How does grounding work if I cut the 5V pads or connect schottsky diodes while keeping the USB plugged in?

liv424

Member
I saw on both the teensy site (https://www.pjrc.com/teensy/external_power.html) and the following post (https://forum.pjrc.com/index.php?threads/external-power-and-usb.55400/) solutions to using an external power supply while keeping the USB plugged in. I am using a teensy 4.1 if that makes a difference.

In both cases, I am confused how ground is referenced. If 5V pads are cut, or the red wire in the USB cable is cut, only the power is cut and the ground from the USB is still connected to the teensy.

Similarly, when the schottsky diode is used, the power and ground lines from the USB are still connected.

In these scenarios, how is ground referenced? For example during ADC measurements will it reference the ground from the external power supply or the USB and when will it reference which? All I can find is that the ADC references the 3.3V reference but this is 3.3V referenced to which ground? The USB ground or the external power supply ground? Would I have a ground loop if the two grounds are different?
 
If they are all connected and lets say the ground from the usb connection is 1 volt higher than the ground from the external, how big of a problem would this be?
 
ALL GROUNDS SHOULD BE AT THE SAME POTENTIAL. IF THEY ARE NOT YOU WILL NOT HAVE A RELIABLY FUNCTIONING SYSTEM.
INPUT AND OUTPUT DATA WILL BE ALL OVER THE PLACE.
 
So there would be no way to have a teensy that can be powered by both USB and VIN independently and at the same time if the two are two isolated. So I would need a removable jumper in this case rather than soldering a schottsky
 
Connect it like below.


z1.png
 
Last edited:
They ARE common, they are connected together inside the Teensy.
Look at the Teensy schematic. See all those earths, they are ALL connected together.
z.png
 
If you need your Teensy to have a different ground potential than the computer (USB GND pin potential), you do need to use an USB isolator.

Analog Devices' ADuM3160 and ADuM4160 full-speed isolators work fine with Teensies. Teensy 4.x/MicroMod automatically limit themselves to full speed, or 12 Mbit/s (or about a million bytes per second of payload data); older Teensies are limited to full-speed and their data rate is not affected by the isolator.

With ADuM3160, you only need four 24Ω 1% resistors (in series with the D+ and D- lines on both sides of the isolator chip), and four 100nF X7R/C0G/NP0 supply bypass capacitors, and a careful board placement as the datasheet application information describes, and it'll work well.

The cheap (<USD $10) ADuM3160-based USB isolators you see at eBay use that circuit with a DIP switch on the SPD pin (low-speed/full-speed selection) and a varying isolated DC-DC converter, and work just fine with Teensies; the only difference is the quality of the DC-DC converters, really.

In your case, the DC-DC converter is a hindrance, as the isolator device/Teensy side is powered externally. (It depends on the exact implementation whether one can simply remove the DC-DC converter or not. The isolated ones have a tendency to not regulate their output well at very low loads.) The device/Teensy side can use either 5V or 3.3V supply; see the datasheet for details.

Analog Devices' ADuM3165 and ADuM3166 and Texas Instruments' ISOUSB211 are similar USB isolator chips, but also support high-speed (480 Mbit/s) like Teensy 4.x/MicroMod do.

ISOUSB211 datasheet figure 10-2 describes a self-powered isolated peripheral case. You need a bunch of capacitors, a ferrite bead for filtering/blocking high-frequency noise from the host USB voltage to the isolator chip (with less than 100mΩ DC resistance), and ESD diodes similar to PESD5V0C1USF on the host/computer side for safety; plus optionally one or two 1.8V linear regulators (that will need to supply up to 100mA) to reduce the heat generated inside the chip (when using its internal LDOs) to provide the 1.8V rails needed for high-speed USB operation.

ADuM3166 requires four 100nF X7R/C0G/NP0 supply bypass capacitors, and a 24 MHz crystal with suitable load capacitors. Host/computer side is powered from the USB 5V line, but you can use either 5V or 3.3V to power the device/Teensy side. ESD diodes on the host/computer side are a good idea here too, but the datasheet does not mention a ferrite, so I assume one is not needed or useful.

