DC12V 12A car window motor - what spec for a 220V PSU?

[marcophys]

Absaar 108 NE/D2 12V Battery Charger
With added capacitor for motor load

Image Notes
The image seems to have been re-scaled.
Even opening in a new tab fails to show the image at full size.

The only unreadable element is the capacitor spec:
SAMXON 680 µF 200v LP (M) 85 deg

Charger Notes
Cable cores are solid 1.5mm D from the secondary winding, and the ammeter circuit.
Multicore:- central switch to rectifier, rectifier -ve output, fuse +ve output
(The capacitor leads used are solid core)

Two Core To Switch
The middle secondary output in grey, is of two solid twisted cores of 1.5mm D.
It connects to the Lo charge switch contact.

Is this simply an interruption in the windings
... the core then returning to the coil to make more turns?

Fuse
Absaar never replied to my email questioning the fuse... so we know nothing about it.
I've marked it as a thermal fuse... it was a reasonable guess.

Switch
There was some resistance across the switch contacts.
This was cleaned up with contact cleaner, and now, both contacts are just 1Ω.

 

[jim hardy]

It's a tapped secondary winding. Low voltage switch setting uses part of the turns high voltage setting uses them all.
So red is one end of secondary, looks like blue is the other end and grey(looks purple on my screen)

is someplace in between .

Your big cap looks to be in exact place i would have put it.

I've marked it as a thermal fuse... it was a reasonable guess.

If the cap unscrews and there's a fuse inside then it's a fuse.
More llkely it's a thermal circuit breaker that's rather slow to open . Its job is to protect the wires and rectifier if it can.
Battery chargers use very stout rectifiers that can withstand a brief short circuit and even survive when somebody hooks up the battery backwards.
I doubt your RBV602 is as robust
from the datasheet you linked

i read that as 40 amps will likely kill it in a second.Nice work.

I'm hunting for my old battery charger and DMM now.

old jim

 

[jim hardy]

My raggedy old charger and el-cheapo DMM gave the following:

Charger alone:
13.02 VDC ,
switching to AC reads 4.45
so unfiltered it has roughly 34% AC content.

With 100 uf across output:
17.52 VDC,
switching to AC reads 0.05
so the 100 uf is an effective filter.
It both raises the DC level to sinewave peak and removes the AC content.

That's why i was surprised by your earlier AC readings of 29 and 44 volts, they were so different from what i expected..

I presume that this confirms that something is amiss with my meter.

Perchance is yours an analog meter ?

If it's a DMM , i noticed something about mine that i'd just always accepted as a "DMM Peculiarity".
When i switched from DC to AC my meter jumped up to about double the DC reading then drifted back down to stable reading over a matter of seconds...
That's because selecting the AC range switches in an internal capacitor to block the DC , and that capacitor takes a while to charge.
During that charge interval the meter is experiencing current at its input namely the internal capacitor's charging current, which it interprets as AC .
One gets so accustomed to that 'oh, it's just another digital meter quirk' he accepts it and forgets to forewarn newbies.
So ---
Solving your meter's AC anomaly might be as simple as waiting for your DMM to settle. Takes a few seconds.Your enthusiasm is both heartwarming and infectious.
Keep on having fun,

old jim

 

[marcophys]

Perchance is yours an analog meter ?

No... it's digital.

Solving your meter anomaly might be as simple as waiting for your DMM to settle. Takes a few seconds.

Sadly, also no!.
I've just measured the charger output on Lo (with capacitor as per drawing) = 39V AC - 17.8V DC
It remains stable within 0.2V for a minute.

On Hi the output is 44.5V AC - 20.3V DC

Note... It's Sunday evening:
Mains supply is fluctuating between 230V - 234V AC (over 2 seconds).
IE. It cycles up and down.

Just a thought... it would be interesting to monitor the mains supply over the course of a day.
However, regardless; we can see why for testing, one needs a regulated mains supply.​

Therefore my DMM is consistently more than doubling the DC voltage, when set to AC.

My DMM

Notice the bar that I resoldered... it had parted.
It leads to the 10A connector.

I have just now resoldered the connections to the board.
I don't think that this should affect the normal volts readings, but maybe the heat (during failure) has caused problems.

