mosfet driver circuit does not work!

[teeeelo]

I tried building this and I get no oscillation from pin 6. what could be wrong?

https://electronicdesign.com/site-files/electronicdesign.com/files/archive/electronicdesign.com/files/29/6230/figure_01.gif

 

[dr pepper]

In my experience mosfet divers are very sensitive to supply variations, you have to make sure the ground is good hefty connection, as well as the power, and you need to have the driver well decoupled, a large electro and a couple of ceramics.
Did you build a pcb to the original designers spec, or did you build it on a breadboard or veroboard, if the latter 2 then thats gonna be your problem, last time I tried using a mosfet driver on a breadboard it went kapow along with the fet.

 

[teeeelo]

explain why this is so important. i am talking about no osc from the get go. can there be so much leakage from perfboard?? i don't understand.

 

[alec_t]

High frequency noise pickup (by the stripboard/breadboard tracks and component leads) and oscillation due to stray inductance/capacitance are potential problems if a properly designed pcb is not used.

 

[ronsimpson]

What a strange circuit! No feed back. No regulation.

You should get a small sawtooth wave form on pin 4. 1.7 volts
The supply voltage (from pin 5 to pin 7) should be between 8V and 30V.

You could get hurt playing with the power line. Please get away from the power line and apply 12V across C4 until you get it oscillation.

 

[alec_t]

Do you have pin 2 connected to circuit 'ground' (pin 5)?

 

[teeeelo]

yes i do. what else could cause this circuit not work?

 

[blueroomelectronics]

Show us a photo of the circuit you built. Also as noted above beginners should stay away from directly coupled mains circuits.

 

[ronsimpson]

TROUBLE SHOOT!
See below. If the supply voltage is too low you will get nothing.

What a strange circuit! No feed back. No regulation.

You should get a small sawtooth wave form on pin 4. 1.7 volts
The supply voltage (from pin 5 to pin 7) should be between 8V and 30V.

You could get hurt playing with the power line. Please get away from the power line and apply 12V across C4 until you get it oscillation.

 

[teeeelo]

here is the driver I am using a uc3844 chip

 

[alec_t]

I don't see a current sense resistor between the FET source and ground?

 

[ronsimpson]

I don't see a current sense resistor between the FET source and ground?

The current sense resistor has little function here. (no feed back, no regulation, no current flow) so if one function is missing it does not matter. (Good eyes to see the missing resistor. Alec_t)

Where is C4?
Where is C2? that makes the oscillator.
Where are the answer to the questions I asked before?
What supply voltage?

 

[alec_t]

The current sense resistor has little function here.

I beg to differ. Without it acting as a pull-down, as soon as pin 6 of the IC goes high the FET source connection to Isense pin 3 (via the resistor from node c58 to node c48) will pull up pin 3. This will be interpreted as over-current (regardless of the actual FET load current) and the pin 6 output will be switched back off.

Where is C2? that makes the oscillator.

At node a49.

 

[ronsimpson]

Where is C2? At node a49.

Yes I see it I was off one row. Then it is C2 that is missing.

The current sense resistor......I beg to differ.

I thought is was shorted but you are right it is open.

With out C2 the current sense could (will) trip early. That will also shorten the duty cycle to about 0%.
With a open current sense resistor the part will trip off very fast. That will also shorten the duty cycle to about 0%.
It looks like the MOSFETs drain is shorted to supply with out a transformer. That will also shorten the duty cycle to about 0%.

Maybe there is no supply voltage. That will also shorten the duty cycle to 0%.

 

[teeeelo]

Your right But in previous iterations there was a current sense resistor.

 

[teeeelo]

why is there a large knee at the fall time? Can the pulses be sharpened somehow?

 

[alec_t]

Your right But in previous iterations there was a current sense resistor.

It's not helpful if the picture you posted isn't a pic of the circuit which had the problem

If the trace is showing the FET gate voltage then gate capacitance plus scope input capacitance plus breadboard capacitance could account for the 'knee'.

 

[Dr_Doggy]

maybe intentional?
according to schematic diagram, pin 1 is floating and pin 2 is grounded, not at all what s going on in the photo's....

 

[teeeelo]

You guys don't understand that the schematic in the first post i could not get to work at all. I spent many hours and alot of chips trying to figure out why. Then it hit me to just use the stuf supplied in the datasheet which i did here. A big problem is that many circuits posted do not work.

 

[blueroomelectronics]

You're right, it's always best to consult the uc3844 data sheet. It does include a schematic that should work just fine, it also shows the part as obsolete. Keep in mind breadboards are not the most reliable way to test circuits, especially when high currents, voltages or frequencies are part of the design.