re: 3 term. voltage regulators....in this case xx7924 (-24v)
A friend suggested that instead of the usual 1n400x diode as the protection diode (between In and Out terminals), that I could:
I do trust him, but just looking for a second opinion, and if this is a practice I should continue to use in future psu's.
=RR=
A friend suggested that instead of the usual 1n400x diode as the protection diode (between In and Out terminals), that I could:
I do trust him, but just looking for a second opinion, and if this is a practice I should continue to use in future psu's.
=RR=
redrabbit said:
There is nothing about 'opinion' in this. It's a straightforward isssue.
This works but is only necessary if there is a danger that the input/output differential will ever be larger than 40V. This can be the case with an input voltage that is higher than 40V and a large output cap. At switch-on, the input voltage may be there before the cap has had a chance to charge up.
Or if you have an output short with full > 40V input.
If that cannot happen in your supply, if the input voltage is never > 40V, it obviously doesn't do anything.
Jan Didden
ditto on jan's comment. Moreover, you have to realize that a 39V 5W zener can only drain 128mA without exceeding its power rating, so that protection is actually only good for transients - and if those are an issue (in car applications they are, for example) I'd rather put a transzorb right across the input. These are essentially zeners with a very high impulse power rating (like 1.5kW for a couple of microseconds).
I agree. But in case of switch-on transients that are essentially one-off events, the zener could take significant more current for the short time it takes the output cap to charge up so that the Vin-Vout difference goes below 40V.
Unless you hang multi-1000 microfahrads at the reg output, which he wouldn't, now, would he?
Jan Didden
Unless you hang multi-1000 microfahrads at the reg output, which he wouldn't, now, would he?
Jan Didden
wine&dine said:
Beg to disagree. If the input/output diff goes above 40V because the output cap is not charged yet, the current through the zener charges the cap, together with the regulator, towards the design output value. Nobody kills any load, where did you get that idea?
As soon as the Input/Output difference goes below 40V as the output rises, the zener stops conducting. That's all there is to it.
Remember, we are talking here about a situation where after the initial conditions, things are safe. For instance: Vin = 60V, Vout = 24V. The zener is there to get the system past initial switch-on, when Vout = 0 (cap not charged yet, that takes a finite time) and Vin is already close to 60V. That would kill the reg if the zener was not there.
After things have settled, the zener just sits there doing nothing. That's why I said that for these typical one-off events, the zener may be rated higher than the DC duty max current/dissipation in the data sheet.
Jan Didden
I do understand what you mean - I still don't like regulators with Vdiff(max) < Vin(max), you can't implement current limiting and there's the startup problem you're talking about.
My point however was that a surge at the input will be seen on the output, only less Vz, and that might still be too much for the load.
My point however was that a surge at the input will be seen on the output, only less Vz, and that might still be too much for the load.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.