Loss of voltage with MOSFET

[Jouellet]

Hi folks,

I need to pick your brain here...

I'm designing a circuit, that will control the speed of a DC motor. it's a geared motor (electric window motor)

The circuit looks like this:



I know that the voltage out of the charger is far from being perfect, but having a battery shall make it look better ! On the battery pole/charger output, without any load, I read 14.5 V.

If I connect the motor directly , the voltage drops to 12.45V, drawing 2,25 amps. If I try to stop it with my hands, I can't get close to do so .

If I use my circuit, the voltage (read at the motor) drops to 11.31 V and I can (fairly) easily stop it with my hands. Current is about 2.5 Amps.

I would have expected the voltage drop to be about 0.055 V, as the spec sheet (included in this post) of the MOSFET (IRLZ44) mention that Rds(ON) is about 0.022 ohm * 2.5 amp

I'm using a 7805 to "emulate" the Arduino that will eventually control all this. The IRLZ44 was selected because of its ability to be driven by a TTL signal.

As of now, this is only half the circuit, as I want to eventually end up with a full H-bridge, allowing forward-reverse operation. If I've lost 1.15V thru the first MOSFET, I'm afraid to lose another 1.15V thru the second one ....

I have tried to put 2 MOSFET in parallel, and it's not any better !

What am I doing wrong ?
Any circuit suggestion for a good H-bridge that would allow as much voltage as possible to my motor ?

I bought a "commercial" bridge off Ebay, and it's not any better...

Any other suggestion ?
Would I get better result with a power transistor ?

 

[CDRIVE]

The 7805 likes to see ~ 3.3uF -10uF from the output pin to the neg rail. See if that helps.

Chris

 

[Jouellet]

Won't change anything.

I tried to connect +12 V to the gate, and I gained maybe .4-.5 V at the motor

 

[Bluejets]

http://forum.arduino.cc/index.php?topic=74718.0

 

[hevans1944]

I see no reason for the excessive voltage drop across the MOSFET. This is a "logic level" device according to the datasheet. Try replacing the MOSFET with a new one.

 

[Colin Mitchell]

have you connected the FET around the wrong way?

 

[CDRIVE]

I see no reason for the excessive voltage drop across the MOSFET. This is a "logic level" device according to the datasheet. Try replacing the MOSFET with a new one.

The only other thing I see there is the absence of a series gate resistor (~100Ω) to limit instantaneous inrush current to the gate. This will increase in importance when driven by the Arduino. However the lack of this resistor should cause no ill effects well after gating it on. So my guess is your guess!

Chris

 

[Alec_t]

Rds(ON) is about 0.022 ohm

That is the rated value when Vgs=10V, but your FET has Vgs <=5V so Rds(on) will be in the 0.025-0.035 Ohm range so will drop more voltage.
What gauge interconnecting wire are you using? Thin gauge and poor battery connections would add significant resistance.

 

[CDRIVE]

That is the rated value when Vgs=10V, but your FET has Vgs <=5V so Rds(on) will be in the 0.025-0.035 Ohm range so will drop more voltage.
What gauge interconnecting wire are you using? Thin gauge and poor battery connections would add significant resistance.

Yeah, I downloaded the datasheet and peeked at the Vgs / Rds curve but my brain isn't working at all in the mornings. Come to think of it,.. Not to good in evenings either!

On another note, I hope he's using a heatsink.

Chris

 

[hevans1944]

Even with the slightly larger Rds(ON) of 0.035 Ω with lower gate drive, we are still only talking 85.7 millivolts drop at 2.5 amperes. I would measure the voltage from drain to source with the MOSFET gated ON to verify it is the culprit here.

As @Alec_t suggested the problem could also be in the wiring or connections... or as @Colin Mitchell suggested, the MOSFET could be connected incorrectly. One final suggestion is to measure the voltage across the battery terminals with the motor running as well as not running. SLA batteries, especially older ones, drop their terminal voltage a huge amount with brief discharges of several amperes and take a long time to recover, if they ever do. Inadequate power supplies are the bane of prototyping. Make sure your power source is rock-solid before looking for problems elsewhere.

 

[Minder]

I agree on the source. Did you try it on a good confirmed source such as a automotive battery?
M.

 

[Kiwi]

"On the battery pole/charger output, without any load, I read 14.5 V. If I connect the motor directly , the voltage drops to 12.45V, drawing 2,25 amps."
This voltage drop with such a small load is a worry.

What is the Gate to Source voltage?

Have you tried connecting the gate directly to the battery?

 

[CDRIVE]

I'm designing a circuit, that will control the speed of a DC motor. it's a geared motor (electric window motor)

The circuit looks like this:

View attachment 23105

I know that the voltage out of the charger is far from being perfect, but having a battery shall make it look better ! On the battery pole/charger output, without any load, I read 14.5 V.

If I connect the motor directly , the voltage drops to 12.45V, drawing 2,25 amps. If I try to stop it with my hands, I can't get close to do so .

If I use my circuit, the voltage (read at the motor) drops to 11.31 V and I can (fairly) easily stop it with my hands. Current is about 2.5 Amps.

Perhaps some clarification of your statements are in order. Your schematic indicates a battery connected to the charger's output but your statement above belies that. Does your circuit include the battery or not? If the answer is no then the problem is obvious. If the battery is absent the whole circuit is being supplied with full wave unfiltered DC.

My guess is I'm just misinterpreting your statement. Or am I?

Chris

 

[Jouellet]

Bluejets: thanks for the link, but it did not bring any solution

Colin Mitchell: connected properly, re-checked

Alec_t: you may have portion of the answer. I changed the wire size, and noticed an increase of voltage. I was using protoboard jumper.

Cdrive: yes, a heatsink was/is installed

Hevans1944: I'm sure the battery is good. I connected the motor directly for 15 minutes and it was still strong ! (yes, I did recharge the battery after that test, in case someone ask !)

Kiwi: Gate to source was 5V to begin, but I also tried directly from the battery

Cdrive: battery AND charger were present when testing


So, my next step is going to use bigger wires. I will probably make a PCB with large trace to avoid any bad contact that alligator clips or jumper may create (I have a few other board to make anyway)

will let you know of the results !

 

[dorke]

Instead of letting everybody guess .

Please measure the voltages as in the pic and post them .
All voltages referenced to GND of battery
MOSFET and Motor and battery directly on pins.

 

[hevans1944]

... will let you know of the results !

Measure the voltage drop from drain to source at the MOSFET to make sure it isn't part of the problem.

When you are ready, MOSFET H-bridges are more efficient than BJTs, and there are integrated circuits available to drive them, including high-side charge pumps for the top legs of the H-bridge and feedback PWM current control... everything you need in one package... just add MOSFETs, heat sinks, and heavy conductors.

 

[CDRIVE]

So, my next step is going to use bigger wires. I will probably make a PCB with large trace to avoid any bad contact that alligator clips or jumper may create (I have a few other board to make anyway)

will let you know of the results !

The first thing I must say is you get 5 gold stars for clearly answering everyone's questions. Honestly,.. it's like pulling teeth sometimes! You're a refreshing change!

Now about alligator clips... EEEEEEEEEK!!! They're a dichotomy. Indispensable and yet the most ill devised or even advised concept ever invented or implemented. When testing circuits like yours they're absolutely, invariably, undeniably and unarguably VERBOTEN!

BTW, what gauge wire are you currently using?

Lastly, please fill in the voltages on the schematic as Dork asked.

Cheers,
Chris

 

[CDRIVE]

I forgot something in my last post. We should have asked for photos of your test setup a long time ago. Please do.

Chris