Current limitation circuit

[OBones]

Hi all,

I'm considering to build my own PWM power supply for my mini drill. My
first idea is to use a NE555 driving an N Type Mosfet and that seems ok.
However, I would like to put some sort of "locking" protection into the
circuit. If I'm correct, when the effort increases onto the drill, the
current goes up, and can go very high if the rotor is blocked.
So this condition can be detected, but I have no clues as to how to do
that. Any leads are welcome.

In the end, I'd like to get the functionnalities of this project:
http://www.web-ee.com/Electronic-Projects/projects/pcb_drill/index.shtml

but using more up to date components, especially Mosfet(s) instead of
transistors to minimize the dissipated heat.

Thanks a lot for your help.
Olivier

 

[Tim Wescott]

Hi all,

I'm considering to build my own PWM power supply for my mini drill. My
first idea is to use a NE555 driving an N Type Mosfet and that seems ok.
However, I would like to put some sort of "locking" protection into the
circuit. If I'm correct, when the effort increases onto the drill, the
current goes up, and can go very high if the rotor is blocked.
So this condition can be detected, but I have no clues as to how to do
that. Any leads are welcome.

In the end, I'd like to get the functionnalities of this project:
http://www.web-ee.com/Electronic-Projects/projects/pcb_drill/index.shtml

but using more up to date components, especially Mosfet(s) instead of
transistors to minimize the dissipated heat.

Thanks a lot for your help.
Olivier

A good way to detect that the motor is stalled is to monitor its
current. The thing that damages the motor when it is stalled is its
current. The thing that damages a motor that's _almost_ stalled is its
current.

So don't try to detect a stall, just limit the current!

If you want to get fancy then let the motor run over stall current for
1/2 second or so before reducing the PWM. This will let it do things
like spin up or get momentarily stuck without shutting the whole thing
down. You can get real fancy with this, but it's hard to do within the
confines of analog circuitry -- and a PIC with a cheezy ADC is almost as
small as a 555.

 

[Joerg]

Hello Tim,

So don't try to detect a stall, just limit the current!
Agree.

If you want to get fancy then let the motor run over stall current for
1/2 second or so before reducing the PWM. This will let it do things
like spin up or get momentarily stuck without shutting the whole thing
down. You can get real fancy with this, but it's hard to do within the
confines of analog circuitry -- and a PIC with a cheezy ADC is almost as
small as a 555.

Analog ain't that bad. It's possible with some cheap logic chip like a
CD40106. I did that for an industrial angle grinder. It would let the
user go through a tough spot for a second but when the load persisted
would regulate it down to safe levels, all the way to "off" in a
complete stall. The circuit also had a "soft start" so if would wind
back up more gently. The formerly common rapid wind up from zero had
placed a lot of strain on the wrists of users.

Possibly something like that might help prevent injuries because a
sudden and fast revving after getting unstuck can send stuff or the
whole tool on a rampage.

Regards, Joerg

 

[Tim Wescott]

Hello Tim,



Analog ain't that bad. It's possible with some cheap logic chip like a
CD40106. I did that for an industrial angle grinder. It would let the
user go through a tough spot for a second but when the load persisted
would regulate it down to safe levels, all the way to "off" in a
complete stall. The circuit also had a "soft start" so if would wind
back up more gently. The formerly common rapid wind up from zero had
placed a lot of strain on the wrists of users.

But your counter circuit is probably the limits of what you would want
with analog circuitry.

If I somehow got a contract that was intended for you and needed to make
a gazillion of the things for less than a penny apiece then I'd take
your approach. If I needed ten or a hundred, or if I needed more
features, I'd think of using the PIC.

Possibly something like that might help prevent injuries because a
sudden and fast revving after getting unstuck can send stuff or the
whole tool on a rampage.

Nick Park's "Wallice and Grommit: A Grand Day Out"; the scene where
Grommit drills a hole...

 

[OBones]

Hello Tim,



Analog ain't that bad. It's possible with some cheap logic chip like a
CD40106. I did that for an industrial angle grinder. It would let the
user go through a tough spot for a second but when the load persisted
would regulate it down to safe levels, all the way to "off" in a
complete stall. The circuit also had a "soft start" so if would wind
back up more gently. The formerly common rapid wind up from zero had
placed a lot of strain on the wrists of users.

Possibly something like that might help prevent injuries because a
sudden and fast revving after getting unstuck can send stuff or the
whole tool on a rampage.

Regards, Joerg

Hey, I like that approach. Are you allowed to post a schematic of that
part of the project?

 

[OBones]

But your counter circuit is probably the limits of what you would want
with analog circuitry.

