Overloading a circuit - Question

Thread Starter

toffee_pie

Joined Oct 31, 2009
215
Hi All,

I am looking into some circuit design regarding the impact of overloading the input, nominal voltage is 24V but want to see what will happen at 40V overload - namely the power path and components situated around this.

The circuit basically has lots of analog blocks so voltages going to ADCs for sensing etc as well as DC Motor so a few FETS also, I already have seen some components that are potentially overloaded (caps not rated sufficently etc) as well as voltage dividers that could supply over 5V into a PIC (18F - Microchip), the ADC inputs. These inputs are voltage driven right? how to best establish the current into the PIC inputs from the voltage dividers?

Is there anything else I can do to establish the impact of overloading? LTSpice or model sim could be used perhaps.?

thanks for suggestions.!
 

dl324

Joined Mar 30, 2015
12,688
A schematic would be helpful.

Simulators aren't all knowing. LTSpice will let you exceed component capabilities with no ill effects. It's up to the user to know when results don't make sense.
 

ebeowulf17

Joined Aug 12, 2014
3,275
Hi All,

I am looking into some circuit design regarding the impact of overloading the input, nominal voltage is 24V but want to see what will happen at 40V overload - namely the power path and components situated around this.

The circuit basically has lots of analog blocks so voltages going to ADCs for sensing etc as well as DC Motor so a few FETS also, I already have seen some components that are potentially overloaded (caps not rated sufficently etc) as well as voltage dividers that could supply over 5V into a PIC (18F - Microchip), the ADC inputs. These inputs are voltage driven right? how to best establish the current into the PIC inputs from the voltage dividers?

Is there anything else I can do to establish the impact of overloading? LTSpice or model sim could be used perhaps.?

thanks for suggestions.!
A schematic would be most helpful.

If the input that may go over is just a signal input, with little real power transfer required, there will probably be easy ways to condition that input right where it enters your system, maybe with clamping diodes or something similar. If you handle it right at the input, you don't have to worry about the specs for every single cap or ADC input, just the few protection components on the way in.

If the input in question delivers power, you can probably use some combination of voltage regulators, bulk capacitors, TVS diodes, and ferrite inductor chokes to prevent transient voltage spikes and/or sustained over voltage conditions.

In either case, if you've got a large group of circuit components all designed around 24V input, I'd think you can protect them as a group systemically instead of trying to get each component spec'd for the higher range.

Of course, I could provide much more meaningful guesses if we had any idea whatsoever what your circuit is. Please share your schematic.
 

Reloadron

Joined Jan 15, 2015
5,897
I am looking into some circuit design regarding the impact of overloading the input, nominal voltage is 24V but want to see what will happen at 40V overload - namely the power path and components situated around this.
The fuse will open and the circuit(s) will survive. Oh, no fuse, no circuit protection? Then depending on the circuit as voltage is increased the current will increase till something along the way fails possibly taking along with it other components. Simply put nothing good will happen only leaving bad things. Less any accurate drawing or schematic, as mentioned, there is no way to predict exactly what will happen or when in an ugly chain of events. You will be creating an over voltage condition to the circuit and something is going to give.

Ron
 

Thread Starter

toffee_pie

Joined Oct 31, 2009
215
Thanks guys, bear with me on Schematic. There are fuses and diodes etc in place, from an initial investigation I can see some passive comonents maybe over stressed as well as some regulators ie not designed to take 40V at the input. Is there anything else I can do other than getting rough estimates of outputs from voltage dividers that feed into ADC inputs of the PIC and seeing some passives that are over stressed, or potentially overstressed? The input will be in the form of 24V from a DC source (SLA cells, 2 x 12V in series) - the idea is to see what will happen if this was to spike to circa 40V.

thanks again
 

recklessrog

Joined May 23, 2013
985
Alternatively, build it and see what gives off the magic smoke when subjected to your specified tests.
A lot of real world testing was actually done this way, in order to pass regulations, each component would be open circuited then short circuited and the results logged. Sometimes stress testing was extreme depending on the purpose of equipment and it's intended operating environment.
 

ebeowulf17

Joined Aug 12, 2014
3,275
Thanks guys, bear with me on Schematic. There are fuses and diodes etc in place, from an initial investigation I can see some passive comonents maybe over stressed as well as some regulators ie not designed to take 40V at the input. Is there anything else I can do other than getting rough estimates of outputs from voltage dividers that feed into ADC inputs of the PIC and seeing some passives that are over stressed, or potentially overstressed? The input will be in the form of 24V from a DC source (SLA cells, 2 x 12V in series) - the idea is to see what will happen if this was to spike to circa 40V.

thanks again
Until we see your circuit, we can't give useful advice.

