Repaired Harbor Freight digital caliper

[Bob Engelhardt]

On 11/23/2011 2:05 PM, Bob Engelhardt wrote:

Not if the PV cell's rated SC current is less than the max LED current
rating.

Isn't the capacitor's rated current MUCH more than the LED's? Bob

 

[Bob Engelhardt]

Nup. It's a 'shunt regulator'.
http://en.wikipedia.org/wiki/Linear_regulator#Simple_zener_regulator

The wiki ckt has a current-limiting resistor in series with the voltage
source. Not so the posted ckt.

Bob

 

[Winston]

[email protected] wrote:

(...)

Yes, but I got one of the HF's and hacked it up with a Dremel tool, to
mount to the lathe.

Can't do that with a Mitutoyo!

And hold one's head up in public, that is.

--Winston

 

[Winston]

The wiki ckt has a current-limiting resistor in series with the voltage
source. Not so the posted ckt.

It *is* in the posted circuit.
Sort of.

As James mentions, the internal resistance of the PV
cell is the current-limiting resistor.

--Winston <--The BSA B50T POS used a single power
zener across the battery as it's only
voltage regulator.

 

[Bob Engelhardt]

It *is* in the posted circuit.
Sort of.

As James mentions, the internal resistance of the PV
cell is the current-limiting resistor.

I meant the capacitor voltage source. Is its internal resistance a
sufficient current limiter? I'm not familiar with super caps, but the
common ones that I am familiar with will supply huge currents, momentarily.

Bob

 

[Jamie]

Yes, but I got one of the HF's and hacked it up with a Dremel tool, to
mount to the lathe.

Can't do that with a Mitutoyo!

Hey, that's one hell of an idea!

Jamie

 

[Winston]

I meant the capacitor voltage source. Is its internal resistance a
sufficient current limiter? I'm not familiar with super caps, but the
common ones that I am familiar with will supply huge currents, momentarily.

This circuit, yes?

D1
.--|>|--+-----+----> (+) to caliper
+ | | |
.------. + | |
| | --- ---
| ---- | --- \ / ~~>
| | C1| ---
| ---- | | | LED (red)
| | | |
'------' | |
- | PV | |
'--------+-----+---> (-) to caliper


Under no circumstances would the voltage across
the LED (and capacitor) go above, say 1.8 V
because the LED turns any additional
voltage into current. In order for C1 to produce
a current large enough to endanger the LED or
the caliper, it would have to be allowed to charge
significantly above the 'zener point' of the LED.
It just cannot.


--Winston

 

[John S]

Isn't the capacitor's rated current MUCH more than the LED's? Bob

What has that to do with it? The voltage is clamped. What are you missing?

 

[Bob Engelhardt]

This circuit, yes?

D1
.--|>|--+-----+----> (+) to caliper
+ | | |
.------. + | |
| | --- ---
| ---- | --- \ / ~~>
| | C1| ---
| ---- | | | LED (red)
| | | |
'------' | |
- | PV | |
'--------+-----+---> (-) to caliper


Under no circumstances would the voltage across
the LED (and capacitor) go above, say 1.8 V
because the LED turns any additional
voltage into current. ...

Oh ... right ... yeah. Dope slap for me. Bob

 

[Winston]

(...)


Oh ... right ... yeah.

We are all here to learn.

--Winston

 

[DoN. Nichols]

I meant the capacitor voltage source. Is its internal resistance a
sufficient current limiter? I'm not familiar with super caps, but the
common ones that I am familiar with will supply huge currents, momentarily.

The capacitor gets its voltage from the PV cell. Assuming that
you don't put a switch between the LED and the cap (there is none shown
in the schematic), the cap will never charge high enough to be able to
damage the LED, because the LED will have already clamped the maximum
voltage based on the current limit of the PV cell. Not sure what would
happen with the PV cell close to an arc welding process like a TIG -- it
depends on the internal resistance of the PV cell and the peak voltage
which the PV cell can produce with such excessive illumination.

Enjoy,
DoN.

 

[[email protected]]

Yes, good site. I linked to it earlier in this thread.


There aren't any ICs with low enough Iq, at least not that I know of.
I used discrete transistors.

There are some pretty good ones, designed for USB applications, but I don't
thing they're quite good enough for this. The TPS6205x Iq is around 5uA to and
in shutdown less than 2uA. You're looking for something an order of magnitude
better than this?

<...>

 

[P E Schoen]

"Jeff Liebermann" wrote in message

That's because of the 12uA typical quiescent current, where the
chip draws about the same current as the caliper load. For equal
currents, that's 50% maximum efficiency. The TPS62054 shows
50% efficiency at 2.7V in and 1.8V out (See Pg 8 Fig 4).

