I don't know if you're aware of it but that flicker can cause seizures. I am currently helping to upgrade an entire building from florescent to LED. The florescent also have that flicker and were contributing to seizures. I probably spent almost a $1000 just testing bulbs and drivers till I found something that would work. For the most part it came down to finding a driver that had a capacitor to smooth out the frequency. We will finish replacing all the bulbs and drivers in the next week or so, but after that there is few of the standard screw-in light bulbs and I am considering just modifying LED bulbs by adding a capacitor. So videos like this are very useful, especially when it comes to understanding the problems I may run into.
240V Mains ➡ FULL BRIDGE RECTIFIER ➡ film in oil motor capacitor _(for heat tolerance and longevity)_ ➡ two of those COBs in series _(assuming the forward voltage is met)_
Clive - If you want to examine individual cells of any configuration of LED - try looking at them using a UV light source ... the phosphor dots flouress quite nicely. Makes it easy to spot burns. I recently converted a torch to use with UV resin glue and it's turning out useful for other things too.
Another great video. I have been wondering for long time if you had ever thought of teaching ? You inspire many people. And your humour makes it so enjoyable.
That was interesting, glad you tried it to see what would happen. It was cool to see that the capacitor did actually remove the flicker and also the side effects it had on the rest of the electronics.
A smaller capacitor might allow for elimination of most of the flicker without pushing the chips that hard. I calculated that the cap should last for about 7ms (from the peak to the next time the voltage enough ( about 185Vish?)), and your 10uF would drop from 330V to 267V in that time, whilst 4.7uF would go to 196V (close enough?). I calculated with crude numbers and assuming constant 0.09A current, but if you like to experiment, maybe try 4.7uF ... EDIT: Also, the sinewave likely exceeds that slope for a while on the top, and caps might be bigger than specified (so good idea to measure them), maybe experiment with even smaller caps too..
Hopefully this is just first generation for driverless, and the second generation will fix it. This technology looks like a good step in making LED technology cheapper and easier to enter more markets.
In the previous video you said that the current was programmed to the chip by a resistor. If you increased the resistor, would it negate the extra power from the capacitor modification?
Jernej Jakob trimming back the silicone is easy enough and the old reflow trick of putting the board on an electric hot plate and use a hot air reflow pen to melt the solder. Don't underestimate Big Clive's persistence.
After adding a cap to smooth out the flickering, could you ad a resistor in series to the LEDs to limit the heat created by the constant on heat buildup of the LEDs? Also could a variable resistor be used as a dimmer controller?
I would be very interested in further ways to stop these LEDs from flickering. What do you think about adding a capacitor in series with the LED with a capacitor across the rectifier as well? Or how about adding a big capacitor between the driver and the LED itself? This wouldn't probably be a lot less effective against flickering since the forward voltage across the LEDs would still ripple a little (depending on the capacity of the capacitor). But at least the drivers wouldn't be getting a peak mains voltage all the time. One might also try to decrease the voltage to 170 VAC, so the rectified voltage is 240 VDC. This may be an overly complicated solution to stop a 2 dollar or so LED from flickering. But I would still be interested in your opinion to different ways fixing the flicker.
The particular driver I was looking at has a temperature rating well above 100c, it may be that with a larger heat sink the LEDs will work just fine even with the wonky power factor and increased power dissipation at the chip. Thanks for the follow up!
Clive, we have a few cheap 140W LED floodlights in the shearing shed which give brilliant light for the shearers but they cannot listen their favourite AM radio station due to the noise the drivers put out. Do you have an AM radio handy you could check the effect these driverless LED's have on the radio? Shearers get very cranky while working without their background music.
V8Snail These should be quieter electrically. The standard high frequency electronic drivers are very poorly filtered if at all. It might be viable to add an extra filter to the existing lights.
+V8Snail put an AM receiver behind the lights (where the radio output from the lights should be blocked by their casings/heatsinks), then draw out the audio wire to a set of speakers where the shearers are?
AM radios are the most susceptible to RF interference. I used to use one to test the effectiveness of the RF suppression on my dimming controllers. Can you switch to FM? I wonder if there's enough room in the lights to add a proper mains filter like this:- www.rapidonline.com/schurter-fmw-41-chassis-mount-ac-line-filters-1-phase-562073
hi Clive loving the videos, I'm on holiday on the island for this week staying just south of Peel would love to pop by and say hello. big thanks Daz in his campervan
similar issue with what clive did, the bridge rectifier inrush current would of been huge when you switched on. that bridge would eventually go pop. with a cap across the LED's the inrush would then hit the regulator chips. pop..
A cap across the LEDs would do nothing. A capacitor tries keep voltage constant and so does a diode. The AC impedance of an LED is very low and that it nullifies that capacitor. To those saying it would increase inrush: No. The driver chips are current sources.
