Thread diversion: While I don't remember the model number, one particular Jensen I measured years ago had better measured response when both the primary and secondary had the polarity flipped.  It could have been an isolated anomoly but the capacitance to ground measurements of the hi-impedance secondary bore out the reasons for doing this.

Anyway, you have to wonder why some transformer manufacturers bring out the Faraday seperately rather than just tie it to the can internally?
I do note however that all of the app. notes on Cinemag's site also show the black Faraday lead tied to audio while the can is tied to chasssis:
http://cinemag.biz/application_notes/application_notes.html

As always, YMMV.

Edit:  Also worth noting that the actual diagram the O.P. asked about shows the black wire connected to audio ground and the white can to chassis. 

CJ said:

now if you leave that black wire ungrounded, what happens to leakage inductance?

Click to expand...

I will defer to you as I have no doubt you've taken apart and examined more transformers than I have 

I wouldn't suggest leaving the black wire un -connected, but, I would say that by having a Faraday in the first place, you've slightly increased the amount of un-coupled flux or leakage inductance.  But would the amount of coupling or inductance change at all depending on the connection or not of that wire? 

Jean Clochet said:

abbey road d enfer said:

Jean Clochet said:

Audio 'Cyclodedia on grounding and wiring has a relevant section

Click to expand...

I'm not familiar with this book; how old is it? For many years, not much distinction has been made between chassis ground and audio ground, although RF engineers certainly knew the difference. 

Click to expand...

The book is old but does make a distinction -  the main point of the chapter is to highlight and show the difference and then to explain the application of a correct scheme. 

I do find it interesting that some of the Jensen schematics aren't consistent,  some showing both Faraday and can tied to the audio ground symbol.  Generally, it looks like the older schematics and app. notes (drawn by Jensen himself or Whitlock?) show the E.S.S. connected to audio but that the later ones are over the map.

Click to expand...

Well, they had to make an effort to be consistent with the "Pin 1 guys"; I think Bill Whitlock has been one of them.

 
Thread diversion: While I don't remember the model number, one particular Jensen I measured years ago had better measured response when both the primary and secondary had the polarity flipped.  It could have been an isolated anomoly but the capacitance to ground measurements of the hi-impedance secondary bore out the reasons for doing this.

Click to expand...

Not really uncommon; one of the members here has been fighting with a Spectrasonics channel that has frequency response inconsistencies (depending on the presence or not of an L-pad) because of that.

Anyway, you have to wonder why some transformer manufacturers bring out the Faraday seperately rather than just tie it to the can internally?
I do note however that all of the app. notes on Cinemag's site also show the black Faraday lead tied to audio while the can is tied to chasssis:
http://cinemag.biz/application_notes/application_notes.html

Click to expand...

You may note that the block that represents the electronics is connected to audio gnd and chassis gnd. I think the person who made the drawing has concentrated on the protection aspect, not much on the rest.

leakage inductance will go down if the shield is left unconnected.

it will go down even further if the shield is left out completely,

why?

because the wires can get closer to the core without that extra layer in the way,

so in this app, moving coil cart, you want to lose as little of that feeble signal as possible, so a transformer wired for lowest pri ohms and e shield left ungrounded...

you can even use a-10 wired for 35 ohms or something like that,

Hoyer told me that trick, the schematic does not show the strapping for the lowest ohms possible,

e shield - response goes up, leackage inductance also goes up,

you trade some coupling for some high end,

what does coupling loss mean?

signal loss.

compromise?

sure, dead folks were smart,

wind a copper wire shield, saves time and money since the wire is right there, already on the winding machine, no foil to mess with, just spin on one layer and tap a lead, you get a partial e shield,

i remember unwinding my first e shield,

a wire on one end, nothing on the other end!

no broken lead or nothin, wtf, over?
then the lite bulb went on after a few days, 

wanna get tricky use wire that has a pitch width equal to 1/4 wavelength of the highest freq. encountered, but thats an rf thing so forget it, 

CJ said:

leakage inductance will go down if the shield is left unconnected.

Click to expand...

I don't see how connecting or not the e-shield can modify the leakage inductance. It certainly modifies the capacitive coupling, the resulting effect may look like a reduction in leakage inductance but technically, it isn't.
And leaving an e-shield unconnected is definitely a waste of resources.

it will go down even further if the shield is left out completely,

Click to expand...

Indeed. Unfortunately, in most professional applications, an e-shield is a necessity.

so in this app, moving coil cart, you want to lose as little of that feeble signal as possible, so a transformer wired for lowest pri ohms and e shield left ungrounded...

Click to expand...

I really don't understand that. If you want the best performance, you start with no e-shield, If you want EMI protection, you connect the shield.
Increasing capacitive coupling works well with moderate ratio xfmrs (bi and quad-filars are very good at it) but with a 1:30 xfmr the difference cannot be more than 0.3dB at the highest frequency.

Hi CJ,
Hmm?  Like Abbey, I also don't understand how, whether an e.s.s. is connected or not changes leagage L. 
As I said earlier, I understand completely how the presence of one, or not, in the first place does so - with a shield there's less coupling due to the extra distance required for the shield.

If I draw in all the C's, R's, and L's on an ideal transformer, I can't get it to compute. 
Care to explain further for my addled brain? 

yes, but the effect is small compared to leaving out the shield completely,

copper foil is not perfect, there are impurities

so instead of being completely non ferrous and thus non reacting to the mag field,

you have in effect, a metal lamination in between the primary and secondary.

so you have the flux from the core getting to the first winding ok,

but it has to go through the slightly magnetic copper foil to get to the next winding.

so if that internal lamination is grounded, any flux reaching it gets shunted to ground,

if it floats, then the mag field will pass through it,

one side of the foil can be positive and the other side negative if it floats, which means the foil can induce a field from one side to the other,

ok, next point,

"e shields are needed for professionals" or something like that, 

UTC A-10 and HA 100 X have no shields, they are humbuckers,

2 coil construction for common rejection,

many other transformers have no shields, depends on the price, the type of transformer, and the construction.

API ouncer input-no shield.other ouncers do have shields,

many high quality bridging transfomers have shields, tamura, jbl/peerless, etc

what was the last beef?

???

CJ said:

copper foil is not perfect, there are impurities...

Click to expand...

OK, thanks CJ.  Sounds reasonable to me.

CJ said:

what was the last beef?

Click to expand...

No beef at all    How about a bacon butty?

Imho the electrostatic shield drain wire can end up carrying a variety of signals - so it is up to the user/designer to determine where in the 'amp' they want the drain signal currents to flow.  If input side emi is substantial then the preference is probably a low impedance connection to chassis.  If there is significant noise signal between chassis and audio ground, then the preference is probably a connection to audio ground, otherwise that noise signal will be between shield and secondary winding.

I think that's a good argument right there.
My preference would be to not use a transformer as a filter (other than its inherent freq. resp. limits of course!) and to eliminate input side interferance before it hits the primary but I realise this isn't always 100% practical.
I've been able to make some odd things happen when e.s.s.'s are connected to chassis/earth rather than audio common (when part of larger systems such as a desk) but I suppose  it's very possible one could go through life and notice no difference at all wherever it's connected.

Cheers.