The chips do have a shutdown pin that cuts the quiescent
current to "less than 2uA". Still high, but much better.

A step-down regulator/converter could be made from a Microchip PIC18LF14K22
(http://ww1.microchip.com/downloads/en/DeviceDoc/41365c.pdf) which has a
quiescent current of 34nA and an operating current of about 10 uA at 1.8
VDC.

And it may be even more efficient to use a low power linear regulator such
as the TPS71501 (http://www.ti.com/product/tps71501) which has 3.3 uA
quiescent current. If the input voltage is, say, 2 VDC and the output is 1.6
VDC at 12 uA, the overall efficiency is (1.6*12)/(2*15.3) or almost 63%.
Even at 3 VDC input it is 42%.

Paul

 

[[email protected]]

Yes, but, in the caliper context, who's going to turn the switcher
off, how, and when?

When the photocell voltage output too low to power the calipers?

 

[Winston]

Mine doesn't. The initial position is apparently stored in the
circuitry. I can also set it to any position, turn it off, and it
will read the same value when turned back on. One of the benefits of
having it draw power continuously.


Retrofits are always more complicated. I tend to think in terms of
new product designs. If this were for a retrofit or modification, the
design philosphy would certainly be different.


Agreed. At 2uA, it probably wouldn't need an on/off switch.


Pancake PM generator which offers the highest velocity at the edge.
Basically a magnetized disk and a stator coil. Spring loaded spool
parallel to the pancake motor and the same size. Ratchet to control
direction. Inertial clutch and some plastic gears between the spool
and the pancake motor. Much like a generator powered flashlight. Pull
on the string and the spool spins, which causes the generator to run,
charging a super-cap. If a pull string is too crude, a spiral
mainspring mechanism can be added. If the pull string and generator
are too sophisticated, a moving magnet inside a coil, that you shake
back and forth to charge the super-cap, much like in some battery-less
flashlights. Maybe a thumb wheel for spinning the pancake generator
might better. Plenty of options.

How about a weighted leaf switch in series with the
cell? It could take the form of a thicker
'battery door'. The bulk bypass caps in the caliper
would hopefully maintain power during vibration
glitches and overhead measurements.

Place the caliper 'display down' and power
is removed.

--Winston<--Cheap! Quick!

 

[Winston]

Jeff Liebermann wrote:

(...)

Retch. A position sensitive on/off mechanism is prone to failure if
the device is parked in an unusual position. For example, if I put it
in it's protective case, and then toss it into my toolbox face up, it
will remain on.

Even if you stored the caliper 'face up' for a year,
it would cost you no more than if you had no
switch at all. Net net, battery life would increase
because, the caliper would be turned off for a far
longer time than it would be turned on.

It would also prevent me from using in the face down
position.

I was appalled that I can't use my blender
upside down too, until I realized that I never
have a need to do that.

Please don't suggest an inertial on/off mechanism (shake to operate).
Bouncing around in my car will probably turn it on.

If you drive your car 24/365, you have much bigger
efficiency worries than a $1.00 cell!

--Winston

 

[Winston]

This is what happens when we lose sight of the original problem that
we're trying to solve. The problem is that the calipers have a short
battery life mostly because they don't really turn off and secondarily
because the dropout voltage is too high to use cheap alkaline
batteries. Turning the calipers off (actually in standby) for a year
won't do much, when the operating current is 14.5uA and the off
current is 13.5uA.

Er. Jeff?

I was talking about the open circuit represented
by a 'position' switch. Off current would be
*far* lower than 13.5 uA.

You need to contact a product liability attorney, and sue for damages
resulting from inverted operation. Be sure to mention that there was
no warning in the user manual about inverted operation.

I've used calipers and other measuring instruments in probably every
orientation possible. If the calipers turned off while I was making
measurements in a contortionists position, I would be rather
irritated.

What proportion of the time do you measure
with the display facing down?
For me it is a very tiny amount
of time. Like seconds-per-decade.

I own seven pairs of calipers, two of which are dial,
two of which are solar, two of which are very
cheap digitals and one is a vernier.
Of them, only the cheap digitals would benefit
from this conversion. I would select
the proper measuring tool for the job.

You do as well, I trust.

True. Also, if I don't drive the car at all, I get an infinitely
large gasoline mileage. Both extremes are worthless.

I dunno. The 24/365 commute places an upper limit
on the amount of time that the calipers would remain
powered up. It probably would cost you say $1.20 a year
to leave them 'rightside up' in your toolbox.
I don't think of this as a huge toll, somehow.

Perspective!

