Full Bridge Induction Heater - first tests at 2.1kW
HTML-код
- Опубликовано: 5 окт 2024
- DE: Heute auf Englisch, mal als Experiment :)
At first sorry for my bad english, the video was unscripted.
Today I will show you the current state of my induction heater prototype.
I have been working on a universal H-Bridge PCB for a while. This will also be used in future Tesla coils and such.
Here are some additional specs:
Resonance capacitor: 8.4µF (84 x 0,1µF all parallel)
Maximum power draw about 2.1kW
Transformer material: N30 from TDK
The idea is to use the finished build for forging knives and tools as well as heat treatment.
It is planned to utilise a µC for controll of frequency and safety functions like over-current-detection in the H-Bridge circuit and over temperature shut-down.
Nice work so keep it going your project looks like it will work good for your application. Keep posting as you upgrade and fine tune your project. Thank you for sharing it with us!
The 0:06 (center-bottom) segment combined to your announced title both captured my attention besides some other details, like a mention of software control and even the heavy-duty task of metal forging.
Well, i once contemplated building my own DiY IH Driver but that was long ago and all which remains of it is an ability to entertain with theorizing...
For example, please allow me to urge you to read Allegro's ACS733 datasheet as it about a hall-effect current-sensor chip with built-in agile resistor-programmable comparator and a 0 Hz (D.C.) to MHz analog output. Noticing the pulse transformer i gather you's have direct access to a local reference clock and hence the means to implement quadrature demodulation, to detect signatures in a susceptor workload...
Relatively to the (5 ~ 6) toroïds stack i'd argue it wouldn't matter what power source(s) drive which copper windings, pleading in favour of magnetic power combination instead of interconnected flux-concentration yokes resulting in *1* single intense electrical stress factor. Only the resonator requires extreme performance as long as each toroïd is driven synchronously...
Now i can only guess what would happen if the number of magnetic vectors (between 2 opposite poles) were raise to 18 in a quadrature phase-coherent model that kicks a workload the same as a submarine sonar listening to echo pulses. Euh... Etc., etc. Though if ambitious and young then the sky is the limit, so good day have fun!!
Keep up your great work!
Your english is great. I'm curious what you're using to drive the control signal for the PWM. Voltage in the LC tank? Current in the tank (between the inductor and capacitors?)
Also, curious, what's the size of your capacitance and inductance in your tank / coil?
What igbt and what value dc blocking cal are you using?
Hello, i have build something similar using a μC that locks in the frequnecy. I went from full bridge to half bridge setup because the power draw was too much. I found that the matching transformer is the critical part of this kind of setup. My ferrite cores were heating up very fast at 2kw and that was a matching problem. I have gone from 1 turn to 2 turns in the secondary and around 18trn in primary and that solved the problem and the power from 2 kw went up to 6-7kw while the ferrite cores were at 80 degrees celcius. The matching transformer should match the impedance of the LC tank. Different materials result into different impedances of the LC tank, correct me if i am wrong. Do you have a ferrite overheating problem with theese cores?
Hello,
I haven't looked at the temperature of the transformer but I intend to measure and regulate the primary current of the transformer in order to not draw to much current and save the IGBTs and Diodes. I also want to measure the power from the grid twice a second with a PZEM-004T module. This way I hope to get good control over the Power draw of the unit.
But if you say I should, then I will also measure the temperature of the Transformer while in development. I don't think I could add another secondary because of the core diameter. I use 8mm copper tubing for the secondary circuit. I will upload some update videos of the induction heater.
How did you program the μC in order to lock the frequency? Do you have a video on RUclips where you show your unit? I would be interested to see it.
