Class D amp using TL494

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I did it already, i ommitted the curr lim circuitry.

I discopvered that 74HCT14N that o bought was only for 5v, so i had to reconfig the TL494 so i get one output inverted from it, it works. The amp as a unit does not work at all yet, the bootstrap diode (1N5820) gets very VERY hot, the fets get alittle hotter than pee and the IR2010 stays alitte over room temp. This far all i get is loud noise but i have managet to get som barely hearable distorted bass.

I have chose Rt/Ct to make it run about 100khz for now.

I take the 2v reference from the TL494 5vref out via a pot, and the have the feedback loop as is, but with a pot instead of resistors.

When you simulated it in switchcad, what values did you have on the output filter cap and coil and what osc freq ?
 
Tekko said:
I did it already, i ommitted the curr lim circuitry.

I discopvered that 74HCT14N that o bought was only for 5v, so i had to reconfig the TL494 so i get one output inverted from it, it works. The amp as a unit does not work at all yet, the bootstrap diode (1N5820) gets very VERY hot, the fets get alittle hotter than pee and the IR2010 stays alitte over room temp. This far all i get is loud noise but i have managet to get som barely hearable distorted bass.

I have chose Rt/Ct to make it run about 100khz for now.

I take the 2v reference from the TL494 5vref out via a pot, and the have the feedback loop as is, but with a pot instead of resistors.

When you simulated it in switchcad, what values did you have on the output filter cap and coil and what osc freq ?

Oh boy, you are very fast indeed. Please try to put the 2v ref at the center of supply. verify if it is really 2v. If it is more than 3v then it is approx. above 90%duty which is not good. If it is 1v is also no good co'z the duty is maybe less than 15....Just make it exactly 2v, this will give you approx. 48% duty..

I did not simulate this to switchcad..I just tried it on my breadboard + headphone and without inductor and without IR2010 and fet..Just bare TL494 and the sound is nicier than my son's discman. the carrier frequency is 150khz. Can you try to put the headphone directly to c7?. This should work coz,this is the configuration that I did before. Please recheck your circuit carefully and I will try to analyze what youve said regarding the 5v for 74hct14n. this may be the culprit. maybe this cannot handle the o/p voltage of TL494..I'll be back

regards
ledmania
 
I did one try with the 74hct14 and then rearranged the TL494 connections to dual output like in a halfbridge smps.

I dont have a split ps yet, i just use a puter ps, for the 2v ref i have set it to 2v, for the feedback loop it does not react at all when i turn the pot back and forth (makes for 0-12v), also when the TL494 is connected to the IR2010 theres no response from the audio signal i inject. After the first try when using the 74hct14 i hooked up the scope and then i got something that looked like a overmodulated pwm signal and even tho the freqs were far from each other it looked like when u do pwm with two signals thats just a few hz from each other.

To be honest this design does not appear to be a working one. all i get is a bootstrap diode hot enuf to fry pork and spikes on the upper output of the ir reaching up close to 50 volts. The fets get about 50deg after a few mins, only once i managed to get deadtime allowing the fets to turn off before the other turns on.

I have tried connecting a speaker and even tho the thing sends out 10v pulses the freq is so high that all u hear is a faint high freq beep, when unkloaded theres a sinewave with about the same amplitude as the power supply peak-peak that with a speaker sounds like distorted mains hum together with white noise and the faint beep.

I have checked the TL494 oscillator and its a really beautiful error free sawtooth waveform, the squarewave output from the outputs has the typical high freq bjt output ringing.

I doubt that i choice of pcb design could cause the thing to not work, i have seen a fully good working class d amp built on phenol paper protoboad here on this forum, a similar design but not with a TL494, it had high speed opamps as comparator and oscillator, it was kindof a VCO with a lowpass filter.
 
Tekko said:
I did one try with the 74hct14 and then rearranged the TL494 connections to dual output like in a halfbridge smps.

I dont have a split ps yet, i just use a puter ps, for the 2v ref i have set it to 2v, for the feedback loop it does not react at all when i turn the pot back and forth (makes for 0-12v), also when the TL494 is connected to the IR2010 theres no response from the audio signal i inject. After the first try when using the 74hct14 i hooked up the scope and then i got something that looked like a overmodulated pwm signal and even tho the freqs were far from each other it looked like when u do pwm with two signals thats just a few hz from each other.

To be honest this design does not appear to be a working one. all i get is a bootstrap diode hot enuf to fry pork and spikes on the upper output of the ir reaching up close to 50 volts. The fets get about 50deg after a few mins, only once i managed to get deadtime allowing the fets to turn off before the other turns on.

