Il 26/06/2022 16:25, Reinhardt Behm ha scritto:> On Fri, 24 Jun 2022 16:25:55 +0200, pozz wrote: > > >>> Of course you could always have had damage from ground loops, spikes >>> when plugging the debugger in and out (if power is on), short-circuits >>> when trying to use an oscilloscope probe with clumsy fingers, etc. If >>> you have access to an X-Ray machine, you might be able to get some idea >>> about what happened. >> >> I admit I don't have professional workbench with anti-static writ strap >> band and so on, but I worked for many years without these kind of >> problems. It's strange I encountered so many problems concentrated in a >> few weeks. >> >> I will understand the problem of ground loops. > > Take the exact setup and before plugging the debugger measure the voltage > between the CPU board ground and the debugger ground with a multi meter > in AC mode. My guess is that you have stray AC (mains) voltage on either > side (by Y-caps in a PSU or just capacitive coupling over the transformer > to the DC side GND). These stray voltages often reach 50% of AC voltage, > in you case probably 60V-AC. When you plug the connector they might > easily kill a CPU port.I made the measure you suggested and I found around 20Vac between debugger GND and board GND *before* plugging them. After adding a good wire connection between PC metallic case (near a mounting screw) and board GND, the Vac measured goes near to zero. Could it be this the cause of issues I observed? I will continue with this connection and see. Thank you for your suggestions.
Cortex-Mx MCUs with SWD access locked
Started by ●June 23, 2022
Reply by ●June 27, 20222022-06-27
Reply by ●June 27, 20222022-06-27
On Mon, 27 Jun 2022 10:07:37 +0200, pozz wrote:> Il 26/06/2022 16:25, Reinhardt Behm ha scritto: >> On Fri, 24 Jun 2022 16:25:55 +0200, pozz wrote: >> >> >>>> Of course you could always have had damage from ground loops, spikes >>>> when plugging the debugger in and out (if power is on), >>>> short-circuits when trying to use an oscilloscope probe with clumsy >>>> fingers, etc. If you have access to an X-Ray machine, you might be >>>> able to get some idea about what happened. >>> >>> I admit I don't have professional workbench with anti-static writ >>> strap band and so on, but I worked for many years without these kind >>> of problems. It's strange I encountered so many problems concentrated >>> in a few weeks. >>> >>> I will understand the problem of ground loops. >> >> Take the exact setup and before plugging the debugger measure the >> voltage between the CPU board ground and the debugger ground with a >> multi meter in AC mode. My guess is that you have stray AC (mains) >> voltage on either side (by Y-caps in a PSU or just capacitive coupling >> over the transformer to the DC side GND). These stray voltages often >> reach 50% of AC voltage, in you case probably 60V-AC. When you plug the >> connector they might easily kill a CPU port. > > I made the measure you suggested and I found around 20Vac between > debugger GND and board GND *before* plugging them. > > After adding a good wire connection between PC metallic case (near a > mounting screw) and board GND, the Vac measured goes near to zero. > > Could it be this the cause of issues I observed? I will continue with > this connection and see. > > Thank you for your suggestions.It could not only be, it will be. Imagine you plug the connector, GND is not contacted first, then the CPU pin sees 20V-AC. That means +20V*sqrt(2) relative to CPU-GND and -20V*sqrt(2). That is far out of spec. It will kill the pin. I had something similar. A colleague had used a 2 wire extension cord for a soldering station, so no earth connection. I did not know this and killed the reset input of my controller. When I found out this I understood why we had several cases of non-functioning boards. The two wire extension cord was 2$ cheaper than a 3 wire cored. the cost of this "saving" was several 100$. -- Reinhardt
Reply by ●June 27, 20222022-06-27
