I'm a self taught hobbyist, not an EE.

I've got a circuit that requires a high DC voltage to drive a bunch of LEDs and a lower DC voltage (12V) to drive the control circuitry. The control circuitry gets hooked up to some expensive equipment, and so it needs to be isolated. I also have some tight space requirements. Current solution rectifies and smooths mains voltage and then uses a DC-DC converter that takes a high voltage input and outputs 12V. This is the converter: http://www.tamura-ss.co.jp/electronics/pdf/EN/SPM/SPM1203SJ_1411BE.pdf

The current circuit works very well. The problem is availability of that DC-DC converter is very limited, so I'm looking for an alternative. After much searching, I've not found any alternative module that is anywhere close to the same price and whose availability won't be a potential problem. So I've decided I need to roll my own solution.

I'd appreciate suggestions on how to essentially replicate the functionality of that converter module. (And how to keep it small in doing so.)

TIA
-j

 

Would this do the job?
http://www.ebay.co.uk/itm/3-3V-5V-9...hash=item1c700e538a:m:mYr76hP4k1xdR4AYwhT4gPA

 

Would this do the job?

Hard to say. I've ordered some though. It'll probably take them a month to get here. Long term, I can imagine problems with availability and quality assurance... but maybe it'll give me a good starting point for rolling my own. Thanks for the link. I'd still appreciate advice on how to go about building a solution.

 

If you haven't done this sort of thing before then don't. The transformer is the most tricky part. It would need to be a custom component.

 

Digi-Key has over 500 of them and says the part status is "active":
http://www.digikey.com/product-detail/en/tamura/SPM1203SJ/MT7388-ND/5056535
It costs less than $7 US.

Where are you located? This can effect availability and cost.

 

Digi-Key has over 500 of them

500 isn't enough. By an order of magnitude. And the lead time is like 5 months. The $7 cost is not an issue. But alternative parts I've found cost much, much more.

 

500 isn't enough. By an order of magnitude. And the lead time is like 5 months. The $7 cost is not an issue. But alternative parts I've found cost much, much more.

Now I see your problem better. Octopart only shows about 600 parts total from all sources.

Does the replacement have to be pin compatible?

 

If you haven't done this sort of thing before then don't.

If everyone followed advice like that, how would anything get done?

The transformer is the most tricky part. It would need to be a custom component.

Please educate me. Why so?

 

Does the replacement have to be pin compatible?

No. Not at all. If I can roll my own solution out of available parts, that would be absolutely fine so long as I can make it all fit on the board. I don't have a lot of room to work with but that's not my immediate concern.

 

If everyone followed advice like that, how would anything get done?

Please educate me. Why so?

If you disagree then go ahead and try it.
If you get it wrong it won't just not work, it will go bang.

 

Perhaps one of these or these would work if you wanted to roll your own.

 

If everyone followed advice like that, how would anything get done?
Please educate me. Why so?

Many people like yourself think that all of electronics is slapping chips on a board and wiring up pins with the same names. That can work ins some environments. Power electronics and DC-DC converters with isolation is not one of those areas. I predict it will take a novice designer such as yourself up to two years to design, prototype, layout, and produce the quantities that you seek. The only way to shortcut this process is to hire competent help that know what they are doing so you can get it right the first time with one and only one board turn.

Believe me or don't believe me -- I don't really care. I do have some mild interest in preventing you from making a colossal mistake.

 

If you disagree then go ahead and try it.
If you get it wrong it won't just not work, it will go bang.

I likely will. I've had things go bang before. That's when things get fun. Just remember to wear eye protection.

While I do appreciate your concern, it's not very informative. And I generally find information more helpful.

Perhaps one of these or these would work if you wanted to roll your own.

Thank you. This was helpful.

Many people like yourself ...

Hmm. I'm less than impressed by your ability to figure out exactly what kind of person one is from a few brief requests for assistance in an internet forum. I said, I'm self-taught. I didn't say I was an idiot. I've got a strong technical background, two decades plus in software and unix systems engineering; I've been dicking around with a soldering iron since before I had hair on my nuts; I have my amateur extra license; I'm cautious; I'm not afraid of math; and I'll simulate, prototype, and test in order to keep my mistakes small and manageable.

If you really have a mild interest in preventing me from making a mistake, maybe you could just point out some of the pitfalls I'll likely have to avoid along the way.

