Allright kewl, thanks. You tried volume control with yours?
Otherwise I'll just have to play around with it to see by how much i can adjust volume..lol
I'm having grounding problems again.
If i powered all the electronic chips from a +12V and -12V dual supply, with the mid-point of those two supplies as Ground.
By this I mean I power the TL494 from the +12V and for ground i pull it down to the mid-point.
The LM361N comparators run on dual supplies of +12V and -12V.
I've connected the Gate Driver to +12V and again for SD and Vss i pull them to ground to the mid-point of the +12V and -12V supply.
And for the MOSFETS i have another +5V and -5V supple, with mid-point as ground.
I reference global ground as the mid-point of the +12V, -12V supplies and have the speaker connected to the Mid-point of the +5V and -5V dual supplies.
Is this correct grounding?
Because when i remove the MOSFET part of the circuit, the current Limiting LED doesn't flash. However when i connect the MOSFET part and the +5V and -5V supply. One of the supplies has its current limiting LED on.
So something fishy going on with the MOSFET part of my circuit. Where i've shorted a connection I'm guessing.
If i powered all the electronic chips from a +12V and -12V dual supply, with the mid-point of those two supplies as Ground.
By this I mean I power the TL494 from the +12V and for ground i pull it down to the mid-point.
The LM361N comparators run on dual supplies of +12V and -12V.
I've connected the Gate Driver to +12V and again for SD and Vss i pull them to ground to the mid-point of the +12V and -12V supply.
And for the MOSFETS i have another +5V and -5V supple, with mid-point as ground.
I reference global ground as the mid-point of the +12V, -12V supplies and have the speaker connected to the Mid-point of the +5V and -5V dual supplies.
Is this correct grounding?
Because when i remove the MOSFET part of the circuit, the current Limiting LED doesn't flash. However when i connect the MOSFET part and the +5V and -5V supply. One of the supplies has its current limiting LED on.
So something fishy going on with the MOSFET part of my circuit. Where i've shorted a connection I'm guessing.
Your multi power supply approach is not fitted to your technical capability.
I can’t figure out that you’re already on the midst of success but suddenly you changed your course and use LM361 instead. If somebody persuades you to use LM361 in between tl494 and fet driver then let him do the thinking. But if it’s only you who decide to use this then I’ll bet that you can’t make it (I hope I’m wrong).
Honest advice, go back to your previous functioning circuit and explore how to implement deadtime. Do not go overboard something that will tax your brain because this will exacerbate you learning process and you will end up to nothing.
Yeah teacher had a look at the LM361N chips i had lying around and asked me to implement those.
The waveform i observed has the dead-time on one side but not the other, which is wierd because the Hin and Lin waveform going into the gate drivers are symmetrical with 250ns dead-time on either side.
Also my original circuit, it's a fall back option only. I don't wanna ruin your working model. Its a final option if everything else fails.
It's week 5 now, I have another 6 weeks left. I will waste this week to try and fix this problem, otherwise I'll take up the advise of using the XOR gate IC's.
The circuit works (sorta). But the modulation waveform I observe from theory is a +5V to -5V, two level modulation. But I don't see that, instead i see a 0 to +10V waveform. So something fishy.
Unlike before the ringing effect on the output waveform this time are small, in comparison to last time.
Again I'll see if the teacher can help me with this, for the remainder of this week. Otherwise I'll seek other options to avoid the multi-supplies for the electronic control.
Ok my teacher had a look at the atrocity i did and explained why i was seeing those waveforms across the CRO.
He's given me ideas on how to implement it now. SO I'll test it out and hopefully come back with good results.
But Ledmania, have you tried volume control in your design ever with TL494?
It'll be a bit settling to know if someone has already played around with it.
Cheers..
He's given me ideas on how to implement it now. SO I'll test it out and hopefully come back with good results.
But Ledmania, have you tried volume control in your design ever with TL494?
It'll be a bit settling to know if someone has already played around with it.
Cheers..
Another simple modulator would be around a (CMOS-) Schmitt-trigger. Such a design of a simple and (very low power) class-d amp can be found in the current issue of German Elektor. AFAIK the issues in the other languages and/or areas usually publish the same projects although not always in the same issue. The Elktor circuit is not very sexy - but it is a good example to demonstrate the working principle.
Regards
Charles
Allright I put a electrolytic Capcitor in series with the speaker to eliminate the DC offset problem i was having (Even though I thought i had 50% Duty Cycle).
Also I put the volume control circuit and the circuit works. The only side effect of this implementation is the distortion as a result of the high pass filter effect that capacitor introduces.
But otherwise I have a working class D amplifier. I still haven't braved to test it higher voltage levels. I'll do so with my teacher, so i can blame him if something happens (lol).
I watched Everybody loves raymond episodes with the speaker today, lol.
My question now is, can I implement your circuit as a Full-Bridge Ledmania?
Because a full bridge you just invert the music signal, such that one of the phase leg is 180 degrees out of phase with respect to the other Phase Leg.
So I put the music signal through an Op-Amp and i feed that inverted music signal into the input of the second TL494 I'll use and after that it's literally just the same circuit I have now in my Half-bridge.
My only hiccup with this idea is, how do i adjust volume then? Since now I'm directly adjusting volume via that 10k pot, but once i go full-bridge. One signal going into the first Tl494 is non-inverting and the other signal going into the second TL494 is inverted.
Don't really understand where I place the Pot and cap then to adjust volume.