Lacking that, I still think you should be able to use ferrites, etc to block high frequency transients that are too fast for other protections, and clamping diodes for the rest, all at the initial input. There's no reason to let all your components see the extreme over voltages.

If your schematic proves that these standard methods won't work in your scenario, then we can discuss the specifics of your schematic.
 

BR-549

Joined Sep 22, 2013
4,938
It will depend on your true purpose of your question.

Prevent the overload?

Survive and function thru overload event?

Or continued function with overload condition?
 

Thread Starter

toffee_pie

Joined Oct 31, 2009
215
apologies for delay - I will post up a schematic tomorrow, as above its to review the design discussed above with a focus on what would occur if back EMF from the DC Motor caused the DC Input to creep up to 40V (its usually 2 x 12V SLA cells). I understand its hard without a schematic, but blocking diodes, TVS diodes etc are already accounted for. Its the ramifications for the increased voltage and potential component damage (including the PIC) that I need to address with some calculations to back it up. I have worked out some voltage dividers will have higher voltage going to the PIC inputs than expected at 40V but not sure about current?
 
if back EMF from the DC Motor caused the DC Input to creep up to 40V (its usually 2 x 12V SLA cells)
The DC resistance of the motor and the motors nominal voltage can give you an idea of the max current. If you know the motor overspeeds (coasting) that can help too.

Your DC motor acts as a generator if it has some sort of momentum or push to keep it going. The series resistance of the motor limits that current.

Some things to remember for a DC motor

Vm is proportional to RPM if the current is zero.

V=Vm-Im(Rm) has been used as a spped control for cassette decks. Vm (voltage across the motor) is not really measureable. Rm is the motor resistance.

Current is generally proportional to torque.
 

Thread Starter

toffee_pie

Joined Oct 31, 2009
215
The DC resistance of the motor and the motors nominal voltage can give you an idea of the max current. If you know the motor overspeeds (coasting) that can help too.

Your DC motor acts as a generator if it has some sort of momentum or push to keep it going. The series resistance of the motor limits that current.

Some things to remember for a DC motor

Vm is proportional to RPM if the current is zero.

V=Vm-Im(Rm) has been used as a spped control for cassette decks. Vm (voltage across the motor) is not really measureable. Rm is the motor resistance.

Current is generally proportional to torque.
thanks,

there is a good article here which reflects what I am trying to work out here

https://opentextbc.ca/physicstestbook2/chapter/back-emf/

how is the best way to estimate worst case scenario back EMF from the motor and peak currents and power distributed back to the source and how to tell if the PIC could be damaged?

lets have a hypothetical scenario using the example in the link above, I replaced the figures to try and associate with my problem - (DC source 24V, total supply potentiall at 40V due to back EMF, assume DC Motor resistance 4Ohm)



Consider, for example, the motor coils represented. The coils have a 4Ohm equivalent resistance and are driven by a 24.0 V emf. Shortly after being turned on, they draw a 24/4 =6A current and, thus, dissipate P=144Watts of energy as heat transfer.


Under normal operating conditions for this motor, suppose the back emf is 40V. Then at operating speed, the total voltage across the coils is 16.0 V, and the current drawn is. 8/4 = 2A Under normal load, then, the power dissipated is 2*16 = 32W

does this make sense in my situation?
 

MisterBill2

Joined Jan 23, 2018
8,286
The initial post talks about overloading a circuit but then you describe an over VOLTAGE situation. That leads to an entirely different set of failure types. Over voltage may often lead to excessive current and heating, but it seldom pops fuses until something fails from over voltage breakdown. Resistive circuits usually run hotter, and may not fail at all unless they are designed way too close to the limits of power dissipation. Semiconductor breakdown is far more catastrophic, with junctions suddenly becoming shorted circuits or else failing open. Capacitors generally fail short circuited from excess voltage.
 

ebeowulf17

Joined Aug 12, 2014
3,275
thanks,

there is a good article here which reflects what I am trying to work out here

https://opentextbc.ca/physicstestbook2/chapter/back-emf/

how is the best way to estimate worst case scenario back EMF from the motor and peak currents and power distributed back to the source and how to tell if the PIC could be damaged?

lets have a hypothetical scenario using the example in the link above, I replaced the figures to try and associate with my problem - (DC source 24V, total supply potentiall at 40V due to back EMF, assume DC Motor resistance 4Ohm)



Consider, for example, the motor coils represented. The coils have a 4Ohm equivalent resistance and are driven by a 24.0 V emf. Shortly after being turned on, they draw a 24/4 =6A current and, thus, dissipate P=144Watts of energy as heat transfer.


