It will drop too much voltage and get a little warm. The warmth is not a problem the voltage drop may not be much of a problem charging a 3.7V Lithium battery but will be rather more of a problem for powering a Pi.
-- Steve O'Hara-Smith | Directable Mirror Arrays C:\>WIN | A better way to focus the sun The computer obeys and wins. | licences available see You lose and Bill collects. | http://www.sohara.org/
You disagree with my previous, possibly over simplified, summary of the situation re USB C power supplies and the Pi 4?
-- Brian Gregory (in England).
Yes. I had a 5 metre one so as to get a TV dongle near the TV socket. Randomly it would stop working. Probably too much V drop Same dongle on 1 meter cable is flawless
--
"First, find out who are the people you can not criticise. They are your
oppressors."
- George Orwell
No, it won't melt, but the voltage at the Pi end will be too low.
As I said I found that about 75% of the cables I tried when powering a Pi 2 B from a very capable power supply didn't work reliably. They didn't melt or anything, just made the red LED on the Pi go on and off at random as the power required varied.
-- Chris Green
Could also be interference from external sources, or other poor quality comms. Using a 5m cable to connect a Topfield PVR to a PC for offloading programmes can be quite unreliable, yet a 1/2M works ok.
Note, this can't be voltage drop related, as the Topfield is not powered by the PC, data link only!
-- --------------------------------------+------------------------------------ Mike Brown: mjb[-at-]signal11.org.uk | http://www.signal11.org.uk
On Tue, 23 Jun 2020 11:19:37 +0100 (BST), snipped-for-privacy@signal11.invalid (Mike) declaimed the following:
You may still have a voltage drop in the 0/1 levels of the data. Assuming CMOS 30/70% points, on a 5V system, 0-bits are 1.5V and under,
1-bits are 3.5V and above. Voltages in between are indeterminate -- circuits likely using Schmitt triggers to prevent signal fluctuations when in that region.NOTE: per a Google, USB 1.1 used 0-3.3V, USB 2 uses 0 to 0.4V (I think the article implies these are minimum signal change); it also indicates that USB uses differential signals, so should be more immune to noise as rather than flipping one line between low and high, two lines flip into opposite states.
In a long enough cable, voltage drop could result in a receiver not detecting transitions to 1-bits reliably (unless you have some odd coupling of a DC signal, the 0-bits are likely reliable)
-- Wulfraed Dennis Lee Bieber AF6VN wlfraed@ix.netcom.com http://wlfraed.microdiversity.freeddns.org/
Voltage drop on data lines due to cable resistance is extremely improbable, since the current is relatively low. Put another way, the impedance of the driving and receiving circuits is high compared to cable resistance.
More likely, capacitance between lines of a longer cable slows signal transitions, eroding signal margins.
An oscilloscope would reveal the truth.
-- -michael - NadaNet 3.1 and AppleCrate II: http://michaeljmahon.com
I understand the problem
but what kind of Chinese crap are you using ?
Please, learn transmission line theory.
David
Is there any evidence that a USB cable behaves as a transmission line? Or that the spec matches source and destination impedance to it?
Hmm it seems they do
There will still be attenuation however over longer cables
-- Ideas are more powerful than guns. We would not let our enemies have guns, why should we let them have ideas? Josef Stalin
I would have been very surprised if the spec wasn't for transmission line behaviour - it's not exactly a new concept. Cheap Chinese cables however...
-- W J G
If it were only the cables :/
All cables behave as transmission lines. So do PCB traces.
There are some situations where it isn't necessary to treat them as such, but USB isn't one of them.
David
... in some cases do not conform to the USB specifications.
Whether it's "some", "many" or even "most" is up for discussion.
David
Well no, they don't. a single signal wire with the earth/return via the other side of the world is so far from a transmission line as to render the notion 'not useful'
-- "A point of view can be a dangerous luxury when substituted for insight and understanding". Marshall McLuhan
I know f*ck-all about transmission line theory, but I know when someone's being a dick.