Best regards
@@dickesf8525 Its a good idea to measure the input power. I currently have an 100A AC meter that measures the line amperage. You mentioned the PZEM-004 module. Does this take into consideration the power factor? The microcontroller is the atmega8 that is programmed with Arduino. The microcontroller measures the phase difference between the capacitor voltage and the Inverter voltage. Theese voltages are fed into the xor gate of the pll ic. Different phases give different duty cycle of the output of the xor gate. The microcontroller reads that and applies a compensated voltage on the pll ic in order to keep the phase 90degrees. Its important to begin scanning from a higher frequency than the Fresonant. My current protection circuit works with a current transformer of the primary fed into a comparator. If the limit is exceeded the output of the comparator triggers a flip flop circuit that shuts down the pll ic. Lastly my only way to reduce the power is to detune the circuit (increasing the frequency above the resonance). I don't have a RUclips video yet, i will upload one though.
@@livewire2404 Yes the PZEM-004T measures the voltage, current and frequency and calculates the power factor, as well as the real power and also keeps track of the used energy. It is basically a complete power meter module with UART interface and Arduino library. I will use an STM32f303 controller though.
So you are using an PLL IC, probably the 4046, and then you are using the µC as the filter between the phase comparator and the VCO? And by adjusting the voltage into the VCO via software you are controlling the power of the system?!
I thought about implementing a software PLL but I am not sure if the 72MHz STM is fast enough here for the ~60kHz of my Oscillator. I will se how this will go.
@@dickesf8525 Yes, and the 4046 is providing the PWM pulses to the inverter. By adjusting the voltage on the VCO pin (pin 9) i can vary the frequency. The μC provides a pwm signal that it is filtered into a dc voltage to the VCO pin. The frequency of the pwm doesnt matter as long as its filtered to pure dc voltage. The duty of this pwm signal is propotional to the output frequency of the pll. You can skip the 4046 ic if you use direct pwm from the stm32. Make sure to calculate if the 16bit timer has good frequency resolution in order to adjust the frequency precicely in order to have a phase lock. I would recomend an external xor gate to measure the phase diference between the inverter voltage and the capacitor tank voltage.
I think you would need more capacitance on the DC after the rectifier. The ripple current is very high and it would destroy theese caps. Also with more stable DC voltage you would get more power out of it.
I am curious about your AC mains filter. Looks like some components of switch mode power supplies. Why is there an additional bridge rectifier next to the large blue capacitor?
Also i had used similar cooper tube iron couplers and they got red hot at power levels more than 3kw. You can change them out to brass ones. The iron ones add some inductance and alot of losses.
@@livewire2404 Isn't it also possible to use two signals, which are already 90° out of phase with the 4046?! If I am correct one could also measure the pahse difference between the capacitor voltage and the resonator current.
The STM32 has also one 32-Bit timer. I have to check again.
An external XOR Gate could work, but I should use one with Schmitt-Trigger, right?!
I only had the space for two of the smaller caps, but i left some space for bigger screw terminals so that i could add some external ones later.
The caps got a littel warm, I noticed that but it was ok. The main AC component should hopefully come via the snubber cap but I can only try and see now. Otherwise this would be an easy upgrade.
The mains filter is something I think of as mandatory and slightly displeasing at the same time because i just used parts from an old device that I still had and soldered them together the same way as before. I don't think this filter is working too well with the currents present. The Chokes got quite warm. The additional FBR is dead and was used before i bought the bigger one. A new Filter PCB is in the making.
The copper tubing would indeed melt but i watercool it and it stays at room temperature this way. The water however heats up quite fast and I need a bigger tank. The couplers are made out of brass and plated with crome for the looks. My father had them laying around and I took them because new ones for the 8mm tubing were quite expensive.
Please give the schematics. Danke.
Plz give schematic
Hello sir i need circuit diagram please sir help me
how to make feedback for this?
There are multiple ways for doing that. You can use a mixture of current and voltage feedback or current only. In general you are trying to get a measure of phase difference between two signals that represents your offset between driver frequency and resonance frequency.
The main difficulty is to measure the resonator current at resonance (or close to it), as it is not only a very high current but also at high frequency.
Toll Camarade. schön. µF contrôle de la fréquence + contrôle du courant électrique + contrôle de la température est essentiel et un projet sur le quelle je travail également. Je suivrai vos avancements en espérant pouvoir m'en inspirer pour ne pas dire de les copiers. Goten tag bis bald. Schuß.