Power A/B Switch System: (2) AC source lines for auto-switching
I did not find anything that would address this question going through all of the topics, otherwise, please redirect. Thank you.
Currently designed is a N+1 redundant multi-PSU system with power balance distribution. Prior to the PSU backplane, additional fault protection needs to be designed to switch between two AC lines. If the primary path fails, it would switch to the secondary line. I wouldn't ask, but I have been researching this for over a month. The bottom line question I would appreciate support on is this,
"Can you run a solid-state relay with DPDT switching between the two AC lines to perform the above-mentioned task WITHOUT the dependence of some sort of controller (micro, PIC, or otherwise)? Can it be done by placing solid-state semiconductors & other circuit components to act as 'short' or 'open' when the main 120 VAC path is blocked?"
If it CAN be done, resource recommendations for this would be appreciated. Thank you!
Comments
Hello, I better understand what you were saying from before. This system works fine now with only solid state components. One problem, I need to have a delay so that AC2 can never be triggered until 8ms after AC1 fails. Basically AC1 & AC2 turn on simultaneously at start-up & AC2 needs some delay components, otherwise stuff will blow up.
The main requirements are,
* All components must be solid state semiconductors
* No Integrated chips/microprocessors can be used for functionality
* Must support 85-265Vac(rms) (120-375Vac) & 17Amps
Is there any suggestion about how to do this? Can a carefully chosen capacitor/transistor combo create that delay? I've attached a spice schematic to illustrate better the design. Thank you!
-Mike
"If there is a power above us, He must delight in virtue, and that which He delights in must be happy."
Hello power_ee_design,
You are on the right track. However, you must be aware of some details. For MOC3082 you are not allowed to apply more than 60mA to the emitting diode. You have to carefully choose the serial resistor R2 in order to limit the maximum current to a safe value, let’s say no more than 50mA. If maximum rectified voltage goes up to 400V, then you must increase R2 to limit the LED Trigger Current under this value (see spec file for max IF, http://www.fairchildsemi.com/ds/MO%2FMOC3083-M.pdf, page 2). This condition forces you to choose for R2 more than 8K.
The second problem is with low input AC, according to your requirements 85V. The minimum rectified voltage will be a little over 110Vdc. Now you must make sure R2 can provide at least 10mA to LED trigger current, see page 3 from the same document. This condition forces you to choose R2 less than 11K.
Combining the 2 conditions you can use a 10K resistor for R2. But there is a second bonus problem. For 400V input you will dissipate 16W into this 10K resistor only. If this huge amount of power wasted is acceptable, by any means, go for it. Anyway, you must take into account the second and the third opto-couplers you’re using and be prepared to burn away 32W of power to maintain two opto-couplers in ON state. You have to check though the specifications for PC817A to see what limits you have for controlling the emitting diode.
A better approach is to use a real power supply (switch mode?) and provide a low voltage at low power to implement the logic control.
And yes, for your actual question, you can use basic transistors + capacitors + resistors to implement your delay of 8ms.
Cheers,
PI_Crusher
Thank you PI_Crusher! I understand what you mean with the resistor values. Also, we plan to use a real power supply at the output of this source switching system. With the 'delay' that you said was possible, I have been having some real trouble with this. I don't know if I'm missing something or doing it wrong, but I can't seem to get it to work even remotely close to a delay. Might you supply a suggestion about that? I understand the RC constant for design values but an having a block in understanding this. Portion of reference schematic attached.
Thank you,
Mike
"...if there is a power above us, He must delight in virtue, and that which He delights in must be happy."
power_ee_design,
We can provide support for design problems related to our products. Unfortunately we do not have unlimited spare time. You have to pay extra attention to your logic control circuit, especially to voltage and current levels employed in your schematic, for any part. I see you use LTSpice, it is easy with this great design tool to check currents, voltages, power, and logic sequence. Do not despair, even if now seems like you have difficult design challenges to overcome, the whole exercise of trying by yourself to figure out what is wrong and how to correct your problems is going to help you more than any advise.
Cheers,
PI_Crusher
Hello power_ee_design,
You can use self powered relays on each AC1 and AC2 lines. The first relay connected to AC1 will be the “master” and will perform two functions: passing AC1 to your supply AND disconnecting the second relay. When AC1 disappears, the first relay will be disengaged automatically. As a result, AC1 will be disconnected AND the second relay will be engaged. This way AC2 is coupled to your supply AND the first relay disconnected. You have to take care of the initial “power-up” of your relays interface, eventually with a timer and a third relay to detect when the unit is coupled to AC1 and/or AC2, and to automatically select AC1 (or AC2) as master. Use a diagram to implement this logic with relays. You can look around for an electrician working with industrial electric machines (milling shop?), they can help you.
Cheers,
PI_Crusher