Router and Modem Question

You were wrong last time you said that, and you are still wrong. Of course, nothing will protect from a transient from a direct lightning strike, but transients of those potentials are not common. Most destructive transients are no where near those potentials and it is those lessor transients a "GOOD" UPS with AVR can quickly compensate for.

You are also incorrect about your power cycling issue. While true, a graceful power cycle on a computer should not damage the hardware - graceful shutdowns are not the point for using a UPS. Data loss/corruption, or preventing it, due to ungraceful shutdowns due to ANOMALIES is the key.

Semiconductors operating between high and low inputs typically deal with differences in potentials of just 5V DC. For example, in a typical integrated circuit, a high (1) is represented by +5VDC while a low (0) is seen when 0VDC is applied. And transistors switching between those low level DC voltages (with µAmps of current) is totally different from powering up with line voltages (110VAC and several amperes) from a total off status.

Any student of electronics knows the power requirements when starting electronics from an off state is much higher than when simply running, and that potentially destructive transients often occur during that time until all circuits compensate and stabilize. This is also why many systems stagger startups of major components as the power demand of everything starting up at once may exceed the capabilities of the supply circuits - circuits that are fully capable after that. Transients from a full start is also the precise reason high power amplifiers in audio equipment delay output to the speakers when first powered on. It is common for electronics to isolate circuits during power up, until stability is reached.

Totally NOT true. Perhaps on a $300 motherboard, but that can not be expected with a typical $50 board. And I note, it is the computer power supply, not circuits inside the computer that must deal with anomalies coming down the AC line. The ATX Form Factor Standard for PC Power Supplies requires a "hold time" of just 17ms - meaning, a power disruption of more than 17ms will result in the PSU dropping power, crashing the computer, and potentially corruption of data. Also, regulation in a computer power supply has nothing to do with regulation with the router or modem - the topic of this thread.

 

Greetings Bill.........Please recommend a few "GOOD" UPS with AVR.

TIA

 

As Bill says in his sig, "Heat is the bane of all electronics!" Consumer level modems, routers and switches do not usually have fans, and users often restrict ventilation through placement. In that situation, running constantly is destructive, because equilibrium temperatures are too high. I have that stuff in small rack with old 120mm case fan running at 7.5V.

I am stuck with dirty and unreliable mains power, so I use online double conversion UPS with generator backup. Both MGE Pulsar and Tripp Lite SmartOnline have worked well.

 

Well, certainly APC is probably the best known brand. I would suggest the BR1000G, 1300G or 1500G. I have the 1500G on my home theater system.

Tripp Lite is also a well known brand. I have also used CyberPower with no problems.

There is a downside to UPS and that is the batteries need to be replaced about every 3 years - unless you get one with Lithium Ion batteries - but they are VERY expensive. Fortunately, this is something the user can do. I buy my UPS batteries from Apex, BatteryWholesale, or Batteryplex, depending on who has the best price at the time (factoring in shipping, which can be significant, but sometimes free).

Good point. For this reason, I have my router, cable modem and switch on top of my desk hutch and not shoved in a cubby hole. I also have a ceiling fan to keep the air moving.

I also use one of these extension cords to plug all the power blocks in, then plug the cord into the battery side of the UPS (most UPS have several outlets on the battery side, and several outlets with just surge and spike protection for non-essential items like a printer). I like these cords because power blocks can get warm too and these cords allow me to spread them out.

And I think mirimir means "ambient", rather than "equilibrium".

 

How often are temperatures inside your electronics exceeding 300 degrees F? Even Intel CPUs operated at 170 degrees C without damage.

Temperature is not hardware destructive as so many are told by hearsay. Temperature causes temporary timing and threshold changes. A system then crashes. Lower the temperature. Those changes return to normal. Then a system works just fine.

Temperature is a diagnostic tool. How to find a defective semiconductor long before its warranty expires? Operate household electronics at an ambient100 degrees F. If the product is defective (fails temporarily at 100 degrees), then it will probably create hard failures months or years later at 70 degrees. Heat does not create the failure. Only identifies it.

Electronics have no fans when a fan is unnecessary. Too many install fans to cure symptoms. Heat identifies a defect. Rather than fix or replace a defect, many would cure symptoms: more fans.