I agree with Mister Hat, a capacitor across the LEDs will not cause a high inrush current and neither will it stop the flickering. When the input voltage drops below the LED voltage (about 220V) the bridge rectifier will become reverse biased, the current in the LEDs will drop to zero and the voltage across the LEDs and the capacitor will drop a bit but essentially just remain constant until the next cycle.
The digits are most likely multiplexed at at least a few hundred Hz. If the (electronic) shutter is open long enough per frame (as in low light conditions), all digits will light up at least once within that single shutter (frame rate and shutter speed are two different things) and thereby look flicker free. (Yes, I'm spamming this of mine comment around)
Hi Clive, hello from Australia. Thanks for your great videos, I put your knowledge to good use when a capacitive dropper in a PIR floodlight failed and the switching relay just buzzed. All sorted now :) The video on your quick test was excellent, are you able to demo some of your other bits n' pieces in future videos? I'm in the process of setting up a workshop and it'd be great to know the details of some of the wattmeters, hand tools and other interesting items of test gear or measurement that I should probably have (whether I need them or not!). Sometimes it's a matter of knowing these things exist so we can buy them. Cheers, Adam
Ok so if I ordered the 240v versions and used them on my 120 mains with the smoothing capacitor would it work better? I mean the rms voltage is much lower so putting that capacitor in there would just compensate for the lower voltage right?
If the peak (not RMS!) value of your supply does not exceed the LED string voltage, it will never light up. Considering there's about 70 LEDs a 3V you'd need at least at around 200V (absolute, not RMS) for it to light up at all.
120V wouldn't hit the 180V needed to light the LED's properly even if smoothed. Now if you swapped out the bridge for a voltage doubler using a cap and two diodes then you would be onto something :)
Great information that you are sharing. It is very useful to me to see you revers engineer and then modify the circuits since to me has been a great help with my own experiments. Thank you for your great work and sharing. You saved me so much time.
If the drivers are self-regulating with temperature, I do not think they will suffer over time. The temperature does not seem higher than your previous, unsmoothed test, soooo... This is quite brilliant.
Hi, how about a external bridge rectifier, smoothing capacitor on the output of the external bridge rectifier, and put (2) LED lamp modules in series? Would that bring the temps down and make it more stable, while illuminating the flicker? There should be room in the case for both modules.
Something I've always wondered about is whether it would be possible to use a phosphor over the LEDs that had a bit of persistence to them and thus would quench the flicker somewhat by glowing more constantly.
Thank you very much, Clive. I highly appreciate your work in exploring and explaining this gadget. I will still have a very hard time, though, soldering anything to the tiny Rectifier legs?
To reduce the flicker without overrunning the chips or LEDs, fit a smoothing cap after the bridge rectifier. Then add a series cap in the mains feed with its value selected to return the power consumption to its previous level.
Butter melts at body temperature, this LED gets up to 185F, so yes. It may take awhile to melt a large quantity of butter since the butter would act as a liquid cooling agent if applied directly to the back of the heatsink.
Clive, I agree with the commenter who recently suggested that you should devise a method of measuring the light output from the many lamps that you review. When we are considering LEDs or other lamps for use in lighting applications what we need to know is how many lumens? Sometimes the eBay sellers quote a value but we all know how much trust to put in that.
You can trust Chinese LED seller specs 1000%, if you move the decimal point twice. Most of them only know what they were told by the manufacturer, and they have no concept of electronics, they're just peddling.
My plan for measuring lumens is to put the lamp inside a perfectly reflecting sphere which has a small hole in it which is covered by a light meter measuring lux. All the light bounces around until it exits via the hole and is captured by the light meter. Does anyone have a 100% reflective sphere handy? Ok, it doesn't have to be 100% reflective a certain percentage of the light will be absorbed so it will have to be calibrated with a known light source like a 60W incandescent bulb. PS:I don't believe any of the conspiracies in that list.
tbh, as long as the testing conditions clive uses are consistent across all the leds, it ultimately does not matter. I would not mind buying leds based on a clive-scale
Hi Clive, I would still like to see what would happen if you mounted a heat sink to the top of these chips to see if the heat decreases from 86º. I would think with the right heat sink they could go down to around 50º
Hi Clive. Have you seen the Gosun floodlights on Amazon yet? I've used a couple of the 30w daylight white ones and they seem pretty good and very competitively priced. They're quite a different design to the normal aluminium troughs and they claim to use cree LEDs. Just thought they might be worth a look.
I wonder... If you add a resistor on mains side to take some load off the current regulating chips would it actually increase the power of this module?
A friend and I bought some of these Chinesium LED lights, they claim 50W. They consume about 35, and get smoking hot. After a while, get a 'black dot of death' in one or two of the LEDs. I put a motor run capacitor, ( about 5 uF ) in series with the incoming power. Dropped the power consumed down to about 17W, and the lights run much cooler. Only been on the test area ( front porch fixture ) for about two months, but have so far outlasted the factory setup. The capacitor lets about 85 volts through.