-- Chris Elvidge, England
You pick an interesting case. When current flows through a wire, a magnetic field is created. We know this is true because, if the current is high enough, we can see it deflect a compass needle.
The magnetic field contains energy. This means the wire has inductance. That's one of the four components in the standard model of a transmission line: series inductance, series resistance, shunt capacitance and shunt resistance.
30-odd years ago I was involved in the design of an audio test set with integral oscillator. The distortion residual of oscillator and analyser was 0.0008% at 1 kHz. At one stage in development, I saw that the distortion measured at the load was a little bit higher than that at the output terminals of the oscillator. The path went through reed relays. I discussed this with the manufacturer of the relays, who sent me some magnetic wire and a couple of little bar magnets. I measured the voltage across a length of wire, and discovered that it was much higher than across a similar length of copper wire - and that it was proportional to frequency, i.e. I was seeing the effect of the wire's inductance, which was much higher for the wire with the built-in magnetic core (the wire was in all cases straight). I measured the harmonic distortion of the voltage across the wire, and found that I could get several percent if I put enough current through it. Looking at the distortion residual, it was all odd harmonics, mostly third harmonic. Then the real eye-opener: put a bar magnet close to the wire, and the second harmonic appeared in large quantity. (The wire was being pushed closer to magnetic saturation when the current flowed in one direction than the other.)All of which is just to remind you that wire does indeed possess inductance. So does cable.
Look at the standard mathematical model of transmission lines, and you'll see the four components I mentioned above.
As for PCB traces: I discovered in the early 1990s, when we were using 74ABT logic devices, which have nanosecond rise and fall times, that devices at the far end of a 150mm PCB trace were being clocked on the rising edge always (as they should be), but often on the falling edge too. Looking at it with a fast enough oscilloscope, both edges had a "dog-leg" or inflection. The inflection on the falling edge was just about the threshold of the receiving device, so it saw a rise there too. The cure was to terminate the PCB trace, which as behaving exactly as the mathematical transmission line model showed it should.
If you can get hold of appropriate kit, these are observations that you can replicate.
David
Around 1980 two of us assembled wire-wrap prototype memory boards (stuffed with 2kx8 SRAM and 74LS logic) the other guy was an experienced wireman who made a lovely job of it, the looms of address and data bus were really clear. I was a student and not so inclined to making things pretty, every wire made the correct connection but they went all over the place in a rat's nest. On first sight the boss commented that his was the way it should be and I needed to do better - fair enough I thought - his was pretty.
Both boards passed a DC buzz test (every wire going where it should be and no extras) and a power on smoke test and so were duly populated with components and tested. Mine worked first time - his radiated at ~30MHz strongly enough to be picked up by an open scope probe across the room purely because of the way the wires were laid out on the board (it didn't work of course) - he wound up hanging terminating resistors at the end of the bus loops to shut it up, mine didn't need them.
Wire wrap looms can behave like transmission lines too.
We all learned something that day - including the boss.
-- Steve O'Hara-Smith | Directable Mirror Arrays C:\>WIN | A better way to focus the sun The computer obeys and wins. | licences available see You lose and Bill collects. | http://www.sohara.org/
On a sunny day (Fri, 26 Jun 2020 15:24:32 +0100) it happened Ahem A Rivet's Shot wrote in :
30 MHz has a wavelength of 10 meter. for distances on an eurocard (100x160 mm) board the transmission line issue is not very significant.I have done some wirewrap but did not really like it,
Here a fast RAM disk from those days:
Things I soldered together in the evening.. At work we had pee see bees made... took a while, Mine worked in a few hours, did prototyping like that at work too, then had the pee see bees made.
I still prototype like that, also at GHz. wavelength
A common misconception is to relate these issues to the repetition rate. It can be one cycle per fortnight, but if the transition times are short (roughly a nanosecond or less), what arrives at the destination may have more edges in it than were sent, unless the line is appropriately terminated.
Whether there is a problem also depends on the bandwidth/speed of the receiver.
David
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.