I am having very unlucky days with xCORE-200 Explorer boards recently. I've burned few boards recently due to my carelessness. I was wondering if there is no protection in the board what so ever? Because, this never happened to me with another board, such as STM32F4 or Arduinos. I think even the 3.3V boards have some sort of protection for reverse current, or 5V. Maybe a diode even for insurance. Why not XMOS?
And if anybody has an idea how to test xCORE-200 Explorer, possibly using the test points, I would really appreciate it. Thanks.
xCORE-200 Voltage/Current Protection Topic is solved
-
mozcelikors
- Experienced Member
- Posts: 75
- Joined: Sat May 07, 2016 11:47 am
-
mon2
- XCore Legend
- Posts: 1913
- Joined: Thu Jun 10, 2010 11:43 am
Hi mozcelikors. Sorry to hear about the raised issues. Suspecting there were cost reductions to offer the XCORE-200 board as-is. With a quick review of the related schematics, there is a single ESD protection array on the USB connector so some forms of transient protection is present on this board. There are also downstream dc-dc regulators which lower this input voltage from 5 volts to 3v3, 2v5 and 1v0. However, these regulators will break down if input voltage is > 7 volts (ST1S06) or > 16 volts (MIC5213).
Care must be taken with respect to:
a) over voltage on the USB connector (>7 volts) used to power the XCORE-200 can destroy the board (there are ST1S06 regulators which are able to support up to 7 volts as input as max)
b) the XCORE-200 GPIO pins are NOT 5 volt tolerant (must limit to 3v3 or 2v5 using respective external protection where applicable)
c) the XCORE-200 GPIO pins are NOT current protected (must limit the current using external protection)
Review this webpage which is a good start on how to make your design more robust:
http://www.rugged-circuits.com/10-ways- ... n-arduino/
* Important to review only the notes related to 3v3 protection !!
Additional comments:
We generally do not prefer the use of zener diodes as the only level of protection as there are better but more costly solutions to perform the tasks with a much better response time. For example, review USB current switches - assorted features from respective vendors that will offer reverse voltage / current protection; over voltage protection, etc. These external components may impose a capacitive load onto the mated pin so you must review if such protection will hinder the required toggle speeds.
If you are using unidirectional data transfers with the outside world, consider optical isolation or digital isolators. This will then create a galvanic isolation with your external interfaces.
We are designing a mix of products that are intended to be powered by low cost AC to USB 5v power supplies. These are typically available from local dollar stores but we are including protection against cases where such cheap and perhaps off-spec power supplies will supply > 5 volts into our board and nuke everything downstream.
Share your details on what led to the destruction of your XCORE-200 kit(s) if the above details do not cover your cases. Suspecting that the XCORE-200 failed due to the out of spec GPIO pin voltages and/or current applied by your external hardware.
Care must be taken with respect to:
a) over voltage on the USB connector (>7 volts) used to power the XCORE-200 can destroy the board (there are ST1S06 regulators which are able to support up to 7 volts as input as max)
b) the XCORE-200 GPIO pins are NOT 5 volt tolerant (must limit to 3v3 or 2v5 using respective external protection where applicable)
c) the XCORE-200 GPIO pins are NOT current protected (must limit the current using external protection)
Review this webpage which is a good start on how to make your design more robust:
http://www.rugged-circuits.com/10-ways- ... n-arduino/
* Important to review only the notes related to 3v3 protection !!
Additional comments:
We generally do not prefer the use of zener diodes as the only level of protection as there are better but more costly solutions to perform the tasks with a much better response time. For example, review USB current switches - assorted features from respective vendors that will offer reverse voltage / current protection; over voltage protection, etc. These external components may impose a capacitive load onto the mated pin so you must review if such protection will hinder the required toggle speeds.
If you are using unidirectional data transfers with the outside world, consider optical isolation or digital isolators. This will then create a galvanic isolation with your external interfaces.
We are designing a mix of products that are intended to be powered by low cost AC to USB 5v power supplies. These are typically available from local dollar stores but we are including protection against cases where such cheap and perhaps off-spec power supplies will supply > 5 volts into our board and nuke everything downstream.
Share your details on what led to the destruction of your XCORE-200 kit(s) if the above details do not cover your cases. Suspecting that the XCORE-200 failed due to the out of spec GPIO pin voltages and/or current applied by your external hardware.
-
mozcelikors
- Experienced Member
- Posts: 75
- Joined: Sat May 07, 2016 11:47 am
Well what I did was measuring TP18 with an oscilloscope, while the other end is in a high GPIO pin accidentally.