The thing is... it seems to work for everything other than AC voltage on a DC circuit.
What to do?
Perhaps it has a capacitor that is failing.

After reassembly, it is no different.
I just have to live with it for the time being.

 

[jim hardy]

That's fine.
For a sanity check try a car battery or 9V battery , something you know has no AC even connected to it.

That'll say conclusively whether Mr DMM is ailing.

 

[marcophys]

Ha!... good idea.
Just checked a 1.3V DC battery... it shows a lovely 2.4V AC.

I guess that confirms it

 

[jim hardy]

I guess that confirms it

Dont you love it when those little confusion factors drop one by one ?

That's troubleshooting... what's left is the truth.

 

[marcophys]

Don't you love it when those little confusion factors drop one by one ?
That's troubleshooting... what's left is the truth.

Yes... if only the search for truth was not a dangerous area of interest.​

However, confusion does have additional benefits, as it stimulates discovery in areas that otherwise might have been missed.

I gained knowledge of my DMM, which is very useful for future metering.
... but the error threw up the question over how the DC voltage was being created, leading to investigation of bridge rectifiers, and then the new wiring diagram.

I do believe that by 'having everything' (particularly early on in adulthood), this condition removes the absolute requirement for innovation, and tends to preclude the development of an innovative mindset.
... and the self belief that no engineering problem is unsolvable.

Of course, once you are hard wired to solve problems... it would be nice to have all the machinery and test equipment at your disposal
... but even with a dodgy DMM, a stopwatch, eyes, and ears, we have created a very useful tool.

It would be nice to have full knowledge of it - perhaps only achievable with an oscilloscope.
Tom raised the issue of the RVB 602 being surplus to requirement.
Clearly the system ran fine without it.
However, with it, the motor ran marginally better at slowest speeds.

I don't know why... I simply observed this.
Similarly the 2nd capacitor.

But at that point, we have reached the limits of my testing capabilities.
I'm fine about that.

It's hard to arrive at a decision to 'stop working' on a project... but it must be done.
I can use the mind trick, and say to myself... we'll have another look later.

So that really is it.
Time to get on with other pressing matters that have been put on hold.
We can reconvene when the next issue arises.

Thanks to everybody for being there.

 

[jim hardy]

Your big capacitor has one other benefit for you. It absorbs transients from the transformer thereby protecting your RVB 50 volt bridge rectifier.

Below is unnecessary, not showing off just want to be thorough. Might as well learn all we can from an experimental oroject.

Recall i said above that battery chargers use robust rectifiers, capable of withstanding short circuits and reverse connected batteries.
They're also capable of withstanding surprising overvoltage.

When you unplug a battery charger from the wall it suddenly interrupts current through the transformer primary.
How much is that current at instant of disconnect depends on where in the sine wave the plug and receptacle break contact .
Size of the "inductive kick" or "spike" you'll get depends on how much was the current at that instant, ie how close to sinewave peak was the current when contact broke.
If the battery is still connected no sweat at all, the battery absorbs the "spike" and you have graceful shutdown.
If the battery is NOT still connected then the rectifier will be subjected to whatever "spike" the transformer produces. Something must absorb the energy.
That's why i say "Battery charger rectifiers are more robust than your garden variety bridges."
I think they are an avalanche type that survives reasonable overvoltage .

To demonstrate that effect,
yesterday while tinkering with my old raggedy charger and DMM
i connected a 10 uf motor run capacitor across output.
AC and DC voltages were about the same with 10uf as with 100 uf , about 13 and 4.5

Then to show the effect of "Inductive Kick" i unplugged the charger power cord.
The meter jumped to 110VDC and bled down over a few seconds.
Tried it again, got only 20 VDC

About 10 'unplugs' showed random DC voltages, many in the 70 to 100 volt range and some showing no spike at all..
That's not surprising because a sine wave statistically spends most of its time near peak . Sin 45 degrees is 0.71 so it spends half its time above 70%. .

I repeated with the 100 uf and never saw above 20 volts if that many, i forget...

What does that mean ? It means your charger could make a spike that exceeds your RBV bridge's 50 volt rating.
But your 680 uf capacitor is surely big enough to absorb the energy and protect the little RBV.

