USB voltage fluctuations

Is it normal for USB voltage to spike now and then? I ask because I have
several of those little LEDs on a chip USB lights, and I put one into one
of the USB3 slots I have in the front of my tower case. The slot in
question is one of four on an add on card I installed where the diskette
drive used to go on older systems. It has four slots, with two USB3 A cords
to supply data and power, and s separate four pin power connection for
additional power. My tower sits beside my desktop, and I wanted to see if
adding this light would help light up my keyboard at night. Shortly after I
inserted it, I notice an occasional flash of light from the LED. There is
no pattern to it that I have been able to discern. As a test I moved the
light to a different USB3 slot that came built into my tower, and it has
not flashed there at all. It would be nice to have a recording voltmeter,
but I no longer have access to one, so that option is out.

Any ideas as to what is going on, or maybe some suggestions for a very low
cost method of recording the voltage on a USB port?

Tim wrote:

Is it normal for USB voltage to spike now and then? I ask because I have
several of those little LEDs on a chip USB lights, and I put one into one
of the USB3 slots I have in the front of my tower case. The slot in
question is one of four on an add on card I installed where the diskette
drive used to go on older systems. It has four slots, with two USB3 A cords
to supply data and power, and s separate four pin power connection for
additional power. My tower sits beside my desktop, and I wanted to see if
adding this light would help light up my keyboard at night. Shortly after I
inserted it, I notice an occasional flash of light from the LED. There is
no pattern to it that I have been able to discern. As a test I moved the
light to a different USB3 slot that came built into my tower, and it has
not flashed there at all. It would be nice to have a recording voltmeter,
but I no longer have access to one, so that option is out.

Any ideas as to what is going on, or maybe some suggestions for a very low
cost method of recording the voltage on a USB port?

You said "add on card ... where the diskette drive used to go". This
USB hub that fits in a drive bay doesn't have a brand and model? What
is the brand and model of motherboard?

That would be a 3.5" drive bay; however, it is unclear if the "add on
card" is actually a daughtercard (in a motherboard card slot) or you are
just mentioning the PCB that fits into the 3.5" drive bay. If the
latter (i.e., you installed a USB hub into a drive bay), to where did
you plug its cables? I would think even if you plugged the USB hub into
USB2 headers on the mobo that the hub's LED would still come on unless
it indicates when the hub is usable as a USB3 hub (which isn't likely if
more than 2 USB ports are in the hub and even then each port would have
to go to a separate USB3 header on the mobo). Did you plug in the hub's
separate power connector (likely a 4-pin Molex)? Does the hub only have
2 USB3 ports (one and only one for each USB3 cable from the hub going to
a USB3 header on the mobo)?

When I've plugged in a card reader which connects to the mobo via a USB
header on the mobo, the LED on the reader is always on. Even when the
computer is supposedly turned off the LED in the card reader in the
drive bay is on. That only indicates there is +5V to the card reader,
not to indicate there was a data transfer. It is not like a drive light
to show activity. Maybe the LED or its logic on the PCB in the USB hub
you put in the drive bay is defective so the LED doesn't come on except
once in a blue moon. Sometimes new equipment is defective. The most
likely failure times are when new, just before a pregnacy period (9
months), or many years after normal use.

Tim wrote:
Is it normal for USB voltage to spike now and then? I ask because I have
several of those little LEDs on a chip USB lights, and I put one into one
of the USB3 slots I have in the front of my tower case. The slot in
question is one of four on an add on card I installed where the diskette
drive used to go on older systems. It has four slots, with two USB3 A cords
to supply data and power, and s separate four pin power connection for
additional power. My tower sits beside my desktop, and I wanted to see if
adding this light would help light up my keyboard at night. Shortly after I
inserted it, I notice an occasional flash of light from the LED. There is
no pattern to it that I have been able to discern. As a test I moved the
light to a different USB3 slot that came built into my tower, and it has
not flashed there at all. It would be nice to have a recording voltmeter,
but I no longer have access to one, so that option is out.

Any ideas as to what is going on, or maybe some suggestions for a very low
cost method of recording the voltage on a USB port?

Take a $0.10 LED and a resistor and connect them to the 5V copper
land on the USB tray PCB (if available) and use the light from
that LED as a "second opinion".

Out of my entire collection of LEDs here, I have just one
white LED with modal behavior. The light output switches
between a higher and lower level at random. And that was
fed from a "clean" DC source, so it wasn't the source doing it.
Other LEDs in the same circuit would remain at a constant
intensity. That LED was purchased singly from our local
electronics shop (Paul "sampling" the product to see if
it's **** or not), and I was surprised to find this defect.

