Not joking at all: if you don't know what to measure and where and how and how it correlates to SQ, then you start calling things 'voodoo' or 'magic' because you can't comprehend it. Re-read the thread and see who used these terms here.
Then by all means, commercialise your own piece of kit. Manufacturers cut corners all the time, and it's sometimes not too bad because it allows for low-budget devices that many could not afford otherwise.
Read about the Regen and see how it's constructed and whether it's done properly.
Its a USB port mate, they have been around for ages and noise on them is nothing mystical, nor is measuring it. Stereophile have measured the device and found no difference, not me, though I am not surprised that they found no difference... so no difference in noise levels yet it still works when it is supposed to reduce noise, how....
Cant comprehend, have you read any of my previous posts, I have spent 30+ years doing PCB layouts for critical situations and always to minimise EMC, now you may not like my recommendations regarding this layout as you seem hell bent to defend it, but I do this for a living and advise quite a few high end electronic based companies on layout or go on site to do their layouts, to far more stringent EMC and noise levels than you find in commercial gear. So I can comprehend what I am on about. You and the others are just blindly following each other like little sheep, as someone has asked on the other thread, how can this device pluged in an unused port have an effect... answer me that one and if you can convince me I will change my view, but the answer has to be based on proved electronic theory (I believe I did mention current through a filter for it to work, there has to be a loop and current passing through the filter..........)
I am talking about the layout, it is bad, this is not cutting corners this is a layout that should have been done properly, there is no cost difference to manufacture a good layout or bad layout.
my design is not going to be commercialised, it was done to illustrate to someone HOW to design for noise immunity, it is an example that shows how it should be done, not some mickey mouse layout thrown down so they can sell a few units to the true believers, with no measurements or any other supporting data.
If anyone wants the Gerber data they can have it for free, its a DIY site and my information a etc. is supplied free.
I've read about the Regen I have not seen enough of the layout to comment on it, but your comment seems to question my ability to pass comment on layouts, sorry but I am very qualified to check and critique layouts as I have said previously and do it every day. I said I would look at the layout and comment if they want on where and if there are any EMC improvements to be made, but why keep bringing the Regen up, we are discussing the jitterbug.....
Cant comprehend, have you read any of my previous posts, I have spent 30+ years doing PCB layouts for critical situations and always to minimise EMC, now you may not like my recommendations regarding this layout as you seem hell bent to defend it, but I do this for a living and advise quite a few high end electronic based companies on layout or go on site to do their layouts, to far more stringent EMC and noise levels than you find in commercial gear. So I can comprehend what I am on about. You and the others are just blindly following each other like little sheep, as someone has asked on the other thread, how can this device pluged in an unused port have an effect... answer me that one and if you can convince me I will change my view, but the answer has to be based on proved electronic theory (I believe I did mention current through a filter for it to work, there has to be a loop and current passing through the filter..........)
I am talking about the layout, it is bad, this is not cutting corners this is a layout that should have been done properly, there is no cost difference to manufacture a good layout or bad layout.
my design is not going to be commercialised, it was done to illustrate to someone HOW to design for noise immunity, it is an example that shows how it should be done, not some mickey mouse layout thrown down so they can sell a few units to the true believers, with no measurements or any other supporting data.
If anyone wants the Gerber data they can have it for free, its a DIY site and my information a etc. is supplied free.
I've read about the Regen I have not seen enough of the layout to comment on it, but your comment seems to question my ability to pass comment on layouts, sorry but I am very qualified to check and critique layouts as I have said previously and do it every day. I said I would look at the layout and comment if they want on where and if there are any EMC improvements to be made, but why keep bringing the Regen up, we are discussing the jitterbug.....
Just to throw a brick into the pond
Releasing ODAC RevB | JDS Labs Blog
The ODAC had some known issues with USB, fixed 3 years on with a new USB receiver. This could be read many ways. I know the way I read it...
(Note for the small scale producer testing every permutation is hard. That is where being able to call on those who have suffered in the past helps).
Releasing ODAC RevB | JDS Labs Blog
The ODAC had some known issues with USB, fixed 3 years on with a new USB receiver. This could be read many ways. I know the way I read it...
(Note for the small scale producer testing every permutation is hard. That is where being able to call on those who have suffered in the past helps).
Who said it's 'mystical' or that measurements can't be done?
