Desktop speakers

[stefannxd]

I am looking for some desktop speakers, i have a Fostex HP-A3 amp dac. I would like some advise which speakers to go for and how i could connect them to my amp dac.

 

Thanks in advance

[OfficialTechSpace]

What's your budget? That's step 1.

 

The HP-A3 can output optical and RCA. Outputting to powered monitors, either would be fine. Just don't rely on it as a power amplifier, it's just a DAC for the most part.

[stefannxd]

My budget would be around 300 euro.

[OfficialTechSpace]

21 minutes ago, stefannxd said:

My budget would be around 300 euro.

Mackie MR524's. They're about ~$150 a piece, times two. So $300.

 

Also beat out everything else in that price category, have since 2017.

[stefannxd]

Thank you for the recommendation, what is the best way to connect them to the DAC?

[OfficialTechSpace]

1 hour ago, stefannxd said:

Thank you for the recommendation, what is the best way to connect them to the DAC?

Being that the MR524's don't have an option for optical inputs, on this DAC we're then limited to RCA's. This should be fine, but being that your DAC only has unbalanced outputs, interference is a possibility. Combatting this would only require an audio isolation module, and I run two of them in my personal setup.

 

I use the RDL AV-HK1 'Hum Killer', but there are cheaper options available. I've only tried this model, but I'm sure a typical $10-$20 variant off amazon or wherever should do most of the same.

 

[stefannxd]

1 hour ago, OfficialTechSpace said:

Being that the MR524's don't have an option for optical inputs, on this DAC we're then limited to RCA's. This should be fine, but being that your DAC only has unbalanced outputs, interference is a possibility. Combatting this would only require an audio isolation module, and I run two of them in my personal setup.

 

I use the RDL AV-HK1 'Hum Killer', but there are cheaper options available. I've only tried this model, but I'm sure a typical $10-$20 variant off amazon or wherever should do most of the same.

 

If i only do 2 speaker setup do i need 1 or 2 hum killers?

 

And is this a right cable to connect the speakers to the hum killer?

https://www.amazon.de/dp/B072V12JWV/ref=cm_sw_r_cp_api_i_9S7673W6CRPEFECZJ4XY?_encoding=UTF8&psc=1

[H713]

1 hour ago, OfficialTechSpace said:

Being that the MR524's don't have an option for optical inputs, on this DAC we're then limited to RCA's. This should be fine, but being that your DAC only has unbalanced outputs, interference is a possibility. Combatting this would only require an audio isolation module, and I run two of them in my personal setup.

 

I use the RDL AV-HK1 'Hum Killer', but there are cheaper options available. I've only tried this model, but I'm sure a typical $10-$20 variant off amazon or wherever should do most of the same.

 

Only if absolutely necessary, and then you spend as much on it as you can afford. The $10 - $20 variant off Amazon is a bad idea. 

 

These are audio transformers, and designing a good audio transformer is neither easy nor cheap. A good one, like those by Jensen will be $50 for just the transformer alone. If it's a Lundahl, double that. Linear, wideband isolation is expensive, no way around it. I've been searching for an economical way to do linear isolation with good bandwidth for years, and so far have turned up nothing. Cheap transformers aren't very linear, and even the expensive ones aren't as ideal as one might hope. Isolation amps aren't great either, and you'll pay through the nose for them.

 

Really cheap transformers will tend to create a lot of low frequency distortion, and in some cases a fair bit of HF attenuation (and definitely HF phase shift). That low frequency distortion can be a good thing (think Neve 1073), but this isn't where you want that.

My stance is that unless you plan to shell out the money to do it properly, don't isolate. Most of the time, you can go unbalanced to balanced without any significant issues. If you look at the schematics for a balanced input on things like those monitors, tying pin 3 to pin 1 basically turns it into an unbalanced input. It's still noisier than an ideal unbalanced input due to the thermal noise from a couple additional resistors, but this is of no consequence in any well-designed gear. 

 

29 minutes ago, stefannxd said:

If i only do 2 speaker setup do i need 1 or 2 hum killers?

 

The answer is you start with none, and only add them if you absolutely need to. They are a problem solver, but unless you spend a good $200 - $300 on them, they'll come at a performance cost.

[stefannxd]

7 minutes ago, H713 said:

Only if absolutely necessary, and then you spend as much on it as you can afford. The $10 - $20 variant off Amazon is a bad idea. 

