Hello everyone. This is a theoretical question.
When my Raspberry Pi is ungrounded, I can reset it by touching RUN with a screwdriver. That doesn't happen if GND from the R-Pi is connected to mains earth (directly or through a capacitor). My question is: is it related to the high impedance (long) path from me (the 50hz antenna) to mains ground, then making the noise level at the RUN pin high enough?
(http://NoisePath)
This particular power supply is not grounded. The same behavior occurs with a portable USB charger, but this time I assume the path is through the HDMI, the monitor PSU (also floating), and some capacitive coupling to ground.
Using a battery pack for a 7-inch monitor and the PI, none of that happens: It does not reset when touching RUN.
80V between R-Pi GND and mains earth (220V ac here). When I said floating I meant no third pin on the ac plug (ground) and no common mode filter inside (no CY capacitors to ground). But there may be parasitic capacitors that do the same effect.
If so, how does this affect the sensitivity of the RUN pin?
Please explain how 80 Volts AC gets discharged.80V between R-Pi GND and mains earth (220V ac here). When I said floating I meant no third pin on the ac plug (ground) and no common mode filter inside (no CY capacitors to ground). But there may be parasitic capacitors that do the same effect.
If so, how does this affect the sensitivity of the RUN pin?
the pi is floating at 80V and you are a big capacitor at ~0V so the pin gets pulled low until the 80V is discharged
Please explain how 80 Volts AC gets discharged.80V between R-Pi GND and mains earth (220V ac here). When I said floating I meant no third pin on the ac plug (ground) and no common mode filter inside (no CY capacitors to ground). But there may be parasitic capacitors that do the same effect.
If so, how does this affect the sensitivity of the RUN pin?
the pi is floating at 80V and you are a big capacitor at ~0V so the pin gets pulled low until the 80V is discharged
It is only "dc" across the smoothing capacitor(s). Each end of those capacitors is at an alternating voltage with respect to actual ground.Please explain how 80 Volts AC gets discharged.80V between R-Pi GND and mains earth (220V ac here). When I said floating I meant no third pin on the ac plug (ground) and no common mode filter inside (no CY capacitors to ground). But there may be parasitic capacitors that do the same effect.
If so, how does this affect the sensitivity of the RUN pin?
the pi is floating at 80V and you are a big capacitor at ~0V so the pin gets pulled low until the 80V is discharged
It doesn't. It's not AC by the time it reaches the smoothing capacitors in the power supply, it's rectified AC, or DC as we like to call it.
It is only "dc" across the smoothing capacitor(s). Each end of those capacitors is at an alternating voltage with respect to actual ground.Please explain how 80 Volts AC gets discharged.80V between R-Pi GND and mains earth (220V ac here). When I said floating I meant no third pin on the ac plug (ground) and no common mode filter inside (no CY capacitors to ground). But there may be parasitic capacitors that do the same effect.
If so, how does this affect the sensitivity of the RUN pin?
the pi is floating at 80V and you are a big capacitor at ~0V so the pin gets pulled low until the 80V is discharged
It doesn't. It's not AC by the time it reaches the smoothing capacitors in the power supply, it's rectified AC, or DC as we like to call it.
It doesn't. It's not AC by the time it reaches the smoothing capacitors in the power supply, it's rectified AC, or DC as we like to call it.
HDMI 7-inch display powered by a cheap 2-prong SM-PSU: ~74V rms in every single pin of the R-Pi. Very easy to reset.
Display powered by an original Samsung charger: ~13.7V rms. Really hard to reset
HDMI 7-inch display powered by a cheap 2-prong SM-PSU: ~74V rms in every single pin of the R-Pi. Very easy to reset.
Display powered by an original Samsung charger: ~13.7V rms. Really hard to reset
I can say it's not only easy to reset, but also easy to damage your RPI or display or even oscilloscope... ;)
It looks like your mains socket missing GND line, and it leads to such issue. In this case you can also notice painful electric shocks if you touch the wires with your hand.
It doesn't. It's not AC by the time it reaches the smoothing capacitors in the power supply, it's rectified AC, or DC as we like to call it.
There is no clean DC or clean AC. Any signal consists of DC and AC components. The question is which component is dominant. If dominant component is DC, we're talking that this is DC voltage. If dominant component is AC we're talking that this is AC voltage.
You're talking about noise, and yes every signal be it AC or DC contains some noise. AC is something else.