Look who's using the word 'mystical' first.
What I wrote is you need to know what to measure, where and how.
If you measure A, but the real measure is B which you avoid measuring because you don't yet know where the important thing to measure for SQ, then you'll be measuring the wrong thing and invoking 'voodoo', 'magic' and 'mystical' reasons.
If you do know, and know what to correlate to SQ then it's easy to further optimise and then you'll stop using 'voodoo', 'magic' and 'mystical'.
However, the assumption that something like the Regen is easy to measure (not talking about the JB here) is not warranted IMO: Swenson did say he'll have to add to his new scope some new measurement device that he will have to build himself.
The JB itself may be somewhat easier to measure. I say 'may' because Atkinson appears to say otherwise (haven't read the September Stereophile but saw some conversations with him and Rankin on FB as well as read other reports to the same).
Jitter and packet noise - sure, these come to mind, ground plane noise too.
But packet noise: you'll have to be more precise than that. Read what Swenson says about his Regen design for more info on what he found.
Not that I have anywhere the equipment needed for such detailed analysis, but even with a crude DIY probe circuit linked to the Mac running a software scope, I could see the 8kHz and harmonics noise of the USB cable + filter was toying with.
There's a comparison article on Audiostream for a set of these devices (although, arguably, the Regen does much more and has different goals) published today.
Lots of great things in both of these if you know where to look, who to trust, and which of the trolls to avoid. At least that's my experience.
That one is a total waste indeed.
You can read about on CA. Use the advanced search feature and look for posts by John Swenson and impedance matching for the Regen.
Not sure it will give you as much as 'looking at the layout' but it will give you an idea of the care taken in the design.
Also looks at the Wyrd review on Audiostream - there's a little controversy with the manufacturer comments - read the manufacturer comment and let us know what you thin about his claim of using a double-layer PCB is enough for the USB speed used as per EMC.
That would be fun.
I haven't mentioned measuring the Regen, apart from when you mention it.
The projects I work on (and all projects I have worked on) have been measured and measured again, its a different ball game with a lot of electronics outside of audio. They have to work otherwise big bucks are lost and businesses go bust or worse. This is not to say things are perfect, but if an interface is going to be worked on then they will get the kit to do the measurements, as well as EMC testing. If I read what you are saying the jitterbug would be hard to measure, so they just designed it and hoped for the best....
I cannot be bothered to search, but I am sure if you do a google search you'll find plenty on measuring noise on USB ports as well as measuring many other types of noise including noise on the ground plane (I once posted that one) so its might not be as hard to measure as you may think.
Now if you suspect USB noise is degrading the sound, where would you measure that noise? On the analogue side initially to confirm the noise is present, would be a start.
I have read all that regarding the Regen and impedance matching (by the way ALL the layout information for USB tells you to lay it out for 90R differential, and most designers do.)
I have also read the two layers Vs 4 layers stuff, and I'm on the side of 4 layers without any hesitation. Maybe you should look at my layer stack up picture that illustrates quite succinctly how I achieved a 90R differential layout for the USB lines on the PCB I put up as an example, and also how I further stated that for a 2 layer board I could only realise a sensible USB layout with a board 1mm thick. The other issue with a 2 layer board is you will not get an unbroken contiguous ground plane for the digital signals return path, this will increase the noise on the design and have a detrimental effect on signal integrity especially if you were going to use the top speed USB 2 is capable of.
I could wax lyrical about planes and planar capacitance, low impedance paths for decouplers etc etc, but thats it for now, I have more important things to worry about for the next couple of days.
Au revoir
I remember this, the regulator under rare circumstances would oscillate as they say, now call me cynical, but the ceramic cap was on the end of a USB cable so there was some parasitic cable resistance, inductance and capacitance as well (and from power supply integrity stuff, the inductance calms the resonant peaks in these sort of situations) so my take on it was, it went crazy when a more bespoke type of audiophile cable was used
I mentioned it because there are a few new devices which deal with the USB issues, the JB is just one of three new ones: the others are the Regen and the Wyrd.
What I am considering here is measurements for SQ in an audiophile setup. It is a particular subject.
I asked you if you measured the jitter of your circuit as compared to the one of a JB because you could think you're improving things by doing a better EMC (and I agree EMC is very important) but if you're not paying particular attention and your circuit adds a lot more jitter, then we wouldn't like that, would we?
