Power Supply Constraints on a PCB

Thread Starter

Jon Tran

Joined Dec 31, 2015
2
Hi AllAboutCircuits,

I'm designing a board at work with some constraints that has me stumped.

I need to supply several voltages, but only have 6 connectors for power that can go into an oven for burn-in. I have another smaller 72-pin connector but that's used for some logic and data signals:

Here are the several power supplies I need:

[] +150 V
[] -150 V
[] +138 V
[] -138 V
[] -6 V
[] 6.5 V
[] 5.5 V
[] 3.3 V

I was brainstorming whether I can 'split' some of the smaller power supplies from one of the smaller supplies... using either Zener diodes or linear voltage regulators. Not sure how to implement it though. Should I use a multiple linear voltage regulator?

I searched through some of these links and forums for some insight:

Can I obtain a new voltage source from outputs of two cascaded linear voltage regulators in practice?

Power Supply Design - Multiple Voltage Regulators

The section I'm referring to is on number '6.'

Supplying 5.5 V and 3.3 V from a 6.5 V supply

  1. +150 V
  2. -150 V
  3. +138 V
  4. -138 V
  5. -6 V
  6. 6.5 V [] 5.5 V [] 3.3 V
Any help and advice would be appreciated.

Thanks
 

MrChips

Joined Oct 2, 2009
30,821
Don't forget your GND pins. Sometimes you need more that one GND pin depending on your circumstances and application.
I think you need more connections.
 

Thread Starter

Jon Tran

Joined Dec 31, 2015
2
Don't forget your GND pins. Sometimes you need more that one GND pin depending on your circumstances and application.
I think you need more connections.
Yes there is already a GND pin.

All the power pins are large gold finger type pads, including the GND pin.
 

AnalogKid

Joined Aug 1, 2013
11,055
For each of the eight voltages you have supplied zero information about the current requirement, regulation tolerance, cross-regulation limits, noise limits, thermal performance, what the load is, etc. Yes, you can derive 3.3 V at say maybe 10 A from the 150 V source with a zener diode. But there might be a better way than 1467 W of heat to reject.
Yes there is already a GND pin.
A ground pin? "a" as in only 1 ground pin for eight power supply outputs? Hmmm... That single pin has to carry the algebraeic sum of all eight output currents. That could be a large problem, but without more information it is hard to say.

ak
 
Top