indulis

You could check out: http://www.frontpanelexpress.com/

Well, you could have an "off-line" or a DC-DC converter... both are considered SMPS. In an off-line switcher, the AC is rectified & filtered (then a PFC stage in most) and then fed to a DC-DC converter. The "switch mode" part is that the DC voltage is "chopped up" (switched on and off) producing a square wave whose duty cycle can be varied and filtered to produce an other DC voltage. A switch mode power supply is generally more efficient than a linear supply. For example using a linear regulator to reduce 12VDC to 5VDC @ 1A the regulator would dissipate (12-5)*1 watts, in this case 7 watts. That's an efficiency of ~41% where as with a SMPS, say a buck converter, you could make the same voltage reduction at an efficiency in the 90's. That's a big reduction in wasted power (heat). It's "not really" the output transistor that's switched on/off and there is always an LC of some type after the "switching device".

And why do you think an inductor is gong to behave differently just becasue you apply a different type of signal (waveform)??? The answer is... it won't... all "the rules" (or math, if you will) that govern their functionality apply all the time! Why do you think: Do you know why the diode is "sometimes" placed in parallel and what happens if you don't?

What is going to be "charged" at 410VDC??? Those spec's seem to be for the front end of a off-line SMPS.

The problem is that there is only one point in the feedback path that contains all the feedback information from both loops, and that would be between the R12, C6 & C9 node and the R11 and U2 pin 4 node... but I'm not sure it's a legitimate injection point. The same rules should apply here as for DC-DC's and per Dr Venable and Dr Ridley, specific criteria have to be met in regards to the injection point, and if not followed, the results are not exactly accurate.

AN920 A quick question... when you generated (simulated) the bode plots, where was your signal injection point?

Simulator results are only as good as their component models, and frankly, most aren't great... at least the ones I've played with...MicoSim, SIMetrics, Intusoft & Workbench. While the second two are "cheap" the first two are not. How does the TIP31A introduce a LF pole? It has a 3MHz gain bandwidth product and if the supply is running a max Iout of 3A, and the 2N3055 is at it's min beta of 5 (On-semi datasheet and that's at a Ic of 10A not 3A), the TIP31A isn't working "that hard" and your no where near it's bandwidth limit. If the high frequency zero due to the cap ESR is an issue, add a high frequency pole a decade earlier. I'm having trouble with your bode's... the form I'm use to is more like this:

Is it really that low? If you are doing a planar, or it's a wire wound non-isolated application, seems that that constant would be unusable. On the bench right now I have a interleaved flyback (current mode) 1/4 brick running at 250KHz with planar's (16 layer PCB with 3oz Cu) that uses RM6 N92 cores (8mil gap) and I'm getting a bit over 120W out if it with efficiencies in the low 90's. While I haven't calculated it out, I don't think it would be in the 40% range.

Don't forget the winding window area as a factor...

No they shouldn't get hot if they are driven properly. Could you provide a little more info... At what frequency are youu running? At what voltage are you driving the MOSFET? A schematic would help... exactly how is your circuit configured "after" the PIC? For 6-7A a MOSFET should work just fine. It's not to hard to find a MOSFET with a Rds on in the 10 milliohm range. If you do the math (I^2*R) for 100% duty cycle, that's only .49W @ 7A. The current you would need to drive it would depend on the switching frequency.

Define what you mean by Most modern switch mode power supplies use MOSFET's as the switching elements (I'm not taking about supplies in the KW range). Yes you could "use a sledge hammer to kill a mosquito". There are other things to consider... switching losses for example, larger MOSFET's tend to have higher "total gate charge" (BTW this is what determines how much current you need to turn a MOSFET on & off in a given time NOT gate capacitance). As your "swiching losses" increase, you will take a hit in efficiency. Anyway, I can't, off hand, think of a reason why it wouldn't work.

You can't, you have to measure it.

Check out http://www.amplepower.com/pwrnews/beer/

Omni Yes tiring if you don't have the answer. Let me reiterate... maybe you didn't read the previous posts carefully enough and missed it before, for the last 15 years of my working career I've been employed as a power supply design engineer, guess what I do... design power supplies!! I am VERY familiar with those tests and many more. I know exactly how to perform them and what the proper equipment is to do it!! I NEVER said I didn't know how to do the test. It is presumptuous of you to assume that just because someone asks a question, that they don

Kevin WHAT?????

Omni Yes, I did "my time" at Northeastern University some 30 years ago (loved that co-op thing... paid for school). I hate to disappoint you, but the IS NOT an industry standard for testing power supplies, it is Agilent's!! While its possible that some schools/courses may have chosen "their" test methodology to use in the classroom, it should NOT be presented as something thats it IS NOT! Yes, there is a certain core parameter set that everyone tests, but by no means is it patterned after something Agilent does. While I'm not familiar with the Astro-Med Dash10... after a google search, that does not appear to be a very good instrument to do power supply testing (I understand having to use what you got, but...). A sample rate of 250K/sec and a bandwidth of 25kHz just won't get you there. It might be acceptable for some basic tests on a linear, but forget ANY off-line switchers or DC-DC's. I don't see any gain/phase modules under conditioners and accessories. Using that instrument, how exactly were you able to generate Bode Plots to determine stability? Did you generate them over temperature? With and without capacitive loading? Where was your "injection point"? Transient testing with a active load unfortunately SUCKS, but everyone sorta does it. I have yet to see an active load, whether running in current mode or resistive mode, that didn't have overshoot... you see the problem with that, right? So one more time, would you care to share your test result's?

Omni I've been designing power supplies for a living for past 15 years. I am VERY familiar with parameters I asked about. I was just curious as to the knowledge level a one semester class in "Advanced Power supply design and circuit application" gets you. It would be QUITE unusual for the average hobbyist, student or even engineer to have access to some of the instruments needed to obtain some of those parameters. This isn't a learning exercise for me, so I'll ask again if you care to share some of the results of your "complete testing"?

Yes, the period of the ringing is related to the time constant of the components causing it. The trick is... what are the components!! Ringing is not considered a first order affect. A large portion of the "values" are probably from parasitics.

Hello Omni Perhaps you could share with us the results of your "complete" testing? What were your results for things like gain and phase margine, cross over frequency, transient response time, short circuit recovery time, start-up time, line and load regulation, thermal derating... etc.

It's for "isolation" of primary to secondary. As isolation voltage levels change so do the creepage and clearances. The opto is often "weakest link" for creepage so you will often find a hole/slit under it.

The capacitor in a "capacitor start motor" is only there to "start" the motor

There are different "kinds" of power

Kevin No!!! A DC-DC converter has a DC primary. You've never heard of a "off-line switcher"? It's still a DC-DC converter on the back end with a AC-DC front end. What do you think the power supply in your computer is??

Guru My comments were intended for Kevin... You mean that the inductive reactance dominates (XL) the impedance (Z). Without a core it's still an inductor (air core) with mutual coupling to another inductor... sure the coupling coefficient is crappy, but it's still "more" than just a piece of wire.

What is the primary current in ANY transformer? The primary current will be the secondary current reflected to the primary by the turns ratio plus the magnetizing current. Don't confuse reactance with resistance!!