I have recently started work on PFC for several projects. The main one being a large Tesla coil. PFC involves a special type of boost SMPS to appear to the incoming power as a purely resistive load. It draws current throughout the sine wave not just at the peaks.
I have one circuit handling 4kw constant with a 400 volt output. This will then become the input to a large full bridge inverter. Most modern "universal input" power supplies use this topology. If people are interested, I have a large supply of 1kw power supplies from the telecom world that are all in good working order just missing the heatsinks. This board is a wealth of parts for people experimenting with smps. It has three different control chips and a bunch of gate drive and current sensing transformers on it. There are also five IRFP460 mosfets and numerous high speed diodes.
The main transformer is an EC70 3c85 material.
I cut the pfc section off of one and with some slight mods, it was putting out 1550 watts with a 120 volt input.
If interested, I will sell them for $10 US each. Buyer to pay the shipping. A group buy for overseas would be great, I will work with you on getting them shipped. I currently have thirty of these in stock. I have a ton more of other supplies with similar parts on them I am going to be selling.
I will be posting my progress on the pfc development and schematics as I get a chance to draw them up.
I have one circuit handling 4kw constant with a 400 volt output. This will then become the input to a large full bridge inverter. Most modern "universal input" power supplies use this topology. If people are interested, I have a large supply of 1kw power supplies from the telecom world that are all in good working order just missing the heatsinks. This board is a wealth of parts for people experimenting with smps. It has three different control chips and a bunch of gate drive and current sensing transformers on it. There are also five IRFP460 mosfets and numerous high speed diodes.
The main transformer is an EC70 3c85 material.
I cut the pfc section off of one and with some slight mods, it was putting out 1550 watts with a 120 volt input.
If interested, I will sell them for $10 US each. Buyer to pay the shipping. A group buy for overseas would be great, I will work with you on getting them shipped. I currently have thirty of these in stock. I have a ton more of other supplies with similar parts on them I am going to be selling.
I will be posting my progress on the pfc development and schematics as I get a chance to draw them up.
Attachments
damn, this is a big unit, since I have 1kw+ unit, and its about 1/5 or smaller... and I don't wan't to compare it to yours... I guess this is 120/230v or is it true "universal input"?
Your design right?
Your design right?
PFC is often a waste of money for US residential users since US residential users are usually not billed by power factor. That means you'll be adding more cost and complexity while (usually) reducing efficiency for no gain. (It does, however, make sense for commercial and industrial users.)
Luka, the pictures are of a board with pfc and three outputs. I am selling them for parts. I rescued them from a recycler, managed to talk them into just unbolting the heatsinks rather than hammering them off. The pfc I have built is 3 by 4 by 2 1/2 inches. I will get a picture up shortly.
Star, the point is to be able to draw more power from an outlet without tripping breakers by averaging out the current draw. For Tesla coils, this is very important. For bench power supplies it helps also if you have limited power available.
Star, the point is to be able to draw more power from an outlet without tripping breakers by averaging out the current draw. For Tesla coils, this is very important. For bench power supplies it helps also if you have limited power available.
True, it does reduce efficiency. But it's the right thing to do. Having seen the harmonic trash that comes out of the wall, I have to say that the poor PF of the many household devices we all have (fluorescent lights are particularly bad) is ultimately degrading our audio experience due to interference and intermodulation.
I'm with you, Ian. The issue for residential users is not one of power factor correction, but the lower harmonic content we demand from the already taxed supply voltage. Those PFC circuits just drive the switching frequency up, though, so we may still have to contend with noise whilst we fix our harmonic footprint.
My local utility is running ragged trying to get my voltage distortion below 5%, which is required by IEEE and PSC. Every little bit helps, and PFC supplies are nearly sinusoidal.
Personally, I avoid flourescents like the plague. Ugly light and nasty on the power grid. I'll take my power hungry incandescents, thank you very much.
My local utility is running ragged trying to get my voltage distortion below 5%, which is required by IEEE and PSC. Every little bit helps, and PFC supplies are nearly sinusoidal.
Personally, I avoid flourescents like the plague. Ugly light and nasty on the power grid. I'll take my power hungry incandescents, thank you very much.
In applications not requiring tight regulation of output voltage, such as audio amplifiers, PFC provides constant output voltage (with some ripple) regardless of input voltage, avoiding a further regulation stage and voltage selection for worldwide operation.
Most modern pfc circuits approach 95% efficiency so they are not too bad energy wise. Two problems with them is they are not isolated from the mains and when there is an over current, the supply will shut down but the catch diode still passes current from the incoming power. If you have a failure post pfc, it will continue to feed current until a fuse (or something else) blows.
I was thinking about winding a transformer on one of the cores and see if I can still get the same throughput with isolation. Any ideas on this Eva?
I was thinking about winding a transformer on one of the cores and see if I can still get the same throughput with isolation. Any ideas on this Eva?
PFC SEPIC? S'pose it's possible, doesn't seem worthwhile though. When your average forward converter fails, it will end up blowing the fuse anyway, it's just going through an extra choke and diode this time. Potentially more to replace, but all the more reason to not make something prone to explosion in the first place.