 

These are audio transformers, and designing a good audio transformer is neither easy nor cheap. A good one, like those by Jensen will be $50 for just the transformer alone. If it's a Lundahl, double that. Linear, wideband isolation is expensive, no way around it. I've been searching for an economical way to do linear isolation with good bandwidth for years, and so far have turned up nothing. Cheap transformers aren't very linear, and even the expensive ones aren't as ideal as one might hope. Isolation amps aren't great either, and you'll pay through the nose for them.

 

Really cheap transformers will tend to create a lot of low frequency distortion, and in some cases a fair bit of HF attenuation (and definitely HF phase shift). That low frequency distortion can be a good thing (think Neve 1073), but this isn't where you want that.

My stance is that unless you plan to shell out the money to do it properly, don't isolate. Most of the time, you can go unbalanced to balanced without any significant issues. If you look at the schematics for a balanced input on things like those monitors, tying pin 3 to pin 1 basically turns it into an unbalanced input. It's still noisier than an ideal unbalanced input due to the thermal noise from a couple additional resistors, but this is of no consequence in any well-designed gear. 

 

 

The answer is you start with none, and only add them if you absolutely need to. They are a problem solver, but unless you spend a good $200 - $300 on them, they'll come at a performance cost.

Thank you for your input, i am a total noob on this subject. Are cables like this the right ones to connect the speakers to my rca output? https://www.amazon.de/dp/B078JPMWTY/ref=cm_sw_r_cp_api_i_WT63RYPY0BG9MAC0HB4W?_encoding=UTF8&psc=1

[H713]

14 minutes ago, stefannxd said:

Thank you for your input, i am a total noob on this subject. Are cables like this the right ones to connect the speakers to my rca output? https://www.amazon.de/dp/B078JPMWTY/ref=cm_sw_r_cp_api_i_WT63RYPY0BG9MAC0HB4W?_encoding=UTF8&psc=1

You just need an XLR male to RCA female cable. Here's an example that will work fine for this.

https://www.fullcompass.com/prod/141362-sescom-sc6xr-xlr-to-rca-cable-6ft-black

 

I'd buy something from a reputable company. No need to get fancy, but I've seen my fare share of cold solder joints and even miswiring on cheap generic crap. An open ground or a pin 3 hot adapter (yes, they still exist... shockingly enough) can be a PITA to troubleshoot.

[cmndr]

Quick question as someone who hasn't fiddled much with active speakers much...

 

Any REAL reason to try to go for XLR over just straight RCA? Two "not bad" RCA cables should do the trick in theory since the target distance appears to only be 1.5 meters and this looks like it's for near field listening, which means quieter signals being output (a relative weak point for XLR).

[H713]

2 hours ago, cmndr said:

Quick question as someone who hasn't fiddled much with active speakers much...

 

Any REAL reason to try to go for XLR over just straight RCA? Two "not bad" RCA cables should do the trick in theory since the target distance appears to only be 1.5 meters and this looks like it's for near field listening, which means quieter signals being output (a relative weak point for XLR).

You could go with XLR to RCA adapters at the speaker, then use RCA cables, but there isn't really a cost advantage to doing so, and it adds one more connector in the chain. 

Some monitors also have an RCA input. You could use that as well, and it would likely work just as well. Saves an adapter.

 

The better solution is to use a device with balanced outputs and eliminate that miserable connector that is RCA from your setup.

[H713]

As for reasons to use balanced audio, the short answer is that it greatly reduces the issue of EMI and ground loops, and it's especially useful for long cable runs. 

 

The reason why it does this is well-documented by Bill Whitlock

https://www.jensen-transformers.com/wp-content/uploads/2014/08/generic-seminar.pdf

As a warning, there are multiple ways of looking at this (but spoiler alert, they all boil down to the same physics), and all of them require at least some understanding of EM fields and magnetic induction. Grounding, ground noise and ground loops aren't an easy subject to understand, and it's a topic that baffles a lot of pretty smart engineers. 

 

 

[cmndr]

On 5/19/2022 at 10:20 PM, H713 said:

As for reasons to use balanced audio, the short answer is that it greatly reduces the issue of EMI and ground loops, and it's especially useful for long cable runs.

This a super short run though...

[H713]

Balanced audio is a bit of an awkward situation to explain, because there are multiple different configurations used, and the terms "balanced" and "differential" are often used interchangeably, when in fact they are not. 

 

The best and easiest way to think about this is the old (and arguably still one of the best) ways of doing this, which is with transformers. You have three conductors in the audio cable - two signal lines and a shield / ground. Strictly speaking, you can operate without the shield.