I have not yet seen any affordable high-speed isolators based on ADuM3166, ADuM3165, or ISOUSB211, unfortunately. Mouser and others do sell the development board for ~ 60€ (+ VAT), but they are for testing and experimenting the chips' capabilities rather than simply using them. I have seen others use these chips to successfully create their own USB 2.0 low/full/high-speed isolators, though, and do intend to make one for myself as well. JLCPCB often has some of these in stock (for ~ USD $16 apiece in singles for PCB assembly), so one could use their assembly service to get almost ready-made boards manufactured and assembled. I personally prefer to use through-hole USB connectors, so the through-hole connectors I'd solder myself by hand.

Even though I am only a hobbyist –– you can find my designs here, they're all Public Domain / CC0-1.0 International: free to use as you wish, just don't blame me for breakage –– I do believe even I can design a functional isolator board, for example in EasyEDA. The thing is to pay attention to the datasheet, really, and the gotchas (like capacitive coupling over the isolation barrier) it points out.

At this point in time, I am using a laptop. Whenever I need the Teensy ground to match whatever external ground, I simply make sure I'm not connected to the mains (making the laptop ground "float"), and ground the USB chassis to that external ground via a high-value resistor (to dissipate the energy in the original potential difference, "slowly" drifting the laptop ground potential to the external ground potential instead of one big static Zap!), before Teensy ground connects to that external ground. So, I've thus far avoided needing the high-speed isolator in practice.

Apologies for the wall of text. :cry:
 
Lets say I get one of the cheapo ebay idolators


and then I cut the 5V cable on the usb cable that connects to the teensy, put a resistor in series with the ground of the external power supply and the ground pin to allow the "floating" ground drift towards the external ground through the teensy ground plane. Would this give me the same effect as floating the laptop and bonding the two grounds?

I think this would allow me to rapidly unplug the cable and use a different cable when I want to transfer data without the external power supply but also would allow me to transfer data while using the external power supply?


Or if i directly buy a not so cheapo isolators that claim high speed with option to directly use an isolated power supply, would these work?





(rudimentary data on the isolator above if you care https://github.com/buelec-tech/EVC9103)
 
Last edited:
Lets say I get one of the cheapo ebay isolators and then I cut the 5V cable on the usb cable that connects to the teensy
That uses K-Cut B0505S-1WR3 isolated DC-DC supply, which has a minimum load of 20mA. So, you need a cable between the isolator and Teensy where the 5V line broken, not connected to anything on the Teensy side, and on the isolator side 5V and GND connected using a 250Ω quarter-watt (or higher-rated) resistor (it will dissipate max. 0.125W = 1/8W).

This will isolate the Teensy ground from the host computer ground.

[Topping HS01 isolator]
I cannot find any teardowns or mentions of which chip this uses, but it supports only high-speed (480 Mbit/s) USB, and might only support USB Audio (UAC2), so I am not sure. Because it does not support full-speed or low-speed USB, it does not seem to be based on Texas Instruments ISOUSB211 or Analog Devices ADuM3166 or ADuM4166; I personally would look for an isolator using one of these chips instead.

[BUELEC EVC9103S isolator]
A commenter on the Amazon page mentioned their isolator did not like externally powered devices, so I don't think it'll work for you.



The EVAL-ADUM4166EBZ datasheet evaluation board (e.g. from Mouser) would work, but it is a bare board and not intended for end users. If you look at Figure 2 in the datasheet, your configuration would be similar: JP1 jumper on 3-4 (BUS), P4 jumper 5V side, but JP2 jumper on 5-6 (3.3VEXT). You'd also need to supply 3.3V at 100mA to the VEXT+ pin on the screw terminal (or the pin hole with square pad marked VBUS2 in P10, or the pad marked VDD2 in P13). Teensy 4.x can supply max. 250mA, so you'd have max. 150mA available at 3.3V for other stuff. (The Teensy side USB GND pin is connected to the VEXT- pin on the screw terminal, and to all pin holes marked GND2; the other pins in P10 and P13 are also GND2. The 3.3V is of course with respect to Teensy GND = GND2.)

