Trying to cheat ohm's law

For a circuit I described here recently, I want to be able to cause a brief, temporary, high current draw on a 12v power supply. I'd like to draw about 2A for a few ms, and at least 0.5A for perhaps 100ms.

I have a resistor + capacitor combination that gives a curve I like. A regulated 12v supply that claims to produce 5A, with a 33000uf cap in series with 3.9 ohm resistor across it, is loaded to 2+A and is still pulling 500ma, 200ms later. +12v--3.9ohm--33000uf--Gnd

The problem is, if I understand things, the resistor would have to handle about 30W for a few milliseconds. I'd like to avoid using a component that physically large and pricey. I'm guessing I could get away with 5W because the high current is brief, but I want this circuit to work reliably for years.

Is there something smaller/cheaper I can use?

 

On Tuesday, July 7, 2015 at 1:46:45 PM UTC-4, scott....@gmail.com wrote:

> Is there something smaller/cheaper I can use?

Never mind. I was put off by 2$ and 3$ resistors, but I found 3.9 ohm 10W for $0.64.

 

[Tim Wescott]

On Tue, 07 Jul 2015 10:46:41 -0700, scott.a.mayo wrote:

For a circuit I described here recently, I want to be able to cause a
brief, temporary, high current draw on a 12v power supply. I'd like to
draw about 2A for a few ms, and at least 0.5A for perhaps 100ms.

I have a resistor + capacitor combination that gives a curve I like. A
regulated 12v supply that claims to produce 5A, with a 33000uf cap in
series with 3.9 ohm resistor across it, is loaded to 2+A and is still
pulling 500ma, 200ms later. +12v--3.9ohm--33000uf--Gnd

The problem is, if I understand things, the resistor would have to
handle about 30W for a few milliseconds. I'd like to avoid using a
component that physically large and pricey. I'm guessing I could get
away with 5W because the high current is brief, but I want this circuit
to work reliably for years.

Is there something smaller/cheaper I can use?

Almost certainly yes. Find a resistor that is specified by it's
manufacturer for transient loads; the ones that I've designed in will
have a graph in the data sheet that plots the load power against duration
(or average power, or energy, or something -- you'll have to do some math
to get it into the form you need).

I can't give you any part numbers or manufacturers off the top of my
head, but they're out there, and for that brief of a load you probably
won't need much more than something rated for your average power
consumption.

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

 

[George Herold]

On Tuesday, July 7, 2015 at 6:34:14 PM UTC-4, Tim Watts wrote:

On 07/07/15 18:46, scott.a.mayo@gmail.com wrote:
For a circuit I described here recently, I want to be able to cause a
brief, temporary, high current draw on a 12v power supply. I'd like
to draw about 2A for a few ms, and at least 0.5A for perhaps 100ms.

I have a resistor + capacitor combination that gives a curve I like.
A regulated 12v supply that claims to produce 5A, with a 33000uf cap
in series with 3.9 ohm resistor across it, is loaded to 2+A and is
still pulling 500ma, 200ms later. +12v--3.9ohm--33000uf--Gnd

The problem is, if I understand things, the resistor would have to
handle about 30W for a few milliseconds. I'd like to avoid using a
component that physically large and pricey. I'm guessing I could get
away with 5W because the high current is brief, but I want this
circuit to work reliably for years.

Is there something smaller/cheaper I can use?


Yes - use a 7-10W resistor.

30W for a few mS is of the order of 0.1J of energy. The question is: how
much will 0.1J heat up a resistor body?

So here's a cool thing, around room temp everything has
about the same volumetric(sp) heat capacity. ~3 J/(K*cm^3).

George H.

Answer - not much. If you want to get technical, look at the mass of the
body and guesstimate a specific heat capacity and work out a pulse
temperature rise.


There are extremes to this - if the pulse is very high for a very short
time, it could heat the resistor material (wire?) to melting point
before the energy dissipates throughout the body of the resistor.

Hence suggesting a 7-10W resistor as these are cheap, not very big and
very common. My gut instinct says that will be fine and in fact you
could go lower still, possibly.

 

[Tim Watts]

On 07/07/15 18:46, scott.a.mayo@gmail.com wrote:

For a circuit I described here recently, I want to be able to cause a
brief, temporary, high current draw on a 12v power supply. I'd like
to draw about 2A for a few ms, and at least 0.5A for perhaps 100ms.

