1. Home
  2. Electronics Design
  3. Feed Forward / Slop compensation

Feed Forward / Slop compensation

1) In switching power supply terms is Feed Forward another name for Slop compensation?

I have been consulting different literature and have got myself confused. I shall try and explain.

I one book, slop compensation is used in continuous output inductor current forward converters employing current mode control, to cure sub harmonic instability. It does this by keeping the average output current constant during line variations.

With current mode control, the peak current remains content whatever line and load conditions, but the peak to peak output ripple current varies with line (without slope compensation) thus the average output current and voltage also vary with line.

Slope compensation cures the resulting regulation problems and sub harmonic instability, by applying a ramp of slope magnitude (half that of the slope of the ramp developed across the primary sense resistor), to the output of the error amp, which is effectively the inverting input of the current sense comparator.

This in effect keeps the average output current constant during line variations.

2) I can see from the books waveforms that this would keep the output average current constant, but I am unsure how to calculate this slope and set up slope compensation in practice.

In another document (circuit diagram in fact) they pull up and down on the current sense pin magnitude depending on the rectified line voltage. I think this would add an extra slope to the top of the current sense trapezoid during higher line conditions thereby bring the fet on time down in proportion with line fluctuations.

3) Is the difference between a graphically fixed compensation technique and a line voltage dependent technique an answer to my first question?

4) Is the objective of both techniques' to keep volt seconds applied to the primary winding constant whatever the line conditions?

Thanks in advance,

Reggie.

Reply to
reggie
Loading thread data ...

Feedforward is a more general term and is one technique used /for/ slope compensation but not synonymous /with/ it. See the literature, this one is good, the author is an expert:

formatting link

Reply to
Fred Bloggs

try this...

formatting link

Reply to
moutou

They are two very different concepts that often show up in power supply design and often share some components.

Feedforward is a method used to reduce the effect of a disturbance on the output without running into trouble with oscillations. A really nice example comes from the case of a heater in an aircraft. I will explain it because it is so clear.

The heater to keep the pilot warm measures the temperature of the air in the cockpit and controls the amount of hot air that is brought from the engine. Such a system has the classic "transport delay" problem. When the circuit orders an abrupt increase in hot air, it take a while for the hot air to get to the point of measurement. When it gets there, the change is still abrupt. This means that there is a large phase lag but no decrease in gain. This means that the servo must act slowly.

Now imagine that the pilot suddenly opens the throttles. The engine temperature rises very quickly so the situation becomes a race between the heater servo acting and the pilot getting Bar-B-Qed by the hot air.

Feed forward in this case is an input taken from the throttle setting and applied to the control of the hot air valve. If the throttle setting's frequency content is filtered just right on the way to the valve, the pilot won't feel a thing.

Slope compensation is a method used to prevent oscillations in converters such as the classic booster where the inductor current is continuous. ACII art:

L1 D1 ---)))))----+---->!------+-- Output ! ! O Q1 === C1 / ! ! GND GND

While Q1 is on, current builds up in L1 but no current goes through D1. This means that turning on Q1 sooner or for longer to increase the output voltage actually reduces the output voltage in the short term. This can lead to oscillation in any servo loop which has a high bandwidth. In a "voltage mode" PWM, this is prevented by reducing the bandwidth.

In "current mode" converters, there is a very fast feedback between the L1 and the Q1. This servo loop will tend to oscillate. Slope compensation adds an artificial ramp to the signal indicating the rising current in L1. This reduces the gain of the fast feedback path from the current and thus reduces the risk of oscillation.

Reply to
MooseFET

I especially like the typo: Slop compensation. I'll probably use that somewhere. ;)

Cheers,

Phil Hobbs

Reply to
Phil Hobbs
[snip]

Yep. I've had a lot of client's circuits that needed "slop compensation" ;-)

...Jim Thompson

-- | James E.Thompson, P.E. | mens | | Analog Innovations, Inc. | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | Phoenix, Arizona Voice:(480)460-2350 | | | E-mail Address at Website Fax:(480)460-2142 | Brass Rat | |

formatting link
| 1962 | America: Land of the Free, Because of the Brave

Reply to
Jim Thompson

Anytime I see something like this

--[messed up signal]>---+ | Summation>-----Output | [correction signal]>--+

I call it feed forward. There is no loop. Kinda like bringing matter and anti-matter together.

However, there is a loop when both feedforward and feedback are used.

I can't say much about slope comp. I'm still learning that myself..

D from BC British Columbia Canada.

Reply to
D from BC

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.