Altor Audio HISO USB isolator says it is based on TI ISOUSB211, and the model III supports both USB-powered and externally powered devices. You'd need to supply (isolated) 5V to the barrel jack (image says center positive); this would be provided to the Teensy VUSB also. Teensy GND and the barrel jack negative would be the same potential.

Deshide/DSD TECH SH-G01L USB isolator and SH-G01B (but NOT SH-G01A) claim to use ADuM3165, and specs match; but it uses a built-in DC-DC converter. So, you should use a cable with VUSB cut between this and the Teensy, and on the isolator side, the VUSB and GND connected with a 250Ω 1/4W or higher-rated resistor, just like in the cheap full-speed USB isolators.

Hifime high-speed USB isolator V2 is reported to use ADuM4166 (and its specs match), but uses a built-in DC-DC converter. The same cable thing as for SH-G01L is needed.

The cheapest seems to be DSD TECH SH-G01L from DSD Tech Direct store at Amazon for about USD $30 right now, but note that you need a modified cable to use with an externally-powered Teensy.
 
I'll test out the EVC9103S since it is coming in the mail tomorrow but I will also mod a cable and try out the DSD TECH.
By the way, the EVC9120S does support different speeds and devices. I think the comment that it does not like external power supply only applies if the 5V USB internal power supply is used instead of a 12V external power supply (which can have a common ground as all other external power supplies).
1715861256630.png
 
That uses K-Cut B0505S-1WR3 isolated DC-DC supply, which has a minimum load of 20mA. So, you need a cable between the isolator and Teensy where the 5V line broken, not connected to anything on the Teensy side, and on the isolator side 5V and GND connected using a 250Ω quarter-watt (or higher-rated) resistor (it will dissipate max. 0.125W = 1/8W).

This will isolate the Teensy ground from the host computer ground
I am a little confused. Correct me if i am wrong. As I understand it, the ground coming out of the DC-DC supply is floating and I need a bleed resistor to make it the same potential as my external power supply? So on the side of the isolator, I need to connect the 5V line to the ground line with a resistor to satisfy the minimum load but do I then need a bleed resistor between the ground wire and the teensy ground pin that slowly makes it the same voltage as the external power supply?
 
I am a little confused. Correct me if i am wrong. As I understand it, the ground coming out of the DC-DC supply is floating and I need a bleed resistor to make it the same potential as my external power supply? So on the side of the isolator, I need to connect the 5V line to the ground line with a resistor to satisfy the minimum load but do I then need a bleed resistor between the ground wire and the teensy ground pin that slowly makes it the same voltage as the external power supply?
If you connect the external power supply to Teensy+isolator when nothing is powered yet, there should not be enough energy in the circuits to cause an ESD event (other than in the same scale that happens ordinarily/all the time), so no bleed resistor is needed. In fact, a resistor between the two grounds may cause other problems, so I would definitely NOT use a bleed resistor myself.

Teensy ground, isolator device side ground, and the external power supply ground, should all be considered "shorted together", always in the same potential when powered. Connecting and disconnecting Teensy from external power should only occur when not powered! The USB cable from the isolator to a host can be connected or disconnected at any time.

So yes, the only purpose for the 250Ω resistor is to "waste" the minimum load needed to keep the DC-DC converter within its operating range; about 20mA in this case (turning 81mW to 121mW to heat, so 1/4W rating or better so it can handle it continuously). Thus, the resistor is put between USB +5V and USB GND on the isolator side. On the device side, the USB +5V is cut, and not connected to anything.

Such a cable works well as a "no-power data cable" as well, when connecting self-powered devices to any host. It just wastes that 20mA on the host side; about the same as a couple of bright indicator LEDs.
 
You have to, actually. They must be directly connected for the USB isolator to work correctly.

It's the 5V output from the isolated DC-DC supply that needs a bit of care. In this case, the Teensy is powered from a separate supply (and its GND must be directly connected to Teensy GND and thus to the isolated DC-DC supply). We cannot leave the 5V from the isolator just floating, because it needs to power the device side of the isolator chip. The manufacturers of these isolators assume there is sufficient load to keep the DC-DC supply within its operating range, but if there is no device side load, that is typically not the case. A 250Ω resistor as a dummy load is the foolproof method.
 
