ProSoundWeb Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: 1 [2] 3  All   Go Down

Author Topic: Wharfdale EQ specifies balanced w/ ground disconnected on one connector?!?!  (Read 12035 times)

John Roberts {JR}

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 17176
  • Hickory, Mississippi, USA
    • Resotune

I don't recall, the exact moment I got religion about pin one, but any pro audio product design engineer who hasn't been living under a rock for the last decade or more "should" know about it by now. That said there's always new engineers coming along to repeat old mistakes.   

JR

PS: Maybe look at publication dates on some of the articles talking about "Pin 1 problem" for a ballpark time period for broad general awareness..
Logged
Cancel the "cancel culture". Do not participate in mob hatred.

Rich Barrett

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 51

There is no simple tweak for legacy gear with pin one problems. Often there are layout and even circuit design issues, not very easy to modify after production.

For now just follow the advice of the manufacturer, and if the product sounds bad, try the easy options, and if nothing easy works consider replacing with something better designed.

As hard as it might be to sell poorly designed legacy gear it will be even harder to sell after you drill holes in it and experiment with grounds.

JR

Well I read some of the tech papers concerning this provided by pin1problem.com and found this in the brief located call " Jensen Paper on Proper Grounding";

*******************************************************************************************************
Figure 12 shows the commonplace pro-audio design practice of bonding cable shields to an input or output amplifier’s 0V reference. Ground loop routed from the outside world and flowing in the finite impedances of a sensitive audio circuit’s 0V system cause noise voltages to arise in circuits’ references, making their audio signals noisy. This problem is known in the EMC world as ‘common-impedance coupling’, and it would not happen if cable shield currents were connected directly to the chassis/frame/enclosure of the equipment, as is best for EMC.

An example: 12mm (˝”) length of a 2.5mm (0.1”) wide PCB trace in ‘1oz copper’ has a resistance of 2.4mW. If this short trace carries shield (ground loop) currents the error voltage its resistance causes in the 0V would be 2.4mV per amp of shield current. 2.4mV is a very high noise level indeed (-50dBu in pro-audio terms), yet the shared length of 0V trace in this example was a mere 12mm. A shared 0.1” wide 0V trace would need to be 0.05mm long for an error voltage of 10µV/A, and much shorter for 10nV/A. Don’t forget that an error in a 0V path has exactly the same effect as the same error voltage in all the signals that use that 0V path for their return currents or reference voltages.

Another traditional pro-audio equipment design technique is to use a ‘chassis’ trace (or wire) to collect all the shield pin connections together and eventually connect them to the chassis. These ‘chassis ground’ traces are isolated from 0V but are almost always routed close to a number of parallel signal and 0V traces (or wires) and magnetically couple with them, giving rise to noise voltages. This problem is known as magnetic crosstalk (or stray mutual inductance coupling) and it would be practically eliminated if cable shields were connected directly to the chassis/frame/enclosure at the point where their cable enters the equipment, as is also best for EMC. Magnetic crosstalk can also occur from the ground currents flowing in the mains lead’s protective ground conductor, so this should be short and routed well away from any signal conductors – better still bonded directly to the chassis/frame/enclosure without using any internal conductors – for example by correctly installing a metal-bodied mains EMI filter (see [4] and [5]).

Sometimes cable shields are connected to a circuit’s 0V, but single-point grounding techniques are diligently applied to prevent common-impedance coupling. It has been shown above how difficult it is to achieve (say) 10µV per amp of shield current in such designs, but if we assume that this problem has been overcome we then find that the stray mutual inductance (magnetically-coupled crosstalk) between the branches of the 0V system, and between 0V branches and signal conductors, can be a very significant source of noise voltage injection.