Thanks for turning my uncomfortable situation with this project into something I can manage now.
Cheers.
Also I put the volume control circuit and the circuit works. The only side effect of this implementation is the distortion as a result of the high pass filter effect that capacitor introduces.
But otherwise I have a working class D amplifier. I still haven't braved to test it higher voltage levels. I'll do so with my teacher, so i can blame him if something happens (lol).
I watched Everybody loves raymond episodes with the speaker today, lol.
My question now is, can I implement your circuit as a Full-Bridge Ledmania?
Because a full bridge you just invert the music signal, such that one of the phase leg is 180 degrees out of phase with respect to the other Phase Leg.
So I put the music signal through an Op-Amp and i feed that inverted music signal into the input of the second TL494 I'll use and after that it's literally just the same circuit I have now in my Half-bridge.
My only hiccup with this idea is, how do i adjust volume then? Since now I'm directly adjusting volume via that 10k pot, but once i go full-bridge. One signal going into the first Tl494 is non-inverting and the other signal going into the second TL494 is inverted.
Don't really understand where I place the Pot and cap then to adjust volume.
Thanks for turning my uncomfortable situation with this project into something I can manage now.
Cheers.
Ok Nvm. Back to drawing board. Something shorted out and my electronic circuit is stuffed.
But basically i had a approx 50% duty cycle PWM out of the TL494.
And that PWM was then split into Hin and Lin, with 250ns dead-time on either side.
The Waveform that I observed at the output was a 0 to 10V switched waveform.
There really wasn't any ringing.
Gonna just stop mucking around with breadboard and move to veroboard now. My schematic works, so no point running around still playing with the breadboard.
But basically i had a approx 50% duty cycle PWM out of the TL494.
And that PWM was then split into Hin and Lin, with 250ns dead-time on either side.
The Waveform that I observed at the output was a 0 to 10V switched waveform.
There really wasn't any ringing.
Gonna just stop mucking around with breadboard and move to veroboard now. My schematic works, so no point running around still playing with the breadboard.
To all of your who build commercial amplifiers for a living. How does one go about testing the distortion in an audio amplifier?
Any particular equipments you can use to look at the FFT?
I don't care about Harmonics beyond 20Khz. But within the 20KHz range.
Any particular equipments you can use to look at the FFT?
I don't care about Harmonics beyond 20Khz. But within the 20KHz range.
Oh one other thing. If you insert between 0V to 3.3V positive into the DTC pin to add dead time. Can you insert negative voltage just below 0V to remove dead time?
Because i was mucking around today with the chip in Push-pull and connected them both as emitter-follower and had complimentary PWM, but the dead time is ridiculously huge (Like 700ns in rising edge and falling).
I would like to know if it you can reduce this dead time lower to say the order of around 250ns or below.
Because i was mucking around today with the chip in Push-pull and connected them both as emitter-follower and had complimentary PWM, but the dead time is ridiculously huge (Like 700ns in rising edge and falling).
I would like to know if it you can reduce this dead time lower to say the order of around 250ns or below.
No replies at all.
It's like I'm talking to myself..lol
My only other thought is to feed both Hout and Lout pulses from the Gate Driver into High Side Mosfet of One phase leg and Low Side Mosfet of the other phase leg.
It's like I'm talking to myself..lol
My only other thought is to feed both Hout and Lout pulses from the Gate Driver into High Side Mosfet of One phase leg and Low Side Mosfet of the other phase leg.
MY Bad I mean the two pulses from the output of my dead-time circuit are fed into 2 gate drivers. Where one pulse is fed into Hin of One gate driver while its fed into Lin of the other gate driver and vice versa.
TL494 CANNOT produce complementary PWM (ie, cannot generate inverted and non-inverted PWM signals). It can produce only usual PWM if you use open collector outputs mode and connect collectors together (with single pullup resistor), as phase_accurate said. You need to use XOR gates then to get complementary signals for mosfets driver (it is the best way to do it, and Eva knows it too 😉).
The dead time control (DTC) in TL494 is used only to LIMIT the maximal duty cycle to avoid shoot-trough effect in output transistors at maximal load. So do not missuse it! 🙂 And this huge (700ns) 'dead-time' is only result of incorrect PWM signal.
Also, I have not seen in this thread a single complete schematic ever. Usually even hand painted schematic helps much more, than many paragraphs of explanations.
Good luck!
The dead time control (DTC) in TL494 is used only to LIMIT the maximal duty cycle to avoid shoot-trough effect in output transistors at maximal load. So do not missuse it! 🙂 And this huge (700ns) 'dead-time' is only result of incorrect PWM signal.
Also, I have not seen in this thread a single complete schematic ever. Usually even hand painted schematic helps much more, than many paragraphs of explanations.
Good luck!
Last edited:
Ya that's how I've arranged it in my circuit.
The single PWM from the TL494 chip goes to a comparator. The PWM is compared against a fixed 2V ref to produce me complimentary outputs.
Those 2 outputs then go into a RCD network and a schmitt trigger circuit built around the LM361N to now give me dead-time adjusted PWM.
I drive that into my gate driver.
The single PWM from the TL494 chip goes to a comparator. The PWM is compared against a fixed 2V ref to produce me complimentary outputs.
Those 2 outputs then go into a RCD network and a schmitt trigger circuit built around the LM361N to now give me dead-time adjusted PWM.
I drive that into my gate driver.
- Status
- Not open for further replies.