Okay revisiting this because I am once again confused. Why do I need an isolator if other people were able to accomplish what I wanted without the isolator as shown in the following?

1717056621743.png


and also in the following threads?



Is this because they are using a battery that has a "floating" ground rather than a power supply that has a defined ground?
 
Why do I need an isolator if other people were able to accomplish what I wanted without ....

I just re-read all 17 messages, and I can't figure out what you really wanted. Maybe I missed something? I see a lot of specific technical questions, but not really any context to understand why you're asking and what you are really trying to accomplish. If you could explain why, maybe we could help better?

But so far you've been asking about 2 different things. Maybe there's some misunderstanding that they're not the same?

The diode circuit, or cutting apart the VUSB-VIN pads, is about using 2 power sources where you wish to prevent current from either from erroneously flowing into the other. It has nothing to do with ground isolation. In fact, the grounds of both power sources must be connected directly together for this to work. Preventing reverse power flow is about avoiding possibly bad scenarios like the USB power accidentally attempting to charge a non-rechargeable battery (as in the photo), or the battery attempting to discharge into your PC when its power is turned off.

Isolation is a completely different thing. Ground and power and signals are all isolated, not just ground, so please keep that in mind when we talk only of ground. Normally ground isolation is used in 2 scenarios. The first is about safety, where something connects directly to AC power lines or might accidentally do so (telco and networking cables, for example). Ground isolation products usually give ratings about high voltages they can withstand because the common use is for safety. The other usage is for protecting sensitive analog signals, usually audio. Wires aren't prefect. They have some resistance and also tiny inductance. If current flows through the ground wire, you can get small voltage difference. For signals like audio which get tremendously amplified and even tiny noise from unrelated ground currents is a problem which ground isolators are sometimes used to solve.

Not sure if this explantion really helps, because I truly just don't understand what you wanted, but I hope it at least helps clear up any possible misunderstanding.
 
Why do I need an isolator
Reread my first message. I wrote, "If you need your Teensy to have a different ground potential than the computer (USB GND pin potential), you do need to use an USB isolator."

It is critical that you understand that "1 V" describes a difference of one volt. In Teensies and other microcontrollers, all voltages are referenced to ground. Thus, in a single electrical circuit, there can only be one "ground" or zero-volt reference, to which all other voltages are measured against. Otherwise, some of the signal levels will be wrong, and can even damage the Teensy or other components.

(There is sometimes a safety ground, which is not a reference voltage, but actually connected via grounded sockets' ground pin to the physical ground. Humans are hurt or killed when they become a part of an electrical circuit with sufficient current and voltage, typically via their hands to their feet to a wet puddle or similar to physical ground. Safety ground is intended to be a better conductor, so that instead of through the human, the current flows through the safety ground instead, hopefully saving the human. It isn't foolproof, of course, especially when someone grabs one thing with one hand and another thing with the other hand, causing them to become part of the electrical circuit without any ground connection. All that depends on what path the current takes through the body: many have survived lightning strikes, but 12V from say a car battery is sufficient to kill in certain circumstances.)

Isolators are devices that "translate" signals from one reference potential side to another. Most of them are galvanic isolators, which means they do not pass any current over the barrier, only the signal in some form. Isolated DC-DC supplies pass current over the barrier, typically at around 80% efficiency: to provide say 5V 200mA = 1 W on the "downstream" side, they draw 1.25 W (for example, 5V 250mA) from the "upstream" side.

Safe USB wall warts are actually isolated AC-DC supplies. The positive output is typically 4.75V to 5.25V above the negative output, but the pair is "floating", not referenced to any specific potential. (There may be a class Y capacitor from the output to the input to keep down the electromagnetic noise generated by the supply, which can cause a "tingle" if you touch either the positive or the negative output, for example the connector on a cable with the other end connected to the supply, because the input is alternating current so the capacitor can pass a tiny amount of current. I prefer so called class II MOPP-rated (Means Of Patient Protection, i.e. "medical grade") supplies myself, without such a capacitor.) Technically, the negative output is the "ground" side, except it is not tied to any specific potential, it "floats".