One of the authors has made some measurements of the inductive coupling between a 100mm square loop and a long straight current-carrying conductor. Both were placed in the same plane (e.g. simulating crosstalk between conductors in a PCB) with the front edge of the square loop at various distances from the current-carrying conductor. The loop picked-up around 18µV per amp of 50Hz current in the long conductor when its front edge was at 9mm distance, 9µV/A at 21mm distance, and 6µV/A at 50mm. When a plate of 1mm thick steel was placed about 4mm away from the conductor and the loop, in parallel with their plane to simulate the base of an equipment under a PCB, the loop’s coupled voltages approximately doubled for each of the spacings between the loop and the long conductor. Closer spacing, or a loop in the current-carrying conductor that overlapped part of the sensing loop, would have increased inductive coupling. Inductive coupling between ground loops and other conductors inside an item of sensitive equipment can clearly be very troublesome.

[6] and [7] both show that the problems caused in pro-audio equipment by not bonding cable shields direct to the chassis/frame/enclosure of equipment were well known to some as long ago as 1995.[/i]
**************************************************************************************************

This paper, as well as others, say the easiest (and best) way to overcome is to terminate the shield directly to chassis and not through the ground plane in the circuitry? I think for right now I'll just take it one step at a time and just connect it up as per the EQ mfg's recommendation.

Thanks for the help guys!
Logged

Craig Montgomery

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 64

Ok, correct me because I'm probably wrong-

The O2R has unbalanced inserts anyway.  Whenever I've made cables to connect balanced gear to unbalanced inserts, I've just connected hot to pin 2, cold to pin 3, and floated pin 1.

(I'd yank all that stuff and just use the EQ in the console anyway ;D )
Logged

Rich Barrett

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 51

Ok, correct me because I'm probably wrong-

The O2R has unbalanced inserts anyway.  Whenever I've made cables to connect balanced gear to unbalanced inserts, I've just connected hot to pin 2, cold to pin 3, and floated pin 1.

(I'd yank all that stuff and just use the EQ in the console anyway ;D )

The inserts ARE NOT balanced, correct. But in the spirit of keeping the ground noise down I'm making sure to keep my unbalanced cabling to as short a distance as possible.

But the EQ mfg confused me by recommending NOT to connect the shield to both sides of the cable / devices even though their EQ is balanced in / out. That's why I posted ...

Yeah I got 4 band parametric through the 02R but after analyzing my vocals on a graph I have some wild frequency fluctuations under 4khz, the most being under 1khz, that I felt I needed a graphic eq to pull them down. My cohost has this rich, natural, smooth, bassy voice that in comparison has a very smooth waveform.

I picked up the eq for $40 and Wharfdale is a solid name in reference speakers so I figured why not? Even if my equipment is faulty and causing the problem, I can compensate with the EQ till I can afford new, excellent gear.
« Last Edit: October 19, 2013, 02:41:08 AM by Rich Barrett »
Logged

Tommy Peel

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1468
  • Longview, Texas

Though I'm not familiar with the 02R I suspect that with some creative soft and/or hard patching you could get more than 4 bands of para eq for your vocals which would eliminate the need for an outboard eq.

Sent from my Nexus 4 using Tapatalk

Logged

Steve M Smith

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 3381
  • Isle of Wight - England

Back in Ye Olden Days it was common practice to eliminate ground loop noise by disconnecting the ground connection from the mains plug.

This was the lazy and totally wrong way to do it.

Having the cable shield connected at one end only is the best way and each piece of equipment should be grounded through the ground connection on its plug.

When I make new guitar leads for myself I use twin core microphone cable and use one core for the ground to ground connection and only connect the shield at one end.  This way, any noise picked up by the shield is not in series with the guitar and does not present itself at the amplifier input as part of the signal.


Steve.
Logged

John Roberts {JR}

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 17176
  • Hickory, Mississippi, USA
    • Resotune

Just like i don't speculate about what other people hear, it is an equally slippery proposition to speculate about what and when they know stuff.

Ok, correct me because I'm probably wrong-

The O2R has unbalanced inserts anyway.  Whenever I've made cables to connect balanced gear to unbalanced inserts, I've just connected hot to pin 2, cold to pin 3, and floated pin 1.
nah... or not generally.

If the console is well designed the insert jack ground will be bonded to the console chassis. However if the insert send/return signals are being sent to balanced or differential inputs the  insert ground should be treated as both audio low and shield for interfacing with the external devices, so audio low, audio high, and shield all get connected.