But if your only advice for me is to hire someone, well then you've entirely missed the point.

 

I likely will. I've had things go bang before. That's when things get fun. Just remember to wear eye protection.

While I do appreciate your concern, it's not very informative. And I generally find information more helpful.


Thank you. This was helpful.



Hmm. I'm less than impressed by your ability to figure out exactly what kind of person one is from a few brief requests for assistance in an internet forum. I said, I'm self-taught. I didn't say I was an idiot. I've got a strong technical background, two decades plus in software and unix systems engineering; I've been dicking around with a soldering iron since before I had hair on my nuts; I have my amateur extra license; I'm cautious; I'm not afraid of math; and I'll simulate, prototype, and test in order to keep my mistakes small and manageable.

If you really have a mild interest in preventing me from making a mistake, maybe you could just point out some of the pitfalls I'll likely have to avoid along the way.

But if your only advice for me is to hire someone, well then you've entirely missed the point.

There are many areas of electronics where you can be entirely successful with that background. I did not call you any pejorative names. I made no vulgar references to parts of your anatomy. What I said was, isolated DC-DC converters are difficult and challenging, and your chances of success in your first design are generally not great. If you have the time to develop the skills, do the research, and toast a few components, then fine -- charge on. There are an incredible number of things that can go wrong with design, layout, fabrication and testing. If you want to succeed quickly, you need experienced help. I cannot get you there with a couple of forum posts. It is your project, your decision, call it however you see it. Once you make the decision you have to own it.

 

Designing a circuit with magnetics is often significantly more difficult because it's a less ideal device than other basic components, with non-idealities such as stray winding capacitance, core saturation, and leakage inductance that can have a significant detrimental affect on circuit operation.
That's why I avoid them in my designs if at all possible.
It's likely that the magnetic circuit's real operation will be significantly different from the theoretical design, and thus the chance of having to do a redesign are high
Thus the caveat about doing a high-frequency design (i.e. above a 100Hz) that contains a transformer.

 

Not wanting to pile on, but...

You have not said where you are. If you are in the US or Euro-zone then you _must_ have the product tested for EMI.
In addition you may want to get the product listed by UL, VDE, CSA or the like for safety.

Both of these add time and money to the project. A lot if you have never done it before.

 

UL testing and compliance is the gift that keeps on taking. A chunk of your wallet every 90 days as long as the certification is active. There is just noting at all in this world like paying the piper and being unable to move product off the shelves.

 

UL testing and compliance is the gift that keeps on taking. A chunk of your wallet every 90 days as long as the certification is active. There is just noting at all in this world like paying the piper and being unable to move product off the shelves.

Interesting. I have only been on the technical side of UL testing. Once a product passed I was done.

Maybe the safety testing without the listing might be an option? That would at least give some warm fuzzies that the product was not going to electrocute people or equipment.

 

I likely will. I've had things go bang before. That's when things get fun. Just remember to wear eye protection.

While I do appreciate your concern, it's not very informative. And I generally find information more helpful.


Thank you. This was helpful.



Hmm. I'm less than impressed by your ability to figure out exactly what kind of person one is from a few brief requests for assistance in an internet forum. I said, I'm self-taught. I didn't say I was an idiot. I've got a strong technical background, two decades plus in software and unix systems engineering; I've been dicking around with a soldering iron since before I had hair on my nuts; I have my amateur extra license; I'm cautious; I'm not afraid of math; and I'll simulate, prototype, and test in order to keep my mistakes small and manageable.

If you really have a mild interest in preventing me from making a mistake, maybe you could just point out some of the pitfalls I'll likely have to avoid along the way.

But if your only advice for me is to hire someone, well then you've entirely missed the point.

I'm not sure I understand your logic, but if you want to worry about 20 parts, assembly, test and compliance, I say have at it.
You might have a look at a company named Powerline.
You probably won't find an exact replacement so what are your specs?

 

Interesting. I have only been on the technical side of UL testing. Once a product passed I was done.

Maybe the safety testing without the listing might be an option? That would at least give some warm fuzzies that the product was not going to electrocute people or equipment.

Every time I wrote a check to UL, I kept thinking to myself: "What a bunch of bloodsuckers".
Testing without listing is worth less than the paper the listing would be printed on.