Under normal operating conditions for this motor, suppose the back emf is 40V. Then at operating speed, the total voltage across the coils is 16.0 V, and the current drawn is. 8/4 = 2A Under normal load, then, the power dissipated is 2*16 = 32W

does this make sense in my situation?
Why not just use a flyback (or freewheeling) diode?

Also, still no schematic?! Your protection needs and options depend on how you're switching it. High side or low side? BJT, MOSFET, or dedicated driver? Etc, etc.

For better or for worse, this is the last you'll hear of me unless you show us your circuit.
 

Thread Starter

toffee_pie

Joined Oct 31, 2009
215
The initial post talks about overloading a circuit but then you describe an over VOLTAGE situation. That leads to an entirely different set of failure types. Over voltage may often lead to excessive current and heating, but it seldom pops fuses until something fails from over voltage breakdown. Resistive circuits usually run hotter, and may not fail at all unless they are designed way too close to the limits of power dissipation. Semiconductor breakdown is far more catastrophic, with junctions suddenly becoming shorted circuits or else failing open. Capacitors generally fail short circuited from excess voltage.
yes, its an over voltage situation - what about voltages higher than 5V going into a PIC input (ADC channels) - current will be limited by the fuses on board but will a high voltage damage the PIC? It says inputs are rated from VSS to VDD but need to check.
 

MisterBill2

Joined Jan 23, 2018
8,286
For any semiconductor device, the voltages given as absolute maximums are those limiting damage. exceeding those voltage limits will probably cause damage of some kind. Do you really choose to risk destruction of the device?
 

Reloadron

Joined Jan 15, 2015
5,897
yes, its an over voltage situation - what about voltages higher than 5V going into a PIC input (ADC channels) - current will be limited by the fuses on board but will a high voltage damage the PIC? It says inputs are rated from VSS to VDD but need to check.
You can't make that statement unless you state the actual voltage, all of the conditions and also the duration of the voltage. A voltage of 5.001 exceeds 5.000 volts so will that matter? Actual testing of the type you are getting at is a complex lab environment type of test along with monitoring and recording everything happening. Starting with ambient temperature, ambient RH (Relative Humidity) thermal rise on the part under test until mean failure and the list goes on. Every parameter also needs well defined.

As was pointed out you want to create an Over Voltage Condition, that said you need to define all of your test conditions as well as your test methods and a detailed procedure. That done you need to lest all of your equipment requirements, that is the equipment you plan to use to conduct your test. What you are describing comes under destructive testing since you plan to destroy the part during the testing. You also need to describe in detail how you plan to acquire the data and where you plan to put the data during the test.

So what is your plan and exact goal, that means well defined. I want to apply a high voltage to a PIC Chip is not well defined.

Ron
 

ArakelTheDragon

Joined Nov 18, 2016
1,353
And my 2 cents:

You should gradually increase the voltage over the ratings to see the proper result. The current and power will increase and the elements rated the closest to the "absolute maximum ratings" will start going first. The ratings are "power, voltage, current, frequency, temperaure". This test needs to be repeated at least 20 times to get some result. But a test can never be perfect (according to the international testing criteria).

You should also perform a test where you suddenly increase the voltage, and compare with the upper test to make sure the same elements go first.

Take caution as this is dangerous and can catch fire, burn you or electrecute you!
 

MaxHeadRoom

Joined Jul 18, 2013
22,707
what would occur if back EMF from the DC Motor caused the DC Input to creep up to 40V
A DC motor generates a BEMF of the same polarity as the applied voltage this opposes the applied voltage, this is proportional to RPM and is almost equal to the applied Voltage, any load will cause a drop in rpm and hence the generated DC will also drop in relation to the applied DC.
Max.
 
Top