Trying to figure out the difference between control methods used in switch-mode regulators - PWM, VFM (variable frequency modulation?). Also, what exactly is 'voltage-mode control' and 'current-mode control'? Does voltage mode control refer to constant output voltage and current-mode control refer to constant current output?
Nope. Current mode refers to the fact that the current through the inductor is monitored and the current level (transferred into a voltage, of course) is the input for the innermost loop. The other loops can do whatever you need them to do, limit the output voltage, the output current, or as in most of my switcher designs limit both of them.
I suggest to read some of the former Unitrode app notes, now on the TI server. This was my teaching material back in the 80's and 90's, in those days in the form of a data book that is now one of the most worn on the shelf. It's more valuable than many of the university text books I ever saw.
-- Regards, Joerg http://www.analogconsultants.com
Okay, thanks for your input!
Any ideas > firebird wrote:
PWM: Pulse Width Modulation VFM: Variable Frequency Mode (usually)
There are lots of other methods such as "pulse skipping". IIRC John Larkin called that "burp mode" a while ago. It works like the control of the rotary engines on really old fighter planes, the ones where the cylinders rotate around a fixed crankshaft. In order to land these without coming in too fast the pilot had to cut off the ignition to some of the cylinders, producing that ratatat ... ratatat sound.
-- Regards, Joerg http://www.analogconsultants.com
You can't cut off just a few cylinders. The pilot switched the magneto on and off to regulate the power, sort of like a modern pwm power supply. This was hard on the engine, but the expected service life of the plane and pilot was only a few hours anyway.
That ratatat ... ratatat sound was from the machine guns of the plane behind you.
Regards,
Mike Monett
Antiviral, Antibacterial Silver Solution:
That sounds more like shooting off its own propeller blades.
-- Thanks, Frank. (remove \'q\' and \'.invalid\' when replying by email)
Hello Mike,
That was on the really primitive ones. But a simple trick could be used to cut the magneto over a portion of crankshaft angle. This was done on the Gnome engines that were AFAIK used in the Sopwith Camel. IIRC they gave you only two selector positions on the Gnomes: All cylinders firing or only one cylinder firing. Since mixture was still being sucked into the others stories have it that if you cut for too long and then re-applied full ignition there could be a big kablouie and the engine went to shreds.
:-)
-- Regards, Joerg http://www.analogconsultants.com
Hello Frank,
The early synchronizers supposedly sometimes allowed that to happen.
We have a runway right here in town, it's an airpark community. Lots of people restore vintage aircraft and then fly them a lot instead of just displaying them somewhere. This is fun. Often when in a conference call and that rat-tat-rat drowns out the phones even older people ask "What on earth was that?". There ain't nothing better than being in an open cockpit biplane with leather helmet, goggles and all and the pilot does aerobatics. Or tells you to take the stick and bring it home. Yeehaw! Afterwards your T-shirt has that genuine aviator look, full of oil that had splattered out of the rotary engine.
-- Regards, Joerg http://www.analogconsultants.com
Here's a report from a pilot actually flying one. Apparently, the procedure is to shut off the fuel and keep the ignition on. This makes a lot more sense:)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Coming down from altitude is usually accomplished by shutting off the fuel, but not cutting off the ignition. With the propeller windmilling, forward pressure on the stick is needed to hold the nose at a fairly steep angle toward the earth to maintain a safe airspeed. It's important to keep the propeller turning so that when you turn the fuel back on, the engine will restart.
[...]If, to descend, you shut off the magneto to the engine but leave the fuel on, you are flirting with danger. While the engine is turning over, fuel is being fed into the engine and exhausted from it into the cowl. Turning the engine back on could ignite the raw fuel collecting in the cowl.
Even with the fuel off, however, there is still a problem to contend with: the oil pump is geared directly to the engine, and even when the engine is off, as long as it is rotating, oil is being pumped into the cylinders.
This could oil up the "sparking plugs" (as they were referred to in 1917), and this could prevent them from firing when the ignition is turned back on.
The Gnome engine manual tells you to keep the ignition on during the glide to prevent this from happening.
In theory, it sounds right, but it really doesn't work because too much oil is pumped into the cylinders, especially on an extended glide.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Regards,
Mike Monett
Antiviral, Antibacterial Silver Solution:
Hello Mike,
[...story in link below...]Thanks! That is an interesting article. Kind of scary to cut the fuel while the oil still flows. Could make the plugs look like uncle Leroy's old barbeque really fast. It probably won't happen to a good pilot but what if he came in a bit short and the thing wouldn't rev up?
