USB cable shield resistance technical measurements

I agree. Those are all valid considerations.

Your cable-only R measurement of Pyst agrees with mine now, so that's encouraging in terms of consistency between cables.

I think a table of C measurements, even if done quickly and at 100-200 kHz, , would be interesting. At least for the two or three best cables .. since you have the setup and the B&K already.

Thanks Bob! I appreciate the goodwill!

 

This is quite fascinating. Other than extrapolation as to the quality of the rest of the cable, what do these measurements signify regarding sound quality? Thanks

 

I was going to ask how is a simple resistance measurement indicative of a cable's shielding effectiveness and started googling a bunch of crap on transfer impedance until I got to the part where Bob said it was just to measure the quality of the connection between the shielding and the connector.

I suppose this is a pretty simple and novel way that pretty much anyone can use to get an idea of a cable's build quality.

Hopefully you can unearth some gems among the dirt cheap garden variety USB cables.

Now that I think about it I'm kinda surprised blue jeans doesn't make one.

 

Part II - USB Cable Shield Resistance measurements

8 USB cables were evaluated for shield resistance to A and B end connectors.

The lineup (01, 05, 08 are repeats from Part I for reference):
01 generic clear small shield - 75 cm
05 Audioquest Forest shield - 75 cm
08 Schiit Pyst shield - 100 cm
09 Schiit Pyst shield - second sample - 100 cm
10 Audioquest Pearl shield - 150 cm
11 Audioquest Carbon shield - 75 cm
12 Wireworld Ultraviolet 7 shield - 75 cm
13 Monoprice shield - 100 cm

Data were acquired using the following:
Keysight 34461A 6.5 digit DMM with Benchvue for data logging
4-wire resistance measurements with sense at the breakout PCB terminals
APRS World USB A breakout board
APRS World USB B breakout board
Optic Bench Plate with risers, posts and clamps to immobilize A and B connectors

Each cable tested had both A and B ends inserted twice into the breakout boards to assure removal of any oxidation on the shield connection. Then A and B ends were secured into the clamps preventing any movement of the USB cables with respect to the breakout board connectors. A positive pressure existed assuring both A and B connectors made positive contact with breakout board connectors. This is different from Part I, which was more representative of actual cable connections with USB devices. As such, resistances will be lower across the cable to breakout board connection.

4-wire resistance measurements is also different from Part I. This effectively eliminates the resistance of the test fixture connectors to the DMM.

Each cable was measured over a period of 10 minutes.
Each cable was subjected to a mild cable disturbance (squeeze test) approximately 5 cm from connector for a period of 3 seconds, 3 times during a 1 minute data acquisition.

Measurement Setup (Part II):



Results:

10 Minute no cable movement:

Monoprice is definitely plays in a space of it's own, not even close to the others. The wandering resistance, even though the cable is motionless and secured, is not reassuring.

Y-axis zoom

These are all much better performers.


1 Minute with three 3-second squeeze tests:

Monoprice did not do well at all. I'd rather not use this USB cable.

Y-axis zoom 1

The generic clear small shield USB cable is on the fence. Pearl is a bit more predictable and appears to return to a stable state after each disturbance. Still, I'd rather have any of those below it as zoomed in the next graph.

Y-axis zoom 2

USB cables on Y-axis zoom 2 are stable and predictable with the Forest showing a little more disturbance than the others. Pyst is a great compromise between stability and price. Carbon shows the least disturbance but is also the priciest of those tested in Part II.

 

Here are the Benchvue raw measurements for the 10 minute data acquisitions.

Please note Benchvue y-axis changes for each measurement to maximize data displayed. What appears to be a large change may be quite small relative to another USB cable raw measurement. Pay attention to y-axis values.

01 generic clear small shield - 75 cm


05 Audioquest Forest shield - 75 cm


08 Schiit Pyst shield - 100 cm


09 Schiit Pyst shield - second sample - 100 cm


10 Audioquest Pearl shield - 150 cm


11 Audioquest Carbon shield - 75 cm


12 Wireworld Ultraviolet 7 shield - 75 cm


13 Monoprice shield - 100 cm

 

reserved for Part II 3/3.

Part II 3/3 will be LCR meter capacitance frequency sweeps when the BK Precision 891 arrives next month.

 

@atomicbob Bob, did you ever get a chance to try some of the Amazon Basic cables? Though if the Monoprice cable is built similar, it's not worth the effort!

I'm revisiting this thread after I came across measurements pertaining to USB length, cable extensions etc found here (circa 2013): http://archimago.blogspot.com/2013/04/measurements-usb-cables-for-dacs.html

Offhand, would you say that length is not detrimental to data transfer quality?