Could you tell me what might have gone wrong, and also for the future, this is the circuit I'm trying to build. Is there any bad connections that you see ?
Thanks for your kind and helpful reply.
Could you tell me what might have gone wrong, and also for the future, this is the circuit I'm trying to build. Is there any bad connections that you see ?
Thanks for your kind and helpful reply.
-
mon2
- XCore Legend
- Posts: 1913
- Joined: Thu Jun 10, 2010 11:43 am
Hi. Taking a measurement with an oscilloscope should be fine unless the probe pin was shorted to ground. On our lab probe, there is a small switch which when you close this switch, the probe pin will short to ground. If this is the case and then the XMOS port pin was high then yes, this could lead to issues and cause a failure.
A basic and simple approach is to insert series resistors to limit the current through the respective port pin. However, you will need to define the max current you wish to permit through this port pin. The pin will not be reverse voltage nor ESD protected. ESD is a fast transient that could surface from an external event like walking across a carpet.
In a quick review of your posted diagram, the I2C lines should be fine since you have a bidirectional voltage level shifter (required) from 5v to 3v3 - as long as the sensors are not very far from your equipment (ie. not hundreds of ft away). If you do then you must review methods to allow for long distance I2C driving methods and most certainly use galvanic isolation. There are assorted solutions on the market for these topics. With galvanic isolation, your ground of the XMOS can be different than the ground potential of your external interface(s). Respectively, in this approach, do not connect the grounds together from the XMOS to the remote sensor or whatever you galvanically isolation.
Moving on, you are dealing with motors. Most definitely recommend that you consider isolation of some sort here to isolate the very noisy motor controllers and motors from the delicate XMOS devices. We have used NVE and SiLabs digital isolators for many many years with success. There are assorted models available and they vary with the chip size and # of ports inside the isolator. You will need an isolated dc-dc convertor to complete the puzzle. Review, Recom (solid company with excellent products - zero defects to date with our deployed modules).
Summary:
5 volts input -> isolated dc-dc power supply module -> isolated 5 volts output.
apply a digital isolator with XMOS side powered by your 5 volt from USB port -> other side of the digital isolator will be powered by the ISOLATED (imported !!) 5 volts output of the Recom (or equivalent) power supply.
You will need to review the power requirements for your project. If using only the digital isolator than 1W is ample. If you intend to ALSO power the motors, motor controller then you will need more (please review the options but 3W is also available).
==
Another approach is to use a separate power supply for the motors and that the ground from the external power supply will mate with the isolated DC-DC convertor. Continue to use the isolated dc-dc module for the isolator. This way, you can use a heftier power supply for the motors + have galvanic isolation with the XMOS device + no worries on the isolator being NOT powered if the external power supply is powered out of sequence. That is, if the XMOS board is powered then so is the Recom power supply and the digital isolator = everyone is happy. Then proceed to power the motors and all is well.
The digital isolator is valid if you are unidirectional with the GPIO pin (ie. only output or only input). Write back if that is not the case.
Next, I would be concerned about the use of the 5 volt power rail for your motor use. Your motors may be exceeded the current supplied by the USB port. Highly recommend that you use a different power supply for the motor use.
You can also consider to review the digital opto isolators from Avago. They have come a long ways and are CMOS compatible (ie. you simply output a logic high and the isolator will do the rest).
If you are cost conscience with the bill of materials, there are other methods and options but typically you will get what you pay for this topic. We have many designs in the military, aerospace and medical fields and working with zero failures but we also like to throw in everything for such designs - including the kitchen sink.
A basic and simple approach is to insert series resistors to limit the current through the respective port pin. However, you will need to define the max current you wish to permit through this port pin. The pin will not be reverse voltage nor ESD protected. ESD is a fast transient that could surface from an external event like walking across a carpet.
In a quick review of your posted diagram, the I2C lines should be fine since you have a bidirectional voltage level shifter (required) from 5v to 3v3 - as long as the sensors are not very far from your equipment (ie. not hundreds of ft away). If you do then you must review methods to allow for long distance I2C driving methods and most certainly use galvanic isolation. There are assorted solutions on the market for these topics. With galvanic isolation, your ground of the XMOS can be different than the ground potential of your external interface(s). Respectively, in this approach, do not connect the grounds together from the XMOS to the remote sensor or whatever you galvanically isolation.
Moving on, you are dealing with motors. Most definitely recommend that you consider isolation of some sort here to isolate the very noisy motor controllers and motors from the delicate XMOS devices. We have used NVE and SiLabs digital isolators for many many years with success. There are assorted models available and they vary with the chip size and # of ports inside the isolator. You will need an isolated dc-dc convertor to complete the puzzle. Review, Recom (solid company with excellent products - zero defects to date with our deployed modules).