Cleanest or dirtiest mains power must result in cleanest low voltage DC. What happens inside computers? 120 VAC is first converted to DC voltages that exceed 300 volts. Then power is 'dirtiest'. Converted to high voltage, radio frequency spikes. Superior circuits inside electronics convert that 'dirtiest' power to clean and stable 3.3, 5, and 12 VDC. Circuits already inside every computer already do better what some recommend a UPS for.

All electronics are so robust as to convert 'dirtiest' power (typically generated by a UPS in battery backup mode) to the cleanest and most stable DC. A need to 'clean' AC mains with a UPS is silly since 1) cleanest power is just made 'dirtiest' inside electronics. 2) Then superior circuits inside electronics convert that 'dirtiest' power to cleanest.

Had he known how electronics work, then he would have never posted:

Another example of myths due to not first learning how things work. Another example of claims by ignoring relevant numbers. Motherboard obviously was not the "circuits already inside every computer".

ATX standards now require computers to withstand a thousand volt spike without damage. Interface semiconductors to meet IEC 61000-4-2 international design standards now withstand 2000 and 15,000 volt transients without damage. These well published requirements are learned from numbers.

Another example: If a UPS is outputting 'clean' power, then the UPS manufacturer provides numbers that say so such as %THD. Anyone recommending a UPS for cleaner power would provide that all so important %THD number. Subjective claims are why so many UPSes can output some of the 'dirtiest' power. Then subjective claims hype that as 'clean' power.

A voltage so low as to cause incandescent bulbs to dim to 40% intensity is perfectly normal voltage for all computers. In fact, the ATX standard says, "The power supply must be able to start up under peak loading" when voltage is that low. Electronics already make most AC mains anomalies irrelevant. Bulb dimming to 40% intensity is an ideal voltage for all computers. Another example of anomalies made irrelevant by existing hardware.

Best is to power off computers when done. So many fears (dirty AC mains, excessive heat, destructive power cycling) are only subjective. Therefore are urban myths based in junk science reasoning.

 

Come on! Get real and stop wasting our time, okay weston?

Yeah right. Even if you meant 170°F, that is 76.7°C which would make any experienced computer user, certainly any technician, nervous. But more importantly, it would cause stability problems and system crashes which are never good as that leaves many open files, lost clusters, and possible corruption of mission critical data on the hard drive.

If you meant 170°C, then that just proves you don't know what you are talking about as that is 338°F - well above the boiling/melting point of many materials.

What? You are just making stuff up!!! Why? You are making yourself look like a total fool! Again!

300V? WRONG! There is nowhere near 300V in a PC, or a PC's power supply - not even in countries with 240VAC mains. PSUs step down, not up.

More made up stuff. And again, you ignore the power supplies for network devices and other peripherals.

OMG! This is getting hilarious! And totally stupid! Really, there's no nicer way to put this. That was a stupid and incorrect comment. If that were true, we could all get by with $10 PSUs.

That is simply an out and out fabrication! I posted the ATX Form Factor standard in a post above so we all can see that no where in it does it say ANY thing of the sort. You are simply making stuff up - and everyone can see that.

Why are you tell falsehoods? I don't get it.

WRONG. That standard deals with static discharge, not transients coming from the AC power source. Do you really not get it, or are you intentionally trying to deceive us?

You clearly have no clue what you are talking about and are attempting to look smart with falsehoods and obfuscations, and I refuse to be dragged down to your level. So I am out of here.

 

@westom

Some years ago, I wanted cheap hardware for Linux servers. Used Gateways were very cheap. So I got one and installed four 1TB SATA drives. After setup, it seemed sluggish, and I saw that drive temperatures were about 70C. After drilling several holes in front cover and adding 120mm rear fan from junked server, drives ran at 35C. Now I have several, and they have all been dependable.

What you say about high temperatures revealing latent hardware defects may be true. However, I prefer workability to seeking perfection.

