As per thread title: needed to power the filaments in a (formerly) battery-powered radio. The mains transformer that is available has a secondary winding of 6.3 v.a.c. centre-tapped @ 1.5 amps. The d.c. O/P needs to be reasonably regulated since it powers 4 valves with 2v filaments. I could easily sit down for a few hours and investigate a few ideas but since I don't believe in re-inventing the wheel, I thought I'd ask here first: to me, this need seems to be a fairly common requirement in the world of vintage radio.
Thanks for your thoughts & ideas.
Al.
Sounds like a job for an LM317. The T version can handle 1.5A direct, more with a bypass transistor.
[
attachment=4468]
- Joe
Using an LM317, you might find that a silicon diode between ADJ and ground gets you to 2V. Well, if you had one that dropped 0.75V at 5mA. The 5mA is set by the recommended 240 ohm resistor between OUT and ADJ, but can be varied considerably, and you might be lucky enough to have a sensible diode that has a slightly higher than normal Vf at modest currents...
Personally, I'd be wary about a single LM317T - presumably the valves will take somewhat more during the first few seconds of operation? Heat sinking might be a pain as well. The die in an LM317T isn't able to run as hot as a power transistor. All of this paragraph is building up to reinforcing Joe's last two words on the subject
Tempting to add an SCR crowbar if those valves are valuable/rare.
All the best,
Mark
Thanks guys - your thoughts are of use and are valued. I must admit that I had overlooked the LM317 - I did think about 3-pin regulators but for some reason my mind was telling me that 5v was the lowest Vout in any device obtainable.
Something to think about then - and investigate - tomorrow . . . err, I see that that's now today, Sat. April 7th.
Time I went to bed!
G'night gentlemen!
Al.
In case anyone else is interested here's a way of using a bypass transistor that's good for 8A, subject to adequate heatsinking.
[
attachment=4469]
Joe
Thank you Joe: another valued contribution which saves me time & trouble.
Al.
Al,
I'd be inclined to put three big diodes and a fuse on the Output just in case it goes bananas.
Alan
I've come to a full stop on this. First, I tried the cct. with the power Darlington - and found that the output voltage was the same as the input voltage and zero regulation. So, suspecting that the (new) Darlington was defective, re-wired the cct. to the basic LM317 mode - as per post #2. Set output voltage to 2v and connected a 100-ohm load. Output voltage fell to 1.3v: regulation non-existent. This was with 7.5v d.c. from the reservoir capacitor, 20,000 uF. So, next: replaced that voltage source with +15v from my bench regulated source (good for 2 amps) and used the same basic LM317 regulator cct. Same result: no regulation. So, suspecting a defective LM317, replaced it: same result. Replaced the LM317 again: same result.
Can I
really have a batch of defective LM317 regulators? I don't understand what's going on here.
(For the purpose of this post, please ignore my 'signature' as below!
)
Al.
Hi Al,
They should be pretty good, even at low voltages. In fact, the IC doesn't know anything about the output voltage; all it tries to do is maintain a fixed 1.25V between OUT and ADJ. Because the internal design is such that hardly any current flows at the ADJ pin, it means that simple Ohms Law applies; the current in the two resistors is equal, and equal to 1.25/R, where R is the resistor between OUT and ADJ. Datasheets recommend 240 ohms, which gives 5mA, near as dammit
. From there, knowing there is 5mA is the resistor between ADJ and ground, it's a simple task to work out which resistor gives a voltage drop of Vout-1.25.
You probably knew all of that
Anyway, you've either got faulty ICs or a mistake with the wiring, an outside chance of parasitic oscillation. The datasheet is excellent and should give plenty of hints about layout, decoupling, protection diodes, etc...
Good luck - and if you get stuck, you can come here and look after the kids, and I'll hide in your "shack" for the rest of the day
Mark
PS: There's plenty of LM317Ts here if necessary...
Thanks Mark. I've concluded that these 317's are defective. I know that I've used one of them successfully in the past; not sure about the others. The point is that they are not unused, virgin stock. Anyway, I'm now investigating alternative approaches.
Al.