 

The two signal lines are 180\degree out of phase. The wanted signal, if the transformers are ideal, is the difference between the two lines. No signal is carried on the ground connection, and any signal common to two lines gets ignored. Since these two signal lines are almost always right next to each other (usually twisted together, often in a twisted-quad configuration), and have exactly the same impedance to ground, interference will be common to both signal conductors and gets completely ignored. Ground noise also gets ignored, because the ground isn't one of the signal lines.

 

Now let's look at a standard unbalanced connection. One side of the output is tied to the system ground of the source, and is probably tied to mains earth. The other side is driven by an amplifier with some finite output impedance. In audio it usually ranges from less than 10 $\Omega$ to about 600 $\Omega$. In instrumentation and RF, it is almost always 50 $\Omega$. 

 

An input has one side (the shield) tied to system ground, and thus probably to mains earth. The other has some finite input resistance, ranging from about 5 k$\Omega$ to over 1 M$\Omega$ for audio equipment. In RF gear, it is almost always 50 $\Omega$ to avoid reflections with a 50 $\Omega$ coaxial transmission line like RG58. 

 

We've got several things going on. By nature, the wanted signal is the signal present on the center conductor (in a coaxial configuration) with respect to ground (the shield). There are multiple problems here. Firstly, we've got two devices, with their enclosures tied to mains earth as NEC requires for safety on equipment that isn't double-insulated. No guarantees that they are plugged into the same outlet, and in some cases they might be on completely separate circuits, so the two enclosures are electrically tied together by a pretty long run of wire.

 

But... when we plug in our interconnect, we've also connected those two enclosures via the shield of the interconnect. We've now got a big wire loop connecting our two systems, and we all know from high-school physics (Faraday's law) that a changing magnetic field passing through a loop of wire results in an EMF. 

 

Ultimately, however, what matters is the impedances to ground. In a balanced system, those impedances are equal. The induced noise then becomes common-mode and is canceled out in the receiver. This is not the case in an unbalanced system, so that induced noise gets amplified and reproduced by the audio system. 

 

If this all seems confusing, read Bill Whitlock's article. It's a lot to cover in a forum post, and some people can explain this in a much more coherent way than me. 

 

 

 

[stefannxd]

On 5/19/2022 at 10:58 PM, OfficialTechSpace said:

Being that the MR524's don't have an option for optical inputs, on this DAC we're then limited to RCA's. This should be fine, but being that your DAC only has unbalanced outputs, interference is a possibility. Combatting this would only require an audio isolation module, and I run two of them in my personal setup.

 

I use the RDL AV-HK1 'Hum Killer', but there are cheaper options available. I've only tried this model, but I'm sure a typical $10-$20 variant off amazon or wherever should do most of the same.

 

I have my mackies in, but they are making this sound, is there any way to fix this?

 

[OfficialTechSpace]

14 minutes ago, stefannxd said:

I have my mackies in, but they are making this sound, is there any way to fix this?

 

IMG_5358.MOV

Yes, of course. I had the same issue, it's prevalent with any similar setup. It basically just means you're getting some level of ground loop interference, or something else.

 

What fixed it for me was buying an audio isolation module. They can be had for pretty cheap, but the one I opted for was an RDL AV-HK1. 

[stefannxd]

7 minutes ago, OfficialTechSpace said:

Yes, of course. I had the same issue, it's prevalent with any similar setup. It basically just means you're getting some level of ground loop interference, or something else.

 

What fixed it for me was buying an audio isolation module. They can be had for pretty cheap, but the one I opted for was an RDL AV-HK1. 

Okay I could do that, but I have my dac connected via xlr to rca and my record player with rca to rca. Do i need 1 or 2 of the audio isolation modules?

 

[OfficialTechSpace]

4 minutes ago, stefannxd said:

Okay I could do that, but I have my dac connected via xlr to rca and my record player with rca to rca. Do i need 1 or 2 of the audio isolation modules?

 

You could use 1 and swap back-and-forth, or get two. I have two, one for my Mackies, and one for my headphone rig.

[stefannxd]

5 minutes ago, OfficialTechSpace said:

You could use 1 and swap back-and-forth, or get two. I have two, one for my Mackies, and one for my headphone rig.

Thank you for the advise, I have ordered the audio isolation modules. Will report back when i get them in

[Heats with Nvidia]

Another way to reduce this noise, is to make sure everything that is electrically connected to the speaker and to what is connected to the speaker (in this case the computer, display etc...) are connected to the same power outlet with rather short cables. The smaller the ground loop, the better. But this sounds a lot more like power supply interference. Try to move the cables, maybe there is a power brick next to the audio cables.