So you pointed out the layout issues with the JB, and that's fine, I pointed you to another device which also deals with USB issues but with a much better layout (at least it seems to me they took particular care in the design). It so happens that the designer of this well-designed new device, had pointed out the jitter issues with the ADUM you seem to like so much in another post/thread/time.
But just because the JB has layout issues you could fix, it doesn't mean it doesn't work, and for many people, they can afford to try one or two in their setup for $49 each.
I agree.
Yeah funny how that happens. One manufacturer announces a product to fix an imaginary problem and others leap on the bandwagon and produce their own.
Not saying that there are not suboptimal products on ebay that might appreciate all the help they can get, but admitting that your own products were not designed properly might indicate a companies view of their punters.
Stop putting words into my mouth, please read exactly what I put, I tend not to be ambiguous.
I am getting a bit peeved...
The design I put up was an example, of layout technique, nothing more. GOT THAT.
The ADUm that I love, read what I put about the use of ADUm and isolation for voltage in industrial/medical applications, I have never said you need to use it for audio, not do I have any particular love for any device!
Every PCB should be designed with care....
This thread is about the jitterbug, no other USB device.
Does it work, we have no hard proof.
I am getting a bit peeved...
The design I put up was an example, of layout technique, nothing more. GOT THAT.
The ADUm that I love, read what I put about the use of ADUm and isolation for voltage in industrial/medical applications, I have never said you need to use it for audio, not do I have any particular love for any device!
Every PCB should be designed with care....
This thread is about the jitterbug, no other USB device.
Does it work, we have no hard proof.
Today I have tried the Jitterbug in an unused usb port as advertized by Audio Quest, no difference. Also tried it on other USB devices. My mouse worked, with my USB HD windows reported there was something wrong with the device with the jitterbug in place an that it needed replacement. Removed jitterbug, worked without problems. Tried it on my printer and the black is much blacker now, no offcoarse not. At least the printer was recognized and worked o.k.
Finally I tested the jitterbug feeding a Qtec USB hub. This did the trick. Everything behind the USB hub was recognized and worked. I haven't got a clue why. I could even use my USB PSU wich did simply not work in any case when used directly with the jitterbug. I'm testing it at the moment. But this is a total cable mess. First the Jitterbug in the USB port of the PC --> USB cable --> Usb Hub --> Usb Cable --> Usb PSU --> Usb cable --> DAC. That's a whole lot of connections just to connect my DAC to my PC, but hey we have all done crazier things in this audio hobby.
So, my conclusion soo far. This is not a plug and play device and does not work as advertized. In the case where it did make a slight improvement it wasn't completely hassle free either. It's a troublesome device. Maybe the addition of a USB HUB is going to make some difference. Don't know yet.
Finally I tested the jitterbug feeding a Qtec USB hub. This did the trick. Everything behind the USB hub was recognized and worked. I haven't got a clue why. I could even use my USB PSU wich did simply not work in any case when used directly with the jitterbug. I'm testing it at the moment. But this is a total cable mess. First the Jitterbug in the USB port of the PC --> USB cable --> Usb Hub --> Usb Cable --> Usb PSU --> Usb cable --> DAC. That's a whole lot of connections just to connect my DAC to my PC, but hey we have all done crazier things in this audio hobby.
So, my conclusion soo far. This is not a plug and play device and does not work as advertized. In the case where it did make a slight improvement it wasn't completely hassle free either. It's a troublesome device. Maybe the addition of a USB HUB is going to make some difference. Don't know yet.
Sounds like a business model. Make it cheap enough so that gullible audiophiles can afford to buy to just try if it works. Heck it works if I spent money on it. No way to make a blind listening test, I can clearly hear the improvement.
Hi.
My 2 cents:
I think the discussion here is kind of weird.
Fact is there's some filtering ongoing with the AQ Jitterbug. Obviously there are more filters (iFi iPurifier,...) out there. AQ is not the one who invented this. They just offer the cheapest one.
At Audio Asylum it was mentioned that the latest Hi-Fi News & Record Review came up with an article that clearly showed the impact on jitter, noise and a related eye diagram.
How such a filter affects a system depends on the system obviously.
In can work, it can be neutral, it can make things worse.