I know this is an old video, but have you tried using a variable capacitor and check power output in varying capacitance. Maybe a small enough value would fix the high rms issue
With the capacitor installed, how would you make this circuit more efficient? If the power consumption is the same as before then how is it dissipating more heat? Does adding the capacitor makes the LEDs dimmer?
*I just got this LED from different seller for 120V, it looks quite different and seems to have testing points, also, the cover on chips is transparent, I could send you a full resolution picture if you want.*
Why not use an external capacitive dropper feeding a bridge rectifier and smoothing capacitor to power the LED with 240V DC or perhaps a bit less? The flicker would be gone and the dissipation of the current regulators would also be better.
Re: Adding capacitor in series @ 8:00 minute mark First, thanks so much for showing us these fascinating products Clive. I bought eight of these on eBay after seeing this and four of them died within an hour running straight from the mains electricity. So, I followed your advice on taming them with some capacitive reactance. The results were amazing but I do have a question. In my case, I'm using 50W "full spectrum" COBs rated for 120V AC 60Hz current so I went with a 1uF and 5uF 250V capacitors in series on the AC line before the COB. According to my calculations the 1uF should have been 3W or so and the 5uF closer to 30W at 120V 60Hz. I don't have a meter to check the light output but it did seem that the COB with the 5uF was much brighter than the 1uF as it should be. However, the strange thing is that they both seem to run completely cool to the touch although they're very bright. It's amazing how cool they run but the 5uF should be warmer it seems because 30W is still quite a bit of heat. Do you have any input on what might be happening?
What about using like a 4.7uF low ESR cap to give some smoothing, but not quite so aggressive? Thinking about a balance between no flicker and a little.
How hard would it be to replace the current sense resistors with a larger value to cut down the current along with the capacitor? You could eliminate the flicker without increasing the power dissipation..
I think I've asked this before but didn't get a reply... How do you like the Hopi meter? I really like the fact that it displays all different values at once. Accurate enough and worth the money?
It's OK, although there are a few design quirks. You can't plug a UK plug straight in because the area the flex comes out is obstructed. It also "cheats" when you put on a really horrifically capacitive load and pretends nothing is there. The most endearing feature is the pop-off battery compartment cover on the back that instantly exposes live connections, all wired in green.
bigclivedotcom Thanks! Too bad the multiplexing make it flicker so much on camera. I think the display's much nicer than the LCD meter you normally use. I notice the flickering totally disappears the moment you lift it up and move it towards the camera though ( 4'13" ) Any idea why?
Would changing that zero ohm link to an actual value help to limit the power so the chips werent running as hard and help, when the capacitor is soldered on the board, to bring it back down to a power similar to before the capacitor was added?
Nepal Plush Yes, but you will be dissipating quite a lot of power in that new resistor, meaning you could not just insert another chip capacitor as it would need to be a high dissipation wire wound type resistor.
Would it be possible to reduce the input voltage with a voltage regulator and THEN connect the electrolytic capacitor that will increase the voltage up to the ideal voltage?
Maybe the answer is a smoothing cap and a series cap to bring the wattage back to 20 watts, not an issue in such a lamp fitting. That is course only if flicker is a problem.
bigclivedotcom is there a better COB solution that doesn’t? I remember you adding a cap to remove the flickering which raised the voltage but I’m curious why you couldn’t/didn’t add in a resister at the mains input to drop the voltage. Too much power?
bigclivedotcom is it 120Hz flicker and does it get more noticeable when a dimmer starts to lob off the front or rear of the wave? The lighting is for a large terrarium.
What if you were to smooth the voltage with a modestly sized inductor (in series) instead of a capacitor? Wouldn't that smooth it without raising the voltage to the peak mains like a capacitor would?
Hi Clive stumbled over a 200W flodlight today, the owner wanted to use it when filmning Wath do you think of get a dropper capasitor before rectifire and e big capasitor on the dc side So dropping 40% ten rectifi and smoth One other option is to build a 160v supply AC by haveing mans in to one transformer with 230/60v ruffly and the secondary in series backwords with the mains giving tuffly 160 v rectifying it to 230v dc wit capasitors
Before using it for filming check to see if it has flicker that is picked up on camera. For high power applications a series capacitor isn't always a good idea.
Interesting, so what would happen if one designed the LEDs and drivers from the start to take a capacitor? even if it doubles the cost, I would view it as worthwhile.
Can you do a video explaining when to use a linear voltage regulator over a buck converter? I'd love to hear it from your view and in your soothing, booming voice. I want to know when I should use my LM317T and when I should go out and buy those cheap break-away buck converters.
The linear regulators are a lot simpler and generate very little electrical noise. But they also dissipate any voltage difference as heat like a resistor. So for low current regulation something like an LM317 or a 7805 is a good choice, but for high current the switching regulators will run a lot cooler.
