Robot Efficiency & other Robot stuff

[Oznog]

Your schematic shows an external diode in parallel with each PMOS.

This should not be necessary. All PMOS that you get off-the-shelf have an internal body diode inside in this direction already. You can read your spec sheet but usually they are capable of carrying substantial current.

Your schematic lacks a capacitor on the output of the reg. This could cause the reg to oscillate and potentially blow your chip.

You need to put a ceramic cap from Vdd to Vss right at the PIC chip, especially on a power circuit like this. I already had this trouble more than once- nearby power consumers, even though they didn't even draw power out of the reg, were causing the PIC to latch up.

 

[williB]

Your schematic shows an external diode in parallel with each PMOS.

This should not be necessary. All PMOS that you get off-the-shelf have an internal body diode inside in this direction already. You can read your spec sheet but usually they are capable of carrying substantial current.

Your schematic lacks a capacitor on the output of the reg. This could cause the reg to oscillate and potentially blow your chip.

You need to put a ceramic cap from Vdd to Vss right at the PIC chip, especially on a power circuit like this. I already had this trouble more than once- nearby power consumers, even though they didn't even draw power out of the reg, were causing the PIC to latch up.

those diodes are allready soldered on the board , wont hurt anything i hope..

You need to put a ceramic cap from Vdd to Vss right at the PIC chip,

ok i'll do that..
0.01 OK ?

 

[williB]

the bottom trace is the pic output.
the top trace is taken from the drain pin of the FET.??
when the gate is turned on by the PIC the drain drops to 0V for a very short time..then rises to half the battery voltage(battery is almost 4V)
it oscillates there until the gate drops then goes back to plus ..
But it should stay at 0v for the whole time the gate is on .
any ideas ???

 

[williB]

one thing i added which is not on the schematic
is a 100k resistor to ground at the gate input

 

[williB]

i fixed the problem..
the diode i placed from the PIC to the FET (to protect the pic) was causing the Fet to not turn on all the way..so that little upward curve and the ripple was caused by the diode.. now the top trace is looking much better

 

[Oznog]

What's the part number for those MOSFETs?

Ones with a threshold of 4V don't turn on very strong with 5V, thus you will only be able to use a fraction of the FET's current rating. Temp variations may make it worse. I had a PWM buck converter working like total crap due to that. Actually, the vgs should have been high enough for the 1 amp current involved, but despite that it wasn't doing even that- I'm not sure why.

The waveform you have suggests just that. The inductance of the coil initially resists the sudden change in current, it appears as a high impedance load so the MOSFET's able to drive vds pretty low. Then the inductor's current increases, you face only the resistance of the stepper coil, and the MOSFET is unable to conduct enough current to keep vds low.

There are ones with thresholds of around 2V which switch much better.

 

[williB]

What's the part number for those MOSFETs?

Ones with a threshold of 4V don't turn on very strong with 5V, thus you will only be able to use a fraction of the FET's current rating. Temp variations may make it worse. I had a PWM buck converter working like total crap due to that. Actually, the vgs should have been high enough for the 1 amp current involved, but despite that it wasn't doing even that- I'm not sure why.

The waveform you have suggests just that. The inductance of the coil initially resists the sudden change in current, it appears as a high impedance load so the MOSFET's able to drive vds pretty low. Then the inductor's current increases, you face only the resistance of the stepper coil, and the MOSFET is unable to conduct enough current to keep vds low.
There are ones with thresholds of around 2V which switch much better.

They are IRF 510s
i removed the diodes(which i was using as PIC protection.)
and I lowered the battery voltage (which drives the FETs) to just under 4V so they turn on much better with the PIC output..

There are ones with thresholds of around 2V which switch much better

which ones?

 

[Oznog]

IRF510 is not a logic level device, not a good choice. Go to Digikey, search for MOSFET, and select "logic level". Try these from **broken link removed**

IRL520N N-ch MOSFET logic-level gate voltage, 10A, 100V TO-220 $0.67
IRL530N N-ch MOSFET logic-level gate voltage, 17A, 100V TO-220 $0.87
IRL540N N-ch MOSFET logic-level gate voltage, 36A, 100V TO-220 $1.17

 

[williB]

IRF510 is not a logic level device, not a good choice. Go to Digikey, search for MOSFET, and select "logic level". Try these from **broken link removed**

IRL520N N-ch MOSFET logic-level gate voltage, 10A, 100V TO-220 $0.67
IRL530N N-ch MOSFET logic-level gate voltage, 17A, 100V TO-220 $0.87
IRL540N N-ch MOSFET logic-level gate voltage, 36A, 100V TO-220 $1.17

Thanks , i didnt know such a devive existed.. :wink:

 

[gurgalesci]

My robot finally moved also!

After 3 days of sitting in front of my robot waiting and praying that it would move, it finally did. I spent 72 hours nonstop sitting in front of it. I also have those brake things that someone was mentioning earlier. They don't work very well. Whenever I try to use them, the robot only goes faster and without a brake. Oh well, the brake cost me $5000 to install and buy the brake. As a result from the problem, my robot was badly damaged and I hired a person from Holiday Inn Express to fix it. He did, but during his repair, the Holiday Inn Express Magic wore off and he made the robot look even worse. Now, the robot just spins around in circles signing "Yankee Doodle." I guess its kinda cool around Independence Day.

 

[l337 8oY]

Re: My robot finally moved also!

It's people like you who distroy the digital world. I spit on your earthian ways! Soon us electronic newbies will take over the world and purge your kind from existance. DIE EVIL BEING. Your life form will heed to the call of micro chips. MUHAHA

BTW i never knew that device existed either!

 

[jrz126]

Re: My robot finally moved also!

After 3 days of sitting in front of my robot waiting and praying that it would move, it finally did. I spent 72 hours nonstop sitting in front of it. I also have those brake things that someone was mentioning earlier. They don't work very well. Whenever I try to use them, the robot only goes faster and without a brake. Oh well, the brake cost me $5000 to install and buy the brake. As a result from the problem, my robot was badly damaged and I hired a person from Holiday Inn Express to fix it. He did, but during his repair, the Holiday Inn Express Magic wore off and he made the robot look even worse. Now, the robot just spins around in circles signing "Yankee Doodle." I guess its kinda cool around Independence Day.

:?: :roll:
Sounds like you are either 8 years old or on drugs

 

[williB]

i thought i would pop this up to the top , but it has been a while since i did any work on it ..