I have a resistor + capacitor combination that gives a curve I like.
A regulated 12v supply that claims to produce 5A, with a 33000uf cap
in series with 3.9 ohm resistor across it, is loaded to 2+A and is
still pulling 500ma, 200ms later. +12v--3.9ohm--33000uf--Gnd

The problem is, if I understand things, the resistor would have to
handle about 30W for a few milliseconds. I'd like to avoid using a
component that physically large and pricey. I'm guessing I could get
away with 5W because the high current is brief, but I want this
circuit to work reliably for years.

Is there something smaller/cheaper I can use?

Yes - use a 7-10W resistor.

30W for a few mS is of the order of 0.1J of energy. The question is: how
much will 0.1J heat up a resistor body?

Answer - not much. If you want to get technical, look at the mass of the
body and guesstimate a specific heat capacity and work out a pulse
temperature rise.


There are extremes to this - if the pulse is very high for a very short
time, it could heat the resistor material (wire?) to melting point
before the energy dissipates throughout the body of the resistor.

Hence suggesting a 7-10W resistor as these are cheap, not very big and
very common. My gut instinct says that will be fine and in fact you
could go lower still, possibly.

 

[Jim Thompson]

On Tue, 7 Jul 2015 10:46:41 -0700 (PDT), scott.a.mayo@gmail.com wrote:

For a circuit I described here recently, I want to be able to cause a brief, temporary, high current draw on a 12v power supply. I'd like to draw about 2A for a few ms, and at least 0.5A for perhaps 100ms.

I have a resistor + capacitor combination that gives a curve I like. A regulated 12v supply that claims to produce 5A, with a 33000uf cap in series with 3.9 ohm resistor across it, is loaded to 2+A and is still pulling 500ma, 200ms later. +12v--3.9ohm--33000uf--Gnd

The problem is, if I understand things, the resistor would have to handle about 30W for a few milliseconds. I'd like to avoid using a component that physically large and pricey. I'm guessing I could get away with 5W because the high current is brief, but I want this circuit to work reliably for years.

Is there something smaller/cheaper I can use?

Just a power MOSFET, an OpAmp and a few resistors and a capacitor.

...Jim Thompson
--
| James E.Thompson | mens |
| Analog Innovations | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| San Tan Valley, AZ 85142 Skype: skypeanalog | |
| Voice480)460-2350 Fax: Available upon request | Brass Rat |
| E-mail Icon at http://www.analog-innovations.com | 1962 |

I love to cook with wine. Sometimes I even put it in the food.

 

[Phil Allison]

scott....@gmail.com wrote:

Never mind. I was put off by 2$ and 3$ resistors,
but I found 3.9 ohm 10W for $0.64.

** 10 watt wire-wound resistors are tough components. Most are rated to accept 10 times rated power for 5 seconds and can usually stand up to 100 times for a few milliseconds.

So in your app there is no problem at all.

Ones like the 6.8ohm example below are often used in line with the AC supply to large transformers to reduce inrush surge currents at switch on - where peak values might be up to 40amps during first half cycles. Normally the resistor is bypassed by a pair of relay contacts after a short time.

http://casamodularsystems.com/images/Restistors/6R8-J-10W-WW-AX-PW10-IRH_640p.jpg


..... Phil

 

[John Larkin]

On Tue, 7 Jul 2015 18:17:32 -0700 (PDT), George Herold
<gherold@teachspin.com> wrote:

On Tuesday, July 7, 2015 at 6:34:14 PM UTC-4, Tim Watts wrote:
On 07/07/15 18:46, scott.a.mayo@gmail.com wrote:
For a circuit I described here recently, I want to be able to cause a
brief, temporary, high current draw on a 12v power supply. I'd like
to draw about 2A for a few ms, and at least 0.5A for perhaps 100ms.

I have a resistor + capacitor combination that gives a curve I like.
A regulated 12v supply that claims to produce 5A, with a 33000uf cap
in series with 3.9 ohm resistor across it, is loaded to 2+A and is
still pulling 500ma, 200ms later. +12v--3.9ohm--33000uf--Gnd

The problem is, if I understand things, the resistor would have to
handle about 30W for a few milliseconds. I'd like to avoid using a
component that physically large and pricey. I'm guessing I could get
away with 5W because the high current is brief, but I want this
circuit to work reliably for years.

Is there something smaller/cheaper I can use?


Yes - use a 7-10W resistor.

30W for a few mS is of the order of 0.1J of energy. The question is: how
much will 0.1J heat up a resistor body?

So here's a cool thing, around room temp everything has
about the same volumetric(sp) heat capacity. ~3 J/(K*cm^3).

Cool, but a 1206 resistor may not have good transient thermal coupling
from the thin resistive element into the alumina substrate.

