Questions about single-phase Automatic Transfer Switches (ATS) (AP7750 & AP4450)

[BillP]

First let me state that I'm not sure, which forum section to post this, so if this is in the wrong place, please move the post and/or advise me, where to repost this.

Recently my 2-year old SMT1500C started failing repeatedly, dropping output power each time and I eventually contacted APC for a warranty replacement. While APC's support was extremely good (I received a replacement unit two business days later and could ship back my old unit in the box of the new unit), the failures themselves were quite annoying, because each time the server/networking gear in my home office had an unexpected shutdown/crash causing lengthy data scrub jobs on my storage server.

So I started looking at an Automatic Transfer Switch (120V, 12A to 15A), and it seems that something like the AP7750 or AP4450 would the right for that purpose. I would connect one input source to the SMT1500C UPS as preferred source and the 2nd to utility power as the backup source. So in case the UPS fails, I'll have at least utility power.

As said I am looking at an AP7750 or an AP4450 unit (possibly even used unit from ebay or a similar source), but before I pull the plug, I have a few questions:

- I understand that the AP7750 is a discontinued model. What else are the differences between the AP7750 and AP4450 and what would be a reason to pick the latter above the former? From the manual I didn't get any indication as to what sets them apart.

- The startup and installation guide of the AP4450 ( https://download.schneider-electric.com/files?p_File_Name=990-5852D.pdf&p_Doc_Ref=990-5852_EN&p_enDo... ) mentions on page 9 (internal document number) in the section "Connect the ATS" that ATS models are only compatible with a double-conversion UPS and not a line-interactive UPS. Since my SMT1500C is of the line-interactive type, I am wondering, what exactly makes the ATS incompatible to that UPS model. I understand that line-interactive UPS have a wider range of voltage outputs, since they essentially forward utility voltage as long as it is in range.

But on the same document page multiple possible source configurations are listed including using utility power from two different utilities. And I would expect that utility power in general has a larger voltage range than even a line-interactive UPS.

So what exactly makes the AP4450 incompatible with the SMT1500C? FWIW the AP7750 manual that I could find ( https://download.schneider-electric.com/files?p_File_Name=ASTE-6Z6K9J_R3_EN.pdf&p_Doc_Ref=SPD_ASTE-6... ), didn't state any such limitations for that model.

Thanks in advance for any advice.

[Brad_C]

All I can add is I have an AP7721 with an SUA3000 in the same configuration and it does what it says on the tin. I can power the UPS off without notice and it switches to line power, and I can shut the main breaker off and it stays on UPS.

I've never dropped the load during transients excepting one server that had a failing PSU that held up less than about 8ms (I'm in 50HZ land and I tested it).

I'm sure there are reasons as to why it's not a great idea, and I could imagine if you were on a gird that was prone to regular disruption you might end up with toasted relays (lots of small and fast changeovers), but I'm not into the 5th year of this config and I've seen no issues. A quick check of the logs of genuine power blips or outages sees the device report "source b is not available" / "source b is available". No inadvertent switching.

One anecdotal data point for you.

[BillP]

Brad, thanks so much for sharing your experience with your ATS. That's along the line of what I would expect and that's why I'm perplexed that they have this "compatibility restriction".

I'm hoping that someone from APC can clarify this.

I could imagine if you were on a gird that was prone to regular disruption you might end up with toasted relays (lots of small and fast changeovers)

Actually that shouldn't be a problem, at least not in our setup, when one source is on UPS and the other on line power. Because the line voltage would be varying just as much as the UPS voltage, so there would be no reason to switch from UPS input to line input. And when the grid has outright disruptions, the UPS would do the switching (as it does now).

If I read the manual for the AP77xx ATS series correctly, you can configure the acceptable voltage range pretty freely, is that correct?

My SMT1500C UPS is set to 120V nominal, with High Transfer Voltage set to 127V (lowest possible) and Low Transfer Voltage set to 106V (highest possible). So as long as I can set the voltage interval for the ATS a bit wider, it would never need to switch, unless the UPS fails.

[Brad_C]

On 26/7/2020 7:57 AM, Bernhard said:

If I read the manual for the AP77xx ATS series correctly, you can configure the acceptable voltage range pretty freely, is that correct?

Yeah, I have mine configured to +/- 35v on 230Vrms (so 195-265V) and the sensitivity set to "low".

