Looking at the photos, you appear to be using countersunk screws on the transistors. These will cause damaging stress on the large plastic packages. Must use cheese heads with bigger load spreading washers.
I fear you have damaged more than just the chipamp.
I fear you have damaged more than just the chipamp.
Cyberzim,
Very very important thing to remember to power up a newly developed amplifier is to insert 10 ohm 1/4W resistors in series with the power supply!!! If there is a problem in the circuit only resistors burn rest will be saved.
Once the voltage drop on these resistors (10 x expected quiescent current) is equal to the expected value, you can remove these resistors and continue further tests!
Hope this technique helps in your next build.
Thanks,
Routhun
Very very important thing to remember to power up a newly developed amplifier is to insert 10 ohm 1/4W resistors in series with the power supply!!! If there is a problem in the circuit only resistors burn rest will be saved.
Once the voltage drop on these resistors (10 x expected quiescent current) is equal to the expected value, you can remove these resistors and continue further tests!
Hope this technique helps in your next build.
Thanks,
Routhun
Great tip! Thank you!
BTW, is there any books you will recommend for designing amplifiers especially with Class-AB? I feel like a newbie in this section or maybe I'm not just experienced enough.
Regards,
Simon H.A.
BTW, is there any books you will recommend for designing amplifiers especially with Class-AB? I feel like a newbie in this section or maybe I'm not just experienced enough.
Regards,
Simon H.A.
Shouldn't that be 10ohms 10W or more ? 10 ohms 1/4 W would burn out in the blink of an eye irrespective of whatever state your amp is in.
Yes, but I'll guess you will have enough time (a few seconds) to read the current so you can conclude whether or not if there is a problem with the amplifier.
NO, no, no,
the 10r resistors drop a small voltage if the amplifier is working properly and the amp does not mind the slightly lower Rail voltage.
If the amp, for whatever reason, draws a higher current, then the 10r drops more voltage and the amp gets a lower rail voltage. This may reduce dissipation in the amplifier devices and thus allow time to notice the excessive current demand. Turn off and find out what is wrong.
If the amp draws very high current then the 10r severely overheats and acts like a very slow blow fuse.
Use a 250mW resistor, not 600mW and definitely not a power resistor.
I would much rather uses a fast blow fuse. But even that cannot protect semiconductors from a gross wiring error.
The best is to test via the bulb tester at every stage of assembly. If the bulb comes on it automatically reduces transformer voltage to a few volts up to a few tens of volts depending on how badly the amplifier has been assembled.
a 230:35+35Vac transformer will give +-50Vdc when the test bulb is not glowing.
If the bulb is brightly lit the PSU is likely to have <+-5Vdc feeding the amplifier.
The fast blow fuses do not need to blow. The amplifier is protected.
There is at least one situation that the bulb tester does not protect.
A working amplifier comes up to voltage and the bulb that initially glowed briefly now goes out. The PSU has charged up and everything is working just fine. Now stick a test probe into the amplifier and let it slip shorting two pads/leads together.
The PSU instantly discharges all it's energy across that short. Bang ! Smoke !
The bulb has not protected against operator misuse.
I've done it! My new layout seemed to be a success. The amp can now play music.
My first impression is that the sound is quite good, but I cannot test it yet until I've hook it up to a decent sound system (it's 2 AM at my place right now).
It only draws 0,5 amps when not amplifying.
It's dead silent when connect to my laptop. And that's with a not so well power supply.
I couldn't find any cheese headed screws so I'll apologies for the countersunk screws for mounting.
Thanks for all you help everyone 😀
Simon H.A.
My first impression is that the sound is quite good, but I cannot test it yet until I've hook it up to a decent sound system (it's 2 AM at my place right now).
It only draws 0,5 amps when not amplifying.
It's dead silent when connect to my laptop. And that's with a not so well power supply.
I couldn't find any cheese headed screws so I'll apologies for the countersunk screws for mounting.
Thanks for all you help everyone 😀
Simon H.A.
good to hear you got it working , btw your power supply capacitance looks a bit on the lower side , I think you need at lest 10000uF X 2 per rail, per amp . Your bridge rectifier should also be rated at 35 amps.
I always use a bulb in series , it gives a very good indication and offers good protection indeed but my series protection resisitors were always 10W types. Although the quarter watt types would offer protection as you said but they would burn out with anything other than no load, you should use a quarter watt type only the during the first power ups and then if you don't find any major faults or smoke switch to 10W types.
No, test with 10r 250mW, If passes set up with 1r0 250mW and set to final output bias.
When all is working perfectly replace the 250mW resistors with fast correctly sized fuses.
I use old fuses with 1r0 or 10r or 100r resistors fitted in lieu of the fuse wire. Can swap between fuse and test resistors at will.
BTW,
200mApk passing the 10r resistor is ~400mWpk the resistor will not blow for a music peak.
That 200mApk allows ~160mW, continuous, of music/sinewave output. Many listeners use that as "normal" listening level.
Ehh, sadly but true now I got some problems. It worked perfectly when I posted my reply and then I just turned it off and went to bed. Today it acts weird. It only draws 10 mA and when I play music you can only hear a very crunchy low sound of the music even at full volume.
Any suggestion what might have caused that?
Regards,
Simon H.A.
Any suggestion what might have caused that?
Regards,
Simon H.A.
Uncleaned PCB, possible solder trace runs, disconnect speakers, and test impendance to make sure they were not damaged, before proceeding.
🙂 Very crunchy ~ highly distorted with limited current draw i.e. 10 mA seem like a driver problem.Check the driver tansistors' collector voltage it should be roughly around 1v less than your rail voltages, if its equal to your rail voltages that means either an open or underbiased driver.
Problem solved! One of the NPN power transistors was damaged so it was always on. I replaced it and also adjusted the bias and now it works!
Only 70mA quiescent current and nice and clear sound 😀
So let's hope I wont get anymore problems 😀
And once again thanks for all your help!
Regards,
Simon H.A.
Only 70mA quiescent current and nice and clear sound 😀
So let's hope I wont get anymore problems 😀
And once again thanks for all your help!
Regards,
Simon H.A.
God damn! It did it again! When it played music it suddenly started getting crunchy in the sound. I've almost used up all of my parts so I'll think I close this case and work with something else. This will cost be too much money.
Regards,
Simon H.A.
Regards,
Simon H.A.
You shouldn't quit now 🙂 , since you've got this far at least find out why you're getting simmilar failures after some time. Do you have a multimeter connected to the amp at all times ? Take bias readings after you've played for some time. May be your Vbe multiplier isn't mounted in the correct place.
Before playing any audio through the amp make sure the bias is stable after 20-30 mins of operation.
that is why you should use a 'Tried &True' Prototyping Path before committing to real
fabrication...
fabrication...
- Status
- Not open for further replies.