Summary:
5 volts input -> isolated dc-dc power supply module -> isolated 5 volts output.
apply a digital isolator with XMOS side powered by your 5 volt from USB port -> other side of the digital isolator will be powered by the ISOLATED (imported !!) 5 volts output of the Recom (or equivalent) power supply.
You will need to review the power requirements for your project. If using only the digital isolator than 1W is ample. If you intend to ALSO power the motors, motor controller then you will need more (please review the options but 3W is also available).
==
Another approach is to use a separate power supply for the motors and that the ground from the external power supply will mate with the isolated DC-DC convertor. Continue to use the isolated dc-dc module for the isolator. This way, you can use a heftier power supply for the motors + have galvanic isolation with the XMOS device + no worries on the isolator being NOT powered if the external power supply is powered out of sequence. That is, if the XMOS board is powered then so is the Recom power supply and the digital isolator = everyone is happy. Then proceed to power the motors and all is well.
The digital isolator is valid if you are unidirectional with the GPIO pin (ie. only output or only input). Write back if that is not the case.
Next, I would be concerned about the use of the 5 volt power rail for your motor use. Your motors may be exceeded the current supplied by the USB port. Highly recommend that you use a different power supply for the motor use.
You can also consider to review the digital opto isolators from Avago. They have come a long ways and are CMOS compatible (ie. you simply output a logic high and the isolator will do the rest).
If you are cost conscience with the bill of materials, there are other methods and options but typically you will get what you pay for this topic. We have many designs in the military, aerospace and medical fields and working with zero failures but we also like to throw in everything for such designs - including the kitchen sink.
-
mozcelikors
- Experienced Member
- Posts: 75
- Joined: Sat May 07, 2016 11:47 am
Well, this is an RC car but the motor is quite powerful. Would a simple solution like a 0ohm resistor work on this? Motor already has a different power supply, this is the motor driver that is shown. And concerning the servo motor, would it be necessary to isolate, if so again would a simpler solution work?
Thank you in advance.
Thank you in advance.
-
Gothmag
- XCore Addict
- Posts: 129
- Joined: Wed May 11, 2016 3:50 pm
What kind of RC car is it. Ive been working on a similar project with a traxxas slash and an explorerkit recently. However I have lines from receiver going directly to explorerkit, and then have it send signal directly to the servo/esc. Where level shifters, which you seem to have, are recommended but I haven't had time to add them.
You probably shouldn't be trying to power the servo through the explorerkit because they can pull alot of current. I also don't know what if any reverse current protection those have or produce. I think total isolation of the power supplies is a good idea, and probably where you're having problems. What I've done is use all the standard components of the car and add a castle battery elimination circuit which supplies 5.1v from battery voltage and used that for power for all additional electronics. Even though I'm using the ~5v signals from receiver straight to the explorerkit kit Ive never had issues. Like said above you definitely shouldn't be running motor ground through xmos if that's what you're doing either.
Even an ok buck/boost converter depending on your battery voltage would be enough to supply power. If you dont have a receiver on the car I would highly recommend using a BEC to power your servo at the very least.
You probably shouldn't be trying to power the servo through the explorerkit because they can pull alot of current. I also don't know what if any reverse current protection those have or produce. I think total isolation of the power supplies is a good idea, and probably where you're having problems. What I've done is use all the standard components of the car and add a castle battery elimination circuit which supplies 5.1v from battery voltage and used that for power for all additional electronics. Even though I'm using the ~5v signals from receiver straight to the explorerkit kit Ive never had issues. Like said above you definitely shouldn't be running motor ground through xmos if that's what you're doing either.
Even an ok buck/boost converter depending on your battery voltage would be enough to supply power. If you dont have a receiver on the car I would highly recommend using a BEC to power your servo at the very least.
You do not have the required permissions to view the files attached to this post.
-
mozcelikors
- Experienced Member
- Posts: 75
- Joined: Sat May 07, 2016 11:47 am
Hello,
Its TBLE-02 RC Car I'm trying to hack. Now, I tested tomorrow that actually motor driver red pin should be not connected for some reason. (it wasnt the case for my previous circuits, I was supplying 5V to that pin)
Now all I need is to isolate sufficient yet not exceed power supplied by 7.2V batteries or 5V 2A powerbank to supply isolated power for Servo motor. Could you help me find a device for that? (Nothing really big)
So, now whats left in the puzzle is that I need to find what comes to where it says ext and its correct ground connections.