Hi Clive I'm not expert in electronics, but would it be possible to use say two of these modules together so they run at alternating sync, thereby no flicker ??
Peter Scharff two legs of power like in a house hold still have the same zero crossing point, but if you had 3 of them connected to different phases of 3 phase there might be an interesting result
Perhaps an odd question - I've been considering getting an insulation tester, seeing you use yours reminded me - Usually I build embedded devices & use someone else's power supply (UL Listing something in the US is costly! But a UL listed "wall wart" power supply handles that UL listing...) For the times I want to test a possibly "iffy" AC power circuit, are there any inexpensive but worth-having "meggers"? Thank you.
Good point. I need to get back working (injured but recovering), still "spend once, cry once" can be good advice. Fluke gear'd be nice, but too costly. I'll find something that's actually US certified, seems wisest, then.
I don't know if you're aware of it but that flicker can cause seizures. I am currently helping to upgrade an entire building from florescent to LED. The florescent also have that flicker and were contributing to seizures. I probably spent almost a $1000 just testing bulbs and drivers till I found something that would work. For the most part it came down to finding a driver that had a capacitor to smooth out the frequency. We will finish replacing all the bulbs and drivers in the next week or so, but after that there is few of the standard screw-in light bulbs and I am considering just modifying LED bulbs by adding a capacitor. So videos like this are very useful, especially when it comes to understanding the problems I may run into.
240V Mains ➡ FULL BRIDGE RECTIFIER ➡ film in oil motor capacitor _(for heat tolerance and longevity)_ ➡ two of those COBs in series _(assuming the forward voltage is met)_
Clive - If you want to examine individual cells of any configuration of LED - try looking at them using a UV light source ... the phosphor dots flouress quite nicely. Makes it easy to spot burns. I recently converted a torch to use with UV resin glue and it's turning out useful for other things too.
Another great video. I have been wondering for long time if you had ever thought of teaching ? You inspire many people. And your humour makes it so enjoyable.
That was interesting, glad you tried it to see what would happen. It was cool to see that the capacitor did actually remove the flicker and also the side effects it had on the rest of the electronics.
Things like this are why I watch big clive. Bad idea? Lets do it!
Dear bigclivedotcom, thank you for answering our comment questions so fast with this video! I learned a lot!
Those things are so cheap that a multi-kilowatt LED array might be amusing to build.
mikeselectricstuff q
h
Roughly 10kW per square metre, say 50% heat would be generated
5kW heat is 1litre of coolant water through 5'C every second?
A smaller capacitor might allow for elimination of most of the flicker without pushing the chips that hard. I calculated that the cap should last for about 7ms (from the peak to the next time the voltage enough ( about 185Vish?)), and your 10uF would drop from 330V to 267V in that time, whilst 4.7uF would go to 196V (close enough?).
I calculated with crude numbers and assuming constant 0.09A current, but if you like to experiment, maybe try 4.7uF ...
EDIT: Also, the sinewave likely exceeds that slope for a while on the top, and caps might be bigger than specified (so good idea to measure them), maybe experiment with even smaller caps too..
Thank you for this update.How about connecting the capacitor directly parallel to the LEDs?!
Hopefully this is just first generation for driverless, and the second generation will fix it. This technology looks like a good step in making LED technology cheapper and easier to enter more markets.
not really... the power dissipation in the LEDs and the current limit resistor/chip becomes too much.
In the previous video you said that the current was programmed to the chip by a resistor. If you increased the resistor, would it negate the extra power from the capacitor modification?
I'd imagine replacing the resistors would be very hard due to the silicone potting and aluminium core PCB. Certainly not worth the effort.
Jernej Jakob trimming back the silicone is easy enough and the old reflow trick of putting the board on an electric hot plate and use a hot air reflow pen to melt the solder. Don't underestimate Big Clive's persistence.
You will need dispass more power on curent regulator too.
I just purchased two 30 Watt flood lamps. Thank you for your tearing apart for me. Very interesting videos.
After adding a cap to smooth out the flickering, could you ad a resistor in series to the LEDs to limit the heat created by the constant on heat buildup of the LEDs? Also could a variable resistor be used as a dimmer controller?
I would be very interested in further ways to stop these LEDs from flickering. What do you think about adding a capacitor in series with the LED with a capacitor across the rectifier as well? Or how about adding a big capacitor between the driver and the LED itself? This wouldn't probably be a lot less effective against flickering since the forward voltage across the LEDs would still ripple a little (depending on the capacity of the capacitor). But at least the drivers wouldn't be getting a peak mains voltage all the time. One might also try to decrease the voltage to 170 VAC, so the rectified voltage is 240 VDC. This may be an overly complicated solution to stop a 2 dollar or so LED from flickering. But I would still be interested in your opinion to different ways fixing the flicker.
