At first I could not get any PC I owned (and I tried several each with different versions of Windows running) to even recognize its existence. It wasn’t even appearing as an unknown device. Nothing. I read up on the internet and a few people said “try a different USB cable”. So I tried a few and it still didn’t work. So then I contacted the eBay seller and they said “try a few different USB cables”. I’d already tried that but just to humor them I tried even more USB cables, and fark me, the fifth one I tried actually worked.
Not only that, it’s consistent. This one cable works every time, the others don’t.
The USB cables that don’t work with this device work perfectly well with other devices, so they aren’t completely defective. As I understand it a USB-A to USB-micro cable is a passive device. They are all just standard, USB-certified cables, all of about the same length. So what might be different about the one cable that works? Is this device maybe very fussy about a clean connection? Or could it be something to do with crosstalk and better shielding? Or what?
I have a faux degree in electrical enginerding from the College of Electronics Being a Hobby so feel free to get technical with your answers.
I must say I really haven’t. I have had this difficulty any number of times with lightning cables because they are not passive devices, plus Apple devices that you connect with them are very fussy. But plain ol’ USB A to USB micro? Nope, generally reliable.
Reminds me of futzing with SCSI devices and terminators and SCSI IDs and SCSI cables back in the day.
I was one of the people with an honest-to-og SCSI chain. External all (Macintosh). SyQuest drive. UMAX Flatbed scanner. External HD. Dayna PC 5.25" drive.
When you say the other cables work with other devices, do you mean actually connect and send data?
Because one big difference between identical looking cables is that some actually only deliver power, not data.
IME most of the cables most people own are power only.
Well it seems you were more or less on the money. Thanks.
But having just spent the last half hour of my life (which I’m never going to get back) testing USBA to USB micro cables, it seems it’s a little more complicated.
I was testing cables with a Kobo e-reader.
There seems to be three categories of cable. I’ve got about two that are purely power. I’ve got about four that are actual data cables.
And then I’ve got about four that cause my Kobo to recognise that it is connected to a computer but then it won’t actually properly connect or transmit data. So when I was testing them yesterday I was seeing the Kobo say that it was connected to a computer and assuming that meant it was a proper cable but apparently not.
Yeah, I’m not sure whether there are three types but it makes “sense” that there might be; the difference between power and data is which connector pins are actually wired up, so it wouldn’t surprise me if there are several permutations of wiring.
It’s a good thing your Kobo shows the difference so clearly. Whenever I buy a device that includes a USB cable, the first thing I do is test it, then put a colored sticker on both sides of the bigger connector to display what kind of cable it is (as well as power vs data, I have colors for USB-C, lightning cable etc, because it’s quicker to see the sticker than find the small connector and orient it to see what kind of cable it is).
People think I’m OCD for doing this, but it saves a lot of time.
BTW another thing to watch out for is that usb ports are not the same; some deliver more power than others. I was close to sending back a new external USB drive that would make the dreaded click of death…until I tried plugging it into the USB port that had the electricity icon on it. It worked fine, and I had no problems since.
Old school USB has 4 pins:
5V
Data+
Data-
0V (ground)
In order for a low-power device to just use the USB connection for electrical power, it may only use the 5V and 0V lines - this may be implemented inside the device (i.e. the middle two pins are not used), or in the cable (the cable only has two wires).
But you can only draw a small amount of current that way. To draw more current, it is necessary for the device to communicate with the host via the data pins. There are ways to do this in a very basic form by ‘pulling up’ or ‘pulling down’ one of the data pins to 5V or ground respectively - just by shorting a resistor across from the 5V/0V to the data pin.
Again, some devices implement this inside the device. Some manufacturers (especially those who made their own cables - like if you see the device logo on one of the USB connectors) may have implemented it inside the cable - so that sort of cable willl work fully with their own device, and will work with other devices for charging only, but may or may not work for comms with a device that has its own/different implementation of pullup/pulldown resistors onboard.
In higher versions of USB, the negotiation between host and device regarding the type of connection and the power supplied is not done so simply and passively - it’s an active conversation between the two.