Energy=1/2 CV²
My 6 amp charger put energy = 1/2 X (10 X 10^-6farads X 110volts² ) = 0.0605 Joules into that ten microfarad capacitor.
That many Joules into 100 uf would be 35 volts if they all went in, some get lost in the wires and transformer...
and into 680 uf it would be only 13 volts.

So your capacitor will prolong the life of your RBV bridge.

I formed the habit years ago of always unplugging the charger's power cord BEFORE disconnecting the battery just to make life easier on my charger's internal rectifier.

There's your trivia for today...

Applying "The Basics" to everyday life is really fun.

old jim

 

[marcophys]

That's fantastic information Jim.
I really appreciate the sharing of that knowledge.
I will never again disconnect the battery before switching off the battery charger.

For the motor control, we now have two capacitors - one in the charger, the other across the RBV DC output.
As it is currently wired (on the motor switch control panel), the RBV and 2nd capacitor is prior to the motor on/off switch.

Theoretically... am I correct in assuming that, ideally the motor switch should be prior to the RBV and capacitor.
... the switch would then cut the current to the RBV, capacitor, and motor, thus allowing the motor to bleed away the residual current?

Note: I have heard you, vis a vis the protection provided by the capacitors... it's just a theoretical question (that can anyway be enacted).

 

[jim hardy]

Theoretically... am I correct in assuming that, ideally the motor switch should be prior to the RBV and capacitor.
... the switch would then cut the current to the RBV, capacitor, and motor, thus allowing the motor to bleed away the residual current?

That connection will be the most gentle one for power off.
But at power on there's substantial current into the second capacitor. Well, i say substantial but with just 13 volts and 680 uf it's hardly high power.

In interest of being thorough
connecting them in order capacitor switch bridge motor gives gentle startup AND gentle shutdown. No sudden charging of capacitor on switch closure, and on switch opening the bridge absorbs inductive spike from motor. But you should include a bleed-down resistor or lamp (12 volt LED will do) across the last capacitor.

I can tell you're sort of a perfectionist. Take that as a compliment. Your wiring is neat and clean looking.
I suffer a similar malady - frustrated perfectionism. We're the type who get a project built and working then go buy all new parts and start all over because it didn't come out looking perfect.

Keep having fun. Enjoy applying basic physics to everyday situations.

old jim

 

[Asymptotic]

But you should include a bleed-down resistor or lamp (12 volt LED will do) across the last capacitor.

Good point.

Going back several posts, the question was how to size the output capacitor.

The violet line added to Jim's waveform is DC voltage with an output capacitor added. If only a little bit of capacitance is added then that line dives deeper towards 0V, and adding more flattens it out towards more nearly connecting each voltage peak. Ripple factor is 1/(4*sqrt(3)*f*c*r). Line frequency f is 50 Hz. Lowering r (resistance), in other words increasing the load, increases ripple, and increasing c (capacitance) decreases ripple. However, no matter how much capacitance is added there will always be ripple, and after a certain point adding more becomes a matter of diminishing returns. Although it doesn't much matter in this case, a whopping huge capacitance makes motor response to (dimmer) speed control setpoint changes sloppy, and (if the motor remains connected to the capacitor on power down) instead of coasting quickly to a stop it'll decelerate more slowly as the capacitor discharges.

680 uF is, in my opinion, a good compromise value in a circuit such as this. The most I've used is 56,400 uF (twelve 4700uF caps in parallel), but under different circumstances; in a 700V DC link supply for a 11 kW servo application that required very fast accel.

Your project does all the things you need it to do, it looks good, and I believe you've wrung out all the performance you are liable to get. Better pick up a different DVM, though.

 

[marcophys]

connecting them in order capacitor switch bridge motor gives gentle startup AND gentle shutdown. No sudden charging of capacitor on switch closure, and on switch opening the bridge absorbs inductive spike from motor. But you should include a bleed-down resistor or lamp (12 volt LED will do) across the last capacitor.

That's the info I was looking for... how it should be wired.
I've drafted the wiring diagram as it currently stands, and added a bridge to the polarity switches.

the question was how to size the output capacitor

This was also the info I was looking for.
It explains everything to me.
I now understand why one side of the bridge rectifier was showing 6V... it was missing the lower wave.
Also the explanation of what the capacitor is doing, and what in fact the ripple is... it's all clear now.