So if you run the 220V version with 110V and smooth it, it should be fine?
I know absolutely nothing about electronics and most of what Clive says goes over my head. But I find these videos oddly entertaining...
The particular driver I was looking at has a temperature rating well above 100c, it may be that with a larger heat sink the LEDs will work just fine even with the wonky power factor and increased power dissipation at the chip.
Thanks for the follow up!
Clive, we have a few cheap 140W LED floodlights in the shearing shed which give brilliant light for the shearers but they cannot listen their favourite AM radio station due to the noise the drivers put out. Do you have an AM radio handy you could check the effect these driverless LED's have on the radio? Shearers get very cranky while working without their background music.
V8Snail These should be quieter electrically. The standard high frequency electronic drivers are very poorly filtered if at all. It might be viable to add an extra filter to the existing lights.
Yeah have tried chokes and various tricks with very limited success. I'll give a retrofit
a trial and see how that fares.
matt bry rural area mate, no interwebs, no mobile phone signal. Old cockies get stuck in their ways.
+V8Snail put an AM receiver behind the lights (where the radio output from the lights should be blocked by their casings/heatsinks), then draw out the audio wire to a set of speakers where the shearers are?
AM radios are the most susceptible to RF interference. I used to use one to test the effectiveness of the RF suppression on my dimming controllers. Can you switch to FM?
I wonder if there's enough room in the lights to add a proper mains filter like this:-
www.rapidonline.com/schurter-fmw-41-chassis-mount-ac-line-filters-1-phase-562073
Hi Clive, in this video you had the capacitor across the rectifier. If the capacitor were across the array of LEDs instead, how would it have fared?
hi Clive
loving the videos, I'm on holiday on the island for this week staying just south of Peel would love to pop by and say hello.
big thanks
Daz in his campervan
Thanks for experimenting! Good follow-up.
What if you put the capicitor over the LEDs instead of the entire input DC source? Would that at all change the results you just demonstrated?
Yep, it wold fry the chips. A cap across the LEDs is essentially a short.
similar issue with what clive did, the bridge rectifier inrush current would of been huge when you switched on. that bridge would eventually go pop.
with a cap across the LED's the inrush would then hit the regulator chips. pop..
A cap across the LEDs would do nothing. A capacitor tries keep voltage constant and so does a diode. The AC impedance of an LED is very low and that it nullifies that capacitor. To those saying it would increase inrush: No. The driver chips are current sources.
I agree with Mister Hat, a capacitor across the LEDs will not cause a high inrush current and neither will it stop the flickering. When the input voltage drops below the LED voltage (about 220V) the bridge rectifier will become reverse biased, the current in the LEDs will drop to zero and the voltage across the LEDs and the capacitor will drop a bit but essentially just remain constant until the next cycle.
I was not talking about inrush current, but the controller chips.
Hi Clive
How strange, when you showed the thermal camera the currant meter stopped flickering.
John Ferguson The video camera used a longer exposure to compensate for a darker scene.
Robert Szasz Thanks mate. That makes a lot of sense.
The digits are most likely multiplexed at at least a few hundred Hz. If the (electronic) shutter is open long enough per frame (as in low light conditions), all digits will light up at least once within that single shutter (frame rate and shutter speed are two different things) and thereby look flicker free.
(Yes, I'm spamming this of mine comment around)
Clive, do you work off of a script or are you just naturally well spoken and knowledgeable?
I usually just go with it. I can talk for hours about technical stuff.
Clive, what about adding a cap to a 240v driver less module and running it at 120v. And do you think it would handle dimming?
Very cool, Thanks for the follow-up tests and explanations!
Hi Clive, hello from Australia. Thanks for your great videos, I put your knowledge to good use when a capacitive dropper in a PIR floodlight failed and the switching relay just buzzed. All sorted now :)
The video on your quick test was excellent, are you able to demo some of your other bits n' pieces in future videos? I'm in the process of setting up a workshop and it'd be great to know the details of some of the wattmeters, hand tools and other interesting items of test gear or measurement that I should probably have (whether I need them or not!). Sometimes it's a matter of knowing these things exist so we can buy them.
Cheers,
Adam
Could you put a couple of high power ceramic resistors heatsinked to the case to drop the peak voltage when using the capacitor?
Ok so if I ordered the 240v versions and used them on my 120 mains with the smoothing capacitor would it work better? I mean the rms voltage is much lower so putting that capacitor in there would just compensate for the lower voltage right?
Quinn Miller I would like to know as well!
If the peak (not RMS!) value of your supply does not exceed the LED string voltage, it will never light up. Considering there's about 70 LEDs a 3V you'd need at least at around 200V (absolute, not RMS) for it to light up at all.
superdau true
120V wouldn't hit the 180V needed to light the LED's properly even if smoothed.
Now if you swapped out the bridge for a voltage doubler using a cap and two diodes then you would be onto something :)
WRONG! Specks call for 86 - 265 vac.
