Film-Tech Forum ARCHIVE: AC ripple

This is topic AC ripple in forum Film Handlers' Forum at Film-Tech Forum ARCHIVE.

To visit this topic, use this URL:
https://ft-forum.com/ft/cgi-bin/ubb/ultimatebb.cgi?ubb=get_topic;f=1;t=001146

Posted by Jon Bartow (Member # 241) on 07-03-2000, 06:20 AM:

Someone showed me how to measure the AC ripple of a Xenon power supply a long time ago and I don't remember quite everthing. Knowing that it can be potentially lethal to test equipment, I want to check to make sure that I am doing it right first. Would someone be kind enough to remind me how?

Posted by Pat Moore (Member # 380) on 07-03-2000, 07:34 AM:

The best way is to use an oscilloscope to get a "picture" of the DC voltage at the power supply output, right across the DC going to the lamp. Ideally you should see a straight line, pure DC voltage, but you probably won't. Whatever "wave" is there should have smooth transition and no peaks. Look at the amplitude of the wave and figure the AC voltage component; that will depend on the volts/division you are set at.
You can also get an estimate using a voltmeter and/or an AC Amprobe. Take a DC voltage reading across the arc, then switch the meter to the AC setting and read again. Same thing with an amprobe.
Take that number (AC reading) and divide that by the DC reading and you get percentage of AC Ripple.
The lower the better. If you're reading 3-6% you're okay, but any higher than that you might anticipate some problems.

Posted by John Walsh (Member # 168) on 07-03-2000, 09:02 AM:

This is the procedure we used to follow to genreally check lamps and power supplies....
------------------------------------------
Before beginning, quickly verify that the power supply is operating at or below the AMPERAGE shown on the power supply nameplate.

1. With a small flat head screwdriver, "zero" the volt meter and ammeter (if analog type).

2.Place the auto-ignition switch to "MANUAL".

3. Open the lamphouse door and defeat the door interlock safety switch.

4.Observing proper polarity, connect a calibrated volt meter to the DC power cables.

5.Power on the lamphouse. The bulb should not ignite while the auto-ignition switch is in "MANUAL".

6. Compare the voltage reading from each volt meter. Record the values. Replace the meter in the lamphouse if the values are more then 15% apart from the calibrated volt meter.

Remove the volt meter leads.

7. Press the "STRIKE" button to ignite the lamp OR: If there is no "STRIKE" button:
Flip the auto-ignition switch to "AUTO", then immediately return it to "MANUAL" after the bulb is lit.

Caution: The auto-igniter must be disabled. If the bulb should go out, it can not be allowed to re-ignite automatically and injure someone connecting leads or damage test equipment.

8. Connect the oscilloscope across the ammeter shunt. Set the oscilloscope to 10 millivolts per division; the timebase to 20 milliseconds per division. Avoid using "X10" probes. There are no special triggering requirements, AC or "AUTO" will work.

Note: Since the oscilloscope will be reading DC, greater accuracy can be obtained by adjusting the scope's baseline to the bottom of the screen, rather then to the center.

Caution: Older oscilloscopes with a "third-wire" ground on the power cord may give erratic or incorrect readings. This may be caused when the shield wire of the oscilloscope probe is connected to one side of the shunt. (Effectivly, one side of the DC output is being grounded when the probe shield in connected.) A solution is to isolate the "third-wire" ground on the oscilloscope using a grounding adapter, but this can create an unsafe condition. If you do this, be aware you are working with ungrounded equipment and act accordingly.

Record Imax (total peak voltage) Record Imin (minimum peak voltage) Calculate ripple and record. The percentage of ripple should be less then 5%.

ripple = {Imax - Imin} over Imax times 100%

9. Record the lamphouse ammeter value.
Because the ammeter shunt reduces the actual current value to the lamphouse ammeter by 1/3000, there will be 3 amps for every 1 millivolt shown on the oscilloscope. Therefore, the amperage can be calculated by multiplying Imax, in millivolts, times 3000. For example:

36 mv * 3000 = 108 amps

Note that this test requires an accurate 150 mv shunt.

10.Compare the readings from the ammeter and the calculated current. Record the values. Replace the meter in the lamphouse if the values are more then 15% apart.

11.Return the door safety interlock switch to normal.

The following information should have been recorded:
The date of testing.
The lamphouse model, serial number, and location.
The power supply model and serial number.
The values: Imax; Imin; operating lamphouse voltage; actual voltage; operating amperage; actual amperage; percentage of ripple.

Posted by John Pytlak (Member # 331) on 07-05-2000, 06:23 AM:

Another method, that does not involve the risk of connecting a meter or scope to the lamp's power supply, is to measure the "ripple" in the light output using a silicon photocell.

Let light leaking from the lamphouse (before the shutter) fall on a silicon solar cell. Use a variable resistor to load the cell to produce an output of about 0.50 volts when fully illuminated. Use an oscilloscope to measure the voltage across the load resistor. If 0.50 volts represents the "steady state" light output, there should be no more than 5% AC "ripple" (0.025 volts peak-to-peak) in the cell's output. Be careful not to let room lights (especially fluorescent) illuminate the cell during the measurement, as they will have considerable AC component.