What a relief!

Apart from Jim's recommended mods... here is the completed panel, and current wiring diagram:

 

[marcophys]

Here it is with an LED added and the diagram updated.
The LED glows while the motor is drawing current, and is just slightly illuminated when live.

... but what's also interesting (and put a smile on my face)
The LED varies the glow according to how hard the motor is working.
If the motor is close to stalling, it glows bright... and brightest if the motor actually stalls.

 

[jim hardy]

@Asymptotic

Nice Job !

@Marcop
here's an approximate way to estimate ripple on a capacitor. Proper calculation requires calculus but this back of the envelope will get you close enough for home projects.

During that bleed down period between peaks, current to your load comes from the capacitor. In power supply filter applications think of the capacitor as a short term battery or reservoir for charge..

Current I into or out of a capacitor is I = C X (Δvolts / Δtime)
and ripple is Δvolts between rectified sinewave peaks.

Where you are is 50 hz so downstream of your rectifier there's a peak every 1 / 100th of a second, 10 milliseconds.

Rearrange I = C X (Δvolts / Δtime) to get Δvolts and you get Δvolts = Δtime X I / C

Now at 1 amp and 680 uf and ten milliseconds i get Δvolts = 0.01sec X 1amp / 680X10^-6Farads = 14.7 volts

so your 680 won't run the motor very long
but is sure helps your dimmer meet holding current..

Really ripple won't be the full 14.7 volts because the actual Δt will be somewhat less than a complete half cycle.

Here's an addition to Asymptotic's great picture:

Asymptotic showed a large capacitor, i showed a small one.
For audio you need a huge one to reduce hum.
You just need enough to keep the triac "Latched", ie meet its holding current requirement.

Knowing slope, Δvolts / Δtime, you could calculate the point where the lines depart. And you could estimate the point where they'll rejoin.
You can use that to build a very fast "Loss of AC detector" . I once had to detect loss of 60 hz AC within twelve milliseconds (for a computer) and that's the approach we used.

sorry for boring everybody

have fun

old jim

 

[marcophys]

sorry for boring everybody

Not boring at all.
Moreover, very pertinent to the discussion.

RE the LED across the capacitor.
With the supply cut to the charger... the LED remains illuminated for a long time - 15 minutes at least.

I'm guessing that this is due to the presence of two 680 uf capacitors.
To drain the capacitors, I switch on the motor.
The LED then goes momentarily bright.

Is the LED doing its intended job, or is its power consumption too low?

 

[Asymptotic]

You can use that to build a very fast "Loss of AC detector" . I once had to detect loss of 60 hz AC within a twelve milliseconds (for a computer) and that's the approach we used.

To serve a similar need - to start an emergency lighting generator when power flickered during electrical storms - I went stone age, and used a 120 VAC ice cube relay (with a 3.3K, 5W resistor across the coil, to speed things up) for line voltage loss sensing, a 12 VDC relay connected to the generator's DC supply to latch in when the other triggered (starting the genset), and a timer relay to reset it once utility voltage had been restored for 5 minutes. :).

Don't know if it's still the case in newer models, but we were using metal halide lighting at the time, and Onan's controller couldn't be set to trigger fast enough to start the generator before their arcs extinguished, and, because conventional MH bulbs require about 5 minutes before they will restrike, plunged the plant into darkness.

 

[Tom.G]

(with a 3.3K, 5W resistor across the coil, to speed things up)

Uhmm... wouldn't that slow things down? R L time constant is t = L/R.

 

[marcophys]

A warning about the CGS HA50

It gets hot... enough to cause a blister.
So it definitely needs a safety guard.

Also, the suggested lamp mod!
It will be a good indicator that the system is powered.

You run some tests, write it up, do something else.
The motor isn't turning... so it is very easy to forget that the system is live.
However, you are quickly reminded in the morning, when you grab the resistor

I found that a PC PSU cage was pretty close in width.
It was very fiddly, with a lot of cutting and folding
... but here is the motor speed control panel, with safety guard:

 

[jim hardy]

It gets hot... enough to cause a blister.

