DC-DC Modules: What am I missing here?!?

[edavid]

OP, I don't know why people are being so hostile.  Your question is completely reasonable.

Have you looked into driving the input of the cheap modules with a higher than rated voltage?

If you monitor the supply current, you should be able to tell if the transformer is saturating.

ETA: You could also try under-driving a 12V-12V module.

[IDEngineer]

OP, I don't know why people are being so hostile.  Your question is completely reasonable.

Thanks! I'm trying to not let their behavior affect mine. 

Have you looked into driving the input of the cheap modules with a higher than rated voltage?

That is a very interesting idea. Since they are obviously completely unregulated, I wonder if they just loosely shoot for a primary-secondary ratio... in which case your idea would work! Need to monitor dissipation too, since running it "hot" might yield precisely that. Our current draw isn't great so no problem there, but if the transformer drivers are running close to their voltage specs we could get some mortality issues.

Excellent idea, thanks!

[Sceptre]

OP, I don't know why people are being so hostile.

I'll take exception to that.  One of the OP's original complaints is that modules with intermediate output voltages don't exist:

But instead of rating these DC-DC modules with that in mind (say, 7V output voltage so you can safely run a 5V regulator), their nominal output voltages are right at the traditional rails of 3.3, 5, 12, etc.

I thought I was being helpful by pointing the OP to available 6V and 9V modules (and even a 7V module, albeit expensive).  Then I find out that there are stringent cost and second source constraints.  And for some reason the OP thought I needed volume production design tradeoffs explained to me.  All in all, not much incentive to take the time and effort to help out.

[IDEngineer]

Jeez, I hate it when simple questions turn into pissing matches.

1) When I asked about DC-DC modules, I was talking about the majority of them. Yes, there are specials with different voltages, but the majority of pin and size interchangeable parts are at the traditional supply rails that I mentioned.

2) As a long time lurker here, I've noticed that when someone is doing a one-off they tend to specifically note that ("I'm building a such-and-such for my car", etc.). So I presumed when that isn't mentioned, folks understand we're talking about Engineering for production. Mea culpa.

3) Since we ARE talking about Engineering for production, things like cost and compatibility and second sources matter. Since your examples didn't seem to take those things into account, I then mentioned these criteria. Again, I didn't mention them before because I presumed we are mostly a community of professional Engineering people who work in Engineering for a living and thus default to the criteria that matter in such an environment.

I continue to maintain that it's odd for high volume parts, that target applications where regulation very clearly doesn't matter, to set their nominal outputs right at common rail values rather than a bit higher to capture regulation sensitive applications too. Yes, such output voltages are available at higher cost and lower compatibility but it's strange that the majority would design themselves out of such a market when there doesn't appear to be a technical reason to do so.

I apologize if I've offended you in some way, that was never my intention. I don't know why my question elicited such an emotional response from you but that too was not my intention.

Can we put the emotion behind us, please? Thanks.

[Damianos]

The question is: why they don't change the specifications of million devices...

My warmest thanks for your helpful and supportive comment.

Nowhere did I suggest they should change the specs. I was simply wondering why they didn't spec them to be more flexible from the start. And my last post before yours specifically outlined our present plan to accommodate their existing specs.

Engineering is an art of compromise. Everything is a tradeoff. My most recent post described how we intend to balance the specs to make these parts work, with both their advantages and disadvantages. How your post then improved things for this community isn't clear to me, but I'll go back and read it again in case I missed some useful nugget of technical insight.

Thanks again.

If you widen a little your view of the things, it will become both helpful and supportive!
In an attempt to be more helpful I will try to explain a little more.

From your first post here and in every next to that you are wondering why these specific devices have these specifications and not something else. As my little knowledge says, they satisfy the specifications that were established about half a century ago. These have not changed up to today and probably not in the near future (at least). There may be reasons for that! Think about it.

It seems that we have a little different opinion of what is engineering. Lets describe what I mean, by taking as example the present case:
I have to put down what I have and what I have to do. This includes things such as inputs, outputs, environment etc.
Then I have to translate the above in specifications.
Search for parts that can satisfy my specifications.
Evaluate what I have found in all of their characteristics; Including cost, availability, reliability etc
If in what I found there is a part that satisfies all of my needs, then the problem is solved. If not I have to continue ...
Examine if a ready made part can be used with an adaptation, so as to match their specifications with mine.
If the above fail, I have to plan to design something ... ...
The above are some draft points, that I think that is a proper way to design a system/device/product.

By the way, you speak about engineering and professionalism but you found great the idea to abuse a little a device! What about your reputation if you have a lot of returns of your product, soon or a little later? Also what you are stated is similar to the trend, when a circuit is not operating as expected, many people are referring to fake components without examining the design and implementation.

To conclude, all of my effort is to attract the attention to the specifications, because, in my opinion, engineering has to do with them.

PS1: to clarify things; I do not know who you are, such as any member of this forum, so there is not something personal. Also take in account that English is not my native language, so if my wording seems offending it is not by intention but from bad translation of my thinking to a language that I barely handle!
PS2: Sorry for the long and seeming as philosophical comment, but it is an attempt to describe how I act practically in the last about forty years.

[IDEngineer]

By the way, you speak about engineering and professionalism but you found great the idea to abuse a little a device! What about your reputation if you have a lot of returns of your product, soon or a little later?

First of all, thank you for the polite tone of your comments. That is sincerely appreciated, and it keeps the forum much more civil.

Secondly, please note that my first comment regarding the "out of spec operation" suggestion was to worry about reliability (running it "hot"). You're absolutely correct about the threat of returns, and my personal #1 Rule is "prevent failures in the field".

