HI all.
I want to make a inverter powered by a 12V wall-wart adapter to power valve heaters with the low side then drive a center-tapped mains transformer in reverse to get the high plate voltage.
Online I see mainly square wave drive designs, but I think this is a bad idea for a high gain valve preamp, so I set up a simulation.
I plan to use a XR2206 or ICL8038 for a basic sine wave generator.
I like the idea of using a sine inverter and can run the transformer at a higher frequency for easier filtering and adjust voltage by a potentiometer after the function generator IC.
This design outputs 130Vac. does this look like I'm on the right track?
I want to make a inverter powered by a 12V wall-wart adapter to power valve heaters with the low side then drive a center-tapped mains transformer in reverse to get the high plate voltage.
Online I see mainly square wave drive designs, but I think this is a bad idea for a high gain valve preamp, so I set up a simulation.
I plan to use a XR2206 or ICL8038 for a basic sine wave generator.
I like the idea of using a sine inverter and can run the transformer at a higher frequency for easier filtering and adjust voltage by a potentiometer after the function generator IC.
This design outputs 130Vac. does this look like I'm on the right track?
Could this be simplified further by just using power op-amps, something like a TDA2030? something like:
Seems to me like it could work well as the chips will have built in protection and very low DC offset, providing they can provide enough current to operate in bridge mode like this. 🤔
Seems to me like it could work well as the chips will have built in protection and very low DC offset, providing they can provide enough current to operate in bridge mode like this. 🤔
Last edited:
You need feedback to level control the sine wave. It's much easier to implement with a chip designed for the purpose. Analog Devices has an entire series of low-noise SMPS chips and gate drivers.
Voltage regulation is not very important for my application and could be done on the high voltage side, or by attenuating the incoming sine wave to the op-amps. Are these chips capable of driving a standard mains transformer in a linear fashion (sine wave)?
Last edited:
One of the advantages of using a SMPS transformer is that the higher switching speed allows for both smaller transformer AND lower value capacitors. You can get a mains transformer to work to a few hundred Hertz.
As long all the switching speed/modulation stays out of the audio band it is a fine solution. But mains transformers are everywhere and much easier to find.
Truth be told I'm not to sure when it comes to SMPS transformer selection, I have only ever built a single coil boost converter.
Truth be told I'm not to sure when it comes to SMPS transformer selection, I have only ever built a single coil boost converter.
The issue is with the rapid current transition --- di/dt -- which causes harmonics. These can have unintended consequences like modulating various PN junctions. If you are using mains frequency you don't get as much efficiency, but you also don't get the harmonics.
Try experimenting with a gate-driver chip like the SG3524 or SG3525. You can post-filter and regulate the rectified DC to eliminate the switching transients.
As always when dealing with HV, be aware of lethal voltages.
Try experimenting with a gate-driver chip like the SG3524 or SG3525. You can post-filter and regulate the rectified DC to eliminate the switching transients.
As always when dealing with HV, be aware of lethal voltages.
This one cannot work without a negative supply for the opamps.
More than 20years ago, I built a multifrequency sine inverter (50/60Hz) using PWM; I may still have some documentation (paper-based obviously).
I'll try to find it, but don't expect too much....
I will include a .33Ohm resistor to the center tap for current sensing. Do I need flyback protection diodes in parallel with the windings?
Worth a shot?
Now, the opamps require a supply voltage twice that of the power section.
Flyback are a no-no, but recuperation diodes across the transistors could be useful
Flyback are a no-no, but recuperation diodes across the transistors could be useful
Can you clarify on this? it will be operating in bridge mode over a 12V transformer.
In operation, the emitters of the power transistors will swing to ~2*Vsupply. Since their bases need to follow them, their drive needs to be able to track them.
Make a sim, it will clarify the situation
Make a sim, it will clarify the situation
I don't follow, simply put, the transformer is a mains EL type 12V with a center tap, meaning 6V across each coil, if anything the input needs to be scaled down. yes there will be losses in the opamp swing and about 1.2V loss in the base to emitter voltage but the supply voltage more than makes up for this.
Frankly... th idea of using a tube pre-amp to add tube distortion to an entirely solid-state and mainly digital audio chain is a strange preference already. But you are not the only one. Then to convert a DC voltage to 12V AC to power those tube heaters to get the characteristic 50 or 60 Hz filament hum seems to be downright off-track. Tube amplifier designers went a long way to avoid this and have even used DC filament supply for pre-amp stages.
To answer your question. Look in 1960-ies and 1970-ies technical articles about sine wave generators for measurement purposes. Those often include provisions for frequency stability (not that important) and amplitude control. In you case important because you need a fairly stable HV. Which can only be achieved using some for of feedback given the Hi-Z of your inverter circuit.
Also be sure not to deviate too much from the transformer specified frequency. Losses are more than you expect.
To answer your question. Look in 1960-ies and 1970-ies technical articles about sine wave generators for measurement purposes. Those often include provisions for frequency stability (not that important) and amplitude control. In you case important because you need a fairly stable HV. Which can only be achieved using some for of feedback given the Hi-Z of your inverter circuit.
Also be sure not to deviate too much from the transformer specified frequency. Losses are more than you expect.
To utilize the transformer properly, you need to apply a negative half-wave of ~12V to each half-primary, but even if you content yourself with 6V peak, those 6V will oscillate about the +supply, because of the cross-coupling between the primaries, and/or the inductive kickback and the need to reset the core.
It is inescapable, and if you are not convinced, convert your schematic into a working sim
You must misunderstand what I am asking. I don't know where you got the notion that I am powering the heaters with AC. It is a 12VDC adapter.
And whats digital about vinyl and a class A output?
I am keeping the SMPS approach open, as long as there's not a problem with noise and/or EMF interference but am struggling to find a suitable transformer, can you maybe point to one on digikey?
This is because I have the center tap wired at +12V or ground and not at 6V? however this will not be the case on my #2 post as both sides of the transformer are at +6V and swing opposing around that center point?
Yes, BTL will work, as will an inductive PP taken from the drains instead of sources (like a tube PP), but it will require some kind of feedback or amplitude control.
I managed to find the remnants of documentation for my 50/60Hz sine inverter, but unfortunately it is too messy to be usable; even I have difficulty piecing it together and making it readable enough to be published would require a lot of work, which would mostly be wasted since it mixes a 50kHz carrier with with the LF modulation
I managed to find the remnants of documentation for my 50/60Hz sine inverter, but unfortunately it is too messy to be usable; even I have difficulty piecing it together and making it readable enough to be published would require a lot of work, which would mostly be wasted since it mixes a 50kHz carrier with with the LF modulation