 

I’ll let Atomic Bob respond since you asked him, but for a quick response, here’s my take: length is detrimental to data transfer. Both because the longer cable acts as a better antenna for noise, and because of LCR effects that degrade the signal as it travels down the cable. However, I would say that as long as your cable is in spec, and the cable length is less than the USB-spec maximum, then you should be fine on the LCR-effects front. What’s left is noise, and that is a matter of shielding, which is where Bob’s earlier measurements come in. A good thick-foil shield will render the noise almost cable-length independent.

 

No, I haven't revisited this topic in awhile. It might be another several months before this one bubbles up in the queue again.

Cable length usually has a distributed capacitance associated which degrades the signal as lengths increase. Standing waves are another issue, especially at RF at which USB is definitely operating.

 

Monoprice has up-ed their audio USB cable game!
Monolith by Monoprice USB Digital Audio Cable - USB Type-A to USB Type-B
3 and 6 feet lengths
https://www.monoprice.com/product?p_id=33464

 

Not too expensive enough! Must sound bad!

 

Thank you for doing this @atomicbob . Very educational for me.

 

Nice stuff!

Would be great to see the tripp lite usb cable included as they seem to be fairly competent and they're affordable.

 

Tim from Holo Autio says great things about the Intona USB cables which are actually tested and have tolerances posted on the site, not just fancy audiophile language. It would be cool to test one of them.

They have 3 grades of cables with tighter and tighter tolerances. the lowest would probably be fine for audio. Would make a nice addition to your lab @atomicbob !

https://www.kitsunehifi.com/product/intona-professional-ids-vna-impedance-controlled-usb-cable/

Also, if you're ever up for testing new cables, I'd love to send you a Cable Matters USB cable to see how good they are. They're my reference non audio data transfer cable. All their cables are double shielded and seem well thought out and made.

 

I would need a usb / gpib converter to do logging but I was just measuring some usb a-b cables out of curiosity. Title says just usb cable shield resistance so I didn't hook up an LCR meter or anything like that... also my LCR meter is some budget junk I wouldn't trust the accuracy, just useful as a component identifier or to check for faults.

Using 4 wire kelvin clips & 6.5 digit dmm. I have no usb breakout boards so clipping directly to the housings. I wonder what the resistance of the usb breakout boards is. I didn't have the dmm on for 2 hrs at the start but it doesn't matter for this level of precision ;/

I've barely got any usb 2.0 a-b cables... just measuring the different cables I had on hand.

6 ft cable, braided cloth sleeving... .079 ohms.
6 ft cable, pu? pe? sleeving... 6~ ohms.
6 ft amazon basics 6.8 ohms so i guess the monoprice was better by this metric. This often wanted to measure higher but I think some of this was my failing to clip onto the outside of the b plug housing properly. If I were really gonna do this I would just buy a 3.0 B female socket and clip directly to the leads or housing on that ;/

usb 3 a-b cables

150 cm .036 ohm, came with $800 monitor. beefy... rubber exterior almost no give to the cable on a squeeze but flexible enough
100 cm .479 ohm, came with 30$ powered usb hub, thinner rubber cable than the other usb3 cables.
30 cm 6 ohms... came with a 50$ optical drive enclosure. rubber exteriror slightly less thick than 150cm cable, compresses a lot to squeezing.

There was far more variation when squeezing the 6 or .5 ohm (shield) cables vs the others. I haven't spent any money on a->b cables specifically... they all came with things besides the amazon basics cable, which by this metric was the worst besides the 30cm freebie. My dacs are old to the point of having mini b plugs and not great implementations... I avoid usb when possible with my devices.

I suppose I could dig out a few more generic cables, I'm sure at least 1 came with each UPS I own. I'm not sure I have any cable matters or tripp lite though.

Edited to add pics for completeness... also measured a 6ft 2.0 A-B cable that came with a $100ish UPS... it measures like total crap.... I can possibly get it as low as 1.5 ohms but often, with clips on the housing so an idealized measurement, but it doesn't maintain any kind of consistent measurement. I made sure the clips were on the housings real good and any sort of movement or squeeze test produced wild swings in the tens of ohms and it could even stay in the 10s-20s until another movement. I thought I was messing up the clips connection to the housing but that was not the issue it's really the cable.

Cables are left to right, top to bottom in the order mentioned. Showing the clips on the uhhh crappiest cable, amazon basics can't compete with this thing's poor construction.

and a picture of trying to measure that awful cable... any amount of squeezing or movement produced wild swings, saw it idling at 44ohms sometimes.