Tim
Tim
And there's the option of not using PFC when first designing it and adding in PFC later.
AC question. Electric expansion valve?? - EcoRenovator
Although the complexity and cost will likely make it impractical, could it be possible to have a DSP measure the harmonics and then shape the input current in such a way as to help cancel them out?
No reason to use regular incandescents either. Halogen lamps are more or less purely resistive loads (at least at 60Hz) and provide better light than regular incandescents.
A similar trick is often used in variable speed A/C compressors.
AC question. Electric expansion valve?? - EcoRenovator
Certainly; they already make them. They are called active harmonic filters.
ABB LV Active Filters - Power Quality Products (Low Voltage Products and Systems)
Just a little pricey, as you might imagine. I think the PFC route is much more accessible and affordable.
Nice, a power filter that isn't a piece of overpriced junk hiding behind the nameplate of audio. Would this give us real hospital-grade power?
The purpose is not 'pure' power, so to speak. It's application is large industrial customers with high harmonic content (arc furnaces, drives, molding, etc) that HAVE to improve their distorted sine wave - and they are the reason it is distorted.
So you might find your supply THD to improve from 10% to 3%, and this would be sufficient. The 7% you corrected was successful because your active harmonic filter injected the necessary harmonic currents into your loads, such that all you draw from the utility is the fundamental. The remaining 3% is due to utility distortion, and there's nothing you can do about that (with this equipment).
Now this type of equipment, to my knowledge, is not available in sizes under a few hundred kVA, so is most unavailable to the residential user, in addition to costing many tens of thousands of dollars. But, with all technology, as it becomes more established, it trickles down to the end user. So who knows, in 10 years, we may all have one of these installed on our panels.
So you might find your supply THD to improve from 10% to 3%, and this would be sufficient. The 7% you corrected was successful because your active harmonic filter injected the necessary harmonic currents into your loads, such that all you draw from the utility is the fundamental. The remaining 3% is due to utility distortion, and there's nothing you can do about that (with this equipment).
Now this type of equipment, to my knowledge, is not available in sizes under a few hundred kVA, so is most unavailable to the residential user, in addition to costing many tens of thousands of dollars. But, with all technology, as it becomes more established, it trickles down to the end user. So who knows, in 10 years, we may all have one of these installed on our panels.
Easy to do passively, just slap on 22000uF of 160VDC motor run caps and a 3.2mH, 100A choke in parallel. Mind startup current!
I don't think a parallel synchronous machine* will do much with harmonics, however a series machine would (i.e. motor-generator). But that would be particularly inefficient.
*These are sometimes used for phase correction, because when the field is overdriven, the machine has a lagging phase, i.e., looks capacitive, cancelling inductive currents. The downside is, as a motor, it's maybe 90% efficient, whereas capacitors are 99.9%.
Tim
I don't think a parallel synchronous machine* will do much with harmonics, however a series machine would (i.e. motor-generator). But that would be particularly inefficient.
*These are sometimes used for phase correction, because when the field is overdriven, the machine has a lagging phase, i.e., looks capacitive, cancelling inductive currents. The downside is, as a motor, it's maybe 90% efficient, whereas capacitors are 99.9%.
Tim
Is that why these things never caught on? I remember when the NASA patent for the Nola design first came out. We were going to see controllers built into every motor. Then nothing. If I remember, they were designed for AC motors in the Apollo command module.
My latest trial. The board was removed from a commercial power supply, I changed components, added some pots so I can adjust frequency and duty cycle. I then added an MIC4420 mosfet driver chip on the output. The reason for using the existing board: It already has a nice ground plane on it. I don't have to etch one to do the experiment.
The scope shows this being driven into a 5820pf gate capacitance on a 75 amp 1200 volt IGBT. There is some ringing on the upper trace, I have a 6.2 ohm resistor in series with the gate. If I lower the resistance, the ringing gets bad and spikes start happening. The resistor gets slightly warm, the driver chip is cool. The whole circuit draws about 50 ma at this setup.
Soon to come: I am going to put some high voltage and a load on the IGBT to see how it behaves at different frequencies. Then I'm going to wind a big boost inductor and see if this will give me the power I want.
The scope shows this being driven into a 5820pf gate capacitance on a 75 amp 1200 volt IGBT. There is some ringing on the upper trace, I have a 6.2 ohm resistor in series with the gate. If I lower the resistance, the ringing gets bad and spikes start happening. The resistor gets slightly warm, the driver chip is cool. The whole circuit draws about 50 ma at this setup.
Soon to come: I am going to put some high voltage and a load on the IGBT to see how it behaves at different frequencies. Then I'm going to wind a big boost inductor and see if this will give me the power I want.
Attachments
Last edited:
Khron, I will check on the cheapest shipping to Romania. The boards have a wealth of parts on them for doing these type of experiments. I will look for a cheaper way, the USPS wants $77.00 to ship two of the to you. They weigh 7 pounds (3.2 kilos) each.
- Status
- Not open for further replies.