The LED was in a T1 3/4 package but was rated at around 100mA
or so (not the normal 20mA industry standard rating), and had
a larger than normal anvil inside and stouter legs, and the
product was not traceable to an actual manufacturing source. It
wasn't branded, so I can't say "a Nichia or Cree did that". I don't
know who made the LED. I even tried matching images of
LEDs to track it down, but nothing I could find on the
Internet in terms of closeup pictures, matched the shape
of the anvil (the thing the LED sits on, under the clear
plastic).

*******

The PC sound card doubles as an oscilloscope, but

1) All inputs and outputs are AC coupled.
You can't "see down to DC" if using the 1/8" jacks.

2) Regardless of sampling rate (44.1, 48, 96, 192KHz),
the input filter is there to prevent aliasing
when sampled. The 192KHz sampling rate doesn't actually
give a 96KHz Nyquist. You can't "listen to bats echo-locate"
with the sound card. All you can really count on,
is the circuit passing 20KHz or so. Anything above
that is "accidental gravy". I have managed to measure
the 25KHz "stimulus" used on Analog Devices ("SoundMax")
for impedance sensing, using a 192KHz RealTek as a "scope".
But I doubt I'll be listening to bats with a piezo, any
time soon. It doesn't go out that far.

Because of the AC coupling, you can attempt to measure
"noise" riding on a DC source. I would probably make up
a little sampling network (attenuator), just to avoid
blowing out the sound card inputs when connecting or
disconnecting to the 5V rail you want to measure.
The sound card has an "input impedance" and you have
to take that into account if designing an attenuator.
It's not a Hi-Z input. It's not a multimeter. All it
has to meet, is probably 10Kohms on input. Using a
calibrated output source, you can probably do experiments
to derive the input impedance of your "sampler".

I wouldn't waste my time, but if you've never done it
before, you might get an evening of amusement from it.
At least it's cheap - a $2.50 resistor assortment pack
has all the materials needed to make an attenuator, if
you choose to protect the input a bit.

Paul

On Fri, 09 Feb 2018 02:45:49 -0500, Paul
wrote:

Tim wrote:
Is it normal for USB voltage to spike now and then? I ask because I have
several of those little LEDs on a chip USB lights, and I put one into one
of the USB3 slots I have in the front of my tower case. The slot in
question is one of four on an add on card I installed where the diskette
drive used to go on older systems. It has four slots, with two USB3 A cords
to supply data and power, and s separate four pin power connection for
additional power. My tower sits beside my desktop, and I wanted to see if
adding this light would help light up my keyboard at night. Shortly after I
inserted it, I notice an occasional flash of light from the LED. There is
no pattern to it that I have been able to discern. As a test I moved the
light to a different USB3 slot that came built into my tower, and it has
not flashed there at all. It would be nice to have a recording voltmeter,
but I no longer have access to one, so that option is out.

Any ideas as to what is going on, or maybe some suggestions for a very low
cost method of recording the voltage on a USB port?

Take a $0.10 LED and a resistor and connect them to the 5V copper
land on the USB tray PCB (if available) and use the light from
that LED as a "second opinion".

Out of my entire collection of LEDs here, I have just one
white LED with modal behavior. The light output switches
between a higher and lower level at random. And that was
fed from a "clean" DC source, so it wasn't the source doing it.
Other LEDs in the same circuit would remain at a constant
intensity. That LED was purchased singly from our local
electronics shop (Paul "sampling" the product to see if
it's **** or not), and I was surprised to find this defect.

In my work I came across a bead diode with the yellow
band on the wrong end.

Paul

VanguardLH wrote in :

Tim wrote:

Is it normal for USB voltage to spike now and then? I ask because I
have several of those little LEDs on a chip USB lights, and I put one
into one of the USB3 slots I have in the front of my tower case. The
slot in question is one of four on an add on card I installed where
the diskette drive used to go on older systems. It has four slots,
with two USB3 A cords to supply data and power, and s separate four
pin power connection for additional power. My tower sits beside my
desktop, and I wanted to see if adding this light would help light up
my keyboard at night. Shortly after I inserted it, I notice an
occasional flash of light from the LED. There is no pattern to it
that I have been able to discern. As a test I moved the light to a
different USB3 slot that came built into my tower, and it has not
flashed there at all. It would be nice to have a recording voltmeter,
but I no longer have access to one, so that option is out.

Any ideas as to what is going on, or maybe some suggestions for a
very low cost method of recording the voltage on a USB port?