How would a cap in series work 8:32, I didn't understand completely. Could someone go into detail, I am interested in trying this. Thank You.
Great information that you are sharing. It is very useful to me to see you revers engineer and then modify the circuits since to me has been a great help with my own experiments. Thank you for your great work and sharing. You saved me so much time.
it worked for flickering problem ? what was the best value for the capacitor of 220v 50W led ?
If the drivers are self-regulating with temperature, I do not think they will suffer over time. The temperature does not seem higher than your previous, unsmoothed test, soooo... This is quite brilliant.
Hi, how about a external bridge rectifier, smoothing capacitor on the output of the external bridge rectifier, and put (2) LED lamp modules in series? Would that bring the temps down and make it more stable, while illuminating the flicker? There should be room in the case for both modules.
Did you put some sort of thermal compound in between the LED backplate, and the casing/heatsink? Might get better thermal performance.
Why does the flickering on the HOPI meter stop when you bring it up and focus on it? I'd be quite interested to know the science behind that!
can you add the capacitor after the regulator chips?
Something I've always wondered about is whether it would be possible to use a phosphor over the LEDs that had a bit of persistence to them and thus would quench the flicker somewhat by glowing more constantly.
That's most informative, thanks!
Automatic thumbs up for taking the time to answer questions.
Wouldn't you able to use 2 with a voltage divider to get the power down so you can use a capacitor on each without overdriving them?
Thank you very much, Clive. I highly appreciate your work in exploring and explaining this gadget. I will still have a very hard time, though, soldering anything to the tiny Rectifier legs?
To reduce the flicker without overrunning the chips or LEDs, fit a smoothing cap after the bridge rectifier. Then add a series cap in the mains feed with its value selected to return the power consumption to its previous level.
Does the heatsink get hot enough to melt butter for the purpose of making trashy two ingredient chocolate though?
Butter melts at body temperature, this LED gets up to 185F, so yes. It may take awhile to melt a large quantity of butter since the butter would act as a liquid cooling agent if applied directly to the back of the heatsink.
So, you're saying that we could forgo the heatsink and just use butter for cooling?
Uhh, no. Things would probably go exciting quick when you have 240v mains buttery liquid exploding everywhere.
"Exciting" is what we're looking for.
What about rectifying, smoothing with a capacitor , and then putting two modules in series, one for 230V, one for 110V, together ~350V?
That could work if their currents were matched. A 110V LED might operate parallel strings at higher current though. Maybe 3 110V LED arrays in series?
Clive, I agree with the commenter who recently suggested that you should devise a method of measuring the light output from the many lamps that you review. When we are considering LEDs or other lamps for use in lighting applications what we need to know is how many lumens? Sometimes the eBay sellers quote a value but we all know how much trust to put in that.
You can trust Chinese LED seller specs 1000%, if you move the decimal point twice. Most of them only know what they were told by the manufacturer, and they have no concept of electronics, they're just peddling.
Measuring light output in lumens is not trivial. Lux is an easy measure however.
My plan for measuring lumens is to put the lamp inside a perfectly reflecting sphere which has a small hole in it which is covered by a light meter measuring lux. All the light bounces around until it exits via the hole and is captured by the light meter. Does anyone have a 100% reflective sphere handy?
Ok, it doesn't have to be 100% reflective a certain percentage of the light will be absorbed so it will have to be calibrated with a known light source like a 60W incandescent bulb.
PS:I don't believe any of the conspiracies in that list.
tbh, as long as the testing conditions clive uses are consistent across all the leds, it ultimately does not matter.
I would not mind buying leds based on a clive-scale
number 33 Google integrating sphere. iirc they're white tio2 inside.
Hi Clive, I would still like to see what would happen if you mounted a heat sink to the top of these chips to see if the heat decreases from 86º. I would think with the right heat sink they could go down to around 50º
I wonder if there's a compromise value for the add-on capacitor that reduces the flicker enough but doesn't affect the power factor too much?
Hi Clive. Have you seen the Gosun floodlights on Amazon yet? I've used a couple of the 30w daylight white ones and they seem pretty good and very competitively priced. They're quite a different design to the normal aluminium troughs and they claim to use cree LEDs. Just thought they might be worth a look.
When running two next to each other in inverted phase (L and N switched over) would it eliminate the visual flicker?
inverted phase would do nothing, both waves would have the peaks at the same time
I wonder... If you add a resistor on mains side to take some load off the current regulating chips would it actually increase the power of this module?
Every time you say 'rectifier,' I hear "rectum fryer." AVE!!! WHAT HAVE YOU DONE TO ME!?
Red Green?!? That's a blast from the past.
AvE is good, but he aint Red Green good. I mean, c'mon, Red Green knew how to fix stuff that wasn't even broken.
I need to buy a rectum fryer
@@eksine Any soldering iron will do.