30 w for a few ms will damage a conventional thickfilm or thinfilm
1206 resistor; trust me on that, we've proven it recently.
Surface-mount wirewounds are the way to go for big power peaks.


--

John Larkin Highland Technology, Inc
picosecond timing laser drivers and controllers

jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com

 

[Tim Watts]

On 08/07/15 04:41, John Larkin wrote:

Cool, but a 1206 resistor may not have good transient thermal coupling
from the thin resistive element into the alumina substrate.

30 w for a few ms will damage a conventional thickfilm or thinfilm
1206 resistor; trust me on that, we've proven it recently.
Surface-mount wirewounds are the way to go for big power peaks.

I don't doubt it - those things are pretty small.

That's why I suggest a 7-10W to the OP

 

[George Herold]

On Tuesday, July 7, 2015 at 11:41:39 PM UTC-4, John Larkin wrote:

On Tue, 7 Jul 2015 18:17:32 -0700 (PDT), George Herold
gherold@teachspin.com> wrote:

On Tuesday, July 7, 2015 at 6:34:14 PM UTC-4, Tim Watts wrote:
On 07/07/15 18:46, scott.a.mayo@gmail.com wrote:
For a circuit I described here recently, I want to be able to cause a
brief, temporary, high current draw on a 12v power supply. I'd like
to draw about 2A for a few ms, and at least 0.5A for perhaps 100ms.

I have a resistor + capacitor combination that gives a curve I like.
A regulated 12v supply that claims to produce 5A, with a 33000uf cap
in series with 3.9 ohm resistor across it, is loaded to 2+A and is
still pulling 500ma, 200ms later. +12v--3.9ohm--33000uf--Gnd

The problem is, if I understand things, the resistor would have to
handle about 30W for a few milliseconds. I'd like to avoid using a
component that physically large and pricey. I'm guessing I could get
away with 5W because the high current is brief, but I want this
circuit to work reliably for years.

Is there something smaller/cheaper I can use?


Yes - use a 7-10W resistor.

30W for a few mS is of the order of 0.1J of energy. The question is: how
much will 0.1J heat up a resistor body?

So here's a cool thing, around room temp everything has
about the same volumetric(sp) heat capacity. ~3 J/(K*cm^3).

Cool, but a 1206 resistor may not have good transient thermal coupling
from the thin resistive element into the alumina substrate.

30 w for a few ms will damage a conventional thickfilm or thinfilm
1206 resistor; trust me on that, we've proven it recently.
Surface-mount wirewounds are the way to go for big power peaks.

Well not many cm^3 in a 1206... (one here is ~0.02 inches tall so
~2.5E-3 cm^3, ~7E-3 J/K, 0.1 J is about a 14 degree rise...
Assuming I punched in all the numbers correctly. )

OK poor coupling... so 30 W for ~3 ms blows up a film 1206?

George H.

OK so

--

John Larkin Highland Technology, Inc
picosecond timing laser drivers and controllers

jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com

 

[John Larkin]

On Wed, 08 Jul 2015 07:45:19 +0100, Tim Watts <tw_usenet@dionic.net>
wrote:

On 08/07/15 04:41, John Larkin wrote:

Cool, but a 1206 resistor may not have good transient thermal coupling
from the thin resistive element into the alumina substrate.

30 w for a few ms will damage a conventional thickfilm or thinfilm
1206 resistor; trust me on that, we've proven it recently.
Surface-mount wirewounds are the way to go for big power peaks.

I don't doubt it - those things are pretty small.

But they have a lot of active resistance material, and it's hard to
damage nichrome wire. The thermal interface surface, wire to epoxy or
whatever, is huge. The steady-state cooling is limited by air-exposed
surface and the small PCB pads, but transient energy absorption is
excellent on wirewounds. They can be had with food TCs, too, unlike
bulk conductors like carbon resistors.

That's why I suggest a 7-10W to the OP

Sure, but big resistors tend to be, well, big. Maybe that's OK.




--

John Larkin Highland Technology, Inc
picosecond timing laser drivers and controllers

jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com

 

[John Larkin]

On Wed, 8 Jul 2015 07:44:08 -0700 (PDT), George Herold
<gherold@teachspin.com> wrote:

On Tuesday, July 7, 2015 at 11:41:39 PM UTC-4, John Larkin wrote:
On Tue, 7 Jul 2015 18:17:32 -0700 (PDT), George Herold
gherold@teachspin.com> wrote:

On Tuesday, July 7, 2015 at 6:34:14 PM UTC-4, Tim Watts wrote:
On 07/07/15 18:46, scott.a.mayo@gmail.com wrote:
For a circuit I described here recently, I want to be able to cause a
brief, temporary, high current draw on a 12v power supply. I'd like
to draw about 2A for a few ms, and at least 0.5A for perhaps 100ms.

