Topics

USB cable quality and QRM in a spyserver setup with an Airspy Hf+ and Raspberry PI3. 6 microUSB cable test. #airspyhfplus #bestpractice #raspberrypi

Graham
 

Some of the better manufacturers put some details on their web site but
at that is usually still somewhat sparse.

for USB micro and apple cables I prefer Anker which are quite well
reviewed and their web page does provide some technical details.

Also, Sabrent and Volutz.

Tyco / TE Connectivity makes some decent USB cables. In their USB A to B
cables they have versions which have 25 AWG for the data pair and 20
AWG for power.

I have destructably disassembled (i.e. cut apart) a number of cables
from cheap to moderately priced. There are differences and you do get
what you pay for.

You can even buy the end parts to make your own cables with the cable of
your choosing if you are particular.

something else to consider is the cable shields. The metal parts at both
end of the cables should not be connected to the ground pin of the
connectors but should be connected to the shield (if there is any) of
the cable. In fact, sometimes it is better to just have the shield
connected to the shell on the end that plugs into the computer and not
connected on the other end. This creates what is called a Farady shield
(or cage if you prefer) over the cable. Connecting both ends can
sometimes create ground loops in the shield. I have run out of fingers
and toes to count the number of times that I have seen where a cable
having the shield connected at one end only solves EMI/RFI issues where
another cable with the shield connected at both ends only made the
problem worse or was in fact the source of the problem to begin with.

Solving EMI/RFI issues is a mixed bag of science and sometimes a bit of
black magic.

This is a could introductory and practical text on the subject:

http://audiosystemsgroup.com/SAC0305Ferrites.pdf

cheers, Graham ve3gtc

On 3/14/2018, "David Ranch" <airspy-groupsio@...> wrote:


For a 2-pack of 10' cables [Volutz], $17 on eBay was a deal.
Wow.. 10'? I'd assume there is some pretty serious voltage drop over
that! Regardless, I'll have to give them a look. Thanks.

--David
KI6ZHD

Graham
 

Another good document that some may find interesting on this subject:

https://www.ti.com/sc/docs/apps/msp/intrface/usb/emitest.pdf

cheers,

PMM
 

One thing I noticed was with my printer, it not grounded through power supply so having a good usb lead was important to kill the noise even though it only in standby mode, in fact printer contributed 20db of noise to the sdr near computer.  Plugging in a decent usb cable pulled noise floor down 20db verse unplugged and against lesser quality usb cable.

So sometimes it can be said a poorly grounded device/non grounded psu can benefit from a good grounding given by usb cable, this seems to validate some of the cable choices due to better shielding/grounding wires in there construction.



On Wed, 14 Mar 2018, 15:11 Graham, <planophore@...> wrote:
Another good document that some may find interesting on this subject:

https://www.ti.com/sc/docs/apps/msp/intrface/usb/emitest.pdf

cheers,

Dana Myers
 

On 3/14/2018 7:35 AM, David Ranch wrote:

For a 2-pack of 10' cables [Volutz], $17 on eBay was a deal.
Looks like there's a better deal on Amazon - free same-day shipping
with Prime, $12. Shop around. It's nice having a ridiculously long
cable in my shack, BTW.

Wow.. 10'?  I'd assume there is some pretty serious voltage drop over that!  Regardless, I'll have to give them a look.  Thanks.
Let's do some math, because we're scientists :-). A graphic at the
Volutz website says the power wires in their USB cables is 22 AWG.
I believe we can safely ignore the data wires here; as long as the
cable isn't dropping frames, it's OK.

Consulting Wikipedia:
https://en.wikipedia.org/wiki/American_wire_gauge#Rules_of_thumb

We see that the approximate resistance of  this wire is 16 ohms / 1000'.
So, 10' of the wire (power and return) has about ~ 10/1000 * 16, or
0.16ohms, resistance per leg, or a total of 0.32 ohms.

Assuming a 200mA load (I don't actually know the Airspy HF+ current
consumption), that means the +5V rail at the HF+ drops by 0.032V,
or 32mV, and the ground rail at the Airspy HF+ increases by 32mV. In
other words, the +5V supply to the Airspy HF+ drops by 64mV.

Given that I bet nothing inside the Airspy HF+ runs from higher than a 3.3V
rail, this is plenty of margin for an LDO (you can do worst case math with USB
host power rail of 4.5V and LDO minimum differential of 0.6V and still have
500mV of margin).
Assume I'm wrong  and the HF+ uses 500mA (the max spec for a
USB port, right?). That's 160mV drop to the HF+; still plenty of margin
for a 3.3V LDO (worst-case above would be > 400mV).