Again some legacy gear may have issues with pin one grounded on both ends but modern well designed gear "should" not.
Quote
(I'd yank all that stuff and just use the EQ in the console anyway ;D )
JR

@Rich Yes "common impedance" or shared ground path is a well known phenomenon that can degrade signal integrity, especially for unbalanced consumer two wire interfaces. The spate of pin one articles a while back, spelled it out in simple terms for those slow to get with the program. Making the customers aware that the problem was design related and not bad venue mojo, put pressure on manufacturers to get their act together.
 
JR
Logged
Cancel the "cancel culture". Do not participate in mob hatred.

Rich Barrett

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 51


If the console is well designed the insert jack ground will be bonded to the console chassis. However if the insert send/return signals are being sent to balanced or differential inputs the  insert ground should be treated as both audio low and shield for interfacing with the external devices, so audio low, audio high, and shield all get connected.

The inserts and EQ are being connected together through the Ebtech LLC-8 line level shifter / noise eliminator which passively pairs unbalanced -10dbu with balanced +4dbm connections and vice versa and even senses TS vs. TRS connections. After posting this thread and all the responses I got I decided to send an email to Ebtech to see if it would make any difference, grounding-wise, if I had the shielding terminated on both ends. Because the Ebtech doesn't use AC power, there is no chassis ground through an electrical plug so I don't know what happens to the shield ground inside the box.

I asked for advice and a diagram - will post when I get an answer.   

Wait a sec, on second thought after reading what I just wrote I think the EQ was designed for connecting to mixer inserts ... on the insert cable, the shield is only connected to the mixer - the in/out leg is unbalanced. I don't need to use the Ebtech!!!! eureka!
« Last Edit: October 19, 2013, 12:22:22 PM by Rich Barrett »
Logged

Tim McCulloch

  • SR Forums
  • Hero Member
  • *
  • Offline Offline
  • Posts: 23743
  • Wichita, Kansas USA

The inserts and EQ are being connected together through the Ebtech LLC-8 line level shifter / noise eliminator which passively pairs unbalanced -10dbu with balanced +4dbm connections and vice versa and even senses TS vs. TRS connections. After posting this thread and all the responses I got I decided to send an email to Ebtech to see if it would make any difference, grounding-wise, if I had the shielding terminated on both ends. Because the Ebtech doesn't use AC power, there is no common ground through the electrical plug so I don't know what happens to the shield ground inside the box.

I asked for advice and a diagram - will post when I get an answer.

Perhaps I missed something, but why the Ebtech?  Do you currently have problems?  Did you try another (borrowed) EQ?

This looks like a big kludge to fix a problem in one device.  IF the EQ is the problem, sell it and buy something else.
Logged
"If you're passing on your way, from Palm Springs to L.A., Give a wave to good ol' Dave, Say hello to progress and goodbye to the Moonlight Motor Inn." - Steve Spurgin, Moonlight Motor Inn

Rich Barrett

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 51

Perhaps I missed something, but why the Ebtech?  Do you currently have problems?  Did you try another (borrowed) EQ?

This looks like a big kludge to fix a problem in one device.  IF the EQ is the problem, sell it and buy something else.

I have the Ebtech because I have a dual mp3 player made for DJ's with unbalanced outs so I picked it up to balance the line levels going into my previous mixer. Bear in mind, I'm new at pro audio rigs with a limited budget. I was thinking I would also need it to connect the unbalanced insert to the balanced EQ - I just realized I do not need it

I haven't used the mixer yet - making my cables today. Switchcraft TRS connectors and Mogami w2549 balanced cable
Logged

ProSoundWeb Community

Re: Wharfdale EQ specifies balanced w/ ground disconnected on one connector?!?!
« Reply #19 on: October 19, 2013, 12:27:41 PM »


Pages: 1 [2] 3  All   Go Up
 



Site Hosted By Ashdown Technologies, Inc.

Page created in 0.031 seconds with 25 queries.