Question about start up current

[Electro Fan]

Is there any general trend or rule of thumb for circuit design with respect to accommodating startup (or "in rush") current?

For example, if a given circuit requires 1 Amp maximum on a normal basis after turn-on, is there some reason to design the power supply to support 2x (2 Amps) of current during startup?  Is there really a need or reason to build a power supply to handle some excess current capacity (beyond the normal current consumed after startup)?  If so, what is happening to the excess current during startup?  Charging capacitors?  Something else?  I'm sure it's somewhat circuit specific but any general guidance would be helpful.

PS, these questions are driven mostly by an interest in DC circuits powered by an AC to DC power supply (typically in the 5 - 30V DC range), so maybe the startup considerations are different for AC, or motors, etc.  For starters, I'm just trying to understand the need to power relatively simple DC circuits.  Thx

Thanks 

[soldar]

No. Each circuit is different and has to be considered on its own. Some circuits can have no inrush current (over normal operating current) and soe can have tens or hundreds times ore and mainly be limited by the resistance of the wires feeding it.

Are you thinking of any circuit in particular?

[Zero999]

There really isn't enough information here to say.

A large DC motor will draw an enormous current on start-up. An LED rope light will draw no more current on start up, than it does in normal operation.

[David Hess]

Safe power supplies are designed to start into any load without damage.

One place there can be trouble however is if a power supply uses fold-back current limiting, it may not be able to start into a heavy load.  Most loads however do not draw high current at low voltages so this is rarely a problem.  If it is a problem, then the fold-back current limiting circuit needs to be adjusted to take this into account.

[Electro Fan]

No. Each circuit is different and has to be considered on its own. Some circuits can have no inrush current (over normal operating current) and soe can have tens or hundreds times ore and mainly be limited by the resistance of the wires feeding it.

Are you thinking of any circuit in particular?

What got me thinking about it is was a GPSDO that I saw spec'd as having 2.5A inrush current but only needing about 0.3 - 0.6A normal operating current.  I realize motors are likely different than non-motorized electronics but I'm trying to understand what are typical non-motorized reasons that startup (inrush) requirements might be higher than normal operating current requirements.  No big project here, just trying to understand the types of things (other than motors
 and mechanical devices) that would dictate additional current overhead capability in a DC power supply.   

[Electro Fan]

Safe power supplies are designed to start into any load without damage.

One place there can be trouble however is if a power supply uses fold-back current limiting, it may not be able to start into a heavy load.  Most loads however do not draw high current at low voltages so this is rarely a problem.  If it is a problem, then the fold-back current limiting circuit needs to be adjusted to take this into account.

Cool - Fold-back Current.  I had no concept of this until now.  Thanks David!

With this heads-up from your post, I just saw/read that fold-back current can have an impact on supply voltage, if I understood that correctly.  I need to read further but I think I will have some more questions / thoughts on this.  Thanks!

[David Hess]

The device will have its own internal decoupling capacitors which need to be charged.  This also applies to devices like computer cards and peripherals which are designed to be hot-plugged and specifications for such include peak currents during charging which the device has to obey.  Sometimes like for USB, this will take the form of a maximum input capacitance if current limiting is not implemented.  For a PCI expansion card, this takes the form of a "precharge" function which charges the decoupling capacitors before main power is connected.

Cool - Fold-back Current.  I had no concept of this until now.  Thanks David!

With this heads-up from your post, I just saw/read that fold-back current can have an impact on supply voltage, if I understood that correctly.  I need to read further but I think I will have some more questions / thoughts on this.  Thanks!

Fold-back current limiting is important when an output short on the power supply would cause excessive power dissipation and eventual failure.  It is a good idea in general for both external and internal power supplies unless the power supply in question is designed to survive a continuous short circuit without fold-back current limiting.

Linear regulator ICs commonly implement some variation of fold-back current limiting which is also tied to temperature in such a way that they can start into a heavy load at normal temperatures but then go into fold-back current limiting mode at higher temperatures.

[soldar]

Every linear power supply I have designed and built had a switch that would select "limit" and  "foldback" so that if it was a load that would just overload the power supply momentarily I would select "limit" but if it was something which needed extra protection I would select "foldback". 

[Electro Fan]

David, Soldar.  Thanks!  That is very helpful, and timely.

I don't have it all grasped just yet but I'm now at least aware of fold-back and some of the possible interactions.

Interesting you should mention USB.

Just today, in the process of trying to figure something else out, I had a small DC motor connected to a linear power supply.  I was vaguely remembering the specs on the motor as 6V and 70ma to 250ma.  As I turned up the voltage and saw Constant Current I gently added some current, some voltage and so forth.  While within or near the specs I heard my computer make the USB startup/quit sound.  It was curious because the power supply and the motor were in no way connected to the computer (or so I thought).  Then I realized that the power supply and the computer were sharing a power strip.  I noticed that as I turned the voltage and current down to some low level the USB sound quit happening; when I turned the voltage and current up the USB sound started happening more often.  My theory is that something I was doing with the power supply (turning it up and down to power the DC motor) was somehow changing the voltage and current at the computer - at least to the level of enabling some USB device or port to toggle on and off.  I looked in the control panel to see if I could find a device connecting and disconnecting but didn't find anything - but I'm pretty sure some USB entity was turning off as my linear power supply was turned up and turned back on as the power supply was turned down.  And the only connection was through the shared power strip - so maybe this was fold-back in action?  Or maybe it was something different and just a coincidence. 