I have tried connecting a speaker and even tho the thing sends out 10v pulses the freq is so high that all u hear is a faint high freq beep, when unkloaded theres a sinewave with about the same amplitude as the power supply peak-peak that with a speaker sounds like distorted mains hum together with white noise and the faint beep.

I have checked the TL494 oscillator and its a really beautiful error free sawtooth waveform, the squarewave output from the outputs has the typical high freq bjt output ringing.

I doubt that i choice of pcb design could cause the thing to not work, i have seen a fully good working class d amp built on phenol paper protoboad here on this forum, a similar design but not with a TL494, it had high speed opamps as comparator and oscillator, it was kindof a VCO with a lowpass filter.

LOL!!!. You are a very adventurous guy! LOL!!. Of coarse it will not work as expected because you split the o/p as double ended. this will happen if you pull up pin 13 to vcc. It must be pulldown to ground to ground to act as "single ended".I think I will encounter difficulty in trouble shooting your circuit co'z you are making a DC to DC converter and not audio amp LOL!LOL!LOL!....almost all of the critical requirments in the sch was violated by by you LOL!LOL!!.....OR beter yet just leave this project for a while and when I'm in Manila I will be the one to act as the shock absorber of this beast. After all, I'm the perpetrator of this mess anyways.LOL!!!!LOL!!LOL!!

Regards
ledmania
 
ledmania said:


LOL!!!. You are a very adventurous guy! LOL!!. Of coarse it will not work as expected because you split the o/p as double ended. this will happen if you pull up pin 13 to vcc. It must be pulldown to ground to ground to act as "single ended".I think I will encounter difficulty in trouble shooting your circuit co'z you are making a DC to DC converter and not audio amp LOL!LOL!LOL!....almost all of the critical requirments in the sch was violated by by you LOL!LOL!!.....OR beter yet just leave this project for a while and when I'm in Manila I will be the one to act as the shock absorber of this beast. After all, I'm the perpetrator of this mess anyways.LOL!!!!LOL!!LOL!!

Regards
ledmania


Tekko said:
That was very unneccessary by you.

Cheer up Tekko! :D

Tekko said:
By the way i tried hooking up headphones to the TL494 output as you said accordingly to the schemo, it souded awful. It was just like badly underbiased class b amp with tons of crossover distortion. On the scope it was just alot of spikes.

interesting.

there is probably a lot of potensial for improvement i imagine.
(like throwing away the 494 :D )
 
Tekko, Honest advice, you are wasting your time and effort in making this project because I can sensed that your level of skill is not fitted with this circuit. Your statement in all your post above indicates that you are making a double ended PWM DC to DC converter but instead of putting a rectifier from the o/p you terminate it with a speaker intead LOL!LOL!LOL!
That is a very funny bug that enters within you LOL!LOL!LOL!

Honest advice again, Open a new thread and put a subject title,,,,"DC to DC coverter with TL494 using ledmanias schematic" and I will participate there also LOL! I'm sure that if someone wants to laugh then they will just visit your thread LOL!LOL!LOL!
 
I gonna make you eat that. Atm i run the thing with audio, without a IR2010 tho, fets directly connected to the 494 via the gate RD circuit. On idle it makes a beeping sound and with music lots of distortion as well as weird noises due to a crappy output filter. Yes its hot enuf to fry an egg.


Btw the TL494 i use is broken so it does not give PWM, just steady pulse train with othyer crap signals around it representing the audio signal.
 
http://user.faktiskt.se/Tekko_Tubeamp/D-Amp/schematic/tl494 class d idea.JPG <-- heres a schematic i have drew together according to the tips so far today, as the cross conduction on the TL494 outputs is so minimal (turning on and off happens exactly at the same time) so i think the IR2010 internal delay is enough to delay the fets timings to prevent cross conduction.

I have already tried this but without the IR2010 and this is the first one that works. I get sound with lots of power, tho its distorted and mixed with alot of switching noise but that is probobly because the TL494 have to work the fets gates by itself.

Apart from the noise and distortion the sound is actually not that bad, i ran fullrange to a 4ohm woofer, the thing runs on 12v from a at ps and i have a 2200µF cap between the output and speaker to take out the 6v dc offset.
 
Eva, it is simpler on a breadboard, no soldering.

I have to undo and redo soldering to make changes, so it takes time, especially when removing an IC when all you have is a 15 watt iron.

What was the circuit and the switching freq on that one ?

I switch my "dc-dc converter" D-Amp @ 100khz, atleast according to the TL494 RT/CT chart.

This is what the thing sound like: http://user.faktiskt.se/Tekko_Tubeamp/D-Amp/Without IR2010.AVI
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.