Needs a bigger heatsink.

i square inch per watt should work , seven inch square 1/16 inch thick aluminum ?
Got an old cookie sheet you can cut up? Two layers of thin aluminum with silicone grease between for heat transfer.. Or heatsink from a junk stereo ?? Silicone grease under the resistor too. ,

A simple heatsink like this should fit under your PC cover. That's a great idea by the way.
I cannot over-emphasize the importance of a good heatsink and thermal grease. The gooey stuff from Ace Hardware is okay.

7" long, maybe 3" tall ?

 

[Asymptotic]

Uhmm... wouldn't that slow things down? R L time constant is t = L/R.

Good catch - you are right. A resistor in parallel with the relay coil is faster than using an RC network for noise suppression, but slows down response versus using no resistor at all.

 

[jim hardy]

With the supply cut to the charger... the LED remains illuminated for a long time - 15 minutes at least.

I'm guessing that this is due to the presence of two 680 uf capacitors.
To drain the capacitors, I switch on the motor.
The LED then goes momentarily bright.

That's very puzzling. Will think on it for a while.

LED's will

 

[marcophys]

With the supply cut to the charger... the LED remains illuminated for a long time - 15 minutes at least.

Actually... it is hours.
I hadn't timed it... but it was switched off for hours.
I looked down and thought hmmm that LED is still glowing.
I flipped the switch to ON... the LED brightened and died.

Development Pause - thoughts and observations
With the product apparently developed - a pause followed by examination anew, casts new light.

The first mistake was seen.
The resistor feed from the MK socket was 'pre-switch'.
Obviously it was, because I had used the standard socket connectors.

Clearly this is fine for the electrical system, however it is counter intuitive for the user.
You could say that the primary dimmer panel switch, is the dimmer itself.
However, normally the ON/OFF feedback is the light that it is controlling.

As a consequence, I rewired the resistor 'post socket switch'.
Thankfully this was relatively simple due to the provision of a lamp in the socket.

I could, and probably should, add a circular gauge to the dimmer switch.
This could indicate OFF (it clicks off when fully counter clockwise)
... also it could be marked to indicate the motor operating range.

However the key element was to ensure that the (dimmed) mains voltage stops at the socket, when the socket is switched off.
That is sensible design.

The second mistake
... is based upon the first mistake.

The foot control motor ON/OFF switch, should control that panel.
It's a separate panel, so the switch should isolate the panel.
It makes sense.
(At the moment the switch is after the rectifier and capacitor)

Locating the lamp (in the charger)
We remember Jim theorized that the lamp should be placed before the rectifier
... I tried a 24V 0.05A lamp and the transformer buzzing didn't change.
We discussed a more powerful small 240V lamp - and didn't have one.

Today I found a beautiful small lamp - 24V 3W (calculated as 0.125A)

It caused me to look again at 'lamp location'.
With the 0.05A already soldered with tails
... the pre-rectifier position was retested - no transformer buzz reduction.
... the post rectifier position was tested - transformer buzz reduction.
The tests were repeated.

I don't know why this is the case, but it is the case.
Therefore the 0.05A lamp can be placed across the output, parallel to the capacitor.
This will smooth the transformer, provide an 'equipment live' signal, and drain the capacitor when switched off.

It looks like a win.

 

[marcophys]

The Lamp Modification
This went well.
There is another gain in motor operation.

I ran the motor down to 7 rpm - I think that is the lowest yet seen.
The difference is marginal, but worthy

More Motor Knowledge
Here's the pre-amble explaining how the knowledge was acquired:

Lamp
The lamp shines brightest with no load
The moment the motor starts, the lamp dims.
As the voltage is lowered, the motor slows, and the lamp dims further.

LED
... is barely visible when the motor is stopped.
It momentarily flashes on when the motor starts.
There is no appreciable increase in brightness until the motor drops to 14 rpm.
It gets brighter as the motor slows.
If the motor stalls it is at its brightest.

Initial Conclusion
1. The LED brightness is driven by current in the circuit.
2. The motor can slow considerably without any significant rise in current draw.
3. There appears to be a point where current begins to appreciably rise.