I monitored the power strip and it is only consuming about 1.25 Amps (and I'm pretty sure it is the only thing on the particular 15 Amp AC circuit) so it's hard to understand why the current should be tripping some threshold with just a small DC motor but the USB sound happened enough as I turned the power supply up and down that I'm pretty sure there was some interaction occurring.  Any possibility that this was fold-back in some manner?  Thoughts?

PS, no doubt the computer uses PCIe.

[ArthurDent]

What got me thinking about it is was a GPSDO that I saw spec'd as having 2.5A inrush current but only needing about 0.3 - 0.6A normal operating current.

A lot of GPSDOs have OCXOs for the oscillator. When cold the oven draws a large amount of current that drops after a relatively short time, typically under 10 minutes, so this isn't what I'd call 'inrush' current. Inrush current is usually a few cycles or a second or so in most cases. That's why you may see your lights dip for a second when the fridge motor starts.

Most power supplies used for circuits like this are designed to handle what might be considered a temporary current overload until the oven reaches steady state. Power supplies with a high enough current rating aren't that expensive or physically much larger so get one with a current rating to handle the cold current rating and you can't go wrong.

[Electro Fan]

What got me thinking about it is was a GPSDO that I saw spec'd as having 2.5A inrush current but only needing about 0.3 - 0.6A normal operating current.

A lot of GPSDOs have OCXOs for the oscillator. When cold the oven draws a large amount of current that drops after a relatively short time, typically under 10 minutes, so this isn't what I'd call 'inrush' current. Inrush current is usually a few cycles or a second or so in most cases. That's why you may see your lights dip for a second when the fridge motor starts.

Most power supplies used for circuits like this are designed to handle what might be considered a temporary current overload until the oven reaches steady state. Power supplies with a high enough current rating aren't that expensive or physically much larger so get one with a current rating to handle the cold current rating and you can't go wrong.

Ok, thanks ArthurDent for the distinction in use cases between more conventional inrush current and the OCXO warm up use case.  That is very helpful.

Arthur - if a GPSDO could operate from 15 Volts to 18 Volts and would consume up to 3.5 Amps on startup and then less than 1 Amp after warm up, would you choose a power supply spec'd for 15 Volts and 4 Amps, or a power supply spec'd for 18 Volts and 3.33 Amps (if those were your only two PS choices)?  Why?  Thx!

[radiolistener]

power supply current specification is ability to provide current. The same as your car can drive at 200 km/h, but you're don't need to drive it always at 200 km/h, you can drive it at any desired speed, but you know that your car limit is 200 km/h. The same thing with power supply. You're don't need to get 2 Amps from your power supply, you can get just 1 Amp, but you know that your power supply allows to provide up to 2 Amps and no more than that.

[radiolistener]

if a GPSDO could operate from 15 Volts to 18 Volts and would consume up to 3.5 Amps on startup and then less than 1 Amp after warm up, would you choose a power supply spec'd for 15 Volts and 4 Amps, or a power supply spec'd for 18 Volts and 3.33 Amps (if those were your only two PS choices)?  Why?  Thx!

18 V 3.3A is bad choice. It cannot provide enough current. Power supply should have some margin for current. It should be able to provide for about 20% - 30% higher current, than your requirements. In your case 15V 5A will be good choice.

15 V 4A technically should works ok, but it has too small margin. So, it is possible that it can provide not so clean power. This may be critical for GPSDO.

[ArthurDent]

radiolistener -

15 V 4A technically should works ok, but it has too small margin. So, it is possible that it can provide not so clean power. This may be critical for GPSDO.

That is correct. While I wouldn't use a power supply at its maximum current rating for a constant heavy load, a GPSDO using an OCXO will only draw the highest current for a short period of time and steady state the current drawn will be well within the supply's rating. I just checked a Trimble Thunderbolt and the +12 for the oven starts out at 0.7A and drops to 0.2A in 1.5 minutes. That +12 internally goes through a 78L08 for the oscillator and other circuitry so the OCXO package would operate properly probably down to +10 but would take longer to reach operating temperature. I'm not saying this is a good idea but in this case operating from a +12V/0.7A would work o.k.. Another consideration is the temperature rating on power supplies. most can output less continuous current at higher temperatures. Here is a photo of a Lambda linear supply with the rating on the case. 

So for many reasons it isn't a good idea to operate power supplies at their maximum ratings. What you do is determined by how well you understand the power supply you plan to use and the circuit you're powering.