You said "add on card ... where the diskette drive used to go". This
USB hub that fits in a drive bay doesn't have a brand and model? What
is the brand and model of motherboard?

That would be a 3.5" drive bay; however, it is unclear if the "add on
card" is actually a daughtercard (in a motherboard card slot) or you
are just mentioning the PCB that fits into the 3.5" drive bay. If the
latter (i.e., you installed a USB hub into a drive bay), to where did
you plug its cables? I would think even if you plugged the USB hub
into USB2 headers on the mobo that the hub's LED would still come on
unless it indicates when the hub is usable as a USB3 hub (which isn't
likely if more than 2 USB ports are in the hub and even then each port
would have to go to a separate USB3 header on the mobo). Did you plug
in the hub's separate power connector (likely a 4-pin Molex)? Does
the hub only have 2 USB3 ports (one and only one for each USB3 cable
from the hub going to a USB3 header on the mobo)?

When I've plugged in a card reader which connects to the mobo via a
USB header on the mobo, the LED on the reader is always on. Even when
the computer is supposedly turned off the LED in the card reader in
the drive bay is on. That only indicates there is +5V to the card
reader, not to indicate there was a data transfer. It is not like a
drive light to show activity. Maybe the LED or its logic on the PCB
in the USB hub you put in the drive bay is defective so the LED
doesn't come on except once in a blue moon. Sometimes new equipment
is defective. The most likely failure times are when new, just before
a pregnacy period (9 months), or many years after normal use.

I can see I didn't describe clearly. What I added in the 3.5" drive bay
was a StarTech 4 port USB3 Hub [35BAYUSB3S4], with a USB3 'A' connector
and a SATA power connector. I used a USB3 20-pin motherboard female to 2
type A femail connectors Y cable to connect the hub to the motherboard,
which is an ASUS F2A85-V Pro.
The LED in question is actually a small circuit board with a USB
connector on one end, and three COB style surface mount LEDs as the light
source, that I picked up from Banggood for a couple bucks. It draws
somewhere around 200ma.

On Fri, 09 Feb 2018 02:45:49 -0500, Paul wrote:

Tim wrote:
Is it normal for USB voltage to spike now and then? I ask because I have
several of those little LEDs on a chip USB lights, and I put one into one
of the USB3 slots I have in the front of my tower case. The slot in
question is one of four on an add on card I installed where the diskette
drive used to go on older systems. It has four slots, with two USB3 A cords
to supply data and power, and s separate four pin power connection for
additional power. My tower sits beside my desktop, and I wanted to see if
adding this light would help light up my keyboard at night. Shortly after I
inserted it, I notice an occasional flash of light from the LED. There is
no pattern to it that I have been able to discern. As a test I moved the
light to a different USB3 slot that came built into my tower, and it has
not flashed there at all. It would be nice to have a recording voltmeter,
but I no longer have access to one, so that option is out.

Any ideas as to what is going on, or maybe some suggestions for a very low
cost method of recording the voltage on a USB port?

Take a $0.10 LED and a resistor and connect them to the 5V copper
land on the USB tray PCB (if available) and use the light from
that LED as a "second opinion".

A couple years ago, I took the lazy way and bought two digital
voltage/current meters from Ebay. They plug into a USB port and provide
a realtime display of the voltage and current. Naturally, they provide a
pass-thru USB port, otherwise current monitoring wouldn't make any
sense.

Just now, I did an Ebay search for "USB voltage" and a bunch of the
little devices came up, although I'm not sure I see the exact ones that
I have. They're all 98% the same, though.

They're fun to play with. I used them to see how much current my phone
versus my tablet would draw during charging.

Char Jackson wrote:
On Fri, 09 Feb 2018 02:45:49 -0500, Paul wrote:

Tim wrote:
Is it normal for USB voltage to spike now and then? I ask because I have
several of those little LEDs on a chip USB lights, and I put one into one
of the USB3 slots I have in the front of my tower case. The slot in
question is one of four on an add on card I installed where the diskette
drive used to go on older systems. It has four slots, with two USB3 A cords
to supply data and power, and s separate four pin power connection for
additional power. My tower sits beside my desktop, and I wanted to see if
adding this light would help light up my keyboard at night. Shortly after I
inserted it, I notice an occasional flash of light from the LED. There is
no pattern to it that I have been able to discern. As a test I moved the
light to a different USB3 slot that came built into my tower, and it has
not flashed there at all. It would be nice to have a recording voltmeter,
but I no longer have access to one, so that option is out.