 

Another possible source of this problem might be a bad ground in the audio cable or a problem with your dac.

 

These isolation modules should be more of a last resort thing because they usually change the sound quite a lot, especially the cheaper ones.

[cmndr]

Noise usually isn't an issue in a well set up system...

Also converting RCA to something balanced for 5' is... REALLY fixing a problem that doesn't exist. Those funds should be spent on room treatment first since reflections off the walls (possibly the desk) are going to literally have hundreds or thousands of times the impact as measured by something like REW.

 

If you do have bad cables, sure, it can be audible (bought a coax cable from Best Buy once... nope awful terrible buzzing, 40 year old RCA cable not technically specced for the use case fixed the buzz)

[H713]

Sigh... I wouldn't have ordered those isolators. They may or may not solve your problem (and even if they do, maybe not for the reasons you think), and they aren't cheap. I do not know why people are so quick to jump to the use of transformer isolators, and I really don't know why people are so quick to immediately recommend them the moment someone has a noise issue. Use your head, not your wallet. 

 

To be clear, they have their place... but they should be a last resort, not the first thing you try. People often use these things as a brute-force solution to a problem caused by doing something idiotic, like creating a massive ground loop the size of a building for a 50 W amplifier three feet from its source. 

 

Isolation has its place. There are times when you're painted in a corner, and you have no choice because of where things are located and what they must be powered from (big amplifiers can be a PITA in this regard). Sometimes one of those "grounds" might be at 3 kV for 500 ns. Sometimes you're dealing with medical or chemical systems where there are strict safety policies demanding isolation. These are all places where it makes sense, but that's a separate issue.

 

Keep in mind, even a $300 audio transformer sucks compared to a $3 op-amp for line-level applications. They aren't very linear, and they are inherently bandwidth limited. To be clear, a good audio transformer will still sound fine, but you're going to pay through the nose for it. Expect to spend $100 (minimum) for the transformers alone, and probably $200+ for an off-the-shelf isolator.

 

This is especially true because this doesn't sound like a normal ground loop to me. It sounds for all the world like switching noise on your ground. First, before anything else, what happens if you disconnect the audio input from the monitors? Does the sound go away? Assuming it does, first make sure you've tackled power and grounding in a reasonable manner. Start by having everything plugged into the same power source (same power strip is fine) with power cables no longer than necessary. This might solve your problem.

 

Next, minimize your setup. One monitor (and try swapping monitors), maybe even try running your DAC from a laptop that's running off battery. I have seen this problem caused by just about everything under the sun. Low-quality computer power supplies, failing power supplies for a computer, failing monitor power supplies, bad USB devices, gaming GPUs (I have yet to see it on a Quadro though), knock-off power bricks being used with an audio device, DACs that do not have USB isolation (digital isolation makes sense here and is built into the DAC most of the time), etc, etc. You want to try and figure out what device is causing this.

 

Now take a look at your audio cables? Are you being sensible and using decent-quality RCA to XLR cables, or are you using 50 feet of untwisted doorbell wire soldered to a BNC? An open ground on one of your cables can cause issues like this, as can an unshielded non-coaxial cable (though that wouldn't be my first bet). 

 

If you're in a really high-noise environment, you could try throwing some ferrites on the cables. A simple clip-on bead isn't a bad place to start, but in some cases I've needed  something beefier, even as much as 15 turns around a 3" toroid. When it comes to RF, any and all I/O is an input. That includes your power cord. I've had to go to some fairly extreme measures in this regard (big ferrites on ALL cables) around big RF power sources (think linear amplifiers) and trigatrons. You'll want to pay attention to the mix used here as well - two identical-looking ferrite cores can have wildly different properties.

 

Now before you spend a bunch of money on a transformer to convert unbalanced to balanced, look at the price of a good DAC with balanced outputs like the Schiit Modius or a Focusrite Scarlett. There may be a cheaper way to solve this problem.

 

A final note on noise / grounding / interference issues is that it does not always make sense. By the time you take all the relevant external and parasitic factors into account, you are dealing with an extremely complicated system. Even for a really good RF engineer, grounding and interference problems can (and often do) require a good bit of trial-and-error. These issues can be caused by all sorts of bizarre things, and it's really easy to chase your tail. Throwing a $200 isolator at a problem caused by a dying switching power supply isn't wise, and it isn't solving the problem.