Il 27/06/2022 17:11, Reinhardt Behm ha scritto:> On Mon, 27 Jun 2022 10:07:37 +0200, pozz wrote: > >> Il 26/06/2022 16:25, Reinhardt Behm ha scritto: >>> On Fri, 24 Jun 2022 16:25:55 +0200, pozz wrote: >>> >>> >>>>> Of course you could always have had damage from ground loops, spikes >>>>> when plugging the debugger in and out (if power is on), >>>>> short-circuits when trying to use an oscilloscope probe with clumsy >>>>> fingers, etc. If you have access to an X-Ray machine, you might be >>>>> able to get some idea about what happened. >>>> >>>> I admit I don't have professional workbench with anti-static writ >>>> strap band and so on, but I worked for many years without these kind >>>> of problems. It's strange I encountered so many problems concentrated >>>> in a few weeks. >>>> >>>> I will understand the problem of ground loops. >>> >>> Take the exact setup and before plugging the debugger measure the >>> voltage between the CPU board ground and the debugger ground with a >>> multi meter in AC mode. My guess is that you have stray AC (mains) >>> voltage on either side (by Y-caps in a PSU or just capacitive coupling >>> over the transformer to the DC side GND). These stray voltages often >>> reach 50% of AC voltage, in you case probably 60V-AC. When you plug the >>> connector they might easily kill a CPU port. >> >> I made the measure you suggested and I found around 20Vac between >> debugger GND and board GND *before* plugging them. >> >> After adding a good wire connection between PC metallic case (near a >> mounting screw) and board GND, the Vac measured goes near to zero. >> >> Could it be this the cause of issues I observed? I will continue with >> this connection and see. >> >> Thank you for your suggestions. > > It could not only be, it will be. Imagine you plug the connector, GND is > not contacted first, then the CPU pin sees 20V-AC. That means +20V*sqrt(2) > relative to CPU-GND and -20V*sqrt(2). That is far out of spec. It will > kill the pin. > I had something similar. A colleague had used a 2 wire extension cord for > a soldering station, so no earth connection. I did not know this and > killed the reset input of my controller. When I found out this I > understood why we had several cases of non-functioning boards. The two > wire extension cord was 2$ cheaper than a 3 wire cored. the cost of this > "saving" was several 100$.Anyway in my case the power cord of my PC is connected to a UPS and the AC/DC that powers the board is connected to earth connection of the grid. Working with two wires power cords is risky and I learned that in the past. However this isn't the case now, both PC and board have a good 3-wires power connection, but there's a 20Vac voltage between the grounds.
Reply by ●June 28, 20222022-06-28
On 2022-06-27 pozz wrote in comp.arch.embedded:> > Anyway in my case the power cord of my PC is connected to a UPS and the > AC/DC that powers the board is connected to earth connection of the grid. > > Working with two wires power cords is risky and I learned that in the past. > > However this isn't the case now, both PC and board have a good 3-wires > power connection, but there's a 20Vac voltage between the grounds.With both grounded, there cannot be a 20VAC voltage between the grounds. (unless there is an extreme amount of current ofcourse, but that will not be the case) Your PC is a standard desktop? Then the 0V is connected to the earth connection of the power plug. This is not the case if you have a laptop. The supply will have an earth connection, but the DC output is floating. Your AC/DC probably has a similar construction as a laptop supply: Earth connection on the power plug and a floating DC output. That there still is a measurable voltage is because of capacitive coupling in the AC/DC between AC input and DC output (assuming PC is really grounded). The voltage you measure depends on the impedance of that coupling and the impedance of your meter. Just measured 0V to earth on a few supplies here with a volt meter with input impedance of > 10 MOhm / < 100 pF: Recent 150W lab supply: 15 mVAC Old analog 2x40W supply: CH1: 15VAC, CH2: 5 VAC Recent 10W switching wall wart: 90VAC Tried to measure short circuit currents as well, but that was not very reliable as I don't have an AC uA meter here. Currents are too small. As the current is very low, this should give no problems. There could be a small surge if you just happen to connect at the peak of that AC voltage. But the capacitance behind that should be very low, otherwise the voltage/current measurements should be different. (Haven't drawn out the resulting schematic and the consequences though). So you could do a current measurement between the grounds. I would expect less than 0.1 mA AC there. But with an isolated supply there also is a chance of static (DC) buildup. Depending an the capacity to earth this could give and ESD discharge on connection. So all this may not be your problem, but connecting an additional earth between PC and your supply will not hurt anyway. And it has the (small) possibility of solvng your problem. -- Stef TODAY IS INTERNATIONAL CAPSLOCK DAY!
