Paralleling audio amplifiers with demo that failed 

John, I enjoy these videos they are very well made and interesting, I love designing audio amplifiers, and building them. The results will very, however, I am just learning and what a great help you are... keep them coming and have a good day....

John, thanks for another great video! Your rationale for the choices you make is very informative. It was especially helpful when you explained why the 1% resistors and matched feedback and gain was important to keep current from flowing from one amp into another. Great work!

Wow, great video! I've learned more from this fail as if there was a straight success because it shows the problems with paralleling power amps. It is a pity that this kind of channels will always strive because of advanced content for a few advanced audiences. Thanks for continuing your great work.

I've been designing and building amps for a while now. Most of my amps are 75W into 8 ohms, a geometric mean Class AB design I developed originally for a line of guitar amps I sell. Early on I used breadboards, but very quickly I found out that it simply wasn't worth the time trying to debug circuits with any amount of current or voltage involved. I'd actually argue that it's much faster to do with perf board because you can be virtually guaranteed of good connections, symmetry, etc... I will say that there's absolutely no reason to go to a full PCB if you're prototyping. Some of my longest running prototypes are built from perf board and I see absolutely no reason to replace them. I have a 150W (2x75W stereo) hi-fi unit on my desk that's been running for 3 years now. Of course my commercial amps are all built using PCBs with mostly 0805 components. But there's no better way to experiment than with a nice stable solid piece of perf board and a good roll of solder. :) If you're wondering how I managed to design a 75W power amp that uses no power resistors and all 0805 resistors, the answer is "with some difficulty" and lots of simulation. Hitting 0.01% THD at 20kHz wasn't too hard once I had the power dissipation in the resistors under control. The power amps are double-sided load with all SMD resistors and caps on the back side, and through-hole parts on the top. Each power amp is around 2"x2" and they're designed so that only the output devices need to touch the heatsink. They use no emitter resistors on the output devices which actually has huge advantages for driving higher current loads and bridging. I recently bridged 2 of my 75W units and measured over 275W into 8 ohms.

11:00 so can it be fixed using diodes at the output of the amp ?

You SHOULD definitely add a gofundme link and/or Paypal donate button! It's not begging or anything. Just add a couple links in the description and mention it in the middle of your videos. People love to be a part of something and as long as it's easy,(easy being the key word) I'm sure a small percentage would like to support you. I would! Come on y'all are you with me? Keep those videos coming John!

I have used multiple output transistors in parallel many times with there own emitter resistors for high power into low ohms, but I have never tried completely separate power amplifiers before. Thinking of stability issues. Two many variables in complete amplifiers. I think it's probably better to use one complete amplifier with multiple output transistors, making sure that you have enough drive available. Maybe a IC is not the best solution.

Hi John love your vids teaching me lots. Just wondered if you ever made another vid on paralleling amplifiers?

21:17 why not used external wire for lower the ohms

Is it more correct to add output high current transistors to one of those existing audio amp ic and we eliminate a lot of unnecessary gain and out of phase issues?

As long as I know, the only chip-amp can be paralleled is TDA7293. One chips act as master and others as slaves. Cheers from Indonesia

this fault cause of power supply.if you use one power supply to drive two chips oscillation occurs.use two one ohm resistor higher wattage and connect two cap like470uf after resistors.and repeat it for other 1875 chip to isolate each chip.in old radio circuits you might see large cap for each chips power supply like head pre amp has one 220 cap followed by 120 ohm and display circuit has anothre 100 cap followed by 22oohm ect.

John, all you have to do is get some copper clad board. With as few components as you have here, it is relatively easy to use an Xacto to define the various nodes in your circuit. After you cut the copper, you heat the metal and peel it up with the Xacto blade to create individual nodes. And copper clad is cheap. It is electrically very good for a single metal plane. We used to do this if we couldn't wait for fabbed boards to come back. In fact if you look at the TI datasheet, you'll see a couple layout examples which were hand drawn based on our copper clad versions.

hey John, did you ever do a video where you go in to how you do all the calculations, and the formulas you use?

It could be failing because the current is feeding back up to the other op amp and creating oscillation via its feedback network. You could try to increase the .1ohm resistors to something that when combined adds to greater than oner resistor + the load resistance, say 5 ohms. The load will "take" more current, even if another amp goes to short. There's a TI application sheet called "Paralleling Op-Amps-is it possible?". Their solution is to use a single feedback resistor tied straight to the load and wire the second op-amp in a slave configuration with unity gain. hth

Theres another good analogy with the trains in parrallel .

How many watts would you recommend those .1R resistors should be?

Hi john, the lm1875 can run @70v VDC (± 35 vdc)? I have a transformer dual 25v rectified ±35v...

It would have been nice if in addition to putting the speaker between the channels to look for imbalance, you had put your o-scope between the channels instead...

just one amplifier +ve signal amplification another one -ve signal amplification so you can connect input push pull amplifier method some time its working...🎓

Could you llist circuit diagram in 2:54

why is it important to use metal film resistors?

try grounding the heat sink.