Where did Google AI get into this? I certainly wasn't using it it, and Waldek Hebisch doesn't seem to have been either. He didn't tie his results back to any specific component that we could go out and buy, and I'd expect Google AI would have done that - if it was digging its data out of a large language model, there ought to be more examples drawn from real life in the data base than academic speculations.
I was asking google AI about some size specifications and satisfied its curiosity about a few details, which I'm beginning to think isn't usually productive.
However, it clearly showed an ignorance when it assumed it knew how the inductor would be included in the process because I never gave any details about that besides that the inductor would be there. Most likely it hooked it up in the worst possible place such as tying the mounted capacitor up to the ceiling.
I just want to briefly explain about the cube example. It was so obvious to me, with my having seen many many cubes what the proper answer was, but I put about ten or twenty questions to Google AI and it continued question after question to display its visual and geometrical ignorance.
Finally, none of you mentioned Google AI. I interacted with it. I hope this isn't a problem. I didn't realize there was some security classification to this material. Sometimes I split my stuff between Google AI and another AI in order to keep them in the dark because of their well known privacy issues.
No it w>
[Spelling errors corrected]If you aren't happy with solder, try Woods metal or some other low-melting-point alloy. In the 1950s, Philips protected their mains transformers with a neat little arrangment of leaf springs , sprung apart but held together with a stirrup made in two parts joined with some sort of LMP alloy. That worked at a much lower temperature than you need for this job.
The chances of a failure of the sort you are worried about are so small that it is not worth using an elaborate thermal switch (with its own failure mechanism), a simple melting alloy wire is foolproof.
The British Standard fool isn't as ingenious as ones you run into in real life.
Even more importantly, if you want to get regulatory approval for a product there are a lot of hoops to jump through. For example, fuses need to have a safety certification from UL or one of the other approvals agencies. A length of solder wire will definitely not have this, no matter how suitable it might be in reality.
When designing products that are to be sold the design has to start with decisions about which standards are to be complied with as this is so integral to guiding design decisions. I have worked in the medical, automotive, consumer and industrial areas and the same principle always applies. Start with the standards, then do the rest accordingly. John
This seems to be ridiculously over-complicated. It probably takes a couple of minutes to undo all the screws holding down the lid of the box. Once the voltage has fallen below 60V it is not considered to be hazardous according to most safety standards. Why do anything complicated when a very simple solution will get the voltage to a reasonable value in the time it takes to get the lid off? Using an inductor to speed up the voltage decay seems totally unnecessary. John
Mechanical safety interlock switch on box lid, applying 50R 50W wirewound or carborundum resistor. Audible or visible indicator across switched load.
Takes about 1 minute. Smaller R value gives shorter discharge time, but watch switch ratings and load surface peak temps.
RL
130 KG is a lot of stuff.
Engineers benefit from quickly discounting designs that are orders of magnitude away from being sensible.
John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics
Yes, the inductor idea is silly.
But even 60 volts in those caps, shorted, makes a gigantic bang.
All I need is a simple, ultra-reliable circuit to discharge the caps to zero volts in under two minutes.
And a lot of LEDs to show when it's not done.
John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics
My engineers and test techs need to measure and probe live, with the top cover off. And they may send a failed unit back to production for a fix. One thing that can fail is the discharge circuit.
200 volts into 50 ohms is 800 watts. I've seen wirewound resistors fail from repeated power spikes.And it's around 4K joules.
John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics
Not if you do this with a 4 pole AC relay:
AC line-relay coil-neutral return AC line-relay N/O contact-your electronics-neutral return
The other three N/C contacts connect the cap bank to the discharge load(s) as Liz mentioned.
If the coil fails its the same condition as AC off, which started the discussion.
We're not talking surface mount for the load elements from domestic appliances, so if that's viable then a big contactor should be, too. I've had to do something similar, but for for big honking industrial stuff where neither space nor budget was a factor. So I'm guessing this is not practical for your situation.
Ed
Don't be silly. The measurement is obviously in kangaroo gibongs.
John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics
As I have pointed out in a new thread, you can allow the inductor to saturate at the start of the discharge process. That lets you get away with quite a bit less iron and copper.
Rejecting the approach too rapidly has blinded you to that particular option. You've compalined about people poisoning brain-storming sessions by being too sceptical too early, though here you are just being intellectually lazy.
Kangaroo's I've seen. "Gibongs" apears to be a nonsense word.
My guess is that John Larin meant kilograms, for which the usual abbreviation is kgm, not KG.
Lazy engineers save their brains from excessive effort by telling themselves that. It isn't always true, but john Larkin lies to himself about that too.
Worrying about the exponential tail seems totally unnecessary too, but that's why John opened the thread. A non-saturating inductor would be too big to be all that practical, but my guess is that you can tolerate an initial period of saturation to get rid of the exponential tail.
John Larin is convinced that the inductor idea is silly - he didn't invent it so it can't be any good.
And a saturating inductor would probably do it (after it came out of saturation).
More demanding. You don't want the LEDs lit in normal operation. You might power them from a winding on the discharge inductor.
Were they rated for the peak power involved? Some data sheets are more comprehensive than others.
Quite a lot of energy, but not exactly intimidating.
Bypassing a safety interlock isn't encouraged and can have unfortunate results. Testing it and ensuring function requires sensible procedures and training. If you've got both, there's no issue. Providing suitable energy-limited test points can make their job SO much easier.
If only part of the installation is hazardous, it can be furnished with appropriate guards with their own interlock.
Both types mentioned are built for it. Consult the literature.
Carborundum, in particular, is used in surge applications.
50R recommended, because it is a common value in HV HF gear.
For 200VDC 4KJ? You're talking automotive 'pyro' .
Have something to add? Share your thoughts — no account required.
Ask the community — no account required