We see fixed crankshaft rotaries very rarely, usually only during air shows. The way they landed was with that regular sputter. There was an old German motorcycle with a monosoupape engine inside the front wheel and they figured out a way to provide a throttle:
-- Regards, Joerg http://www.analogconsultants.com
Some other articles mentioned it might take 5 seconds or more to clear the plugs and start producing power. If you were low and slow, the engine torque might spin you into the ground if you weren't watching for it. They lost a lot of pilots that way. My Malibu had a similar effect. It had dual exhaust-driven turbochargers that spooled up on takeoff, so when the torque kicked in, you had to be ready with your foot on the rudder pedal to keep from going off the runway. I let my friend try a takeoff one time. Even though I warned him about the torque, he still couldn't keep it centered and we almost went into the weeds. I had to put it back on the runway to let him continue the takeoff. Scared the sh*t out of him. He thought it was fantastic:)
Amazing. Looks like some of the design approaches we see here from time to time:)
Regards,
Mike Monett
Antiviral, Antibacterial Silver Solution:
If I may bring up the original topic in the thread...please look at the following link:
Here, VFM stands for voltage-to-frequency modulation. This sounds like another name for variable frequency modulation, except it is more descriptive in that it tells you what changes the frequency (voltage...output voltage that is fed back?) But I am not sure. Is this the same as pulse skipping?
Mike M>
Ok, pulled the ST5R00 data sheet. Tell the guys at ST to hire more competent web site designers and we would use more of their parts. I don't believe the wording voltage-to-frequency modulation is quite accurate here. Maybe a marketeer coined that phrase but I won't say that loud because I married a marketeer :-)
The way this converter works is similar to pulse skipping. You could call it "pulse stretching". The schematic in fig 1 doesn't seem to accurately depict the innards since a classical oscillator could not do what is outlined under "Operation":
Ton is fixed while Toff is determined by the error amplifier...
This means that you would only see very few pulses under light load, spread far apart. Just like you would with pulse skipping, except that pulse skipping often can't do this gradually but in digital steps. When the load increases then the pulse rate increases for both cases.
Voltage-to-frequency is something entirely different. Look at the data sheet of the National Semiconductor LM331. That is a true V/F converter where the frequency it puts out is proportional to the input voltage. This is not the case for regulators such as this ST series. Look again at the ST5R00 data sheet, figure 1: There is a logic gate driving the block they call VMF control (probably a typo). Such logic gates feature very high gains, tens of thousands. I cannot see anything in there where a voltage would control a frequency. What it does is trying to maintain the same output voltage via changing the frequency, which in turn changes how many chunks of energy are delivered into the output capacitor per second.
-- Regards, Joerg http://www.analogconsultants.com
Okay. I think I am beginning to understand the difference b/w PWM and VFM now.
It seems that output voltage seems to be equally well regulated/controlled by PWM and VFM, but when it comes to efficiency - VFM wins. I can't figure out one thing from the datasheet...how much better efficiency would VFM provide when compared with PWM + pulse skipping?
Joerg wrote:
Hard to say without knowing the device capacitance and geometries, especially of the FET in the power path. Companies don't release such data anymore. It boils down to this: At very low power a PWM keeps switching because its frequency is fixed. Ton will be very low so that next to nothing in energy is pumped. But all the switching losses continue. A large part of the total power consumption in CMOS devices can be attributed to switching losses and these increase almost linearly with switching frequency. Pulse skipping greatly reduces that at the expense of higher output ripple, no free lunch here.
You could compare this to a car: A long time ago Volkswagen made models that would stop the engine if the traffic light remains on red for more than a couple seconds. Hence the car would not consume any gas during the minute or so you had to wait. When it turned green a wee step on the pedal would start it again. A normal car would idle all that time, consuming gas while not delivering any transport function to you. Wouldn't work here in California because everyone wants to keep the A/C running ;-)
-- Regards, Joerg http://www.analogconsultants.com
Hello Mike,
Happens in older turbo cars as well. One guy entered a long curve and at the end the turbo kicked in. Engine gyro plus torque jolt on the wheels made him spin out, big time.
But it won races which is what counted. It wouldn't work as a street bike though.
-- Regards, Joerg http://www.analogconsultants.com
Okay. Thanks...things are a bit clear now!
Joerg wrote:
Yes. I had a 380SEL which I loved. So I gave it to my girlfriend and got a
500D. Picked it up in Saarbrücken. What a blast to go on the freeway and pass a Polizei in a white and green Porsche at 125 mph. And wave at him and have him wave back at you!Then I had it AMG'd. Big mistake. They put in high-lift cams and made other mods to the engine. The result was zero torque up to about 3,500 rpm, when all hell broke loose. It was impossible to control on a flat, straight street, let alone hills or curves. So I had them put the engine back the way it was. Too much power on wheels can get you in trouble real fast. (But it's ok in planes:) Regards,
Mike Monett
Antiviral, Antibacterial Silver Solution:
Hello Mike,
Cool. My Audi over there wouldn't do more than 115mph but often I zoomed from Cologne to Ulm on Sunday afternoons when the rest of the nation was watching Fussball. Kept it at 110mph almost the whole time.
The only time I got in trouble passing a Polizei was on my bicycle. A racing bike and I was a bit late for an engineering exam. Flew by a whole lot of cars and saw too late that one of them had that blue hump on the roof. Oops. Got a speeding ticket.
Next time ask if you could test drive one first, or rent one somewhere. AMGs are meant for racing, not so much for our 65mph world out here. A Finnish race car driver once said that he never really shifts past 3rd gear and that they might as well take out 4th and 5th to save weight.
-- Regards, Joerg http://www.analogconsultants.com
LOL! Regards,
Mike Monett
Antiviral, Antibacterial Silver Solution:
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.