@@statusquo9520 thanks, you saved me a bunch of money!
From now on I'm going to call explosions "one big flicker"
A friend and I bought some of these Chinesium LED lights, they claim 50W. They consume about 35, and get smoking hot. After a while, get a 'black dot of death' in one or two of the LEDs. I put a motor run capacitor, ( about 5 uF ) in series with the incoming power. Dropped the power consumed down to about 17W, and the lights run much cooler. Only been on the test area ( front porch fixture ) for about two months, but have so far outlasted the factory setup. The capacitor lets about 85 volts through.
Good idea using the motor capacitor in series.
for 50w driveless, what capacitor spec e use? thanks
I know this is an old video, but have you tried using a variable capacitor and check power output in varying capacitance. Maybe a small enough value would fix the high rms issue
What about adding a capacitor for smoothing and a series resistor to shift the dissipation off the board?
With the capacitor installed, how would you make this circuit more efficient? If the power consumption is the same as before then how is it dissipating more heat? Does adding the capacitor makes the LEDs dimmer?
Why does the flickering of the test instrument stop when you place the thermal camera in front of the camera you are shooting with?
Exposure speed adjusting.
*I just got this LED from different seller for 120V, it looks quite different and seems to have testing points, also, the cover on chips is transparent, I could send you a full resolution picture if you want.*
Why not use an external capacitive dropper feeding a bridge rectifier and smoothing capacitor to power the LED with 240V DC or perhaps a bit less? The flicker would be gone and the dissipation of the current regulators would also be better.
Re: Adding capacitor in series @ 8:00 minute mark
First, thanks so much for showing us these fascinating products Clive.
I bought eight of these on eBay after seeing this and four of them died within an hour running straight from the mains electricity. So, I followed your advice on taming them with some capacitive reactance. The results were amazing but I do have a question.
In my case, I'm using 50W "full spectrum" COBs rated for 120V AC 60Hz current so I went with a 1uF and 5uF 250V capacitors in series on the AC line before the COB. According to my calculations the 1uF should have been 3W or so and the 5uF closer to 30W at 120V 60Hz.
I don't have a meter to check the light output but it did seem that the COB with the 5uF was much brighter than the 1uF as it should be. However, the strange thing is that they both seem to run completely cool to the touch although they're very bright. It's amazing how cool they run but the 5uF should be warmer it seems because 30W is still quite a bit of heat. Do you have any input on what might be happening?
how about running the 220 volt version on 120 volt with the cap
What if we add diodes after the rectifier? Will it smooth out the leds flicker?
What about using like a 4.7uF low ESR cap to give some smoothing, but not quite so aggressive? Thinking about a balance between no flicker and a little.
How hard would it be to replace the current sense resistors with a larger value to cut down the current along with the capacitor? You could eliminate the flicker without increasing the power dissipation..
You could definitely change the sense resistors to make the whole device lower power.
whats the equivalent to a non-led lamp per watt? (candles - lux) . can you compare this to an ordinary lamp?
What about using an inductor?
I'd be interested to see the result of the cap across the LED's instead of across the rectified mains.
I think I've asked this before but didn't get a reply...
How do you like the Hopi meter? I really like the fact that it displays all different values at once.
Accurate enough and worth the money?
It's OK, although there are a few design quirks. You can't plug a UK plug straight in because the area the flex comes out is obstructed. It also "cheats" when you put on a really horrifically capacitive load and pretends nothing is there. The most endearing feature is the pop-off battery compartment cover on the back that instantly exposes live connections, all wired in green.
bigclivedotcom Thanks!
Too bad the multiplexing make it flicker so much on camera. I think the display's much nicer than the LCD meter you normally use.
I notice the flickering totally disappears the moment you lift it up and move it towards the camera though ( 4'13" ) Any idea why?
Ni5ei When the area the camera is looking at is bright it effectively has a faster shutter speed that catches the multiplexing flicker.
Could you limit the power before or after the capacitor so lower the voltage?
Hi. Does the flickering affect the heating of the LED?
Can't we put a ceramic cap in series to decrease current, and also a cap in parallel? Reactance? If I'm wrong, please correct me.
What would happen if you put the capacitor on the other side of the current limiting chips?
Would changing that zero ohm link to an actual value help to limit the power so the chips werent running as hard and help, when the capacitor is soldered on the board, to bring it back down to a power similar to before the capacitor was added?
Nepal Plush Yes, but you will be dissipating quite a lot of power in that new resistor, meaning you could not just insert another chip capacitor as it would need to be a high dissipation wire wound type resistor.
Eating some trash chocolate, watching the new Big Clive episode..
It was edible, but somewhat foul-tasting. Perhaps because I used non-alkalized cocoa powder. I'll try adding milk next time I make it.