There might be plenty of filtering already applied inside your DAC. You'd basically overdampening the data stream by introducing another passive filter. This can make things worse. Or your USB quality is already that bad, that by introducing the filter
the red line is passed. On a poorly powered RPI that might happen sooner as you think.
In many cases though there is a fair chance that things do improve.
By introducing a hub or a USB Regen you do some reclocking and refreshing of the datastream. You also supply a usually better 5V to VBus -- if your need that. ( I cut the VBus for all DACs I have - In my case, it makes a difference - antenna effects, crosstalk,... - who knows)
It's been known and reported since years that an active hub improves USB DAC performance in many cases.
USB Regen just added a filter, better clock and a better PS to a standard hub.
Basically you won't really face the potentially intrusive performance of a passive filter.
People reported that daisy chaining a passive USB filter and the Regen improved things
even further. That makes sense too. You apply double filtering, but you reclock and refresh the signal with the active hub in between. This way you won't run into the overdampening trap. It's the other way around. Your filtering gets more effective.
The Regen btw also does some filtering by applying a small (a couple ohms) resistor to the ground to reduce the impact of nasty groundloops.
Just to mention it:
There's in interesting patented device from Wuerth for the DIY crowd: WE-EPLE USB 2.0 A Connector - Product Catalog Passive Components
The datasheet tells about the performance.
Bottom line:
All this is not a black or white issue. There are a lot of grey tones in between.
IMO the main problem are the actual DAC interfaces. The vast majority of them is not able to
properly cope with the noise and data jitter that comes in over USB.
Obviously it's all but sufficient to build an interface that follows basic XMOS implementation guidelines.
Dacs like the Soekris DAC and other high quality commercial DACs try to fight these issues. They introduce isolators and relock the data stream.
They usually do so on the I2S stream. By then the noise and other crap made it into your device already. That's probably why devices still show slight
changes by improving the USB side.
Cheers
My 2 cents:
I think the discussion here is kind of weird.
Fact is there's some filtering ongoing with the AQ Jitterbug. Obviously there are more filters (iFi iPurifier,...) out there. AQ is not the one who invented this. They just offer the cheapest one.
At Audio Asylum it was mentioned that the latest Hi-Fi News & Record Review came up with an article that clearly showed the impact on jitter, noise and a related eye diagram.
How such a filter affects a system depends on the system obviously.
In can work, it can be neutral, it can make things worse.
There might be plenty of filtering already applied inside your DAC. You'd basically overdampening the data stream by introducing another passive filter. This can make things worse. Or your USB quality is already that bad, that by introducing the filter
the red line is passed. On a poorly powered RPI that might happen sooner as you think.
In many cases though there is a fair chance that things do improve.
By introducing a hub or a USB Regen you do some reclocking and refreshing of the datastream. You also supply a usually better 5V to VBus -- if your need that. ( I cut the VBus for all DACs I have - In my case, it makes a difference - antenna effects, crosstalk,... - who knows)
It's been known and reported since years that an active hub improves USB DAC performance in many cases.
USB Regen just added a filter, better clock and a better PS to a standard hub.
Basically you won't really face the potentially intrusive performance of a passive filter.
People reported that daisy chaining a passive USB filter and the Regen improved things
even further. That makes sense too. You apply double filtering, but you reclock and refresh the signal with the active hub in between. This way you won't run into the overdampening trap. It's the other way around. Your filtering gets more effective.
The Regen btw also does some filtering by applying a small (a couple ohms) resistor to the ground to reduce the impact of nasty groundloops.
Just to mention it:
There's in interesting patented device from Wuerth for the DIY crowd: WE-EPLE USB 2.0 A Connector - Product Catalog Passive Components
The datasheet tells about the performance.
Bottom line:
All this is not a black or white issue. There are a lot of grey tones in between.
IMO the main problem are the actual DAC interfaces. The vast majority of them is not able to
properly cope with the noise and data jitter that comes in over USB.
Obviously it's all but sufficient to build an interface that follows basic XMOS implementation guidelines.
Dacs like the Soekris DAC and other high quality commercial DACs try to fight these issues. They introduce isolators and relock the data stream.
They usually do so on the I2S stream. By then the noise and other crap made it into your device already. That's probably why devices still show slight
changes by improving the USB side.
