SMPS, control loop?

Maybe it is just the state I live in, but I see "shed" smokers all the time: probably in the neighborhood of 500 cubic feet.

You don't BBQ lox, you eat it with capers, drizzled with wedge of lemon. Bagel and cream cheese optional.

Best regards, Spehro Pefhany

In a real analog system, everything happens simultaneously. A mental walkabout tends to be sequential, essentially introducing delay that's not there in real life.

But it can correct it as it happens or, in the case of oscillation, keep it from happening in the first place. It's all in the numbers.

John

Hello Martin,

My very first time in America long ago our barbeque coals were starting rather slowly, more or less not at all because it had been raining. The campground owner came by and said "Nah, that's not how we do it in America". He took the light fluid bottle, poured half of it over the coals and flicked a match into them. KABOOM!

Regards, Joerg

We initially had NG in a long black pipe going to the pool heater. We placed the BBQ over that pipe for obvious reasons and after the plumber did his thing and passed inspection we began to cook. The flame was weak so I removed the NG pressure (voltage) regulator between the black pipe and the flame. The improvement was enormous, we went from jerky to sear in one step. No wood, charcoal nor animal dung. So JT, sounds like you are getting ready for the annual hosted SED Beef + Grog at your place. Harry

Nice, and imaginary numbers. I like the simultaneously not sequential concept. I sometimes add lead networks that aren't really needed. My brain needs parallel processing. Harry

The black pipe I had to put in when I lived in Scottsdale was (IIRC)

Made that offer several years ago, but only Spehro and "C" attended.

...Jim Thompson

Oh, there's "smokers" and there's "smokers". I saw a cooking show on TeeVee, where some guy put a handful of tea leaves in a little aluminum foil boat, put it into a skillet, put a baking rack over the tea in the boat in the skillet, put slabs of meat or fish or whatever on the baking rack, and covered the skillet. Put it on high enough heat to make the tea smoke, and smoke it until it's done. I imagine you could use hickory, mesquite, apple, sage, any combustible herb, and smoke stuff.

;-P

Sometime before I die, I want to smoke a fresh salmon. I wouldn't smoke walleyed pike, because it's just too delectable just the way it is, but I'd smoke a northern pike in a heartbeat. ;-) Heck, I'd even smoke a carp! My Grandpa once[1] told me how to eat a carp: You get a shingle, and put it over the fire, throw the carp on the shingle, cook it until the carp is done, throw away the carp, and eat the shingle!

Cheers! Rich [1] Several times, actually, but hey! This was Grandpa! :-)

I just had a thought occur to me - you wouldn't want to smoke a cut of beef like that, would you? I can see putting a couple of raw hams or turkeys in it - my Dad used to tell about when he worked at Hormel he and his buddies would grab hams off the line at the outlet of the smoker, and run them through the smoker again; melt in your mouth, I was told - but back to my point, I've never heard of smoked beef, other than the smoke from a grill.

Any discussion on that? (and please don't bring up frying a turkey! ;-) )

Thanks! Rich

But, is it _THE_ one? ;-)

So, are those anchovies that come rolled up around a caper supposed to be served as individual w**re's davores? Like, on a Ritz cracker with creamcheese?

Thanks, Rich [my inner imp wants to say, "Creamcheese and snot?" but I'm suppressing him at the moment. :-) ]

Hey, I'd be happy to show up, if you'll EFT me about $500.00 to make the trip. :-)

Heck, I'd even try your homemade wine! ;-}

Cheers! Rich

PS: JIC you have a burning desire to email me, I have a spamdump at snipped-for-privacy@yahoo.com; to make a real email simply elide ard.

Hello Rich,

I can't say much about beef. We didn't do much of that but we smoked fish and pork in it. Both very delicious.

Regards, Joerg

You do that on this newsgroup fairly often! ;-)

I've always found that they're way too hard to keep lit. ;-)

Cheers! Rich

Hello Tim,

Ask a politician. Many claim that they can :-)

Maybe it helps to think of it the following way and that's actually the way most of my discrete switcher designs work. There is an oscillator whose duty cycle can be controlled. Three or more things can control that duty cycle and they can do so instantaneously if you allow them to, IOW if there were no loop filters:

1. Inductor current
2. Output voltage
3. Output current limit
4. Whatever else the client wants regulated...

All these "throttle" the duty cycle whenever a limit has been reached, kind of like the governor on steam engines or on stationary gasoline engines (wood splitters etc.).

Regards, Joerg

Ok, so it's like my induction heater currently: the circuit naturally drives towards maximum, and the most significant of the various error signals is the one that throttles it back?

Then to apply that to an SMPS: at idle, voltage would rise but the error amp keeps it regulated to the set voltage; the current feedback reacts immediately and proportionally to load changes, so as to change inductor energy as best possible to reduce over/under-shoot; and, the difference between the two is the current loop is HF response (above the filter's F) and the voltage loop is LF response (below)?

Hmm...that might just work...

Tim

Hello Tim,

That's it in a nutshell. Again, that is how most of my designs work but there are many others.

Usually the inductor always charges up to the current limit you have set. This point is often determined by a sense resistor and the threshold of detector is fixed. So the inductor current feedback doesn't really react to load changes directly. When one or more of the other error feedback paths (load voltage, load current etc.) signal a demand the regulator goes into "full bore", pumping the maximum current allowed through the inductor. When a limit is reached these pump cycles might briefly be cut short but many regulators then don't settle for a lesser inductor peak current. Instead they'd go into a "staccato" mode by stretching the repeat rate of those charge events. IOW the frequency drops. On really light loads they often begin to skip pulses and this is also referred to as "pulse skipping mode".

Again, that is just one example. There are many ways to do it. I have also built regulators where the switching frequency is held very tightly and synchronous to something else, for noise reasons. Then the current ramping on the inductor is indeed cut short so that it does not reach the selected maximum. But then the inductor current sense feature in essence is relegated to a safety measure, preventing a saturation event. Inductor saturation is typically followed by an impressive pyrotechnic show and that is to be avoided ;-)

Regards, Joerg

Is that within a cycle, or is the limit equal to the load current? In which case, the current loop would just speed up the inductor's current response, NFB around a time constant.

Is that like the UC384x, outside of the adjustable current limiting way it regulates? That is to say, the current won't rise more than x, where x is of course determined by the sense resistor.

Hit n' miss engine, eh?

That would be driven by ripple, huh? A positive feedback sort of deal at any rate, which since it's strictly a Negnative loop, would have to be reactive: filter delay or such. Which also means there's a maximum time before the phase gets reasonable again and the circuit responds (something like a relaxation oscillator, but er, LC), which also makes sense.

Ah yes, I'm familiar with this. I found a surround sound amplifier from a DVD decoder thing, or something like that. It has two power supplies and six TDA-(something four digit) 10 or 20W class AB amplifier chips. One of the power supplies appears to be a one-chip flyback circuit for standby, the other powers the amps, something like +/-15V. Well when you plug in the bare board (that is to say, outside of the plastic(!) case), it just starts ticking. With the standby line patched in, the other power supply pulsed

Frequency can't vary much if I'm using an SG3524 or variant, at least.

So is this thing you speak of more of an inductor charge thing, where you put pulses into the filter, unlike the somewhat integrating function the choke input filter provides. Basically, L < Lcritical (for choke input)?

Heheh, I intend on a current transformer in series with the power transformer, it'll get a peak detector for that sort of stuff. Don't think I'll bother with a desat detector though (would be annoyinh anyways what with hard switching going on :-/ ).

Tim