[Electro Fan]

Ok, so moving on from theoretical to slightly more specific.

A particular Star-4 GPSDO is spec'd as working from 15-24 Volts.  It is unclear depending on what day the seller specifies the current specs as to what the current requirements are.  Best guess so far is 0.3-0.6 Amps in normal steady state running conditions, but supposedly it could be anywhere from 2.5 to 3.5 Amps as recommended for "startup".  Setting aside the current specs for a moment, the question is, why is this (or any GPSDO) spec'd for a range such as 15-24 Volts rather than say 20 Volts?  Just easier to find a PS laying around that will work, or something else? 

Back to the the current.  Let's say the peak requirement for startup is 3.5 Amps.  Let's add 1 Amp for some margin (nearly 30% as suggested in this thread), so the total is 4.5 Amps.  4.5 Amps x 15 Volts = 67.5 Watts.  4.5 Amps x 24 Volts = 108 Watts.  For peak performance, it makes sense to go with the 24 Volt, 4.5 Amp, 108 Watt PS, yes?  Only reason to go with the 15 Volt, 4.5 Amp, 67.5 Watt PS is to save a few $ on the PS and the electric bill?  Or is there some other reasoning to be applied?  Which gets back to the question of why the GPSDO would be spec'd as operating at 15-24 Volts?  Is it to appeal to a user's sense of economics (ie save some $ at 15 Volts), or some other reason?  I guess if it's just economics, and the user doesn't care about how long it takes for the OCXO to heat up and for the GPSDO to lock, then the user might as well go for the lower voltage PS.  On the other hand it seems like both example PS's (15V and 24V) can handle 4.5 Amps, so why should the GPSDO operate differently with 15V vs 24V?  Unless maybe the resistance of the circuitry is such that the GPSDO would be stuck in constant current mode due to an insufficient voltage supply?  Is this like a car that runs ok on regular gasoline but will run somewhat better if the user is willing to spend more on premium gasoline?  (Only analogy I've come up with so far.) 

In summary, what is the best guess as to why the GPSDO is spec'd for 15-24V?  And, $ aside, would 24V/4.5A likely provide better performance than 15V/4.5A, or not necessarily?

(Disclaimer, I'd like to squeeze some extra performance out of the GPSDO if possible but I'm primarily interested in using this as a way to learn some electricity 101.  Thanks)

[ArthurDent]

Electro Fan –

”… On the other hand it seems like both example PS's (15V and 24V) can handle 4.5 Amps, so why should the GPSDO operate differently with 15V vs 24V?”

 The simple answer is-it doesn’t. Almost anything designed to operate over a wide voltage range today uses a SMPS. If you look at the back of your computer, or on your laptop power supply, it will probably say ‘100-240 VAC input’. This does not mean that either device will draw twice as much power at 240 VAC as 120 VAC.
 
This explanation may be overly simplified but by using pulse width modulation, switching mode power supplies transfers only enough power from the input to the output to maintain the required output. In math that is integrating the area under the curve. That means as the input voltage goes up, the input current goes down and the input power remains constant and the output voltage remains constant, controlled through a feedback loop.

 Most Telco and a lot of other battery operated equipment have to operate over a wide voltage range to compensate for the charge state of the batteries. If you have a frequency standard it must operate from a regulated source to insure it is as stable as possible.

Here is a photo of the DC-DC power module inside a Telco 24VDC GPSDO device to deliver the various regulated voltages it needs. You can see that this supply is designed to operate from 18-36 VDC which would be supplied by the Telco backup batteries during a power failure. 

[Electro Fan]

Electro Fan –

”… On the other hand it seems like both example PS's (15V and 24V) can handle 4.5 Amps, so why should the GPSDO operate differently with 15V vs 24V?”

 The simple answer is-it doesn’t. Almost anything designed to operate over a wide voltage range today uses a SMPS. If you look at the back of your computer, or on your laptop power supply, it will probably say ‘100-240 VAC input’. This does not mean that either device will draw twice as much power at 240 VAC as 120 VAC.
 
This explanation may be overly simplified but by using pulse width modulation, switching mode power supplies transfers only enough power from the input to the output to maintain the required output. In math that is integrating the area under the curve. That means as the input voltage goes up, the input current goes down and the input power remains constant and the output voltage remains constant, controlled through a feedback loop.

 Most Telco and a lot of other battery operated equipment have to operate over a wide voltage range to compensate for the charge state of the batteries. If you have a frequency standard it must operate from a regulated source to insure it is as stable as possible.

Here is a photo of the DC-DC power module inside a Telco 24VDC GPSDO device to deliver the various regulated voltages it needs. You can see that this supply is designed to operate from 18-36 VDC which would be supplied by the Telco backup batteries during a power failure.

Ok, that explains a lot

- SMPS can determine the area under the curve to balance out voltage and current.
- telco equipment is designed to work with a range of battery voltages for power failure.

Thanks!