The keen observer might have noticed I suggest the DC ground
in the HF+ is ~32mV higher than the USB host ground. Note that
this doesn't matter because the USB signalling is differential data,
not analog.

OK, maybe we're not scientists, we're ham radio operators. In which
case you can disregard the above and say "it works great" (like the way
J-Pole antennas are 3/4-wave verticals :-)).

73,
Dana  K6JQ

Dana Myers
 

On 3/14/2018 7:53 AM, Graham wrote:
You can even buy the end parts to make your own cables with the cable of
your choosing if you are particular.
Is this really practical, especially from a mechanical ruggedness perspective?

73,
Dana  K6JQ

Dana Myers
 

On 3/13/2018 9:32 PM, Ian DXer wrote:
Hi Dana,

Pleasing to read your report on your success with the Volutz USB cables.
Thank you.


I was looking at the Volutz Facebook page & I note that Volutz have just released a new Apple Lighting Cable.
For noise immunity I like the metal shroud that links the foil shield of the cable & surrounds the connector & components
underneath. One also gets a better look at the dual cable shielding from the drawing on the Facebook page.
Are these SMD components part of the lightning spec or are they part of Volutz's attention to EMI/EMC or impedance
matching?
I do not know anything really about Lightning (the only Lightning cables in
my house are associated with my wife's iPhone).

I see a similar shroud on their USB C cable from their Facebook page.
What's not clear to me is if Volutz use a metal shroud on their connectors for their USB 2.0 to micro USB cables or
any filtering SMD components (if that's what they are?) similar to their new Apple Lightning Cable.
The Volutz website is scant with close up pics & tech data.
I do not know the details; I just know I found one graphic that says the power
lines in the USB cable are 22 AWG :-)
A further improvement to their USB 2.0 to micro usb cables would be to have a foil shield between the data & power pairs.
Perhaps, but what are you seeking to address with this? The data lines are a
differential pair and surely twisted (but I don't know for a fact). Given that the
data lines are very low current, I'd be surprised if they induce meaningful noise
into the power lines. On the other hand, I'd expect noise from the power lines
to be induced as common-mode into the data lines and thus disappear from
the data.

73,
Dana  K6JQ

Graham
 

it depends on how particular you are and your attention to detail. It is possible to "build your own" that will be better than what you can buy. Likely at four times the cost however.

Audiophools are group that will build their own in their never ending quest for superior sonic performance. You can find many such hand made usb cables for sale, on the low end for $50, middle range a few hundred, and on the high end thousands. Gold plating, silver wire, cryogenic treated and oxygen free copper and other assorted snake oil ingredients but that is getting a bit too carried away. ( I hope I didn't offend anyone )

In the end you do get what you pay for (within reason considering the above) and buying cheap on eBay or Amazon is not always the place. Unfortunately low price is a selling point and many shop based on $$ alone.

cheers, Graham

On 2018-03-14 17:39, Dana Myers wrote:
On 3/14/2018 7:53 AM, Graham wrote:
You can even buy the end parts to make your own cables with the cable of
your choosing if you are particular.
Is this really practical, especially from a mechanical ruggedness perspective?

73,
Dana K6JQ



Dana Myers
 

On 3/14/2018 3:05 PM, Graham wrote:
it depends on how particular you are and your attention to detail. It is possible to "build your own" that will be better than what you can buy. Likely at four times the cost however.
Well, OK. You'll never see me doing it ;-)

Audiophools are group that will build their own in their never ending quest for superior sonic performance. You can find many such hand made usb cables for sale, on the low end for $50, middle range a few hundred, and on the high end thousands. Gold plating, silver wire, cryogenic treated and oxygen free copper and other assorted snake oil ingredients but that is getting a bit too carried away. ( I hope I didn't offend anyone )
Because those NRZI-encoded data frames are so fragile :-)

<remainder of messaged redacted>

73,
Dana  K6JQ

jdow
 

On 20180314 07:35, David Ranch wrote:

For a 2-pack of 10' cables [Volutz], $17 on eBay was a deal.
Wow.. 10'?  I'd assume there is some pretty serious voltage drop over that! Regardless, I'll have to give them a look.  Thanks.
--David
KI6ZHD
My AirSpy HF+ is on the end of a 5 meter active extension cord. It works. I'd not put any MORE wire than that on it without an externally powered hub feeding the AirSpy HF+ end of things. Distance from the computer is your friend.
{^_^}

jdow
 

Jim knows ferrites and grounding very well, with a many years level of experience running audio in a theater in Chicago building that also housed most of the local network affiliate transmitters.