Cheers
Last edited:
The issue is simple. If Audioquest knew that their implementation was susceptible to noise on USB, then the filtering should have been there in the firefly from the get go. Selling another add on to get your box to the performance level it should have been at to start with shows a lack of proper design in the first place.
I don't think this discussion is so weird at all.
I'm the one who owns one of these and have been messing around with it for the last two weeks at my own place and at some friends places and all I can tell is that with 4 different desktop computers and 9 laptops of various kinds and with 7 USB DAC/Spdif converters in total there where only 2 occasions where it worked flawlesly. So maybe it is because Apple, HP, Asus, Acer, Fujitsu siemens, Shuttle and Averatec all supply there computers with shitty USB implementations and that either Nuforce, Apogee, RME, Focusrite etc provide their DAC's with shitty USB implementation. That's all possible, yes, but the Jitterbug is being advertised as an universal cure for these problems witch it ain't, period.
I'm the one who owns one of these and have been messing around with it for the last two weeks at my own place and at some friends places and all I can tell is that with 4 different desktop computers and 9 laptops of various kinds and with 7 USB DAC/Spdif converters in total there where only 2 occasions where it worked flawlesly. So maybe it is because Apple, HP, Asus, Acer, Fujitsu siemens, Shuttle and Averatec all supply there computers with shitty USB implementations and that either Nuforce, Apogee, RME, Focusrite etc provide their DAC's with shitty USB implementation. That's all possible, yes, but the Jitterbug is being advertised as an universal cure for these problems witch it ain't, period.
It's not that simple.
There's is the same designer (Gordon Rankin - Wavelength Audio) involved in the
Jitterbug and Dragonfly design.
Rankin was the guy who has been pushing hard on asynchronous USB being the holy grail of USB Audio. I think he was one of the first coming up with async USB Audio.
That's been his business case for years. In fact he probably sold thousands of licenses of his implementation. That very same implementation also went into the Dragonfly.
Over years the industry tried to talk down other issues associated to USB Audio. That's been my impression. That's IMO why there havn't been many improvements.
(Though it wasn't the industry only. I think I had similar discussions almost 10 years back in the Linux Audio The Way to Go thread - with phofman and other inmates oposing my standpoint)
Anyhow. Meanwhile there are quite some active and passive USB filters in the market.
(Almost) Nobody questions the mostly positive effect of these filters anymore.
For Audioquest it was a nice move to go ahead and offer the cheapest solution.
Beside being a nice and cheap upgrade for the Dragonfly, they also serve the rest of the market. I havn't seen a direct comparison between e.g. Jitterbug and iPurifier. I'd expect though that they don't differ much. If so, the iPurifier will have a hard time to stay in the market at double the price. AQ for sure shuffles the passive USB filter market for the moment.
I wouldn't be surprised if soon we'd see a competing active product to the Regen. The Schiit Wyrd lost that battle already as far as I see it.
But as I said. It IMO would be best to have the filtering done inside the DAC, Obviously the audio industry will charge the consumer an arm and a leg for doing so. Probably it's better to stay with a Jitterbug or a $10 DIY filter from Wuerth instead.
Cheers
Last edited:
As I said. Sometimes it works, sometimes it's neutral, sometimes it makes things worse.
There are several factors - and it's always the sum of these - leading to this or that result and conclusion.
If they advertise the device as the universal cure - as you say - they are most probably wrong.
As much being wrong as advertising asynchronous USB as universal cure for every USB Audio associated issue.
But that's how marketing works!
I do also own a RME Fireface UCX. In my case a USB filter and cutting VBus makes quite a difference. And that applies no matter if I run the RME from my ARM streamers (Cubitruck/PI) or my desktop PC. (I'm running the UCX into Adam A5x)
I do have to admit that the improvements became more prominent after I swapped the
rather mediocre power supply of my UCX. -- They' would have killed me if I'd post something like that in the RME forum.
Not to forget. USB cables can also mask tuning effects. You won't get around identifying all the bottlenecks in your chain(s).
Obviously we need to put this discussion into perspective.
Nobody expects such a device lifting your $50 DAC with a $2,5 cable into audiophile spheres.
You might end up with a bit less grain, slightly better dynamics, better separation, blacker black.... ... the ususal.
The more resolving your system, the more you're focussd, the more you'll notice. The usual.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.