I have a resistor + capacitor combination that gives a curve I like.
A regulated 12v supply that claims to produce 5A, with a 33000uf cap
in series with 3.9 ohm resistor across it, is loaded to 2+A and is
still pulling 500ma, 200ms later. +12v--3.9ohm--33000uf--Gnd

The problem is, if I understand things, the resistor would have to
handle about 30W for a few milliseconds. I'd like to avoid using a
component that physically large and pricey. I'm guessing I could get
away with 5W because the high current is brief, but I want this
circuit to work reliably for years.

Is there something smaller/cheaper I can use?


Yes - use a 7-10W resistor.

30W for a few mS is of the order of 0.1J of energy. The question is: how
much will 0.1J heat up a resistor body?

So here's a cool thing, around room temp everything has
about the same volumetric(sp) heat capacity. ~3 J/(K*cm^3).

Cool, but a 1206 resistor may not have good transient thermal coupling
from the thin resistive element into the alumina substrate.

30 w for a few ms will damage a conventional thickfilm or thinfilm
1206 resistor; trust me on that, we've proven it recently.
Surface-mount wirewounds are the way to go for big power peaks.

Well not many cm^3 in a 1206... (one here is ~0.02 inches tall so
~2.5E-3 cm^3, ~7E-3 J/K, 0.1 J is about a 14 degree rise...
Assuming I punched in all the numbers correctly. )

OK poor coupling... so 30 W for ~3 ms blows up a film 1206?

One specific case was a 1206, 50 ohm precision resistor pulsed at 30
volts, which is 18 watts. Pulsed for 2 ms at 1 Hz is 36 mJ per shot,
36 mW average power. Average lifetime to hard failure was around 30
minutes, but just a few shots changed the value out of the 0.1%
tolerance. On an IR viewer, each shot made a central thermal hotspot -
poor coupling! - which has to be doing mechanical damage. We think
that laser trimming creates hot/weak spots on the surface. It looks
like four of those resistors connected series-parallel would be OK,
but we'll probably go with with one wirewound. For the 5 ohm range, we
have to go wirewound.

I'd expect that 30 watts for 3 ms will kill a normal 1206 resistor in
one shot. Try it!


--

John Larkin Highland Technology, Inc
picosecond timing laser drivers and controllers

jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com

 

[Tim Wescott]

On Tue, 07 Jul 2015 23:34:08 +0100, Tim Watts wrote:

On 07/07/15 18:46, scott.a.mayo@gmail.com wrote:
For a circuit I described here recently, I want to be able to cause a
brief, temporary, high current draw on a 12v power supply. I'd like to
draw about 2A for a few ms, and at least 0.5A for perhaps 100ms.

I have a resistor + capacitor combination that gives a curve I like. A
regulated 12v supply that claims to produce 5A, with a 33000uf cap in
series with 3.9 ohm resistor across it, is loaded to 2+A and is still
pulling 500ma, 200ms later. +12v--3.9ohm--33000uf--Gnd

The problem is, if I understand things, the resistor would have to
handle about 30W for a few milliseconds. I'd like to avoid using a
component that physically large and pricey. I'm guessing I could get
away with 5W because the high current is brief, but I want this circuit
to work reliably for years.

Is there something smaller/cheaper I can use?


Yes - use a 7-10W resistor.

30W for a few mS is of the order of 0.1J of energy. The question is: how
much will 0.1J heat up a resistor body?

Answer - not much. If you want to get technical, look at the mass of the
body and guesstimate a specific heat capacity and work out a pulse
temperature rise.


There are extremes to this - if the pulse is very high for a very short
time, it could heat the resistor material (wire?) to melting point
before the energy dissipates throughout the body of the resistor.

Hence why I suggested finding a resistor whose data sheet lists pulse
power vs. time -- because the manufacturer has tested it for all the
various thermal time constants, and you don't have to do things b'guess
and b'gosh.

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

 

[George Herold]

On Wednesday, July 8, 2015 at 12:19:11 PM UTC-4, John Larkin wrote:

On Wed, 8 Jul 2015 07:44:08 -0700 (PDT), George Herold
gherold@teachspin.com> wrote:

On Tuesday, July 7, 2015 at 11:41:39 PM UTC-4, John Larkin wrote:
On Tue, 7 Jul 2015 18:17:32 -0700 (PDT), George Herold
gherold@teachspin.com> wrote:

On Tuesday, July 7, 2015 at 6:34:14 PM UTC-4, Tim Watts wrote:
On 07/07/15 18:46, scott.a.mayo@gmail.com wrote:
For a circuit I described here recently, I want to be able to cause a
brief, temporary, high current draw on a 12v power supply. I'd like
to draw about 2A for a few ms, and at least 0.5A for perhaps 100ms.