I'm glad somebody else knows of his site. I get evangelistic about it from time to time.
{^_-}

On 20180314 07:53, Graham wrote:
Some of the better manufacturers put some details on their web site but
at that is usually still somewhat sparse.
for USB micro and apple cables I prefer Anker which are quite well
reviewed and their web page does provide some technical details.
Also, Sabrent and Volutz.
Tyco / TE Connectivity makes some decent USB cables. In their USB A to B
cables they have versions which have 25 AWG for the data pair and 20
AWG for power.
I have destructably disassembled (i.e. cut apart) a number of cables
from cheap to moderately priced. There are differences and you do get
what you pay for.
You can even buy the end parts to make your own cables with the cable of
your choosing if you are particular.
something else to consider is the cable shields. The metal parts at both
end of the cables should not be connected to the ground pin of the
connectors but should be connected to the shield (if there is any) of
the cable. In fact, sometimes it is better to just have the shield
connected to the shell on the end that plugs into the computer and not
connected on the other end. This creates what is called a Farady shield
(or cage if you prefer) over the cable. Connecting both ends can
sometimes create ground loops in the shield. I have run out of fingers
and toes to count the number of times that I have seen where a cable
having the shield connected at one end only solves EMI/RFI issues where
another cable with the shield connected at both ends only made the
problem worse or was in fact the source of the problem to begin with.
Solving EMI/RFI issues is a mixed bag of science and sometimes a bit of
black magic.
This is a could introductory and practical text on the subject:
http://audiosystemsgroup.com/SAC0305Ferrites.pdf
cheers, Graham ve3gtc
On 3/14/2018, "David Ranch" <airspy-groupsio@...> wrote:


For a 2-pack of 10' cables [Volutz], $17 on eBay was a deal.
Wow.. 10'? I'd assume there is some pretty serious voltage drop over
that! Regardless, I'll have to give them a look. Thanks.

--David
KI6ZHD

jdow
 

The figure 6 "Spring Contact Clips" are what I cobble together with very fine braid, careful positioning, and scraped off paint. It helps.
{^_^}

On 20180314 08:11, Graham wrote:
Another good document that some may find interesting on this subject:
https://www.ti.com/sc/docs/apps/msp/intrface/usb/emitest.pdf
cheers,

jdow
 

Loren and I call them "audiofeelies". They are sheep to be shorn of their excess cash. Be sure to sell them a few gold plated wall sockets and gold plated power cords to match. Also find that magic color of green for the felt tip pens for their CDROMs to make them sound better. But make sure they have empty state electronics for their amplifiers. The added noise and distortion is OK. It's softer than the hard edged digitally introduced noise. And and and - infinite potential to suck money from their wallets into yours. You are a good wood craftsman? Then turn some nice custom wood knobs for their equipment to reduce resonances from the metal knobs you're replacing.

{^_-}

On 20180314 15:05, Graham wrote:
it depends on how particular you are and your attention to detail. It is possible to "build your own" that will be better than what you can buy. Likely at four times the cost however.
Audiophools are group that will build their own in their never ending quest for superior sonic performance. You can find many such hand made usb cables for sale, on the low end for $50, middle range a few hundred, and on the high end thousands. Gold plating, silver wire, cryogenic treated and oxygen free copper and other assorted snake oil ingredients but that is getting a bit too carried away. ( I hope I didn't offend anyone )
In the end you do get what you pay for (within reason considering the above) and buying cheap on eBay or Amazon is not always the place. Unfortunately low price is a selling point and many shop based on $$ alone.
cheers, Graham
On 2018-03-14 17:39, Dana Myers wrote:
On 3/14/2018 7:53 AM, Graham wrote:
You can even buy the end parts to make your own cables with the cable of
your choosing if you are particular.
Is this really practical, especially from a mechanical ruggedness perspective?

73,
Dana  K6JQ




iz5dkm@...
 

  Also I had usb noise problems and I solved by self-building a cable with two rg174 for I / O signals and powering the sdr with a second stabilized power supply (no switching).
Much noise has disappeared.

Roberto Zinelli
 

Yes, a traditional stabilized power supply solved also mine noise problem
73 IW4ENS

On Thu, Feb 21, 2019 at 7:13 PM <iz5dkm@...> wrote:
  Also I had usb noise problems and I solved by self-building a cable with two rg174 for I / O signals and powering the sdr with a second stabilized power supply (no switching).
Much noise has disappeared.

zennix2014@...
 

Can you guys elaborate on the power supply wiring, and also the cable changes? 

I'm heading down the same path (airspy HF+, and raspberrypi), and have had noise issues in the past with other SDRs picking up noise, so more details on connection, power supply make/model, schematics / wiring diagrams, etc., would be very helpful.

I'm unclear about the "self-building" cable part - if you split the USB cable apart or what, so any details help.

Doug