Alternating Current (AC) is a type of electrical current, in which the direction of the flow of electrons switches back and forth at regular intervals or cycles.
Direct current (DC) is an electric current that is uni-directional, so the flow of charge is always in the same direction
If the positive and negative don't swap places, that is DC no matter how much noise there is.
AC means alternating current. If the current that is actually flowing or would flow if an appropriate resistor were placed between the two relevant points of potential difference is not changing direction from time to time, then it is not AC. And vice versa,
if polarity is constant and don't change, it is still AC with frequency = 0 Hz
V=sqrt(2)*220*min(0,sin(2*pi*50*time))AC means alternating current. If the current that is actually flowing or would flow if an appropriate resistor were placed between the two relevant points of potential difference is not changing direction from time to time, then it is not AC. And vice versa,
Then there is logical question - where is border limit at which you can make sure that this is AC or not AC?
For example if current change direction once per 1000000 years, is it AC or not? :)
No. You're wrong when you say this. The Alternating part of Alternating current refers to polarity, + changes to - and vise versa.
If the polarity stays the same, it's DC. Whether it is pulsing DC that pulses at a frequency, DC that has a lot of noise, or the most stable DC possible, they are all DC.
Border limit? Are you just making up your own terms now?
That is 50hz in the UK and 60hz in the US. As it is alternating current, that means that every 100th of a second + and - swap places, which is the definition of alternating current, alternating polarity.
To disprove that DC is DC, you're now trying to claim that a signal that is stable for 100,000 years can be AC...
Let's take sine waveform which has amplitude 100 V and DC offset 100 V. It never crosses zero, so it never changes polarity and never change current direction. But it still changes in time by sine law. There is no noise at all, just a clean sine.@radiolistener.
This is AC current. But in your terminology such sine waveform signal is DC. Do you understand it?
No it is not AC. It is varying DC. As you wrote yourself, the current never changes direction. Ergo, it is not alternating.
2. The current through the secondary does not alternate. It varies between zero and some peak value. It never changes direction. It is DC.
3. The voltage across the capacitor does not alternate. It changes slightly around some mean non-zero voltage. It is DC.
4. The current in the capacitor does alternate. Current in when the diode conducts and current out when the diode does not conduct. It is AC.
5. The voltage across the resistor does not alternate. Same as the voltage across the capacitor. It is DC.
6. The current though the resistor does not alternate. Always flows inthe same direction. It is DC.
Virtually all of @radiolistener's reply is so incorrect it is clear that his understanding is seriously flawed. I'll give three examples.No it is not AC. It is varying DC. As you wrote yourself, the current never changes direction. Ergo, it is not alternating.
No, it is AC+DC. It is not change polarity and direction because there is DC offset. If you remove DC offset with capacitor you will get clean AC.
In order to measure AC voltage we're needs to know its zero level. Then we needs to measure AC voltage relative to that zero level (average value). The polarity and current direction changes relative to AC zero value.
In my example zero level is 100 V, the same as amplitude, so there is no polarity/direction change. But you're needs to measure AC relative to it's average value, not relative to it's DC offset. ;)2. The current through the secondary does not alternate. It varies between zero and some peak value. It never changes direction. It is DC.
No, the current on the secondary coil of transformer is clean AC.
Probably you you mean voltage on rectifier output, but it also not DC, this is still AC + DC and you can turn into clean DC by adding capacitor in series.3. The voltage across the capacitor does not alternate. It changes slightly around some mean non-zero voltage. It is DC.
No, it is not DC, it still AC+DC. The capacitor in parallel just reduce AC component, but it still here.4. The current in the capacitor does alternate. Current in when the diode conducts and current out when the diode does not conduct. It is AC.
5. The voltage across the resistor does not alternate. Same as the voltage across the capacitor. It is DC.
6. The current though the resistor does not alternate. Always flows inthe same direction. It is DC.
there is no DC in real world, because DC needs to be constant for infinite time, it cannot in real world. We name it as DC just to notice that this AC frequency is so small that we cannot take its frequency into account. In addition there is always present some AC ripple, so its always AC+DC
Virtually all of @radiolistener's reply is so incorrect it is clear that his understanding is seriously flawed. I'll give three examples.
The first is the statement "No, the current on the secondary coil of transformer is clean AC." The current in the secondary flows through the diode. The function of the diode is to let cuurent flow in only one direction. Certainly neither AC nor "clean AC" whatever "clean" is supposed to mean.