Yet...
I then checked the ammeter on the battery charger.
... and the current drops as voltage is reduced.
It shows 2.9 - 3 amps at fastest and 2.3 - 2.5 amps at slowest.

The LED is of course nearer the motor
Anyway, that apart, I am presuming therefore that it is best not to run the motor at its absolute slowest speed.

But it makes you think about 'torque'.
That ability to deliver power at low revs.

Clearly, lower gearing would allow the motor to spin faster, and deliver the power at a slower speed.
We can't change the gearbox... can we increase the torque of the motor?

Here's the updated battery charger wiring:

 

[marcophys]

Needs a bigger heatsink.

Here's one that will fit... found in the PC PSU.
The alloy is 5mm thick, 70 x 70.

Re the heat paste.
I found it amazing that 30 year old paste around a stereo chip was still a paste under the skin.
However, I'm glad to know that silicone grease works, as I have plenty of that.

 

[jim hardy]

found in the PC PSU.
The alloy is 5mm thick, 70 x 70.

Great !

 

[jim hardy]

Behavior of your lamps and motor can be somewhat counter-intuitive.

Realize that the dimmer knows nothing about the load, it is aware only of voltage across itself.

If load changes perhaps because motor sped up, that changes voltage across the dimmer and it will react.

You're a mechanical, right ?
Imagine a control valve set to hold constant differential pressure across itself (the dimmer)
in cascade with another control valve whose % opening is proportional to ratio of pressure across it and flow through it.
They'd interact with one another, right ?

So lacking an osciloscope your empirical approach is best analytical tool you have.

I'd say you're doing great.

Keep your eyes peeled for a windshield wiper motor, though. I think you'll find it's well behaved and doesn't mind continuous duty.

Youtube shows how to get them out.

 

[marcophys]

Keep your eyes peeled for a windshield wiper motor,

It is very possible that a windscreen wiper motor would be better.

Of course, I am very happy with our project as it is... the difficulty is deciding when to stop learning.
For the brain, it's no longer a question of the specific 'here & now'.
... it's more like "oh... so below a certain rotation, it doesn't work as well... why's that, and is it written in the windings of the motor"?

It's not like we are now under any pressure - the turntable works, and at 12 - 14 rpm it is fine.
The working project is finished.
Anything learned now is just for sharing, and maybe incorporated into a future version.

For that... the windscreen wiper motor would be the better option, as they are available everywhere.
They're also designed to operate at different speeds.
But that's for the future.
I have an engine to rebuild now :)

 

[jim hardy]

the difficulty is deciding when to stop learning.

Isn't this a great time to be alive? So much is known and the communication revolution( Gutenberg to Morse to Bell to Marconi to Internet ) has placed it all at our fingertips.

I have an engine to rebuild now :)

So many projects, so little time.

 

[marcophys]

Isn't this a great time to be alive? So much is known and the communication revolution( Gutenberg to Morse to Bell to Marconi to Internet ) has placed it all at our fingertips.

I do agree.
The knowledge at our fingertips is... enabling.
I struggle for words to describe this dramatic change that we have witnessed

I make sure that I contribute... the engine rebuild is also being shared.

 

[jim hardy]

the engine rebuild is also being shared.

Good !

i just bought four Detroit Diesel 6V53's for 30 cents a pound
guess i'll be learning something about them.

See you in the mechanical engineering threads ! old jim

 

[marcophys]

Ha!
Not this time

I do provide general engineering knowledge - particularly specific tips for universal application.
Here's one I put up two days ago that I did live.
Putting it all on the line helps students believe in themselves


So as you can see, the project is a small capacity two stroke.

These engines have their own unique problems that are generally left unsolved.
I'll be showing people how to solve those problems.
There is always crossover, so you might be interested.
I've completed 3 parts so far.

I never envisaged the project for the physics forum.
it's a lot of work, simply documenting everything on youtube... but my vids typically get thousands to tens of thousands of views, so I believe that it is worth the effort.

I will be dealing with some advanced physical aspects (introduced in part 3), relating to our turntable project, but whether it's physics forum stuff, I don't know.
You can always subscribe, and let me know what you think, if you see something that is pertinent.

Here's an electrical one that is probably right up your street, but it's 50 minutes in length.
My excuse is that it was necessary