------------------
John P. Pytlak, Senior Technical Specialist
Worldwide Technical Services, Entertainment Imaging
Eastman Kodak Company
Research Labs, Building 69, Room 7419
Rochester, New York, 14650-1922 USA
Tel: 716-477-5325 Fax: 716-722-7243
E-Mail: john.pytlak@kodak.com

Posted by Gordon McLeod (Member # 33) on 07-05-2000, 08:19 AM:

Osram published a paper in the early 60's on the ripple factor and it is not directly readable either in light or voltage method but in current
THe proper way is to use a Non Inductive shunt (similar to a meter shunt but has no inductive qualities) and connect a scope accross it. THe scope must have either a balanced input or be decoupled from the line via a isolation transformer and a lifted ground.

As the lamp lights the inrush current rise can be graphed and inrush is one of the more distructive items the lamp faces

The ripple is then meausered as the precent fluctuation visible in the waveform

The ushio group at the same time working with chrisitie determined that the standard rectifiers large output capacitance caused far to much inrush current and still failed to reliably get the lamp to light so they worked out a output stage that supplies the boost in two humps thus making ignition easier on many lamps and in fact allowed for lower noload voltage.
A chart recorder is also connected for most tests

Posted by John Pytlak (Member # 331) on 07-05-2000, 08:37 AM:

Gordon is correct that Osram defines the current pulsation as P = ((Imax - Imin) / Imax) x 100%, which requires reading the current across a shunt. Isolating the scope input and not having a "ground loop" interfere with the readings can be tricky if you are not sure what you are doing. Using the silicon solar cell to measure any AC component in the "steady state" light output is less risky, and can be used as a relative measure of ripple.

BTW, a scope usually reveals alot of "hash" in the current or voltage waveform being supplied to an operating xenon lamp. The plasma of the arc is not as well behaved as a tungsten lamp.

Posted by Francis Casey (Member # 514) on 07-05-2000, 08:40 AM:

Gordon
For many years now i have been using a BELL model CG100D AC/DC clamp on ampmeter to measure ripple . I clamp it on a DC lead after the lamp is lit , measure the DC current , then switch it to read AC and measure the AC current ;i then calculate the percent ripple . Do you think this method would be as accurate as your SHUNT method ?
Francis

Posted by Gordon McLeod (Member # 33) on 07-05-2000, 11:36 AM:

I don't know. The reason for the non inductive shunt is to remove the stray inducatance and capacitance in the test equipment itself. I very rarley have to actualy test for ripple since in most cases a reactance rectifier typically relies on a bridge and cap if it is an electronic one that is a different story and they can be a pain. Also there has never been much published on the effects of the hash high frequency noise that is on the dc especially with the electronic ones
I have always found that the Sanrex and the Christie rectifiers have provided the best D.C. and longest lamp life.
But some lamps (ORC) don't like the lower boost voltage that the Sanrex provide
The number osram quoted normaly is each ignition was one hour of life due to inrush and in a lot of cases bulb failures have been due to inrush more than ripple
Especially in the XPS45 styled rectifier (from any of the cloneing manufacturers) and that is from people replaceing the nichrome wire with regular wire

Posted by Mark Gulbrandsen (Member # 72) on 07-10-2000, 03:38 AM:

I have one of the ORC ripple/inrush meters and have used it for quite a few years. It has been compared to the standard shunt that Osram uses and is literally as accurate. The basis of it is also an FW Bell digital DC ampprobe with its output feeding a specially designed ORC meter. This unit also measures inrush current as well with a peak measurement-hold type of circuit. I don't know if ORC still makes these but it has been very handy on a number of occasions. I always check all the supplies with it when I go through a plex on maintainance.

Mark

Posted by John Walsh (Member # 168) on 07-10-2000, 07:24 PM:

Osram specifically stated somewhere (of course, I can't find it now) that they did not recommend the use of clamp-on amp meters for measuring ripple. This was in 1980 or so. They specifically stated "Amprobe" meters, which then were not as complex as today, so perhaps this no longer applies.

Posted by System Notices (Member # 2357) on 07-21-2004, 09:18 PM:

It has been 1472 days since the last post.

Posted by Matt Manalis (Member # 2367) on 07-21-2004, 09:18 PM:

Hi

i am using an irem recitier and i do believe that they have quite a bit of ripple even they are new but i think the is excessive ripple and was just wandering what the best method to check this would be i have a scope but tha is about all?? thanks

Posted by John Pytlak (Member # 331) on 07-22-2004, 05:29 AM:

The safest method is using the technique of letting some of the light leaking from the lamphouse shine on a silicon photocell, and looking at the output on a scope. The steady state DC represents the average total light output, and any AC is the flicker from current ripple. The OSRAM xenon technical manual discusses the technique in detail:

http://www.sylvania.com/business/photo/pdfs/ENGR_BLTN11.pdf

quote:
See "Current Ripple", Pages 17-18

[ 07-22-2004, 08:33 AM: Message edited by: John Pytlak ]