Second is the statement: "Probably you you(sic) mean voltage on rectifier output, but it also not DC, this is still AC + DC and you can turn into clean DC by adding capacitor in series." A series capacitor blocks DC. After the capacitor is AC not "clean DC".
this is AC + DC and you can turn it into clean AC by adding capacitor in series.
Third is the statement: "DC needs to be constant for infinite time". Wrong on two counts. It does not need to be constant. It does not need to be infinite time.
Connect a resistor across a battery. Direct current flows. It does not change direction.
The current is not constant - it reduces as the battery discharges and its voltage falls.
Eventually - but long before infinite time has elapsed - the battery voltage falls to zero and no current flows. is there anyone else in the world who says the current is not DC?
V=sqrt(2)*220*max(0,sin(2*pi*50*time))So in this one the Y capacitor goes to the positive side of the primary DC bus. That changes the 'bv' source for your sim to:Code: [Select](changing min() to max() for positive half cycles) and of course use the actual 1nF Y capacitor value, not the 4.7nF I previously recommended.V=sqrt(2)*220*max(0,sin(2*pi*50*time))
However there is also the monitor to consider and its Y capacitor so you may want to compare the results as you step the Y capacitor between 1nF and lets say 5nF
Now I have the circuit of the PSU mentioned in the topic. You guys could continue the AC vs DC struggle on top of the schematic ;D
Now I have the circuit of the PSU mentioned in the topic. You guys could continue the AC vs DC struggle on top of the schematic ;D
Now I have the circuit of the PSU mentioned in the topic. You guys could continue the AC vs DC struggle on top of the schematic ;D
I might be a bit old school with my ideas on power supplies, but personally if I had to convert 220V down to 5V for a pi, I'd use a step down transformer at the first point to get something like 18V then rectify that, smooth it using caps, convert it to 5V using something like a LM60440DRPKR buck converter, followed by a coil and some caps to clean up the power. You can fine tune the voltage by adjusting a resistor value in the example circuit (23.7K gets you 5.2V). The LM chip can cope with a wide range of input voltages so you don't need to be that careful with circuit design before it so long as you smooth it fairly well and the efficiency is very good, so almost no heat. Depending on your power needs, that chip can supply 4A, there are less powerful chips available. I regularly use this chip to power a raspberry pi (and screen) with no issues, although I use a very cheap 12V brick power supply which I then convert down to 5.2V, for an easy life.
Or you could check out this WeBench circuit, 2A 220V AC - 5.2VDC optoisolated for ideas.
https://webench.ti.com/appinfo/webench/scripts/SDP.cgi?ID=AF1132CC6B0C04BB
Now I have the circuit of the PSU mentioned in the topic. You guys could continue the AC vs DC struggle on top of the schematic ;D
I might be a bit old school with my ideas on power supplies, but personally if I had to convert 220V down to 5V for a pi, I'd use a step down transformer at the first point to get something like 18V then rectify that, smooth it using caps, convert it to 5V using something like a LM60440DRPKR buck converter, followed by a coil and some caps to clean up the power. You can fine tune the voltage by adjusting a resistor value in the example circuit (23.7K gets you 5.2V). The LM chip can cope with a wide range of input voltages so you don't need to be that careful with circuit design before it so long as you smooth it fairly well and the efficiency is very good, so almost no heat. Depending on your power needs, that chip can supply 4A, there are less powerful chips available. I regularly use this chip to power a raspberry pi (and screen) with no issues, although I use a very cheap 12V brick power supply which I then convert down to 5.2V, for an easy life.
Or you could check out this WeBench circuit, 2A 220V AC - 5.2VDC optoisolated for ideas.
https://webench.ti.com/appinfo/webench/scripts/SDP.cgi?ID=AF1132CC6B0C04BB
The LM60440DRPKR looks nice. The package is a bit of a pain though.
Anyway, even using an isolation transformer, an old school transformer to unregulated DC, etc. the problem remains there (I guess I'm just adding series -parasitic- capacitors), but it can be easily solved grounding the Raspberry pi.
I was just chasing how the reset happened, and I think my last video shows so.
Edit: BTW: Using the metallic side panel of a PC case flat on the floor and connecting the R-Pi to it was enough to reduce that 50Hz leakage to half!