Wawa:
Don't know about that. It's 2016.
I thought lineair regulation, with it's inefficiency and heat, was a thing of the past.
How much current are we talking about.
Nowadays you can get 2Amp 3mm x 3mm smd MPPT chips (e.g. LT3652) that put more current into the battery than the solar panel delivers.
A 15Amp MPPT charger in a case is available from ~US$7.50 on ebay, shipped.
Leo..
I don't have the space on my board for that many components, and its a bit pricey, but certainly the most efficient solution.
The part is NOT out of spec. It is the WRONG part.
So far you are the only one who refuses to accept that. Of course a TL431 VOLTAGE REFERENCE would work better. ( that's what it is designed for.).
You should change your post title to
"[SOLVED] (wrong part)"
If you googled "SOT" ,why did you ask what it meant ?
@Resinator,
How come you onl come out of your hole when I post ?
I suppose I should take that as a compliment (that you find me more interesting than the rest of the forum. Who'd have thunk? 8)
As you wish
Seriously though, if you ever do want to change your post title you have to do it in the ORIGINAL POST, and that way it is refelected in any further posts that occur after you change it. Also, FYI, when your post is solved, it actually is the forum custom to insert [SOLVED] in your Original Post, so it shows up in the title on the list of posts so people won't waste their time with it if they are looking for "UNSOLVED MYSTERIES" (so to speak). It is , of course , entirely up to the OP to decide if they want to change anything else about the title.
Apparently nobody else got the memo.
Here's the list of Electronics Industry Abreviations.
Small Outline Transistor is on it.
Select On Test is not. Is that a UK thing ?
I'd just like to thank you all again, especially raschemmel for his input. By hashing it out here in the forum, I have come to a resolution without having to involve my board manufacturer, which potentially could have soured our relationship. I do have a work around for my current boards and I will implement the TL431 on my next run.
Cheers.
I'll try and work out how to changed the original title to SOLVED.
No - one prefers not to have to use aot/sot or a pot unless really necessary - it slows down production. And I don't like pots because their temperature drift isn't very good - and idiots can tweak them when they shouldn't.
But if they're only used for say +/- 5% of the range that can be tolerated.
You can buy 0.1% references.. but they're dear
It's all a trade-off between production time , stabilty , and the cost of high precision components.
I recently had 100 MCPCB's made that had Vishay 15V zener P.no. BZG03C15 on board. The zener is passing 25mA and does not even get warm to touch, yet the voltage across them is 15.6V or more. According to the data sheet, 15.6V is the tolerance maximum with a test current of 50mA, I'm only passing half that. As such my boards, which cost me $3500 are scrap. What recourse is available to consumers when components are supplied out of tolerance?
Cheers.
Did your BOM specify the manufacturer and distributor? We require that for all boards we build where we buy the components.
Paul_KD7HB:
Did your BOM specify the manufacturer and distributor? We require that for all boards we build where we buy the components.
Paul
Yes the BOM was specific to the part number.
As was mentioned in an earlier reply the tolerance values relate to pulse testing so heat is taken out of the equation. This is pretty frustrating because I imagine the majority of implementations for zeners are under constant conduction, in which case you can't really use the tolerance specifications for design. (Unless of course you have a Ph.D.)
The irony is that my work around for my 100 boards is to put another 5% zener across the battery. This is counter intuitive because the original problem was that 5% is too broad a tolerance for a regulator. But my experience with zeners is that despite being rated at 5% they are usually much closer in tolerance within a particular batch. For example during prototyping, I bought ten 15V zeners and they were all +/-1% nominal voltage at 25mA. But my production boards are all 15.6V +/-1% at 25mA. The 5W 13V 5% zeners I am now using as a bandaid all zener within 0.1V of each other when passing 100mA, closer to 1% than 5%. So I wonder why they just don't bin them into tighter tolerances?
The confidence I had in my design was due to my experience, not the data sheet, and it came back to bite me. Live and learn.
I think you're missing the point that a TVS diode is not meant to be used as a zener.
A 15volt TVS diode, as you have used, is made to NOT conduct at the rated 15volt.
It is made to protect a 15volt supply rail, and not use too much current from that rail in the process.
Can't you solder a common 15volt/1watt zener across that TVS diode?
Leo..
