Aldi £59 petrol generator and television

Oh, how silly of me not to know that ...

Where did I ever dispute that such designs don't care about frequency or input voltage ? I merely stated that if you abuse them, they have a predisposition towards catastrophic failure - and I base that on many years of repairing the things every week for a living. Even if they do have a PFC front end, that in itself is just another SMPS, and the control IC and switching FET are just as prone to failure as in any other design.

Arfa

Switchers fail gracefully. The output voltage goes to zero. Analog power supplies are not so graceful when they fail. The output usually goes to the maximum of whatever the unregulated part of the power supply will deliver, which destroys everything connected to it. I like switchers.

--
Jeff Liebermann     jeffl@cruzio.com
150 Felker St #D    http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann     AE6KS    831-336-2558

"Jeff Liebermann"

** Only sometimes, in many SMPS designs the output goes high.

If that is on a 5V rail, it destroys most of the ICs.

Very few have a "crowbar" to stop it .

** Unregulated ones do no such thing. ** Just like SMPS do then, but far less commonly. ** But you are an utter ass.

.... Phil

Arfa Daily wrote

But clearly can't work out for yourself that they will work fine on that generator.

And using one of those on that generator doesn't qualify.

But clearly can't work out for yourself that they will work fine on that generator.

Wrong when the only downside of that generator is that it may well not control the voltage and frequency of what it generates very accurately. The SMPS won't give a damn.

Jeff Liebermann wrote

Not always.

Yes.

Not necessarily.

Me too, particularly when used on that sort of cheap generator.

No.

Both of these are an extraordinarily rare ocurrence in the real world.

Gareth.

.

You don't know that. For all you know, the output might be covered in crap like ignition noise. In which case, a lot of switchers will fail sooner or later. I don't know what your background is - I've never seen you on S.E.R. before so I guess you're here as a result of the cross-posting. For all I know, you might be the best SMPS designer in the world, and your efforts might be bomb-proof. But for every one of you, there are a thousand Chinese sticking their shit-street designs in everything from light bulbs to vacuum cleaners. I see this crap every day (as well as some good ones) and many of the designs do, as I said, work by the skin of their teeth. If you throw mains at them that's covered in spikes and other garbage, they *will* fail. It's not at all rare to see items such as TV sets that have been used on the end of cheap inverters and generators, with either blown power supplies, or corrupted EEPROMs.

I would also point out that many cheap Chinese designs without a PFC front end rated to 260 volts AC, have a single main filter cap rated at 400 v. Even with mains at the high end of 'normal', that doesn't allow a lot of overhead. If the output of this generator is poorly regulated, the input voltage to the power supply could easily take this cap close to its limits. In which case such a supply may well 'give a damn'.

So to the O.P., if you want to believe this guy and ignore Aldi's warning, go right ahead. You might get away with it if if the TV or whatever has one of Rod's designs of PSU in it. Otherwise you may well not ...

Arfa

Not likely. See

Yeah sorta. I could probably contrive a design that would do that. However, I haven't seen any like that since the days of the original Compaq portable, which did have such a PS. I've seen a fair number of computah power supplies in the last 30+ years. Very few that I've seen fails to full output voltage.

However, there are exceptions:

I think this one failed to full output voltage. It was in a no name PC. No evidence of lighting or power mains problems. It simply blew up. Proof positive that it is possible to design a lousy switcher.

Yep. Oddly, I haven't seen many such motherboards with all the IC's blown. Plenty of other failure modes, but no volcanic eruptions in the middle of chips. That's what full power over-voltage does. It blows out a small piece of the epoxy-B package.

None of the schematics that I've seen have such a feature. Most of them rely on a "power good" circuit inside the regulator chip to provide the necessary protection. That's not as good as an external crowbar, but then, I don't think any PC power supply manufacturers are going to add 30Amp SCR's for protection.

ATX power supply specification.

See Section 3.4 for the multitude of output protection circuitry required. Output faults are suppose to shutdown and latch.

True, but nobody runs unregulated power supplies these days. By analog, I meant linear power supplies with analog (non-switching mode) regulators.