Reply by ●June 28, 20222022-06-28
Il 28/06/2022 11:23, Stef ha scritto:> On 2022-06-27 pozz wrote in comp.arch.embedded: >> >> Anyway in my case the power cord of my PC is connected to a UPS and the >> AC/DC that powers the board is connected to earth connection of the grid. >> >> Working with two wires power cords is risky and I learned that in the past. >> >> However this isn't the case now, both PC and board have a good 3-wires >> power connection, but there's a 20Vac voltage between the grounds. > > With both grounded, there cannot be a 20VAC voltage between the grounds. > (unless there is an extreme amount of current ofcourse, but that will > not be the case)The PSU of desktop PC and AC/DC of board are connected to the earth conductor, i.e. I'm using 3-wires power cord. But their separate GND aren't connected.> Your PC is a standard desktop?Yes.> Then the 0V is connected to the earth > connection of the power plug. This is not the case if you have a laptop. > The supply will have an earth connection, but the DC output is floating.No, it's a desktop PC, not a laptop. I don't know why I have 20Vac between GND of J-Link and board. Maybe the problem is in the UPS that has some failure.> Your AC/DC probably has a similar construction as a laptop supply: Earth > connection on the power plug and a floating DC output.Yes, but I connect GND (0V of DC output) to earth.> That there still is a measurable voltage is because of capacitive > coupling in the AC/DC between AC input and DC output (assuming PC is > really grounded). The voltage you measure depends on the impedance of > that coupling and the impedance of your meter. > > Just measured 0V to earth on a few supplies here with a volt meter with > input impedance of > 10 MOhm / < 100 pF: > > Recent 150W lab supply: 15 mVAC > Old analog 2x40W supply: CH1: 15VAC, CH2: 5 VAC > Recent 10W switching wall wart: 90VAC > > Tried to measure short circuit currents as well, but that was not very > reliable as I don't have an AC uA meter here. Currents are too small. > > As the current is very low, this should give no problems. There could be > a small surge if you just happen to connect at the peak of that AC > voltage. But the capacitance behind that should be very low, otherwise > the voltage/current measurements should be different. (Haven't drawn out > the resulting schematic and the consequences though). > > So you could do a current measurement between the grounds. I would > expect less than 0.1 mA AC there. > > But with an isolated supply there also is a chance of static (DC) > buildup. Depending an the capacity to earth this could give and ESD > discharge on connection. > > So all this may not be your problem, but connecting an additional earth > between PC and your supply will not hurt anyway. And it has the (small) > possibility of solvng your problem.
Reply by ●June 28, 20222022-06-28
On 2022-06-28 pozz wrote in comp.arch.embedded:> Il 28/06/2022 11:23, Stef ha scritto: >> On 2022-06-27 pozz wrote in comp.arch.embedded: >>> >>> Anyway in my case the power cord of my PC is connected to a UPS and the >>> AC/DC that powers the board is connected to earth connection of the grid. >>> >>> Working with two wires power cords is risky and I learned that in the past. >>> >>> However this isn't the case now, both PC and board have a good 3-wires >>> power connection, but there's a 20Vac voltage between the grounds. >> >> With both grounded, there cannot be a 20VAC voltage between the grounds. >> (unless there is an extreme amount of current ofcourse, but that will >> not be the case) > > The PSU of desktop PC and AC/DC of board are connected to the earth > conductor, i.e. I'm using 3-wires power cord. > > But their separate GND aren't connected. > > >> Your PC is a standard desktop? > > Yes. > > >> Then the 0V is connected to the earth >> connection of the power plug. This is not the case if you have a laptop. >> The supply will have an earth connection, but the DC output is floating. > > No, it's a desktop PC, not a laptop. > I don't know why I have 20Vac between GND of J-Link and board. > Maybe the problem is in the UPS that has some failure.Measure voltage between PC case and earth. If that is not zero check the UPS and the power cables.>> Your AC/DC probably has a similar construction as a laptop supply: Earth >> connection on the power plug and a floating DC output. > > Yes, but I connect GND (0V of DC output) to earth.If your PC is connected to earth as well, you should measure 0V between them. If not there is something wrong. -- Stef Original thought is like original sin: both happened before you were born to people you could not have possibly met. -- Fran Lebowitz, "Social Studies"
