I am not strong in electronics, so please explain yoru responses. I have a 24VAC electronic controller that is monitoring something critical. When power drops out, the generator picks up after a few seconds, but the controller loses power and has to reboot, causing several downstream things to fail-safe. To alleviate this, we are going to add a small 120VAC UPS to the 120VAC->24VAC transformer that provides 24VAC to the controller. Unless I purchase a REALLY expensive 'always on' UPS, here will be a 3 millisecond switchover time, and from experience I know that it's not uncommon for the controller to behave erratically during this time.
The controller only draws around 50VA, is there something simple I could do that would keep the 24VAC voltage up during that 3 millisecond switchover time?

 

24V AC disappears down to zero a hundred times a second. Are you sure it's not a DC supply you want to keep operational?

 

24V AC disappears down to zero a hundred times a second. Are you sure it's not a DC supply you want to keep operational?

There's probably bridge in the controller with a cap or two.
In this case I would try and modify the controller, tap into an area that is the DC part, then feed DC power to that in a fail scenario. After that making a small backup PSU (24vdc or whatever is needed) with a relay as a switch should be easy.

 

There's probably bridge in the controller with a cap or two.
In this case I would try and modify the controller, tap into an area that is the DC part, then feed DC power to that in a fail scenario. After that making a small backup PSU (24vdc or whatever is needed) with a relay as a switch should be easy.

If it's rectified AC, there will already be a capacitor to keep it operational for 10ms.

 

If it's rectified AC, there will already be a capacitor to keep it operational for 10ms.

Maybe that's why 5ms is the goal with switching in a backup power source? A little buffer.

If the controller has any IC logic in it, the input AC is rectified to DC.

 

Update it's 50VA-80VA

 

Is it possible for you to increase the DC filter capacitor value in the controller as suggested, as that should solve the problem.

But it's strange that it would drop out from a 3ms transfer time, which is much less than one AC cycle.

 

Since you start with 120 VAC I would just do as you mentioned and include a UPS in there. Let the UPS drive your 120 VAC to 24 VAC transformer. When considering a UPS I would make sure the UPS outputs a TSW (True Sine Wave) and not the low end MSW (Modified Sine Wave) output. Don't make this complicated. Since you mentioned 120 VAC I assume 60 Hz North American power. Just know your load and use a UPS which will deliver enough power. I also don't understand the 3.0 mSec dropout? As mentioned a full cycle of North American 60 Hz power is 16.6 mSec and a half cycle is 8.33 mSec.

Ron

 

24V AC disappears down to zero a hundred times a second. Are you sure it's not a DC supply you want to keep operational?

Yeah, the controller needs 24VAC. Honestly I think it will ride through, but the owner wants added assurance. But they want a $100 UPS, not a $1050 UPS.

 

Since you start with 120 VAC I would just do as you mentioned and include a UPS in there. Let the UPS drive your 120 VAC to 24 VAC transformer. When considering a UPS I would make sure the UPS outputs a TSW (True Sine Wave) and not the low end MSW (Modified Sine Wave) output. Don't make this complicated. Since you mentioned 120 VAC I assume 60 Hz North American power. Just know your load and use a UPS which will deliver enough power. I also don't understand the 3.0 mSec dropout? As mentioned a full cycle of North American 60 Hz power is 16.6 mSec and a half cycle is 8.33 mSec.

Ron

Reloadron, There are three levels of UPS, offline, line interactive, and double conversion.
Offline or Standby - keeps the battery charged but passes input power direct to the output. If power drops, it shifts a relay and applies the inverter output to the UPS output. That switch time is 3-10 milliseconds depending on manufacturer.
Line Interactive - same as above but has a voltage regulator on the UPS input power to better protect downstream equipment. Still has that 3-10 millisecond switch time.
Double Conversion - converts 120V UPS input to 24V DC on a DC bus, then converts that DC back to 120V AC for the UPS output. Batteries are connected to that DC bus in the middle, so there is no switching time if input power drops out, the batteries supply the output inverter instantly. Zero switch time.

 

Is it possible for you to increase the DC filter capacitor value in the controller as suggested, as that should solve the problem.

But it's strange that it would drop out from a 3ms transfer time, which is much less than one AC cycle.

@crutschow You just turned on light bulbs in my head... you are right, 3ms really should be negligible. I was off a couple orders of magnitude in my head. I am going to go back and push for the cheap UPS and some field testing. Thank you!

 

Lets ask the basic, what controller is it?

If you can wire in 24vdc as a source, all that's needed is 24v lithium batt, a small batt maintainer, a fuse, and a 24vac SPDT relay. When the AC drops out the relay switches and voila, instant 24vdc from a batt. When AC comes back on the relay switches back to AC power source and the batt maintainer charges batt back to full. Can probably do all that under $100.

If you want 24vac, need to add a DC-AC chopper between relay and device, then just parallel the two 24vac sources.

3ms should be an "eternity" for a small Omron mechanical relay.

Side note, if the controller input stage is a bridge rectifier, then you can likely just feed it 24vdc.


Many controllers ask for vAC so that the source is a basic xfmer, no PSU needed outside of the controller, etc.

 

AT this point two things are clear: First, that the hold-over time needs to exceed the UPS takeover time, and second, that we need more information. "IFF, and only if," the controller only uses the 24 volts AC rectified, THEN, a large capacitor storage scheme can work. BUT if the controller is also using the 24 volts to feed outputs, then probably you need a 24 volt AC backup. THAT backup could be an inverter with a 24 volt AC output, powered directly by the rectified mains power, and supported by a capacitor bank. That functionality will be very similar to the high cost UPS.
So the TS needs to examine the controller circuit to discover which it is.

 

So the TS needs to examine the controller circuit to discover which it is.

Which controller is a key question.

 

Which controller is a key question.

I am referencing the important controller that the UPS is supplying backup power for. The one that must be protected from any power loss of any length of time.
I was not aware of any other "controller" in this discussion, so I did not bother to identify it more specificly.

 

I am referencing the important controller that the UPS is supplying backup power for. The one that must be protected from any power loss of any length of time.
I was not aware of any other "controller" in this discussion, so I did not bother to identify it more specificly.

Yes, we know that. The question is for OP, what controller make/model is it?

 

OK, now other questions, because, SOMETIMES, the controllers use some of that input power for internal circuits. AC Phase control that regulates AC outputs. So certainly more information about the controller is required. If there is a block diagram in the manual that came with it, THAT will be very handy. And if the very most prefect backup is required, use an inverter to supply the 24 volts at 3 1/2 amps, and power the inverter from a battery that has a charger that provides both the float charge and the inverter power, so that the effective battery load is always zero. Then, if the AC mains power fails, they have a day or so to grab a backup generator to keep the battery charged.
The actual cost depends on what the controller is controlling. As an example, if it were the factory producing nitroglycerin, I would have redundant controllers PLUS a backup generator ON HAND. With some processes there is no second chance!!

 

use an inverter to supply the 24 volts

I hear all your concerns.

24vac it is. Do as I said, but just inline this device (or like it) between the 24v batt & relay and the controller.
https://www.powerstream.com/inv-24vdc-24vac.htm

Voila, DIY UPS on a budget, and it should work very well, and cost little.