What is it with british people not liking bitter chocolate xD The bitter is what makes chocolate good :)
Would it be possible to use an inductor/magnetic ballast and some diodes to reduce the ripple? So it gets mains waveform and a phase shifted waveform?
clive, where to buy that power meter..
Would it be possible to reduce the input voltage with a voltage regulator and THEN connect the electrolytic capacitor that will increase the voltage up to the ideal voltage?
would adding a power resistor work to limit the voltage after smoothing?
Maybe the answer is a smoothing cap and a series cap to bring the wattage back to 20 watts, not an issue in such a lamp fitting. That is course only if flicker is a problem.
what is that big red button device clive has the mains wires going into?
Tighe Donnelly It's called a Cliff Quicktest.
How about a half bridge rectifier and a fatter cap? Maybe use a 120V version on half bridge rectified 230V?
Has the flickering been fixed in the two years since your analysis?
No. That type still flickers.
bigclivedotcom is there a better COB solution that doesn’t? I remember you adding a cap to remove the flickering which raised the voltage but I’m curious why you couldn’t/didn’t add in a resister at the mains input to drop the voltage. Too much power?
bigclivedotcom is it 120Hz flicker and does it get more noticeable when a dimmer starts to lob off the front or rear of the wave? The lighting is for a large terrarium.
@@sh91899 Yes it is at 100/120Hz and the use of a dimmer makes it much stronger.
@@sh91899 If you used a resistor it would get very hot. The capacitor increases the power/heat dissipation from the regulator circuitry.
Do you think this configuration might work better at 120V instead of 240V?
What if you were to smooth the voltage with a modestly sized inductor (in series) instead of a capacitor? Wouldn't that smooth it without raising the voltage to the peak mains like a capacitor would?
Hi Clive stumbled over a 200W flodlight today, the owner wanted to use it when filmning
Wath do you think of get a dropper capasitor before rectifire and e big capasitor on the dc side
So dropping 40% ten rectifi and smoth
One other option is to build a 160v supply AC by haveing mans in to one transformer with 230/60v ruffly and the secondary in series backwords with the mains giving tuffly 160 v rectifying it to 230v dc wit capasitors
Before using it for filming check to see if it has flicker that is picked up on camera. For high power applications a series capacitor isn't always a good idea.
I also found those questions interesting so I'm very glad you explored it, Clive!
But what if you added the cap after the chip, in parallel to LED string?? Flicker should be gone with no negatives perhaps?
Can you mitigate the flickering by increasing the frequence?
Do you tell the power company to spin the generator faster? Transition the U.K. from 50Hz to 120Hz to match modern television refresh rates
Great for a Floodlight conversion
Interesting, so what would happen if one designed the LEDs and drivers from the start to take a capacitor? even if it doubles the cost, I would view it as worthwhile.
Can you do a video explaining when to use a linear voltage regulator over a buck converter? I'd love to hear it from your view and in your soothing, booming voice.
I want to know when I should use my LM317T and when I should go out and buy those cheap break-away buck converters.
The linear regulators are a lot simpler and generate very little electrical noise. But they also dissipate any voltage difference as heat like a resistor. So for low current regulation something like an LM317 or a 7805 is a good choice, but for high current the switching regulators will run a lot cooler.
Wonderful, thank you!
So If you wanted to add a capacitor , would you have to change the resistors that set the current?
It could be useful, but is tricky to do on the aluminium core PCBs.
Hi Clive I'm not expert in electronics, but would it be possible to use say two of these modules together so they run at alternating sync, thereby no flicker ??
Peter Scharff two legs of power like in a house hold still have the same zero crossing point, but if you had 3 of them connected to different phases of 3 phase there might be an interesting result
No it would just be three phase power. You know.... What is used by industry EVERY DAY. (and they have the same problem.)
Alright, now these driverless LEDs have me quite fascinated. Can you put a capacitor on the LEDs themselves? As in after the drivers?
Perhaps an odd question - I've been considering getting an insulation tester, seeing you use yours reminded me - Usually I build embedded devices & use someone else's power supply (UL Listing something in the US is costly! But a UL listed "wall wart" power supply handles that UL listing...) For the times I want to test a possibly "iffy" AC power circuit, are there any inexpensive but worth-having "meggers"? Thank you.
+Mr Sheesh You can get cheap Chinese insulation testers, but they probably wouldn't be certification grade outside of China.
Good point. I need to get back working (injured but recovering), still "spend once, cry once" can be good advice. Fluke gear'd be nice, but too costly. I'll find something that's actually US certified, seems wisest, then.
Where did you get the cast housing used in this video? Is it a standard size? I have been unable to determine what the package size specs are.
Greg Cooper The package is 60x40mm.
Do you have a link for the housing? Thanks
What is the emc of thosed led's and how much rf to they put out and at what frequency, a shortwave radio will give a good idea of any problems.
I just ordered a small pile of these chips, and I'm hoping to pull DC off the board to power a 12v pc fan for cooling. Any ideas?