Any ideas as to what is going on, or maybe some suggestions for a very low
cost method of recording the voltage on a USB port?
Take a $0.10 LED and a resistor and connect them to the 5V copper
land on the USB tray PCB (if available) and use the light from
that LED as a "second opinion".

A couple years ago, I took the lazy way and bought two digital
voltage/current meters from Ebay. They plug into a USB port and provide
a realtime display of the voltage and current. Naturally, they provide a
pass-thru USB port, otherwise current monitoring wouldn't make any
sense.

Just now, I did an Ebay search for "USB voltage" and a bunch of the
little devices came up, although I'm not sure I see the exact ones that
I have. They're all 98% the same, though.

They're fun to play with. I used them to see how much current my phone
versus my tablet would draw during charging.

The reason I didn't suggest that, is the sample rate
of some of that stuff is too low to waste money on it.

You'd think after all these years had passed, that someone
would make a "single chip DSO", but it doesn't seem to have
happened yet. We can have some guy in China make a coin
mining chip (i.e. something having no significant utility),
but we can't have any really useful stuff.

Paul

Tim wrote:

ASUS F2A85-V Pro

That mobo has 10 USB2 ports: 2 at the backpanel and 8 onboard.
Hopefully you didn't connect the USB3 drive bay hub to a USB2 header on
the mobo. You mentioned the mobo end of the cable was 20-pin which
would match with a 20-pin header on the mobo for a USB connection.

https://www.startech.com/Cards-Adapt...ay~35BAYUSB3S4

There are already 4 LEDs in the hub, one for each USB port in the hub;
however, the 20-pin USB3 header is only for 2 USB ports so a pair of
USB3 ports in the hub must be shared to one mobo USB3 port or one of the
USB3 ports in the mobo header is wasted and only 1 mobo USB3 port gets
shared amongst all 4 USB3 ports in the hub. With the SATA power
connection, it becomes a powered hub so you don't have to share power
from a single USB3 mobo port amongst 4 USB3 hub ports - but it still
seems bandwidth would suffer in having to share 4 USB3 hub ports to a
singles USB3 mobo port. That's how hubs work.

The LEDs in the hub are apparently only as device-link indicators. That
is, they light up when an active device is plugged into a hub port and
the hub is properly communicating with the controller on the mobo. If
they were simply indicating "I've got power from the mobo's USB port"
then only 1 LED would be needed. All the manual says is "LEDS: 4 x USB
Device Link". Not sure why you need an always-on LED just to say some
other USB port on the mobo can supply power.

So how does the Banggood PCB with LED come into play here? You already
consumed an entire mobo USB 2-port header to use only one of them to
route to the drive bay hub. Where does the this separate PCB get its
USB power for its LED? It has a USB connector but did you connect that
PCB to a USB2 header on the mobo? A USB2 port should deliver 500 mA
which would be sufficient for the 200 mA load of separate the LED on
PCB. 500 mA is the theoretical or rated current load. I've seen USB
ports go bad. I've got one now in my desktop PC that cannot supply
sufficient power for anything plugged into that port. The port went
bad. Took a while of diagnosing the cause of some USB ills to track to
that port which now has some self-stick black velcro covering its
opening (and I added a USB daughtercard to give me more ports, anyway).

Not sure why you want a LED to tell you there is power available but on
a different USB port on the mobo than is used for the USB drive bay hub.
Ah, maybe that's why you mentioned using a Y-adapter cable. You have a
20-pin mobo connector to a USB3 port on the mobo. Half of that
connection goes to the USB3 drive bay hub using one of the type A USB
cable ends. The other type A USB cable end goes to your LED PCB. Okay,
but that still has the LED PCB showing power is available for a
different port than what the hub is getting. I'm assuming the Y-adapter
is to have separate A-type ends - one for each USB connection in the
20-pin mobo header - not that the A-type ends come from the same one
USB3 mobo connection (where only one port gets used to both the hub and
LED PCB and the other USB3 mobo port gets wasted). Does each A-type
cable end come from the same USB3 port on the mobo (not the same header
on the mobo but the same controller port on half of the mobo's header)?
Or do the A-type cable ends connect separately to both USB ports in the
mobo's header (that is, you don't end up wasting a mobo port)?

With a digital voltmeter, you should be able to probe the voltage on the
PCB pads where the LED is soldered. What voltage do you read there?

With the USB3 hub occupying a 3.5" drive bay, I'm curious as to where
this LED PCB contraption is mounted so you can see its LED.