I viewed the suggestion similar to the technique of offseting the reference pin of a three-terminal voltage regulator... at first glance you're running the device at a "higher" voltage but strictly speaking it's not being run out of spec. That was the reason for my comment regarding their possible ratiometric design, where the output is a certain percentage of the input - it may be (subject to testing, as I mentioned) that the device would be perfectly fine operating that way, as long as overall power dissipation or max input voltage wasn't exceeded. I wouldn't commit to a design like that without a call to the manufacturer's Applications Engineering staff to discuss what was really going on in those little packages. Your caution is well warranted and thank you for emphasizing it for future readers of this thread.

[Sceptre]

1) When I asked about DC-DC modules, I was talking about the majority of them. Yes, there are specials with different voltages, but the majority of pin and size interchangeable parts are at the traditional supply rails that I mentioned.

OK, let's look at those parts on Mouser a bit more.  Here are the first 20 normally-stocked isolated 1W, 4.5V min Vin, 6/9/12V Vout converters (onesies prices, but higher volumes track):
Mean Well  SBT01L-12  $3.57
Mean Well  SBT01L-09  $3.57
CUI  PDM1-S5-S12-S  $4.24
Mean Well  SFTN01L-12  $4.27
CUI  PDS1-S5-S12-M-TR  $4.43
Mean Well SMU01L-12  $4.46
Mean Well SMU01L-09  $4.46
Mean Well SMA01L-12  $4.46
Mean Well SMA01L-09  $4.46
Delta  PB01S0509A  $4.58
Delta  PB01S0512A  $4.58
Mean Well  SFT01L-09  $4.64
CUI  PEM1-S5-S12-S  $4.84
Delta  PA01S0509A  $4.91
Delta  PA01S0512A  $4.91
Cincon  EC1TA02N  $4.91
Murata  MEU1S0509ZC  $5.08
Murata  MEU1S0512ZC  $5.08
Delta  PE01S0512A  $5.20
Delta  PE01S0509A  $5.20

Notice how there are seven 9V converters that, far from being specials, are simply variants of the 12V devices? And that they are priced the same as their 12V counterparts? Only five 12V converters in that list don't have identically priced 9V brethren.  What about 6V?  Well, those are indeed rarer:  down the list is the Murata MTE1S0512MC/MTE1S0506MC at $6.68.

2) As a long time lurker here, I've noticed that when someone is doing a one-off they tend to specifically note that ("I'm building a such-and-such for my car", etc.). So I presumed when that isn't mentioned, folks understand we're talking about Engineering for production. Mea culpa.

I understood that just fine ("We're working on a project here..."), but 'Engineering for production' can mean many different things.

3) Since we ARE talking about Engineering for production, things like cost and compatibility and second sources matter.

To varying degrees, depending on circumstances.  We have no idea what this product's volume, sale price, development cycle time, market lifetime, margin, etc. are, so it's unrealistic to expect us to know what level of BOM cost sensitivity you face.  And there are organizations, believe it or not, that have loose or even no second source requirements (or have exceptions that may cover these modules as well as FPGAs, micros, and certain high-performance analog devices).

Since your examples didn't seem to take those things into account, I then mentioned these criteria. Again, I didn't mention them before because I presumed we are mostly a community of professional Engineering people who work in Engineering for a living and thus default to the criteria that matter in such an environment.

I agree that we are mostly a community of professional Engineering people who work in Engineering for a living, but your presumption that we should all default to the criteria that matter to you in your particular environment is spurious.

[IDEngineer]

And if you dig just a bit deeper into the spec sheets, you find the following packaging and pinout varieties:

SBT01L-09/12: SMD, pins 1, 2, 4, 5, qty on hand 50 and 15
PDM1-S5-S12-S: Thruhole, SIP, pins 1,2,4,6, qty on hand 60
SFTN01L-12: SMD, pins 1, 2, 5, 8, qty on hand 0, 16 weeks out
PDS1-S5-S12-M-TR: SMD, pins 1, 2, 4, 5 (at last, something compatible), qty on hand 262
SMU01L-09/12: Thruhole, pins 1, 2, 3, 4, qty on hand 16 and 22

...and that's just across your first seven examples.

Why care about package and pinout compatibility? Because you'll notice that none of these parts are deeply stocked. Even a very small production run of 500 units can't be satisfied by any one of your first seven part numbers, even IF they were compatible. In fact, there aren't 500 pieces in stock of ALL of your first seven examples, despite them being different in packages, voltages, and pinouts.

There's more to selecting parts than just matching a few of the numbers on the spec sheet. Unless you're doing a one-off, that is. But we've already discussed that.

Are you done now? I'd like to be. Thanks.

[tszaboo]

So what if there is no stock? Either your application is not cost sensitive, or you wait for production times. Its really either or. It doesnt work, that you want to make something and order all of them from digikey, and somehow tell yourself that it is cost sensitive. 16 weeks of lead time is quite normal recently. MOQ will be low for these kind of devices.  I tell my bosses, that either accept it, or go into the bakery business, there everything is ready in 30 minutes.
If it is a critical component, and you want 500-ish production numbers, then just simply stock it at your favorite distributor. Its really not rocket science. You just need to understand how the sourcing of electronics components work.

[IDEngineer]

I understand that, but we're getting off topic. That's a discussion of stocking, specifically "safety stock", which you can set up anywhere in the supply chain from the manufacturer to your own stockroom. Instead, my most recent post was addressing the assertion that there are any number of equally applicable part numbers available with various output voltages. The off-rail voltages are out there, but just because you can compile a list of various voltages doesn't mean the parts themselves are mix-and-match interchangeable.

On your topic: Another useful thing you can do on some distributor's websites is an inverse sort by stock on hand. This is a clue to the consumption that particular parts enjoy. Highly popular parts are more likely to be stocked in higher quantities, stocked by more distributors, and less likely to be near end of life, all of which buys you some insurance for your designs.