I non-cleverly
installed a "shake to unlock" (iShakeLock) application on my iPhone
3G. The standard method of unlocking the phone requires both hands. I
figured that one handed unlocking would be better. The problem was
that it kept unlocking in my pocket. Battery life deteriorated
rapidly. So, I dumped the app.

You keep your iPhone next to your calipers in your pocket?

Don't the pointy 'inside' anvils smart a bit?



--Winston<--Tell me about your pocket lint.

 

[Winston]

[email protected] wrote:

(...)

Decent! How about this as a retrofit to the Harbor Freight cheapies:

push a pushbutton switch, which charges a supercap when you need it,
giving an hour or so runtime per press?

Iq = 0uA.
Battery life: Same as shelf life.
Parts list: homemade switch, supercap. $0.20.

Caliper starts blinking? Press the button again to keep it going.

I like that! But I can reduce the BOM size.
Just a (rotary?) switch in the battery cover.

If you leave it on, you get 10 months
out of the battery. If you use the power
switch religiously, you get 'shelf life'.

* A switch handling ~40 uA does not have to
be nearly as stout as one handling 'short circuit'
current into the super cap.

* Most super caps don't tolerate high current well.

* This mod is dead simple, reversible and does not
require access to a ground connection.

--Winston

 

[Winston]

[email protected] wrote:

(...)

Low-current switches can be a bear--the contacts oxidize. Gold fixes
that generally, IIRC.

D.C. switches have their own complications, though
I suspect ~40 uA wouldn't be a problem, given the
proper contact plating and wiping action.
I think gold is gonna be problematical if there will
be much in the way of capacitor charging 'inrush'
needed.

TATOO: Look boss, deplate! Deplate!

No problem here. The low battery voltage and high internal resistance
of these tiny cell-phone super caps limits the current nicely.

The super cap is in parallel with the low-z bulk
bypass caps in the caliper, though. Your
pushbutton will have to be properly sized and
snubbed. The folks at Palm found out about that
much too late.

I like some sort of dead-man timer better. It could be just a low
threshold mosfet with a gate cap which you charge, and a *large*
discharge resistor(*) so it times out. That guarantees the battery-
saving feature even if you forget.

I hear LTSpice calling.
We need a pass element that has a gate voltage
saturation point in the 200-300 mV region.
I just don't see a MOSFET in that role, somehow.

(*) (From the jellybean / junkbox standpoint, the discharge "resistor"
might be a reverse-biased rectifier's leakage.)

But yes, just adding a real switch is a huge improvement over not
having it.

A real switch also lets you hold the zero setting over night if you
want to. I sometimes do that with the lathe, if, for example, I'm in
the middle of something when it's time to turn in.

You *do* need access to the underside of the cell, since that's where
the (-) contact is.

You can switch the positive side as easily if your
interposer board is thin enough.

Insert a very thin piece of double-sided printed
circuit mat'l, wire a tiny slide switch to both sides of that, and
Bob's yer uncle.

Yup. that is how we do it.
I used very thin double sided stock, though
most of the time the 0.062" stuff worked fine.
It is great for measuring current too.

That's reversible, and if you're a brute, you can even hang the switch
outboard by the wires. That way there's no modification of the
caliper needed at all.

That's good enough and simple enough that I'll put 'er on the list.
I've got maybe a dozen of these (two on the lathe alone), and it'd be
nice not pulling the batteries (as I do now).

For your 'fixed' installations, you could solder
some small 'earphone' wire to the battery contacts
and use a huge, cheap external cell. (Huge = AA)

--Winston

 

[P E Schoen]

"Winston" wrote in message
[big snip...]

For your 'fixed' installations, you could solder
some small 'earphone' wire to the battery contacts
and use a huge, cheap external cell. (Huge = AA)

I didn't expect my post to generate such a big discussion, but it's been
quite interesting. Since I only use my calipers once in a while (it had been
years until I once again tried to use them and found out how to fix them),
it seems that just removing the cell is best, and that guards against
possible damage if it leaks. My machine shop is in an unheated building and
the temperature and humidity extremes may make leakage more likely.

Otherwise I think we have established that the normal lifetime of a 175
mA-hr cell, even with the typical 10-15 uA of on or off current, still
provides well over 1 year of life. Replacing it annually for $0.75 compared
to once every 3-5 years of shelf life does not warrant any expenditure of
time or materials for a high-tech (or even low-tech) solution. I've already
spent an hour or more in this discussion. At my normal consulting rate, that
would buy a hundred batteries!

But if removing the battery is too annoying, you might be able to cut a thin
slot in the battery compartment, and just insert and remove a thin, stiff
piece of mylar or similar insulating material to isolate the negative
terminal of the battery from the contact.

Paul
www.pstech-inc.com