Phantom pwr from a DC to DC converter?

:shock:

I'd prefer just to scope it and see what(if any) noise/ripple etc is there and adjust the passive components accordingly to nullify any effects.. of course i speak of all this implying a clean DC going into the converter to begin with as this is a dc to dc converter, not an offline converter.

but that is just me, the eternal minimalist.

A tube and transistor power trans, both with a 48 tap. Sorely missed. Wonder what the holdup is.

adjust the passive components accordingly

Click to expand...

You have to be careful when tuning a dc-dc converter. There is only so far you can push the values before you really start to mess up the control loop or other performance characteristics. For example, it seems to make sense to just keep making the inductor larger and larger until the spikes or triangular ac current waveforms disappear. Nope. The control will go to pot. It will go unstable. Same sort of thing goes for output capacitance. You will get much better performance by just adding an extra output filter.

power trans, both with a 48 tap

Click to expand...

The Hammond 300 series has a 50V tap.

jh

oh no hagtech i meant passive filtering after the circuit, a series R, another L maybe, tuned decoupling Cs etc.

[quote author="hagtech"]You have to be careful when tuning a dc-dc converter. There is only so far you can push the values before you really start to mess up the control loop or other performance characteristics. For example, it seems to make sense to just keep making the inductor larger and larger until the spikes or triangular ac current waveforms disappear. Nope. The control will go to pot. It will go unstable. Same sort of thing goes for output capacitance. You will get much better performance by just adding an extra output filter.[/quote]
Nobody says you have to adjust values arbitrarily and just hope for an improvement rather than a puff of smoke. The data sheets for these switcher chips have detailed information about how to optimize the circuit, including formulas for calculating necessary inductor values, compensation networks, and input/output filters specifically for the purpose of keeping the circuit stable within your defined range of operation. For example, if you look at the circuit Moamps made using the LM2577, you can see he used the inductor value shown in a generic circuit in the data sheet. He's lucky to have gotten reasonable performance out of it with that value. Rather than randomly tossing in bigger inductors til something bad happens, he could use the formula in the LM2577 data sheet and calculate the value he needs, which is considerably larger than the 100uH he's currently using. This isn't about voodoo or trial and error. It's about math. That's why they call it engineering.