My servo looks like this: http://blue-bird-model.com/en/servos/bms-622.html
Edit: What about this one http://www.digikey.de/product-detail/en ... ND/2318142
Its TBLE-02 RC Car I'm trying to hack. Now, I tested tomorrow that actually motor driver red pin should be not connected for some reason. (it wasnt the case for my previous circuits, I was supplying 5V to that pin)
Now all I need is to isolate sufficient yet not exceed power supplied by 7.2V batteries or 5V 2A powerbank to supply isolated power for Servo motor. Could you help me find a device for that? (Nothing really big)
So, now whats left in the puzzle is that I need to find what comes to where it says ext and its correct ground connections.
My servo looks like this: http://blue-bird-model.com/en/servos/bms-622.html
Edit: What about this one http://www.digikey.de/product-detail/en ... ND/2318142
-
Gothmag
- XCore Addict
- Posts: 129
- Joined: Wed May 11, 2016 3:50 pm
Do you still have the radio receiver for the car? Its not on the schematic but doesn't mean it isn't. You can more or less use it as a small power supply. Typically the motor controller small wires go to it, and the steering servo. Even without using radio control, but instead bluetooth, I'd plug red and black of servo and controller into it, just running signal wires to xmos. If the voltage from that connector isn't too high you can directly attach power/ground to xmos 5v/ground via headers on right and left of board(they have 5v/3v/gnd) If the power is B+ you need a different supply, which will allow al grounds to be tied. Without doing that the ground voltage is usually not the same for all componenta and you will get erratic behavior. However that motor controller only appears to offer 1.5a and because of that I'd only recommend using it for servo power. You could even directly attach controller and servo red/black wires. If.you had a small 2a battery that could power xmos but you need to make sure grounds are tied using another ground connection on the xmos to either B+ at batter or on one of the small connectors.
I would not recommend that supply, it can only manage 0.5a, not appropriate for anything. If you search amazon or ebay for a small buck supply that could easily power your xmos/bluetooth for a small amount of money. They aren't perfect and have their limits but used within reason they're ok. You could also design your own supply. On another project car I use wires attaching straight to battery terminals to a power supply and leave everything else in tact except motor controller and servo signal lines.
Typically none of the vehicles power lines should route directly to the xmos unless you're choosing to use a seperate battery to power it in which case you only add a TAP to vehicle ground, don't try to ground vehicle through xmos it can't handle that. Depending on your level shifter tying grounds isn't necessary either, and you just need to stop routing power through xmos, hooking up motor controller and servo signal wires only straight to xmos from level shifters(which typically have vcc, gnd connections for both sides of circuit)
I would not recommend that supply, it can only manage 0.5a, not appropriate for anything. If you search amazon or ebay for a small buck supply that could easily power your xmos/bluetooth for a small amount of money. They aren't perfect and have their limits but used within reason they're ok. You could also design your own supply. On another project car I use wires attaching straight to battery terminals to a power supply and leave everything else in tact except motor controller and servo signal lines.
Typically none of the vehicles power lines should route directly to the xmos unless you're choosing to use a seperate battery to power it in which case you only add a TAP to vehicle ground, don't try to ground vehicle through xmos it can't handle that. Depending on your level shifter tying grounds isn't necessary either, and you just need to stop routing power through xmos, hooking up motor controller and servo signal wires only straight to xmos from level shifters(which typically have vcc, gnd connections for both sides of circuit)
Last edited by Gothmag on Mon Oct 03, 2016 3:57 pm, edited 1 time in total.
-
mozcelikors
- Experienced Member
- Posts: 75
- Joined: Sat May 07, 2016 11:47 am
Hello,
1. We didnt order the radio receiver.
2. That power converter im thinking only for servo motor. Actually, now I'm thinking this one:
http://www.digikey.de/product-detail/en ... ND/4488257
Still 0.5A.
What do you think?
1. We didnt order the radio receiver.
2. That power converter im thinking only for servo motor. Actually, now I'm thinking this one:
http://www.digikey.de/product-detail/en ... ND/4488257
Still 0.5A.
What do you think?
-
Gothmag
- XCore Addict
- Posts: 129
- Joined: Wed May 11, 2016 3:50 pm
I still think since you're using level shifters on signal lines you use the 2a battery for xmos and connect red/black wires from 3 pin to red/black of 3 pin servo. The remaining two wires got to level shifters. Xmos 3v/gnd to lv of level shifters, red/black tap for hv on shifters and that should work. You want a 1-2a supply for a servo that size.