Thus, if I want to create a self-powered –– as opposed to USB-powered –– Teensy project, all I need to do is either cut the VUSB trace on the Teensy, or replace it with a (Schottky) diode with sufficiently high current capability, and power it using one of my safe USB wall warts. In that case, the ground potential is set by whatever is connected to the USB ground, including both my Teensy project and even the wall wart!

I sometimes use my Teensies to examine the operation of my routers and single board computers (SBCs), streaming data to my host computer. The ones I know are powered from isolated AC-DC supplies, I can connect directly, because then my host computer sets the ground potential all the way including the router or single board device. This is the majority of cases, especially with legitimate (and safe) products.

For the others, I can either use a laptop (which is not connected ot mains) so the target device ground sets the ground potential all the way including my host laptop; or, I can use an USB isolator, making a ground barrier at the USB connection. The cheap ~ USD $10 USB isolators off eBay and elsewhere, based on ADuM3160/4160, work just fine with Teensies, although you do need to check which isolated DC-DC converter they use: everything except the DC-DC converter providing power over the isolation barrier is straight off the datasheet, but the amount of current and other properties vary among DC-DC converters quite a bit.

(There are now Analog Devices' ADuM3165/3166/4165/4166 and TI ISOUSB211 chips, that are only slightly more complicated, that provide isolation for USB 2.0 low speed, full speed, plus high speed (480 Mbit/s), i.e. all that Teensies themselves support, that require about USD $20 in parts or so when bought from Digikey/Mouser in singles. Unfortunately, these haven't become common yet, and those that exist, seem to have pretty "dumb" isolated DC-DC supply circuits. Many users would prefer one completely without.)

One of my favourite isolation schemes with Teensies, however, is when I want to connect my host computer to the serial port on a router or SBC I'm investigating and not sure yet: TI ISO6721 digital isolator. On the Teensy side, it is connected to 3.3V, RX, TX, and GND. On the other side, it is connected to VCC (1.8V or 2.25V – 5.5V), TX, RX, and GND. (On both sides, you also need a 0.1µF X7R bypass capacitor between the supply and GND. See here for my break-out board design for this.) In this case, the digital isolator is the barrier. It does need to be powered from both sides, but it doesn't need much current. On the router/SBC side, the supply also sets the RX and TX signaling levels. This way, even if the router/SBC is somehow grounded to a specific potential, everything is referenced to a single ground potential, with the ISO6721 as an isolation barrier; and it all works safely.

Whenever two points at different voltage potentials are connected with a conductor, current will flow until the voltage difference vanishes. When you have a circuit where two ground points are somehow tied to a specific but different potential, you get a ground loop. Depending on the voltage and current, this can be irrelevant, annoying (causing noise), troublesome (causing problems), or even damage your circuits. If you follow the rule of only tying the ground to a specific potential at a single point, this will not happen.



I suspect the confusion stems from a misunderstanding; not realizing that voltage is always a relative measurement. You physically cannot say or measure that a point in a circuit is at X volts, without either explicitly or implicitly specifying a point or potential it is measured relative to. This is why multimeters have two connectors: one is for the reference potential ("ground"), and the other is for the point or thing you want to measure the voltage at/of.

In digital circuits and many analog circuits, the signals are defined by the voltage levels. For USB, this is the USB ground. For Teensies, this is the USB ground, or if not connected to USB, the power supply ground (negative side). When connected to both, they must have the same ground potential.

Because all safe USB wall warts are isolated AC-DC supplies, whose positive output is about +5V with respect to their negative output, but the pair can float (within limitations, of course!), powering Teensy from such while connected to USB typically only requires one to cut the VUSB trace or add a couple of diodes (two, so that neither USB nor the external supply will try to power the other, or "backfeed" the other; only power the Teensy). In my part of the world, all legal/safe DC wall warts are similarly isolated, so diodes will suffice for normal self-powered projects.
(LED lighting supplies are a typical example of non-isolated, mains-referenced DC supplies. They may or may not be legal in your jurisdiction, so do not just blindly assume everything should be isolated. Similarly, cheap eBay/AliExpress/Temu USB wall warts may be mains-referenced, no matter how dangerous or illegal that might be.)