 

 

 

 

 

[OfficialTechSpace]

9 hours ago, H713 said:

Sigh... I wouldn't have ordered those isolators. They may or may not solve your problem (and even if they do, maybe not for the reasons you think), and they aren't cheap. I do not know why people are so quick to jump to the use of transformer isolators, and I really don't know why people are so quick to immediately recommend them the moment someone has a noise issue. Use your head, not your wallet. 

 

To be clear, they have their place... but they should be a last resort, not the first thing you try. People often use these things as a brute-force solution to a problem caused by doing something idiotic, like creating a massive ground loop the size of a building for a 50 W amplifier three feet from its source. 

 

Isolation has its place. There are times when you're painted in a corner, and you have no choice because of where things are located and what they must be powered from (big amplifiers can be a PITA in this regard). Sometimes one of those "grounds" might be at 3 kV for 500 ns. Sometimes you're dealing with medical or chemical systems where there are strict safety policies demanding isolation. These are all places where it makes sense, but that's a separate issue.

 

Keep in mind, even a $300 audio transformer sucks compared to a $3 op-amp for line-level applications. They aren't very linear, and they are inherently bandwidth limited. To be clear, a good audio transformer will still sound fine, but you're going to pay through the nose for it. Expect to spend $100 (minimum) for the transformers alone, and probably $200+ for an off-the-shelf isolator.

 

This is especially true because this doesn't sound like a normal ground loop to me. It sounds for all the world like switching noise on your ground. First, before anything else, what happens if you disconnect the audio input from the monitors? Does the sound go away? Assuming it does, first make sure you've tackled power and grounding in a reasonable manner. Start by having everything plugged into the same power source (same power strip is fine) with power cables no longer than necessary. This might solve your problem.

 

Next, minimize your setup. One monitor (and try swapping monitors), maybe even try running your DAC from a laptop that's running off battery. I have seen this problem caused by just about everything under the sun. Low-quality computer power supplies, failing power supplies for a computer, failing monitor power supplies, bad USB devices, gaming GPUs (I have yet to see it on a Quadro though), knock-off power bricks being used with an audio device, DACs that do not have USB isolation (digital isolation makes sense here and is built into the DAC most of the time), etc, etc. You want to try and figure out what device is causing this.

 

Now take a look at your audio cables? Are you being sensible and using decent-quality RCA to XLR cables, or are you using 50 feet of untwisted doorbell wire soldered to a BNC? An open ground on one of your cables can cause issues like this, as can an unshielded non-coaxial cable (though that wouldn't be my first bet). 

 

If you're in a really high-noise environment, you could try throwing some ferrites on the cables. A simple clip-on bead isn't a bad place to start, but in some cases I've needed  something beefier, even as much as 15 turns around a 3" toroid. When it comes to RF, any and all I/O is an input. That includes your power cord. I've had to go to some fairly extreme measures in this regard (big ferrites on ALL cables) around big RF power sources (think linear amplifiers) and trigatrons. You'll want to pay attention to the mix used here as well - two identical-looking ferrite cores can have wildly different properties.

 

Now before you spend a bunch of money on a transformer to convert unbalanced to balanced, look at the price of a good DAC with balanced outputs like the Schiit Modius or a Focusrite Scarlett. There may be a cheaper way to solve this problem.

 

A final note on noise / grounding / interference issues is that it does not always make sense. By the time you take all the relevant external and parasitic factors into account, you are dealing with an extremely complicated system. Even for a really good RF engineer, grounding and interference problems can (and often do) require a good bit of trial-and-error. These issues can be caused by all sorts of bizarre things, and it's really easy to chase your tail. Throwing a $200 isolator at a problem caused by a dying switching power supply isn't wise, and it isn't solving the problem.

 

 

 

 

 

Nice novel, are you gonna publish that? Lol. I've heard Penguin Random House is pretty easygoing.

[H713]

On 5/28/2022 at 12:16 PM, OfficialTechSpace said:

Nice novel, are you gonna publish that? Lol. I've heard Penguin Random House is pretty easygoing.

Grounding, ground loops and EMI is a subject that has seen countless research papers and textbooks. Most of them are incomplete. There is no "1 paragraph solution" to what can be a very complicated problem. Trying to summarize the topic into one post is nearly impossible. My 3:00 AM post probably could have been more coherent, but I still cut out a lot of information. 

 

I have spent weeks on end troubleshooting ground loops and EMI issues. Even the troubleshooting procedure could fill a suitably long application note. 

 

I would love to see your one-post explanation on how to troubleshoot noise/interference/grounding issues.