In ham radio, the RF noise produced by switchers makes weak signal work rather difficult. Most ham radio power supplies use fairly simple linear power supply designs. The leading company is Astron:

Schematics can be found at:

Note that *ALL* of the linear power supplies have an SCR crowbar across the output. The slightest hint of overvoltage, and it will shut down and latch the regulator. This is rather irritating when the power supply is located at a mountain top radio site. Various reset circuits have been contrived. For example:

I don't know which flavor will best survive a crappy generator. I've seen switchers run on a much wider range of voltages and fequencies than linear power supplies. However, during ham radio Field Day, it seems that the switchers are the ones that are most susceptible to protective shutdown from generator spikes and transcients. Most of the power supplies running during last months Field Day were linear, to keep the RFI down, except for the switchers used to charge the laptop batteries.

Repeating again, I don't see any evidence of an overvoltage problem in the multitude of switchers that cross my bench.

Of course. But I'm still right.

--
Jeff Liebermann     jeffl@cruzio.com
150 Felker St #D    http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann     AE6KS    831-336-2558

Or to Vin.

So do boost regulators, not that it's necessarily a bad thing.

I have no preference. There is rarely a choice (I guess if there is it's an automatic "choice" for the linear - so no real choice).

If you want my honest opinion, I wouldn't use any crap devices that couldn't handle a noisy or shaky power line.. Do you really think the commercial power is a 100% ?

If the device can't handle a little frequency and voltage variation with a little noise on it, then the device was designed as a cheap, get it out the door garbage money maker.

P.S.

I have the 2000 Watt Peak generator that Aldi's has or did have. Yes , it is a little noisy however, frequency only varies +/- 3 Hz from what I could see and voltage was with in reason under load changes..

The wave form on it looked almost as good as my 5000 watt Honda generator..

So putting that into prospective, I guess if you think the Aldi generate is crap, maybe I shouldn't use my Honda either!

jamie

No, you need a bigger one to electrocute yourself, Maynard. Those small generators put out a crap waveform, and are intended for resistive loads. That Honda has a cleaner output but nothing like the AC mains used to, before SMPS became common.

Arfa Daily wrote

Yes I do.

And a SMPS won't give a damn.

Fantasy.

Leaves yours for dead.

Fantasy.

Its nothing like as common as you originally claimed.

But it doesn't just curl up and die when that happens.

Unlikely.

e

to

y

ld

hing

Maybe. A power line surge proved to be too much for the surge protector my old Toshiba CRT TV was plugged into.

Could it be a private labelling of this Einhell generator? I found a polyglot owners manual for it.

(...)

I've read a fair quantity of your messages on this topic. All of them are much like this one. It states your opinion, position, or pontification. There is no corroborating information, no examples, and no useful information. While you are certainly entitled to an opinion, I believe that you will be more effective if you kindly explain your position.

--
# Jeff Liebermann 150 Felker St #D Santa Cruz CA 95060
# 831-336-2558
# http://802.11junk.com               jeffl@cruzio.com
# http://www.LearnByDestroying.com               AE6KS

Jeff Liebermann wrote

I've read all of yours.

Yours in spades.

Everyone can see for themselves that that is a lie.

And you aren't, particularly with that absolute claim you made at the top.

I did that with the basics on whether SMPSs give a damn about the voltage within limits and the frequency in spades.

It is very common for the cheapo Chinese switchers used in items like DVD players, to fail in this way. They are extremely susceptible to secondary side filter cap failure - typically bulging and high ESR. This usually results in the rail voltage dropping. Most of these supplies take their control loop reference from just one of the output rails, and if it happens to be the one that's got the bad cap on - often the case - then the supply tries to crank it back up to the correct level, which results in all the other rails going way up. This often results in destruction of LSIs elsewhere in the item.

As a matter of interest, better quality multi-rail switchers used in some TV sets, do have a form of 'crowbar' in that all of the rails are monitored for output voltage, and if any become excessive, they trip a shutdown circuit on the main supply control IC. The Vestel 17PWxx series that were used in a number of manufacturer's TVs, are a good example of such.

Arfa

OK. You can shut up and piss off now. I've lost all interest in you and anything you have to say. Clearly, you are one of those know-it-all dickheads that pop up on usenet from time to time.

Feel free to pop back for another conversation some time, though.

After you've learnt to spell would be good ...

Arfa

Arfa Daily wrote

You have no idea that they did fail by overvoltage.

Doesn't mean that they failed by overvoltage, just bad caps.

Another sweeping statement that can't be substantiated.

Not even possible when it's the rail that powers the LSIs that has the bad cap.

That's not a crowbar. A crowbar shorts out the rail like a real crowbar would do.

But its not a crowbar.