VanguardLH wrote in :

Tim wrote:

ASUS F2A85-V Pro

That mobo has 10 USB2 ports: 2 at the backpanel and 8 onboard.
Hopefully you didn't connect the USB3 drive bay hub to a USB2 header
on the mobo. You mentioned the mobo end of the cable was 20-pin which
would match with a 20-pin header on the mobo for a USB connection.

https://www.startech.com/Cards-Adapt...3-Front-Panel-
4-Port-Hub-35-5-25-inch-Drive-Bay~35BAYUSB3S4

There are already 4 LEDs in the hub, one for each USB port in the hub;
however, the 20-pin USB3 header is only for 2 USB ports so a pair of
USB3 ports in the hub must be shared to one mobo USB3 port or one of
the USB3 ports in the mobo header is wasted and only 1 mobo USB3 port
gets shared amongst all 4 USB3 ports in the hub. With the SATA power
connection, it becomes a powered hub so you don't have to share power
from a single USB3 mobo port amongst 4 USB3 hub ports - but it still
seems bandwidth would suffer in having to share 4 USB3 hub ports to a
singles USB3 mobo port. That's how hubs work.

The LEDs in the hub are apparently only as device-link indicators.
That is, they light up when an active device is plugged into a hub
port and the hub is properly communicating with the controller on the
mobo. If they were simply indicating "I've got power from the mobo's
USB port" then only 1 LED would be needed. All the manual says is
"LEDS: 4 x USB Device Link". Not sure why you need an always-on LED
just to say some other USB port on the mobo can supply power.

So how does the Banggood PCB with LED come into play here? You
already consumed an entire mobo USB 2-port header to use only one of
them to route to the drive bay hub. Where does the this separate PCB
get its USB power for its LED? It has a USB connector but did you
connect that PCB to a USB2 header on the mobo? A USB2 port should
deliver 500 mA which would be sufficient for the 200 mA load of
separate the LED on PCB. 500 mA is the theoretical or rated current
load. I've seen USB ports go bad. I've got one now in my desktop PC
that cannot supply sufficient power for anything plugged into that
port. The port went bad. Took a while of diagnosing the cause of
some USB ills to track to that port which now has some self-stick
black velcro covering its opening (and I added a USB daughtercard to
give me more ports, anyway).

Not sure why you want a LED to tell you there is power available but
on a different USB port on the mobo than is used for the USB drive bay
hub. Ah, maybe that's why you mentioned using a Y-adapter cable. You
have a 20-pin mobo connector to a USB3 port on the mobo. Half of that
connection goes to the USB3 drive bay hub using one of the type A USB
cable ends. The other type A USB cable end goes to your LED PCB.
Okay, but that still has the LED PCB showing power is available for a
different port than what the hub is getting. I'm assuming the
Y-adapter is to have separate A-type ends - one for each USB
connection in the 20-pin mobo header - not that the A-type ends come
from the same one USB3 mobo connection (where only one port gets used
to both the hub and LED PCB and the other USB3 mobo port gets wasted).
Does each A-type cable end come from the same USB3 port on the mobo
(not the same header on the mobo but the same controller port on half
of the mobo's header)? Or do the A-type cable ends connect separately
to both USB ports in the mobo's header (that is, you don't end up
wasting a mobo port)?

With a digital voltmeter, you should be able to probe the voltage on
the PCB pads where the LED is soldered. What voltage do you read
there?

With the USB3 hub occupying a 3.5" drive bay, I'm curious as to where
this LED PCB contraption is mounted so you can see its LED.


You're not listening. The indicator LEDs have nothing to do with this
discussion. Visualize a very small circuit board about 1/2"x3". On one
end there are the USB comtacts. On the other end are three COB LEDs on
one side, along with the associated electronics. When plugged into a USB
port, the LEDs light. What I noticed when I did this was that there was
an occasional flash of brighter light. That led to my posting. The
interesting thing is I no longer see any bright flashes. Hmmm. As the
song says, "Strange things happen in this world!" [50s teen ballad song
actually]

Yes the mobo has four 10-pin USB2 sockets and one 20-pin USB3 socket. I
purchased a cable with a USB 3.0 20-pin connector on one end, and a USB
3.0 Type A female connector on the other. I then used that Type A
connector to feed the ScanTech hub.

My reason for posting my question is that there appears to be several
very knowledgable electronic and computer gurus watching this group. My
concern was if this is something I should be worried about or not. I can
now answer my own question that it appears to have been a transitory
phenomenom, since the flashes no longer appear. I suspect some sort of
oxidation on the contacts of the circuit board that was removed as it was
inserted and removed.