If I somehow got a contract that was intended for you and needed to make
a gazillion of the things for less than a penny apiece then I'd take
your approach. If I needed ten or a hundred, or if I needed more
features, I'd think of using the PIC.

Yeah, but for my project, it feels like using a hammer to splat a fly...

 

[Tim Wescott]

Yeah, but for my project, it feels like using a hammer to splat a fly...

Microprocessors are like that. But they're also often the smallest,
least-cost way to go for a surprisingly large (and growing) set of problems.

If you can do it reliably with analog circuitry, meet all your
performance, price and manufacturability goals _and_ not be too big then
that's the way to go.

 

[Joerg]

Hello Tim,

Most of my stuff is in the gazillion category. Maybe that's how I got to
be a cheapskate when it comes to electronics. Other times I don't like
simple apps to return "the program has performed an illegal
operation...". But that's rare nowadays with uCs, which is why I started
looking at the MSP430. Before it ends up in a design though they have to
make good on their promise "starting at 49 cents". At the current Dollar
something a pop, no chance.

Hey, I splatted a fly with a wood splitter a couple weeks ago. It got
sucked into the intake of the engine.

Microprocessors are like that. But they're also often the smallest,
least-cost way to go for a surprisingly large (and growing) set of
problems.

Problem is, often that's only financially feasible with Chinese
four-bitters. Anything eight bit seems to be 30c or more.

Regards, Joerg

 

[Joerg]

Hello OBones,

Hey, I like that approach. Are you allowed to post a schematic of that
part of the project?

Unfortunately not. Also, this was not for DC but for 240V AC. Anyway,
all you need to do for jobs like this is figure out what kinds of ramps
and delays you need. Then take a Hex-Schmitt and create those. "Poor
man's logic" with dual-diodes and transistors does the interaction. Most
of the time I don't even need all six inverters.

Regards, Joerg

 

[Tim Wescott]

Hello Tim,



Most of my stuff is in the gazillion category. Maybe that's how I got to
be a cheapskate when it comes to electronics. Other times I don't like
simple apps to return "the program has performed an illegal
operation...". But that's rare nowadays with uCs, which is why I started
looking at the MSP430. Before it ends up in a design though they have to
make good on their promise "starting at 49 cents". At the current Dollar
something a pop, no chance.



Hey, I splatted a fly with a wood splitter a couple weeks ago. It got
sucked into the intake of the engine.



Problem is, often that's only financially feasible with Chinese
four-bitters. Anything eight bit seems to be 30c or more.

Regards, Joerg

For you.

Most of my customers so far build around 100/year/board and would like
to be able to do field upgrades or at least have flexible manufacturing
capabilities; for them the engineering time makes a big factor in the
amount they have to charge per board. Add that to the fact that with a
digital system you only have to sweat the precision of the ADC and DAC
circuits and microprocessors start to look very attractive.

 

[Joerg]

Hello Tim,

Most of my customers so far build around 100/year/board and would like
to be able to do field upgrades or at least have flexible manufacturing
capabilities; for them the engineering time makes a big factor in the
amount they have to charge per board. Add that to the fact that with a
digital system you only have to sweat the precision of the ADC and DAC
circuits and microprocessors start to look very attractive.

Yes, for 100qty the uC makes perfect sense. Sometimes even for large
qties. Like a strobe that has to double flash at very precise intervals.
A uC with a versatile timer can run circles around an analog solution
for such designs.

Regards, Joerg

 

[OBones]

For you.

Most of my customers so far build around 100/year/board and would like
to be able to do field upgrades or at least have flexible manufacturing
capabilities; for them the engineering time makes a big factor in the
amount they have to charge per board. Add that to the fact that with a
digital system you only have to sweat the precision of the ADC and DAC
circuits and microprocessors start to look very attractive.

I agree with that, but thing is, this is a once in a lifetime project
(most probably), and this would require me to have a PIC programmer,
which I don't have. I could build one (I have the schematics), but it
increases the costs for something that I consider to be simple.
I don't need high precision, just some sort of a protection. Thanks for
the help, I appreciate it.

 

[OBones]

Hello OBones,



Unfortunately not. Also, this was not for DC but for 240V AC. Anyway,
all you need to do for jobs like this is figure out what kinds of ramps
and delays you need. Then take a Hex-Schmitt and create those. "Poor
man's logic" with dual-diodes and transistors does the interaction. Most
of the time I don't even need all six inverters.

Ok, that's what I thought, NDA and all that stuff coming into play. I'll
look around and think hard about it.
Thanks.