I still have some 188 source change events over the last 5 years, a good proportion of which I can't correlate to UPS events or manual source changes (13). I figure given the sources are always in sync it won't me much of an issue, and I'm sure the relays are rated for more than a couple of hundred cycles.

[BillP]

On 7/26/2020 2:14 AM, Brad said:

Yeah, I have mine configured to +/- 35v on 230Vrms (so 195-265V) and the sensitivity set to "low".

On my SMT1500C UPS the sensitivity doesn't affect the voltage range at which the switch to battery is triggered, only the speed with which it is triggered (so a "low" setting would skip very short voltage drop-offs). Is that the same on your ATS?

I still have some 188 source change events over the last 5 years, a good proportion of which I can't correlate to UPS events or manual source changes (13). I figure given the sources are always in sync it won't me much of an issue, and I'm sure the relays are rated for more than a couple of hundred cycles.

That's pretty weird. Assuming you have set your source from the UPS as the preferred source, I would have expected the ATS to never change sources, unless the UPS dies. But as you said: 188 times in 5 years should be well within the rating of the relays (at least I hope so).

[Brad_C]

On 27/7/2020 1:30 AM, Bernhard said:

That's pretty weird. Assuming you have set your source from the UPS as the preferred source, I would have expected the ATS to never change sources, unless the UPS dies.

I originally thought it might be boost/trim events but I can't find any that correlate in the NMC logs on the UPS. I can certainly imagine scenarios where a quick transient of some sort might trick the ATS into switching over and then back. Those wouldn't ever occur on the output of a full on-line UPS. Anyway, I see the warning is in the manuals for the new 44 series, but not in the manuals I have for the 77 series.

[BillP]

Hey Guys,

The main difference that a user would notice between the AP44xx and the AP77xx is that the AP44xx uses an integrate management card based on the NMC2, it's also a newer design with better electronics but from a day to day perspective the management is the main difference. Also the AP77xx is will no receive any further firmware updates, we'll endeavour to provide technical support for the older units as long as they are functional but it will be best effort.

As for why ATSs only support online (double conversion) UPSs, I think it's best run through a scenario, I'll lay out some assumptions first, you only have one mains feed for both ATS and UPS, an ATS had a transfer time of between 6-16ms, a line-interactive UPS has a transfer time of about 5-12ms. The setup goes, Mains >> UPS >> ATS A >> Load and Mains >> ATS B >> Load. ATS is set to use A as its primary source.

Mains go out, after 8ms the ATS switches to source B (mains) as the UPS has not kicked, 8ms later ATS switches back to source A as the UPS now has become active. Now the ATS and the load have both been 16-25ms without power, there is now a good chance of the load dropping, also all capacitors (and anything else that requires a charge) in the system will now be empty, there will be a massive inrush of current which can blow fuses in the ATS, UPS and the devices, it can cause UPS overloads and I have seen cases where it trips breakers. Even if that doesn't happen your shortening life of your UPS, ATS and power supplies.

TLDR, it's too unpredictable, will reduce the life of your equipment and not supported.   

[BillP]

Gavan,

Thanks so much for your response and taking the time to provide this explanation. There are two aspects that aren't quite clear to me, yet and I hope you don't mind two follow-up questions:

1. In the scenario you describe you state:

Mains go out, after 8ms the ATS switches to source B (mains) as the UPS has not kicked,

Why would the ATS switch to source B here, since ATS source B is also connected to Mains (which just went out in this scenario)? So there really isn't a "good" source to switch to.

2. The AP7750 / AP7750A models don't appear to have that limitation of "no compatibility with line-interactive UPS" listed in the manual, as far as I could tell. Is that just an oversight or are those models not as affected by this than the AP44xx series?

[BillP]

Hi Bernhard,

No worries, that's what we're here for.

On point one, at this stage this ATS won't know that source B isn't live as the power supply connected to source B will still have a charge, it was ready to take over but it won't have anything drawing on it to reduce its charge.

On point two, that's roughly what happens with any ATS regardless of the type or model, (or possibly brand) I believe there was a white paper that detailed this when that model was on sale but I can't find it now.

-Gavan

[BillP · Accepted solution]

It's in the following FAQ:

https://www.apc.com/us/en/faqs/FA156201/