Char Jackson wrote in
:

On Fri, 09 Feb 2018 02:45:49 -0500, Paul
wrote:

Tim wrote:
Is it normal for USB voltage to spike now and then? I ask because I
have several of those little LEDs on a chip USB lights, and I put
one into one of the USB3 slots I have in the front of my tower case.
The slot in question is one of four on an add on card I installed
where the diskette drive used to go on older systems. It has four
slots, with two USB3 A cords to supply data and power, and s
separate four pin power connection for additional power. My tower
sits beside my desktop, and I wanted to see if adding this light
would help light up my keyboard at night. Shortly after I inserted
it, I notice an occasional flash of light from the LED. There is
no pattern to it that I have been able to discern. As a test I moved
the light to a different USB3 slot that came built into my tower,
and it has not flashed there at all. It would be nice to have a
recording voltmeter, but I no longer have access to one, so that
option is out.

Any ideas as to what is going on, or maybe some suggestions for a
very low cost method of recording the voltage on a USB port?


A couple years ago, I took the lazy way and bought two digital
voltage/current meters from Ebay. They plug into a USB port and
provide a realtime display of the voltage and current. Naturally, they
provide a pass-thru USB port, otherwise current monitoring wouldn't
make any sense.

Just now, I did an Ebay search for "USB voltage" and a bunch of the
little devices came up, although I'm not sure I see the exact ones
that I have. They're all 98% the same, though.

They're fun to play with. I used them to see how much current my phone
versus my tablet would draw during charging.


I have one of those but unfortunately the response time is not fast
enough to register a possible voltage spike like I suspected this was. To
be sure of registering such a spike properly I would assume a sampling
rate of around 250-300 samples per second would be required.

And yes, they are fun to play with. That is how I determined that the LED
light in question draws 200ma of current.

Tim wrote:

VanguardLH wrote in :

Tim wrote:

ASUS F2A85-V Pro

That mobo has 10 USB2 ports: 2 at the backpanel and 8 onboard.
Hopefully you didn't connect the USB3 drive bay hub to a USB2 header
on the mobo. You mentioned the mobo end of the cable was 20-pin which
would match with a 20-pin header on the mobo for a USB connection.

https://www.startech.com/Cards-Adapt...3-Front-Panel-
4-Port-Hub-35-5-25-inch-Drive-Bay~35BAYUSB3S4

There are already 4 LEDs in the hub, one for each USB port in the hub;
however, the 20-pin USB3 header is only for 2 USB ports so a pair of
USB3 ports in the hub must be shared to one mobo USB3 port or one of
the USB3 ports in the mobo header is wasted and only 1 mobo USB3 port
gets shared amongst all 4 USB3 ports in the hub. With the SATA power
connection, it becomes a powered hub so you don't have to share power
from a single USB3 mobo port amongst 4 USB3 hub ports - but it still
seems bandwidth would suffer in having to share 4 USB3 hub ports to a
singles USB3 mobo port. That's how hubs work.

The LEDs in the hub are apparently only as device-link indicators.
That is, they light up when an active device is plugged into a hub
port and the hub is properly communicating with the controller on the
mobo. If they were simply indicating "I've got power from the mobo's
USB port" then only 1 LED would be needed. All the manual says is
"LEDS: 4 x USB Device Link". Not sure why you need an always-on LED
just to say some other USB port on the mobo can supply power.

So how does the Banggood PCB with LED come into play here? You
already consumed an entire mobo USB 2-port header to use only one of
them to route to the drive bay hub. Where does the this separate PCB
get its USB power for its LED? It has a USB connector but did you
connect that PCB to a USB2 header on the mobo? A USB2 port should
deliver 500 mA which would be sufficient for the 200 mA load of
separate the LED on PCB. 500 mA is the theoretical or rated current
load. I've seen USB ports go bad. I've got one now in my desktop PC
that cannot supply sufficient power for anything plugged into that
port. The port went bad. Took a while of diagnosing the cause of
some USB ills to track to that port which now has some self-stick
black velcro covering its opening (and I added a USB daughtercard to
give me more ports, anyway).

Not sure why you want a LED to tell you there is power available but
on a different USB port on the mobo than is used for the USB drive bay
hub. Ah, maybe that's why you mentioned using a Y-adapter cable. You
have a 20-pin mobo connector to a USB3 port on the mobo. Half of that
connection goes to the USB3 drive bay hub using one of the type A USB
cable ends. The other type A USB cable end goes to your LED PCB.
Okay, but that still has the LED PCB showing power is available for a
different port than what the hub is getting. I'm assuming the
Y-adapter is to have separate A-type ends - one for each USB
connection in the 20-pin mobo header - not that the A-type ends come
from the same one USB3 mobo connection (where only one port gets used
to both the hub and LED PCB and the other USB3 mobo port gets wasted).
Does each A-type cable end come from the same USB3 port on the mobo
(not the same header on the mobo but the same controller port on half
of the mobo's header)? Or do the A-type cable ends connect separately
to both USB ports in the mobo's header (that is, you don't end up
wasting a mobo port)?