Sorry for spamming but I remembered another anecdote
I bought an mp3 player that works underwater (for swimming). I liked it so much I bought a second one – one to keep at home, one for the office if I want to swim at lunchtime. They came with a USB to 3.5mm headphone jack for charging.
One time I was charging one of them, and I touched the device and burned my finger; it was red hot. To cut a long story short, I eventually discovered that the cables were not interchangeable. Even though they were the exact same model, the packaged cables only worked with that specific device and with the other device it would neither charge nor share data, just get hotter and hotter…
I still have no idea what the difference is between those cables, but luckily since I bought different colors of device at least, I was able to label the cables such that I won’t burn my house down.
Recognition is a three step process. Negotiation of the power level is done at the hardware level: the host recognizes that something is connected, but does not know what it is. Then the device sends identifying information over the data channel. Then the host recognizes that information, and ‘names’ the device.
A non-recognized device could be just a device that is not recognized, or a device that can’t communicate properly.
That is tangential to the original question. USB is a high-speed serial protocol, and it’s always been the case that some cables are flaky, and some devices are flaky, and some devices are flaky with some cables. It could be cross-talk or poor impedance matching. Inside the case was also a common location of the problem, even when power levels were lower: the extra cable from the M/B to the front of the case, with the extra junctions, was often a problem.
Ethernet managed to work at even higher speeds with no problems: either a lot of USB cables were really very very badly made, or the transformer coupling on Ethernet connections really did make a big difference.
I have Ampere installed on all of my Android devices that it is compatale with. It tells you how fast the device is charging. It keeps me from finding out that the device is taking forever to charge the hard way (by waiting, and waiting.) The best cables for my current phone charge it at around 1800 mA. But when the cables go bad (as they always do) the rate plunges down to near zero (or, of course, to zero.) My current cables cap out at around 900 mA. But (on point for the quote) when you first plug something it, it usually charges at a very low rate for a number of seconds, then ramps up to the best the cable can give over a period of a few seconds after that.
(As an aside, with my two latest cables, which are USB-C, one of them was good but the other one just rattled around in the port without making a connection. It confused me how it could possibly be too big, until I compared the connectors on both cables side-by-side and the bad connector was actually too short. I trimmed back some of the plastic with a knife, and it fit properly after that.)
If that was by design rather than a fault, there’s an electrical engineer out there somewhere who deserves to be shot.
The mildly interesting thing is that with what I have described above as the third type of cable I seem to own, the device (the Kobo) recognised that it was connected to a host but the host did not recognise that it was connected to anything at all.
I have had that happen with a Garmin cycle computer and a Tomtom navigator but even worse is that the cables will carry data from other devices but just not to those.
Yes, some cables are data and some are power-only, and yes, this accounts for some fraction of the issues the OP described. But there are still plenty of cases of a cable that verifiably carries data with some devices, but doesn’t work at all (not even to charge slowly) with other devices. There, the problem is that the connectors are slightly the wrong shape, or possibly have non-conductive corrosion on some spots of the contacts, and some combinations of shapes make good contact, while others don’t.
The one working may be defective cable. It may connect pin 4 to pin 5 at the B end… This will repair the defect with your board, in that its using pin 4 instead of pin 5 for power earth ??
The mistake has happened because original USB had 4 pins, and mini,micro USB has 5 pins, AND, the earth moved along to pin 5.
I see the schematics for the tinyISP circuit says to use pin 4… of the 4 pin classic USB connector !
Just academic , because I think this is the pin 4,5 issue… Long USB cables do fail… the limit isn’t soft…its true and hard… but you don’t use 5 metres long do u ? . As to “purely passive device”… since you using the 5 volt line for digital circuits, its possible that humungous capacitance is a good thing on the power rail… also there can be voltage drop due to current and the circuit could be very fussy on getting 5 volts ??
I still do, with my circa-1989 Amiga 3000. External CD-ROM, external hard drive, external DDS tape backup. You bet I don’t move those boxes or screw with the cabling without outstanding reason and lots of time to troubleshoot.
My guess with USB cable sensitivity is a slightly out of spec receptacle on the board. Picky about making good contact.
