Amps vs PCB track width using TO-220 package

[ruffrecords]

It could very well do 50A, the test was likely done with ~100us pulses

Except the datasheet says 50A continuous. IIRC it is even higher for a pulsed current. The key factor is the 50 amps is only specified when the transistor is fully turned on when its Rds is less than 0.02 ohms. The other problem is getting rid of 50W from a TO220 package without exceeding a Tj of 175 degrees. That probably means a chunky heatsink with fan or water cooling.

Cheers

Ian

 

[Analog_Fan]

Those are switching devices, the expected use case is very little voltage across the device when the current is that high. Linear amplifiers are a very different condition, you have to design for the device dissipation limit.

I'm also working on "function generator" PCB that generates a triangle @ line level audio.
Pretty neat circuit, using a integrator and some logic.

IC icl8038 (Harris) seams obsolete, just now I'm simulating it's transistor Sine wave circuit from the datasheet.
Falstad: https://tinyurl.com/28dnh7z2
Some weird resistor values, i can't seam to dial it in.
Otherwise i take the sine circuit from the Buchla 259.

But will also use static voltages to find it's maximum, i can't exceed the maximum of BD137, BD140.

 

[Analog_Fan]

Except the datasheet says 50A continuous. IIRC it is even higher for a pulsed current. The key factor is the 50 amps is only specified when the transistor is fully turned on when its Rds is less than 0.02 ohms. The other problem is getting rid of 50W from a TO220 package without exceeding a Tj of 175 degrees. That probably means a chunky heatsink with fan or water cooling.

Cheers

Ian

I designed the PCB in such way, huge metal "chunks" can be mounted.
You'll will see.
The guy around the corner gave me some coolers for free, he probably got from an amplifier, seams like 100 grams a piece.

I also wanna try to build a small induction heater with that board and another thing i wanna probe with transformers from switch mode psu's i got for free.

 

[RogerAF]

There's an old-school trick of stripping #12 Romex and using lengths of the bare copper wire on the PCB traces and soldered (usually done with a gun before adding any components), Not recommended for mass-market manufacture, but for a one-off? Works great.
Another old engineering trick is to never exceed 70% of any device's capacity. If you want to have 70A of current, then use devices with 100A rating. Given that your devices are rated 50A, limit your design to 35A.
A trick I saw Vari-Lites use was Boost Converters to get voltages up to +500VDC for routing around in the chassis (these were for the motor drive circuits). That way they could use smaller wires at lower currents. Then, when they needed the high currents to drive the motors, they would Buck Convert back to lower voltage. They used Boost and Buck converters with DC like transformers work for AC.

 

[Analog_Fan]

There's an old-school trick of stripping #12 Romex and using lengths of the bare copper wire on the PCB traces and soldered (usually done with a gun before adding any components), Not recommended for mass-market manufacture, but for a one-off? Works great.
Another old engineering trick is to never exceed 70% of any device's capacity. If you want to have 70A of current, then use devices with 100A rating. Given that your devices are rated 50A, limit your design to 35A.
A trick I saw Vari-Lites use was Boost Converters to get voltages up to +500VDC for routing around in the chassis (these were for the motor drive circuits). That way they could use smaller wires at lower currents. Then, when they needed the high currents to drive the motors, they would Buck Convert back to lower voltage. They used Boost and Buck converters with DC like transformers work for AC.

That sound like a experiment i though of some time ago to drive tubes from a +/- 12V main psu.
I haven't dived in to boost / buck converters yet, but i think there is a way from salvaged transformer from 5V USB chargers. a coupled of days ago, i drawn a new circuit witch is better than the one i made like one year ago.
Witch was somewhat inspired by a YouTuber called: HyperSpacePirate, now i refined it again although virtual.

This pcb is there for.

 

[vintageampfixer]

Power amplifiers for many applications use bus bars to supplement PCB tracks. A bus can be a simple piece of wire, or an intricate structure stamped from heavy copper sheet.

I've also seen processes where selected tracks are thickened using electrolysis to deposit additional copper onto unmasked areas of a PCB after its etched.

No one solution is perfect for all situations and good results are possible in prototypes and small volume production using TCW soldered flat onto the tracks, off-PCB wires to carry the higher currents, or direct wired connection to the transistor(s) in question to completely avoid very high currents in a PCB.

In any case, remember there’s a difference between volt drop in a short length of track or cable and that introduced by a longer run – or to put it another way, massive speaker cables might be great/necessary over runs of a few metres but you probably won’t need to use such big CSAs to wire inside the amp itself because the cable runs are significantly shorter.

The bigger issues I've tended to observe are keeping Tj low enough* to prevent early device failure and or performance loss (due to resistance) in the PCB tracks, switches, etc. rather than literally burning tracks off the PCB … which can and does still happen when things go wrong!

(* more to the point, finding / designing a heatsink large enough to be effective, small enough to fit and cheap enough to satisfy the bean counter ... I’ve found they're often sold with spare bubbles for spirit levels from the shelf next to anti-static wrist straps knitted from unicorn fur!)