With a digital voltmeter, you should be able to probe the voltage on
the PCB pads where the LED is soldered. What voltage do you read
there?

With the USB3 hub occupying a 3.5" drive bay, I'm curious as to where
this LED PCB contraption is mounted so you can see its LED.


You're not listening. The indicator LEDs have nothing to do with this
discussion. Visualize a very small circuit board about 1/2"x3". On one
end there are the USB comtacts. On the other end are three COB LEDs on
one side, along with the associated electronics. When plugged into a USB
port, the LEDs light. What I noticed when I did this was that there was
an occasional flash of brighter light. That led to my posting. The
interesting thing is I no longer see any bright flashes. Hmmm. As the
song says, "Strange things happen in this world!" [50s teen ballad song
actually]

Yes the mobo has four 10-pin USB2 sockets and one 20-pin USB3 socket. I
purchased a cable with a USB 3.0 20-pin connector on one end, and a USB
3.0 Type A female connector on the other. I then used that Type A
connector to feed the ScanTech hub.

My reason for posting my question is that there appears to be several
very knowledgable electronic and computer gurus watching this group. My
concern was if this is something I should be worried about or not. I can
now answer my own question that it appears to have been a transitory
phenomenom, since the flashes no longer appear. I suspect some sort of
oxidation on the contacts of the circuit board that was removed as it was
inserted and removed.

I figured an add-on card in a drive bay that has USB was you adding a
USB hub. Then you mentioned the Startech USB hub in another reply. How
could I get confused you were talking about a USB hub? Instead of
bringing in the USB hub, you should've said you weren't asking about a
USB hub, not mentioned a USB hub in a drive bay, and merely said you
were asking about just a custom PCB with LED(s) that connects to USB
ports. So, we'll go from there now ...

Since the introduction of LED light bulbs to supplant incandescents,
flicker has become a pervasive problem in the lighting industry. I
photo labs with LED lighting, the momentary flicker can result in
underexposed photos. The flicker can cause headaches. However,
lighting runs off of A/C which can vary its frequency, like to adjust
for number of cycles per day for A/C-synchronized clocks, can have
brownouts or surges, incur spikes, and suffer from RF. You can hope the
USB power (at whatever USB port into which you plug your PCB kit)
doesn't suffer any of those problems and has a constant +5V (or very
close to that but with fluctuation, ripple, spikes, or RF). You said
the flicker is noticed with one USB port but not with another. Could be
the power from one port is well regulated where the other may not be.
Since USB ports are paired to a controller, it's possible the other
paired USB port is connected to equipment that irregularly sucks down a
varying power load. LEDs flicker when dimmed too much, so perhaps the
resistor for the LED was improper chosen so a variance in USB port
voltages could be just enough to trigger the LED dimming flicker.
Circuitry on the PCB could be very simplistic. For example, no ballast
or regulator to control the current to the LED.

https://www.banggood.com/0_2W-WhiteW...r_warehouse=CN

Don't know if that's what you have. There is almost nothing on its PCB
other than contacts, one chip and the COB LEDs. No way to identify what
that surface mount chip does. Seems something designed to simply
provide some light but not guarantee level of output or complete
eradication of flicker due to any source variation. It's just a light,
not a metering gauge to reflect state of the power source. It's like
using an analog voltage meter that has never been calibrated for many
years or decades: just shows you relative differences in voltage and a
general reading. Most users don't get their DVMs calibrated, either.

https://www.adafruit.com/product/1852
http://tinyurl.com/ybdtosb3
https://www.amazon.com/dp/B013FANC9W/?tag=hotoge-20

Those might be something you'd use to see just what voltage and load was
on a particular USB port. However, because of its low cost, I'm not
sure how accurate it would be. A USB power meter with logging would cost
more and require software to do the logging but it's possible its
polling interval could miss when there was a momentary power fluctuation
that would make flicker your LED PCB.

https://www.indiegogo.com/projects/p...wer-analyzer#/

Pretty expensive to get logging added to the USB power monitor.

http://www.electronicdesign.com/test...tocol-analyzer

I didn't look much into that solution but is cheaper ($200). From the
screenshot, looks like voltage gets logged. Voltage is shown in the
display and since logging is supported then supposedly voltage gets
logged, too. I'd have to dig into that one to see if it was appropriate
to logging the unloaded and loaded voltage on a USB port and if sampling
rate can be shortened to something really tiny to catch those infrequent
momentary flickers you notice.