There are other scenarios, however, especially connecting two separately powered devices (like a desktop computer connected to mains, and a router or single-board computer connected to mains, for example) where isolation can be desired or needed. To determine for yourself, you must check if any (and how many) points tie your ground to a specific potential.
 
Last edited:
Thank you. I guess one last question. When you say USB wall wars are isolated AC-DC supplies, are they also floating? What about a laptop charger? Is the center prong of the plug purely for safety and does not actually set the ground potential? Lets say I have two laptop chargers, can I connect the two negatives together to make them the same ground potential? Earlier, you said that you make sure the laptop is not plugged in and therefore not connected to mains. Would that be a contradiction telling me that no, I should not connect the two negatives of two laptop chargers? Why is this the case?

What about the common dc barrel power supplies? Some forums mention a laptop USB not having a stable ground, if I use a floating power supply, does this mean the ground is not "stable?" Does plugging in a laptop make the ground any more stable?
 
Thank you. I guess one last question. When you say USB wall wars are isolated AC-DC supplies, are they also floating? What about a laptop charger? Is the center prong of the plug purely for safety and does not actually set the ground potential? Lets say I have two laptop chargers, can I connect the two negatives together to make them the same ground potential? Earlier, you said that you make sure the laptop is not plugged in and therefore not connected to mains. Would that be a contradiction telling me that no, I should not connect the two negatives of two laptop chargers? Why is this the case?

What about the common dc barrel power supplies? Some forums mention a laptop USB not having a stable ground, if I use a floating power supply, does this mean the ground is not "stable?" Does plugging in a laptop make the ground any more stable?
We're going round in circles here. I personally believe that you are making things more difficult for yourself.
Can you PLEASE answer Paul's comment in message #18.
I suspect that you are chasing a problem that does not really exist, but without further information along the lines of what Paul has suggested I fear that we are just going round in circles.

We want to help but just feel frustrated at the moment, as I suspect you are.
 
When you say USB wall wars are isolated AC-DC supplies, are they also floating?

In this situation (low power, only 2 pin AC power input), "floating" and "isolated" mean pretty much the same thing, that there is no metal connection between the output wires and any of the input wires.


What about a laptop charger? Is the center prong of the plug purely for safety and does not actually set the ground potential?

It depends. You really need to check the specs of each power supply, if the input has an earth ground. Nobody can give you an certain answer for unknown or hypothetical power supplies because at least 3 different types exist. This page has details and pictures of the symbols usually printed on the label for each type.


Even if the power supply has no reliable specs or datasheet, look for the 2 squares icon which means "Class II" (isolated).

Your latest message has several more questions, but I'm going to stop here (at least for now) because I really don't understand why you're asking all this and what you're really trying to accomplish.
 
But just another quick comment on the words "isolated" and "floating". While they have similar meaning, usually you would use the word "floating" to describe the device or equipment receiving power. When something is said to be floating, typically the inference implies that thing isn't intrinsically made that way. The word "floating" is usually understood to mean the "floating" thing could be grounded or connected to some other non-ground reference, but is floating because of the way it is installed or connected. A battery powered USB device would normally be floating, but could become grounded rather that floating when its USB cable is plugged in. Whether it remains floating or becomes grounded depends on the power source of the PC you plug the cable into.

The word "isolated" normally is used to describe a device designed to deliver power or signals without conductive connection. When something is said to be "isolated", usually that word implies the intrinsic design of that particular device.
 
I fully agree with Paul above, both in intent and definitions.