 

There is no disputing this - whether mechanical, electrical, or electro-mechanical, excessive heat taxes all systems. It is a simple matter of physics. When matter gets hot, it expands. When it gets too hot, it expands beyond tolerances. In mechanical devices, that causes excessive wear (friction) which can compound the problem with more heat, and eventually, a breakdown occurs. A defective part will fail prematurely - hence the reason for stress tests and "burn-in" testing. In motors, like drives and fans, the bearings can seize.

In electronics, there is a general property that says the larger the conductor, the less the resistance through the conductor. There is also Ohm's Law that says E=IR. That is, electromotive force (aka voltage) = current x resistance.

So if R goes down (a "short" for example), and voltage remains the same, current must goes up. And when current goes up, so too does heat because there are more electrons running through (and banging into the walls) of the conductor. If this cycle of more heat causes more current and more current causes more heat is not controlled, "thermal runnaway" occurs, until the weakest component fails, or until some circuit protection device kicks in.

But there's the rub. More advanced (and expensive) electronics (like CPUs) have advanced circuit monitoring and protection circuits that will shutdown the device or circuit BEFORE the heat becomes excessive. Cheap devices, like budget power supplies, inexpensive DC power blocks on routers, are lucky to have a fusible link - at best.

In any case, for electronics, unless you hit the device with a hammer, it is always heat that takes it out. Whether it be excessive heat due to poor cooling management, too much current, a design or manufacturing defect, or simple age weakening the device, eventually the device can no longer handle it, and fails. Thus, "heat is the bane of all electronics!" For this reason, you will extend the life of your electronics if you keep it cool.

 

I have all my devices on 24/7 all the time. The bill is not that high.
The steady state is good for electronics, plus there's nocturnal activity going on.
Mrk

 

A nice tutorial on computer switching power supplies:

-http://www.hardwaresecrets.com/article/Anatomy-of-Switching-Power-Supplies/327/2

 

I forgot about that tutorial. Nice find. Note how it emphasizes that not all PSUs are equal, especially in how they handle, if they handle, transients. As noted in the article, many cheap "generic" supplies don't have transient filters. And it links to several other good PSU articles, including Why 99% of Power Supply Reviews Are Wrong. I have used this one several times to justify my position that testing a PSU with a multimeter is useless. You need an O'scope or power analyzer with the PSU under a realistic load, and this article explains why.

 

I am a firm believer in using a UPS for both the battery back up and the surge protection. A few years ago I did some research on surges. Then I installed a whole house surge protector. I still use individual surge protectors and UPSs.

 

Whole house protection is great for protection from outside the home - anomalies coming in through the grid. But it does nothing from anomalies coming from internal sources.

For example, I have not yet turned on my central heating for the coming winter. But just this morning, it got down into the low 50s outside, and the bathroom was a bit chilly. So we use the supplemental "space heater" when doing our "business" in the bathroom. Each time the heater cycles on and off, the bathroom lights flicker due to the heater sucking up the power - until the supply stabilizes. Flickers are a series of surges and sags (opposite of surges). Spikes and dips (opposite of spikes) generally happen too quickly (much faster rise and decay times) to be seen by the human eye as flickers, but still must be compensated for. And spikes and dips tend to reach much greater extremes.

ATX power supplies are expected to handle "expected" anomalies, as long as the sags are no longer 17ms, or the voltage drop is not below 90VAC (in countries where 120VAC is standard). And the better supplies handle the surges and spikes IF equipped with MOV devices. But there is NO requirement for ATX PSUs to have MOV devices to protect from incoming anomalies.

It is important to note that any high wattage appliance can, if malfunctioning (or on the same circuit as other demanding devices), dump excessive anomalies on the line. These include hair dryers, refrigerators, toasters, toaster ovens, microwave ovens, water coolers. And it these excessive anomalies that can cause instability issues with a computer.

And again, even if a PSU is capable of detecting an anomaly, shutting down out of self-preservation still causes the computer to come crashing down. It does not signal the computer to "gracefully" save your data, shutdown Windows, and power down the computer. Only a "good" UPS can do that. And of course, modems, routers and the like don't have sophisticated power supplies like a computer does.

As long as you don't plug one into the other. Most UPS manuals warn against that as the AC power from a S&S protector (if the sinewave has been "clamped") may confuse the AVR/Line condition monitoring circuits.