At this point, can't tell if there's a problem with the USB port or with
the LED PCB. Unless something bad is happening, I wouldn't bother
investigating based on the flickering of an LED on a $2 gizmo.

Tim wrote:


You're not listening. The indicator LEDs have nothing to do with this
discussion. Visualize a very small circuit board about 1/2"x3". On one
end there are the USB comtacts. On the other end are three COB LEDs on
one side, along with the associated electronics. When plugged into a USB
port, the LEDs light. What I noticed when I did this was that there was
an occasional flash of brighter light. That led to my posting. The
interesting thing is I no longer see any bright flashes. Hmmm. As the
song says, "Strange things happen in this world!" [50s teen ballad song
actually]

Yes the mobo has four 10-pin USB2 sockets and one 20-pin USB3 socket. I
purchased a cable with a USB 3.0 20-pin connector on one end, and a USB
3.0 Type A female connector on the other. I then used that Type A
connector to feed the ScanTech hub.

My reason for posting my question is that there appears to be several
very knowledgable electronic and computer gurus watching this group. My
concern was if this is something I should be worried about or not. I can
now answer my own question that it appears to have been a transitory
phenomenom, since the flashes no longer appear. I suspect some sort of
oxidation on the contacts of the circuit board that was removed as it was
inserted and removed.

Did you examine the wiring pattern of the LEDs ?

The white LEDs I have here, run on about 2.5V.
If you were to put two in series, and put them
across the USB 5V rails without a series resistor,
that would make the brilliance of the LEDs
very voltage dependent.

On the other hand, if the LEDs each had a series
resistor, then 5-2.5 = 2.5V across the resistor.
If the voltage changed to 2.6V across the resistor,
the change in current flow would be quite small.
An inefficient implementation that relies on
series resistance, should be a bit more immune to
intensity variation.

Since your design has three COB LEDs, I don't see
a good way to hook up that number of LEDs. Series
resistor would be an easy way to do it, but wasteful.

On the other hand, some LEDs consist of multiple LEDs
inside a single package, either uncommitted, or hooked
in series or parallel. (This is possible because they've
become very good at binning or matching them.) Then,
even if you see three objects, there might be an
opportunity for weird wiring.

You can also spot the wiring pattern, by applying
a supply to the PCB using a bench supply. If the
intensity is strongly linked to supply setting,
then it's probably the kooky option of sticking
two LEDs right across the rails.

I have one LED lighting solution here, where the
circuit has no resistors in it, but that's possible
because the power source feeding it is a current
source, and cannot burn out the LEDs.

Paul

On Fri, 09 Feb 2018 11:48:21 -0600, Char Jackson
wrote:

On Fri, 09 Feb 2018 02:45:49 -0500, Paul wrote:

Tim wrote:
Is it normal for USB voltage to spike now and then? I ask because I have
several of those little LEDs on a chip USB lights, and I put one into one
of the USB3 slots I have in the front of my tower case. The slot in
question is one of four on an add on card I installed where the diskette
drive used to go on older systems. It has four slots, with two USB3 A cords
to supply data and power, and s separate four pin power connection for
additional power. My tower sits beside my desktop, and I wanted to see if
adding this light would help light up my keyboard at night. Shortly after I
inserted it, I notice an occasional flash of light from the LED. There is
no pattern to it that I have been able to discern. As a test I moved the
light to a different USB3 slot that came built into my tower, and it has
not flashed there at all. It would be nice to have a recording voltmeter,
but I no longer have access to one, so that option is out.

Any ideas as to what is going on, or maybe some suggestions for a very low
cost method of recording the voltage on a USB port?

Take a $0.10 LED and a resistor and connect them to the 5V copper
land on the USB tray PCB (if available) and use the light from
that LED as a "second opinion".

A couple years ago, I took the lazy way and bought two digital
voltage/current meters from Ebay. They plug into a USB port and provide
a realtime display of the voltage and current. Naturally, they provide a
pass-thru USB port, otherwise current monitoring wouldn't make any
sense.

Just now, I did an Ebay search for "USB voltage" and a bunch of the
little devices came up, although I'm not sure I see the exact ones that
I have. They're all 98% the same, though.

They're fun to play with. I used them to see how much current my phone
versus my tablet would draw during charging.

They are good and cheap. I also use them to check charging rates.