By "floating", I specifically meant that the supply does not tie its outputs to any specific potential; it only keeps the positive output a specific voltage above the negative output, up to a maximum current draw; and only up to a limit (difference to input voltages, often in the range of 1000 V or so). I hope I haven't caused any misconceptions! :oops:

When you say USB wall wars are isolated AC-DC supplies, are they also floating? What about a laptop charger?
The legal wall warts here in Finland are isolated, with their output not tied to any potential, as long as it is within about 500V of the mains voltage. Many of the cheapest ones sold at eBay and elsewhere would be illegal here. I recommend you take a look at DiodeGoneWild and BigClive USB charger teardowns at Youtube, because it shows how dangerous some can be.

Cheapness does not necessarily mean bad or dangerous, however, because for example Ikea Småhagel 5V 2.4A (BigClive's teardown video at Youtube) are well made and safe, even though they only cost 3€ apiece here right now.

I have both kinds of laptop supplies: class II ones where their output is not tied to any potential, as well as high-power proper ground-referenced ones that are only supposed to be used with grounded mains outlets (using Schuko F plugs and sockets here in Finland, similar to Germany), with their negative output voltage ordinarily at a specific potential with respect to safety ground. I am not sure, but I suppose this is because of efficiency and safety reasons. Simply put, it varies.

The one major difference here is non-user-accessible LED supplies: these may be non-isolated, and their current-limited DC output is referenced to mains voltage. This means that if you see "12V DC supply" here, one must still check what kind of supply it is. No shortcuts, no rules of thumb!

Is the center prong of the plug purely for safety and does not actually set the ground potential?
Not necessarily. I have seen both kinds of laptop chargers: ones where the safety ground was only connected to the supply and not output, and ones where the output negative side is (loosely) tied to the safety ground potential. (The connection was not direct, as very little current would pass from the output to the safety ground, but it was enough to show that with a simple load like a LED and a resistor, the negative output was within a fraction of a volt to the safety ground potential.)

Supplies where the output is not referenced to any particular potential, are class II. Not all isolated supplies are class II. You need to check.

Lets say I have two laptop chargers, can I connect the two negatives together to make them the same ground potential?
If at least one of them is class II then yes. Otherwise no.

Earlier, you said that you make sure the laptop is not plugged in and therefore not connected to mains. Would that be a contradiction telling me that no, I should not connect the two negatives of two laptop chargers? Why is this the case?
I don't know what kind of laptop chargers you have!

But, when your laptop is NOT connected to mains, and aside from a wired mouse and/or keyboard, the only cable it is connected to is to your project, then we are 100% sure that the laptop ground –– and thus USB ground –– is not tied to any specific potential on the laptop end. It is the safe, simple configuration!

Disconnecting your laptop from mains (and all electrical cables to other equipment except your project) makes things safe because it eliminates all laptop-side causes of ground-related issues. It does not mean it is always necessary; just that it is sufficient.

What about the common dc barrel power supplies?
The same thing. If class II –– their output not tied to any specific potential, just positive output a specific voltage above the negative output –– then their negatives can be connected together.

Some forums mention a laptop USB not having a stable ground, if I use a floating power supply, does this mean the ground is not "stable?" Does plugging in a laptop make the ground any more stable?
In this context, the "not stable" refers to the USB GND not having a fixed voltage (potential difference) with respect to something else, for example a safety ground or external true ground. In most cases, this is completely okay, and has no effect on your USB circuits: it's like believing that the neighboring village gets a different amount of sunlight than we do, without never visiting there.

If the ground gets coupled to a periodic signal, it can cause issues, because things like capacitors and inductors within the circuit can detect the change. Even then, it is the ground being coupled to some external signal that is the problem, not the "not stable" ground.

Whether plugging in a laptop charger affects that, depends on both the laptop and the charger. Most laptops with a barrel jack connector, the negative side is the common ground across the laptop, including on the USB sockets. In any case, some laptop chargers' output is referenced to (safety) ground, and some aren't, so it varies.
 
@Nominal Animal I think you are complicating/confusing what @live424 needs/wants.

I am sure that he is new to Teensy and electronics and has become confused by all the complexities.

If he can come back to us answering Pauls question in #18 I am sure that we can help him, but it does not help when we bamboozle him with technicalities which are unnecessary and over his head.

Sorry @live424 is I am putting you down I don't mean to. We have all been where you are at the moment!
 
Back
Top