 

[Luhan Monat]

Hi all,

I'm considering to build my own PWM power supply for my mini drill. My
first idea is to use a NE555 driving an N Type Mosfet and that seems ok.
However, I would like to put some sort of "locking" protection into the
circuit. If I'm correct, when the effort increases onto the drill, the
current goes up, and can go very high if the rotor is blocked.
So this condition can be detected, but I have no clues as to how to do
that. Any leads are welcome.

In the end, I'd like to get the functionnalities of this project:
http://www.web-ee.com/Electronic-Projects/projects/pcb_drill/index.shtml

but using more up to date components, especially Mosfet(s) instead of
transistors to minimize the dissipated heat.

Thanks a lot for your help.
Olivier

For the current limit function, I like 10:1 self-adjusting resistors.
These are sold in many venues marketed as 'light bulbs'. A 100 watt
bulb has a cold resistance of about 10 ohms, goes to about 100 ohms hot,
and, oh yea, dissipates 100 watts if necessary.

 

[Joerg]

Hello OBones,

I agree with that, but thing is, this is a once in a lifetime project
(most probably), and this would require me to have a PIC programmer,
which I don't have. I could build one (I have the schematics), but it
increases the costs for something that I consider to be simple.
I don't need high precision, just some sort of a protection. Thanks for
the help, I appreciate it.

In case you want to learn about micro controllers, and I guess only then
it would make sense: Check out the Texas Instruments MSP430 series. They
can be programmed via an RS232 protocol and the interface is rather
simple. Of course, you'd also have to make sure your computer still has
one of those. Unfortunately many "modern" laptops don't.

If you go that route the MSP430F1232 is a pretty decent one because it
has ADC capabilities.

Regards, Joerg

 

[ehsjr]

Hi all,

I'm considering to build my own PWM power supply for my mini drill. My
first idea is to use a NE555 driving an N Type Mosfet and that seems ok.
However, I would like to put some sort of "locking" protection into the
circuit. If I'm correct, when the effort increases onto the drill, the
current goes up, and can go very high if the rotor is blocked.
So this condition can be detected, but I have no clues as to how to do
that. Any leads are welcome.

In the end, I'd like to get the functionnalities of this project:
http://www.web-ee.com/Electronic-Projects/projects/pcb_drill/index.shtml

but using more up to date components, especially Mosfet(s) instead of
transistors to minimize the dissipated heat.

Thanks a lot for your help.
Olivier

Put a current limiting chip/circuit between the DC supply and
the mosfet. For complete turn off if too much current is
drawn (such as would happen in a rotor locked condition), add
a double pole relay and a pushbutton. One set of contacts
on the relay is in series with the supply, and, when the relay
is energized, connects the relay coil to the output side
of the current limiter. The other set of contacts is in series
with the input to the current limiter. The pushbutton momentarily
connects the relay coil to the power supply. When too much current
is drawn, the current limiter drops the voltage and the relay
drops out, de-energizing everything. I'll try a drawing:
__
+--o o-------------------------+---relaycoil---+
| | |
| +--o/ o--+ |
| | |
+ --+--o/ o--+--Limiter----+-----mosfet-----motor---+
| | |
+--------------------PWM---------------+
|
- --------------------------------------------------+

Ed

 

[OBones]

Hello OBones,



In case you want to learn about micro controllers, and I guess only then
it would make sense: Check out the Texas Instruments MSP430 series. They
can be programmed via an RS232 protocol and the interface is rather
simple. Of course, you'd also have to make sure your computer still has
one of those. Unfortunately many "modern" laptops don't.

If you go that route the MSP430F1232 is a pretty decent one because it
has ADC capabilities.

Oh but I know about microcontrollers, I used them when I was a student,
two years ago. But I don't have access to the uni resources anymore...

 

[OBones]

Put a current limiting chip/circuit between the DC supply and
the mosfet. For complete turn off if too much current is
drawn (such as would happen in a rotor locked condition), add
a double pole relay and a pushbutton. One set of contacts
on the relay is in series with the supply, and, when the relay
is energized, connects the relay coil to the output side
of the current limiter. The other set of contacts is in series
with the input to the current limiter. The pushbutton momentarily
connects the relay coil to the power supply. When too much current
is drawn, the current limiter drops the voltage and the relay
drops out, de-energizing everything. I'll try a drawing:
__
+--o o-------------------------+---relaycoil---+
| | |
| +--o/ o--+ |
| | |
+ --+--o/ o--+--Limiter----+-----mosfet-----motor---+
| | |
+--------------------PWM---------------+
|
- --------------------------------------------------+

Thanks.