Thanks for the schematic, looks good. You must be working on this for quite some time !
400Hz is no problem for most equipment, like you mentioned, the core of the transformer can handle it easily when designed for 50/60Hz. Regulation is also better as the caps get charged more frequent.
As I work in the aircraft industry, I have some experience with 400Hz, the disadvantages are :
Some transformers or chokes can produce some 400Hz tone, which can be annoying over time.
Don't touch it ! While I can handle 115V 50/60Hz fine, it hurts a little, but 115V 400Hz hurts alot! Did not try 230V 400Hz
400Hz is no problem for most equipment, like you mentioned, the core of the transformer can handle it easily when designed for 50/60Hz. Regulation is also better as the caps get charged more frequent.
As I work in the aircraft industry, I have some experience with 400Hz, the disadvantages are :
Some transformers or chokes can produce some 400Hz tone, which can be annoying over time.
Don't touch it ! While I can handle 115V 50/60Hz fine, it hurts a little, but 115V 400Hz hurts alot! Did not try 230V 400Hz
Those designs look quite complicated and inefficient. You would be better served by purchasing a Class D module for the amplifier and driving a step up transformer. I built a line synchronized oscillator with lower distortion and variable phase shift (which I needed for other reasons) as attached. Extremely low distortion. The most distortion is added by the transformer, even though it is running at 50% of the volt-sec product, no where close to saturation.
If you are daring and want to build your own class-D amp, there is a forum for that. They are relatively cheap and ridiculously efficient, so I think you are better served just buying one. Either way, you can build any frequency power supply you want with this design; just substitute a bare bones RC oscillator up front, and you get a clean sine wave output better than what your amplifier can provide.
If you are daring and want to build your own class-D amp, there is a forum for that. They are relatively cheap and ridiculously efficient, so I think you are better served just buying one. Either way, you can build any frequency power supply you want with this design; just substitute a bare bones RC oscillator up front, and you get a clean sine wave output better than what your amplifier can provide.
Attachments
A self generated power source is already isolated from the Mains.
The output could be any useful voltage and any useful frequency, to suit the loads.
BUT the REALLY big safety feature of an isolated supply is that you have to touch TWO Live points for current to flow around the high voltage circuit.
With Mains Power, one only has to touch ONE Live point for there to be a hazard.
That is why Mains voltage is dangerous and that is why the rules insist on isolated supplies.
The output could be any useful voltage and any useful frequency, to suit the loads.
BUT the REALLY big safety feature of an isolated supply is that you have to touch TWO Live points for current to flow around the high voltage circuit.
With Mains Power, one only has to touch ONE Live point for there to be a hazard.
That is why Mains voltage is dangerous and that is why the rules insist on isolated supplies.
The input called "AC in" is simply a tap off the 6.3V winding, which feeds a comparator to produce a square wave synchronized to the line voltage. My pcb comes with a jumper that allows me to either use line voltage or any external source, such as a function generator. If your AC line is 50Hz, this device will produce a very low distortion sine wave at 50 Hz; there are no changes to make. And it will be phase locked to the AC line, and you have a full cycle of phase shift at your disposal to displace this phase lock in time (which is what I am working on testing with my DHT amp).
If you wanted to produce 400 Hz, you would need to provide a 400Hz input signal to the "AC in" points. This could be sine, triangle, square, whatever - the comparator transitions at the zero crossing. Keep in mind the PLL was optimized for 40-70 Hz range, so I don't know how fast and smooth it will operate at 200Hz+ frequencies.
I might have a few bare pcb's left over if you are interested in playing around.
Unless you want to run your 1970s alarm clock from a generated sine wave, I do not think it is important to phase lock to the mains at all. I like the idea of higher frequency a lot. I also like the idea of a class D amplifier, but I must be honest that I don't know much about class D.
What I do know from previous experience is that the transformer primary was in the order of 0.4 ohms and needed a few mosfets to cope with this low impedance, but the system works alright but could use a bit of refinement.
I have been looking at using the simple PIC to generate 400 Hz square wave as well as the a clock for utilising a MAX291 8th order switching capacitor filter. This would take the 2nd harmonic down to -80dB which is an order of magnitude or more better than the mains utility. In basically two components one can generate any frequency you please and would be easy to run and find the sweet spot for the application.
From the filter output we would need a voltage amplifier and then followed by either class D or B amplifier and a suitable transformer.
Andrew, it looks like we are going to cook with GAS here!
When experimenting I did find that the inrush current from a toroidal transformer was very high compared with an e-core and I think either will perform equally well at 400 Hz.
What I do know from previous experience is that the transformer primary was in the order of 0.4 ohms and needed a few mosfets to cope with this low impedance, but the system works alright but could use a bit of refinement.
I have been looking at using the simple PIC to generate 400 Hz square wave as well as the a clock for utilising a MAX291 8th order switching capacitor filter. This would take the 2nd harmonic down to -80dB which is an order of magnitude or more better than the mains utility. In basically two components one can generate any frequency you please and would be easy to run and find the sweet spot for the application.
From the filter output we would need a voltage amplifier and then followed by either class D or B amplifier and a suitable transformer.
Andrew, it looks like we are going to cook with GAS here!
When experimenting I did find that the inrush current from a toroidal transformer was very high compared with an e-core and I think either will perform equally well at 400 Hz.
Last edited:
Much to the contrary, there is a difference between requiring tight frequency control versus phase lock. I don't care if my power is 59.9 Hz or 60.1, as would be important to drive a clock. However, I do need it to be in sync with the line. I heat my DHT's with this generated sine wave, and need to phase lock to the mains in order to maximize 120 Hz cancelation between the DHT residual and the ripple on the power supply. By adjusting the phase shift I have even further control of cancellation. Without this I would get varying degrees of hum at some low beat frequency, which could be annoying (at least psychologically, knowing it's there).
Correct, although I suppose I could use batteries if that was my only available source. Then I would just need to add a bare bones oscillator on the front end, which would defeat the purpose of obtaining a phase lock.
Anyway, the OP was regarding power conditioning, which generically I take as cleaning up the AC mains and/or reconstructing with double conversion. Seems it headed in the direction of RF filtering for a while. But generating a 400 Hz signal for local use is certainly a valid goal. My concern would be about 800 Hz artifacts if any supplies are not regulated, which would be potentially more audible than 120Hz. Fun project either way.
Anyway, the OP was regarding power conditioning, which generically I take as cleaning up the AC mains and/or reconstructing with double conversion. Seems it headed in the direction of RF filtering for a while. But generating a 400 Hz signal for local use is certainly a valid goal. My concern would be about 800 Hz artifacts if any supplies are not regulated, which would be potentially more audible than 120Hz. Fun project either way.
good lord > class D to generate a clean source of power, what are u trying to correct / just the fundamental, what about all the EMI / RFI trash class D itself makes?
I rather live with a distorted sine wave, it doesn't matter much on good engineered power supplies eg looks a little like a slightly lower line condition. EMI /RFI an the other hand youll play hell with trying to filter that.
I rather live with a distorted sine wave, it doesn't matter much on good engineered power supplies eg looks a little like a slightly lower line condition. EMI /RFI an the other hand youll play hell with trying to filter that.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.