| Kai Klaas 01/31/10 16:23 Read: 99 times Germany |
#172824 - It is very difficult... Responding to: Brian Reid's previous message |
...to safely handle this enormous involved power with a "tiny" protection circuit!
- 640V (450VAC) does not allow the use of Polysfuses! - A normal melting fuse must be able to handle 640V and to stop huge currents. So, a fast acting, sand filled 1000V fuse should be used, like those used in Fluke DMMs for instance: http://docs-europe.origin.electrocomponents.com/webdocs/0...279c60.pdf - A fat wirewound power resistor, like this here http://www.vitrohm.com/download/series/24/ could be used, which can withstand huge currents for a few milliseconds (when properly chosen!) until the fuse breaks. Eventually more than one must be connected in series to withstand a short to 640V. - A very very fat PTC could also be used, eventually in combination with wirewound power resistor(s). - One or several 5KP30A Transzorbs in parallel http://www.vishay.com/docs/88308/88308.pdf could be used to clamp the input voltage to a safe level. This 5kW Transzorb can handle a few kW during a few milliseconds! - A LM2574HV switcher (see the "HV"!) should be used, though. Make no mistake, I didn't say that it is simple to design something like this!! An example for a scheme working up to 120VAC can be seen here (figure 11): http://cds.linear.com/docs/Datasheet/178591fc.pdf But remember, a Polyfuse will not work in your application, as there doesn't exist any that can withstand 640V, as far as I know. Kai Klaas |
| Topic | Author | Date |
| LM2574-5 AC voltage protection | Brian Reid | 01/29/10 16:58 |
| of course not | Erik Malund | 01/29/10 17:06 |
| no AC protection | Stefan KAnev | 01/29/10 17:11 |
| Thank you both | Brian Reid | 01/29/10 17:34 |
| fire brigade | Stefan KAnev | 01/29/10 17:40 |
| suggestions? | Brian Reid | 01/29/10 23:45 |
| something like this | Stefan KAnev | 01/30/10 04:27 |
| already made | Erik Malund | 01/30/10 07:29 |
| just one more thing | Erik Malund | 01/30/10 20:06 |
| thank you | Brian Reid | 01/31/10 15:31 |
| It is very difficult... | Kai Klaas | 01/31/10 16:23 |
very small components are not recommended | Erik Malund | 01/31/10 16:41 |