NE555-to Supply Hi-Drive N-MOSFET

[Kerim]

I couldn’t get ICs for driving a high N-MOSFET (whose source is connected to the battery 12V+).

First, I tried designing the following circuit (its LTspice files attached) to source about 20mA with a voltage about 10.5V (above the battery positive terminal).

The approximated formula of this voltage source is (based on the simulator tests):

V = 13.257 – 136.7*I where I is the DC load current [in Amp]

I wonder if there is a better idea for this by also using low-cost discrete components (no special IC or ferrite).

 

[crutschow]

Why not drive it with the 555 operating from 12V battery with the output connected to a 5k ohm pull-up resistor to V+?

 

[danadak]

Page 28 some possibilities :

https://www.ti.com/lit/ml/slua618a/slua618a.pdf?ts=1676282389514&ref_url=https%253A%252F%252Fwww.google.com%252F

Regards, Dana.

 

[Kerim]

Why not drive it with the 555 operating from 12V battery with the output connected to a 5k ohm pull-up resistor to V+?

If I remember well, I think I also tried something like this, first.

By the way, the 12V battery is of the acid type. Its voltage may vary from 10V to 16V. So, in order to have a stable well-defined voltage source, I chose 8V (10V - 2V needed by LM7808). And by adding 2 transistors, as a rail-to-rail buffer, allow the use of 555 ICs in case they are not genuine, and their output are somehow weak.

Do you think I can get 20mA at 10V without these two transistors while supplying the 555 IC with 8V?
Thank you.

 

[Kerim]

dana,

Thank you for the link.
Unfortunately, I am not privileged to access any site of TI. Thank you anyway for your care.

 

[danadak]

dana,

Thank you for the link.
Unfortunately, I am not privileged to access any site of TI. Thank you anyway for your care.

Here is doc attached.

Regards, Dana.

 

[crutschow]

If I remember well, I think I also tried something like this, first.

So why didn't it work?

the 12V battery is of the acid type. Its voltage may vary from 10V to 16V.

The 555 should tolerate that range of voltages.

 

[Kerim]

Here is doc attached.

Regards, Dana.

Thank you, Dana.
As you know, on figure 23, Vdrv is supposed to be generated by a driver IC as IR2210 for example. I try generating it by using discrete components instead and the timer 555 which I have many.

For instance, I will likely use the same configuration (on figure 23) to turn the hi-NMOS on and off.

Regards,
Kerim

 

[Nigel Goodwin]

Thank you, Dana.
As you know, on figure 23, Vdrv is supposed to be generated by a driver IC as IR2210 for example. I try generating it by using discrete components instead and the timer 555 which I have many.

For instance, I will likely use the same configuration (on figure 23) to turn the hi-NMOS on and off.

Regards,
Kerim

If you want a discrete driver, how about this:

Which I used in some PWM experiments I did, and use in various products since. I compared it with the IR2210 etc, and it had similar switching speeds.

The extra FET Q6 was on my test boards, so I could select it (or not) using a jumper across PL2, so the output FET was OFF when the PIC feeding it wasn't running (or wasn't present).

For a high side FET, you simply replace the low side one with a PMOS one at the top - the collector of Q8 of course goes to the positive supply rail. This was another reason for Q6, so it gave options - my test boards used both NMOS and PMOS, two lot's of drivers, and you could simply select either for the PWM, by switching the unused one ON with the PIC.

If you're wanting to use an NMOS at the top as well?, then go with the IR2210 type chips, which generate an extra voltage to switch it properly.

 

[Kerim]

So why didn't it work?

I will try to repeat it to give you (and myself) the answer because I tried many methods.

The 555 should tolerate that range of voltages.

This is likely true for genuine ICs of NE555. I try my best to avoid stressing a component at any of its limits mainly if I have a doubt that it is not made by a trusted manufacturer (you know, not all 555 are made in USA ).

Since in the circuit I used, the charging/discharging currents are not regulated in any way, their values vary appreciably from 10V to 16V supply (assuming at 10V, Vdrv is about 10V for 20mA load).

I don't say what you proposed is wrong, I just prefer doing things with minimum risk and as stable as possible (though this costs me more work time and components in general ).

 

[Kerim]

If you're wanting to use an NMOS at the top as well?, then go with the IR2210 type chips, which generate an extra voltage to switch it properly.

Thank you for sharing your work.

I am afraid that I can't get IR2210 (or equivalent). This is why I try to generate Vdrv (10V higher than the battery voltage) by using discrete components.

 

[Nigel Goodwin]

Thank you for sharing your work.

I am afraid that I can't get IR2210 (or equivalent). This is why I try to generate Vdrv (10V higher than the battery voltage) by using discrete components.

Do you have to use an NMOS at the top?.

 

[Kerim]

Do you have to use an NMOS at the top?.

You are right, driving PMOS is easier for supplies up to about 20V (Vgs limit).

But the power PMOSs available at the local retailers have an Rson 10 times, if not higher, of their equivalent NMOS (having about the same voltage and current limits). On the other hand, I was looking for a solution that works for 24V acid battery as well.

 

[Nigel Goodwin]

You are right, driving PMOS is easier for supplies up to about 20V (Vgs limit).

But the power PMOSs available at the local retailers have an Rson 10 times, if not higher, of their equivalent NMOS (having about the same voltage and current limits). On the other hand, I was looking for a solution that works for 24V acid battery as well.

Is that Rson actually a 'problem'?, have you calculated what it might cause, or just assumed that 'lower is better'?.

Part of my tests was using both PMOS and NMOS high side devices, and using IR2110 (or similar) for the NMOS versions - and for my application the PMOS are perfectly fine, so I decided against using the IR2110 devices.

 

[Kerim]

Is that Rson actually a 'problem'?, have you calculated what it might cause, or just assumed that 'lower is better'?.

It causes a higher voltage drop (about 10 times higher at best) and needs therefore a much better heatsink.
For instance, I am talking about currents around 100A.
For example (below are available locally):
Rson of IRF9540N (100V-23A) equals 117mR @Vgs=-10V and -11A
Rson of STB75NF75 (75V-75A) equals 11mR @Vgs=10V and 40A

Perhaps on your side, you have a wider and better options.

Part of my tests was using both PMOS and NMOS high side devices, and using IR2110 (or similar) for the NMOS versions - and for my application the PMOS are perfectly fine, so I decided against using the IR2110 devices.

As you said, this depends on the application. In fact, the role of a professional engineer is to choose the optimum solution/devices for every application of interest, as you did... and as I try doing.