I have a resistor + capacitor combination that gives a curve I like.
A regulated 12v supply that claims to produce 5A, with a 33000uf cap
in series with 3.9 ohm resistor across it, is loaded to 2+A and is
still pulling 500ma, 200ms later. +12v--3.9ohm--33000uf--Gnd

The problem is, if I understand things, the resistor would have to
handle about 30W for a few milliseconds. I'd like to avoid using a
component that physically large and pricey. I'm guessing I could get
away with 5W because the high current is brief, but I want this
circuit to work reliably for years.

Is there something smaller/cheaper I can use?


Yes - use a 7-10W resistor.

30W for a few mS is of the order of 0.1J of energy. The question is: how
much will 0.1J heat up a resistor body?

So here's a cool thing, around room temp everything has
about the same volumetric(sp) heat capacity. ~3 J/(K*cm^3).

Cool, but a 1206 resistor may not have good transient thermal coupling
from the thin resistive element into the alumina substrate.

30 w for a few ms will damage a conventional thickfilm or thinfilm
1206 resistor; trust me on that, we've proven it recently.
Surface-mount wirewounds are the way to go for big power peaks.

Well not many cm^3 in a 1206... (one here is ~0.02 inches tall so
~2.5E-3 cm^3, ~7E-3 J/K, 0.1 J is about a 14 degree rise...
Assuming I punched in all the numbers correctly. )

OK poor coupling... so 30 W for ~3 ms blows up a film 1206?


One specific case was a 1206, 50 ohm precision resistor pulsed at 30
volts, which is 18 watts. Pulsed for 2 ms at 1 Hz is 36 mJ per shot,
36 mW average power. Average lifetime to hard failure was around 30
minutes, but just a few shots changed the value out of the 0.1%
tolerance. On an IR viewer, each shot made a central thermal hotspot -
poor coupling! - which has to be doing mechanical damage. We think
that laser trimming creates hot/weak spots on the surface. It looks
like four of those resistors connected series-parallel would be OK,
but we'll probably go with with one wirewound. For the 5 ohm range, we
have to go wirewound.

It's hard to know what's going on at the surface.
The different TEC (Thermal Expansion) could rip things apart,
with each pulse.

I'd expect that 30 watts for 3 ms will kill a normal 1206 resistor in
one shot. Try it!

Well I'm not sure I have time. (I'm using electronics
to find my math mistakes in a sqrt circuit. :^)
But I can believe that the volume of the resistive material
is pretty small.

Having the resistor taking up all the volume..
like old carbon composites, makes a lot of sense for pulsed power.

Or wire wound, as you said.

George H.

--

John Larkin Highland Technology, Inc
picosecond timing laser drivers and controllers

jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com

 

[Phil Allison]

John G wrote:

It happens that Phil Allison formulated :

http://casamodularsystems.com/images/Restistors/6R8-J-10W-WW-AX-PW10-IRH_640p.jpg


Phil Have you moved across the ditch? lol

** John is alluding to the gold coloured coin shown alongside the 10W resistor in the JPG, the Kiki bird depicted indicates a New Zealand dollar.

Australia is separated from new Zealand by the Tasman Sea - aka "the Ditch".

The 10W resistor has "IRH Australia" printed on it, so it was likely made here in Sydney.

The resistor is 48 x 9.5 x 9 mm - which seems to be standard size.


.... Phil

 

[John G]

It happens that Phil Allison formulated :

scott....@gmail.com wrote:



Never mind. I was put off by 2$ and 3$ resistors,
but I found 3.9 ohm 10W for $0.64.

** 10 watt wire-wound resistors are tough components. Most are rated to
accept 10 times rated power for 5 seconds and can usually stand up to 100
times for a few milliseconds.

So in your app there is no problem at all.

Ones like the 6.8ohm example below are often used in line with the AC supply
to large transformers to reduce inrush surge currents at switch on - where
peak values might be up to 40amps during first half cycles. Normally the
resistor is bypassed by a pair of relay contacts after a short time.

http://casamodularsystems.com/images/Restistors/6R8-J-10W-WW-AX-PW10-IRH_640p.jpg

> .... Phil

Phil Have you moved across the ditch? lol

--
John G Sydney.