Ok, look my opinion, even it doesn't worth anything because of those damn tags which will kill my email somewhere.... For someone who need to learn something, this is a great email. But for a product which must become a mass product made in Taiwan (in reality in The People Republic of China because in Taiwan there are only some poor offices [The political phrase was democratical removed from here by the own author, try to guess what where about] :) will be never a better H bridge than let say BA6209 from Rohm, KA3100D from Fairchild, 2xMC1472 from Motorola, LB1838 from who knows and maybe another dozens of specialised IC for H or H/2 bridge. The weak point of this design is the RL amortizing circuit (22 ohm in parallel with the motor, because it's sourcing current from the batteries when it should not). Thar 22 ohm resistor may be connected in a different way with just one transistor which will be ON only when the H command stops. I let the lighty brains from this list to show the way, even I know you already see it. best regards, Vasile On 7/16/05, Russell McMahon wrote: > The "better H bridge" didn't get used in the product for the initial > run so there's time to have another crack at it. > I'll revise broaden and simplify the specs based on input here and > recent experience. > > This is very similar to prior requirement BUT also different enough to > be worth resubmitting. > Note that the MUST MEET specs MUST be met. There's no point in a > design that fails on any point. > > _________________ > > Short version: > > A battery powered device uses an H bridge to drive a small geared > motor. The present H bridge design drops a minimum of several volts > across the bridge components. The motor has a desired terminal > operating voltage of from 3 to 4 volts and maximum of 4.5 volts. > Object is to design a circuit that maximises battery usage and allows > motor to start (high surge current) when the battery is very flat. > > Long version > (necessary to read for proper understanding). > > A small DC motor with gearbox is used in a piece of consumer > equipment. The motor is operated at essentially constant mechanical > load. For practical purposes the motor may be considered to implement > a "servo". The output is a rotary shaft which can rotate through about > 270 degrees. A 5k pot is mounted on the shaft to indicate position. > When operated with 3v across the motor the pot can be swept from end > to end in 20 seconds. At 4v end to end it takes 15 seconds. At 2.5v it > takes about 25 seconds. When running the motor draws nominally 100 to > 200 mA > at 3.5v for an effective motor load of 3.5v/200 mA =~~12 ohms or 24 > ohms at 100 MA. Current draw varies with loading variations. A 22 > ohm resistor is mounted in parallel with the motor to assist stopping > times > > When starting the motor presents a very brief, *VERY* low impedance > load. If battery is low then limiting current at startup may cause it > to stall, draw high current (300 mA plus) and not turn. This is not > nice :-). Voltage and speed limitation schemes need to address startup > issues. > > The motor can be controlled to stop at any of 16 positions. A > controller (not part of this design) provides 2 x 0v/5v signals to > control motor direction. Up to 3 mA may be drawn from the drive > signals if essential although less is better. Signal voltage droop at > 3 mA is 0.5v max. > > Motor is on or off, there is no PWM. When signal A is high motor runs > clockwise. When signal B is high motor runs anticlockwise. Signals A & > B are never high simultaneously and there is a substantial delay > between direction changes (say 500 MS minimum). The controller moves > the motor randomly (for our purposes) between any of the 16 positions. > Sometimes it may take a single "step" and sometimes go all the way fro > min to max, or any other combination. > > Operation is from 4 x C cell batteries. While these are specified to > be "Alkaline" customers may use any available "1.5v" battery available > on the retail market including eg Le Clanche or Zinc Chloride. Battery > voltage when new is therefore as much as 6.5v. A nominally 6v mains > adaptor may be plugged in - when it is the battery is disconnected. On > light loads (motor not running) the adaptor voltage may be as high as > 10 volts. With motor running adaptor voltage is PRESENTLY about 8v or > 6v under certain circumstances (optional backlight turned on). This is > only PRESENTLY as the existing H bridge has substantial voltage drop > (2 to 3 volts) and this is a factor in motor speed and adaptor > loading. > > The existing driver is essentially as per attached circuit. This uses > a emitter follower driver in both high and low sides. *As a > consequence of the existing design* the low side transistor drops at > least 1 x Vbe plus saturation voltage and the upper 1 x Vbe + > Vbase_resistor_drive_voltage plus saturation. > > An Alkaline C cell is rated at about 7500 mAH and has an endpoint > voltage of about 1 v (arguably 0.9v but there is little difference in > energy content between these points). A Zinc Chloride / Le Clanche > call has maybe 1500 - 3000 mAH capacity. While the "endpoint" is say > 1v/cell the actual terminal voltage drops severely under load as the > battery ages and rises when load is reduced. A battery with 5 volt low > load (10 mA) terminal voltage may drop to say 3.5v at 250 mA load. > > Will be made using auto-loading where possible. Probably SMT but MAY > be through hole. > > FWIW: Present design has NO inductive spike suppression and is > reliable. > > REQUIREMENTS > > When motor is operating operate at from 3 to 4 v at nominal 200 mA > supply where reasonably possible. 22R parallel resistor current is > extra. > > Must never operate motor above 4.5v across motor. > > Must aggressively minimise all up cost of production (made in Taiwan > or > China). (Target component cost would be a few tens of cents US). > > Use of existing and common components is preferred BUT, subject to > cost > requirement, may use any components desired that are nearly certain to > be attainable in Taiwan at a few weeks notice. Importation acceptable > as long as instantly available and subject to cost restraints. (eg > Digikey sourcing is fine if its fits the budget) > > Should start and operate on as low a battery voltage as reasonably > possible. Mustn't stall and fail to start in any situation where old > circuit would have started motor. > > Must operate on up to 8.5v operating and 10 volt instantaneous > starting voltage while not overspeeding. > > Must remove power from motor as soon as control signal goes low. > > Ask if anything unclear or contradictory. > > > > > > > Russell McMahon > > -- http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist