5 volt 1000 mA regulator or reduction to

Thread Starter

Schidtztorm

Joined Oct 1, 2019
44
Here's a sim of a crude buck converter.
A standard multivibrator comprising two L6s drives a P-MOSFET to chop the 12V supply . The chopped supply has an average voltage set by the duty cycle of the multivibrator and is smoothed by L1 and C3 to give 5V out to the load. The duty cycle is adjusted by varying R2 or R3.
View attachment 187395

I would not recommend this for driving critical loads, since there is no negative feedback to regulate the output voltage.
OK. I'm familiar with the multi vibrator. It wouldn't be practical for charging things. But I like it though! I could also half wave rectify it so it pulses. I think it would be great if I was using a 12 volt battery.
 

Thread Starter

Schidtztorm

Joined Oct 1, 2019
44
They should.
The current through the transistors is quite low (less than a mA).
Yes.
For 12.5V, you should increase the value of R3 from 14kΩ to 15kΩ.
Thank you! I also believe I found a zener. It's a small glass one with 5V on it. I would think it has to be.
Do I run it strait to the base from the positive in reverse? Should I still use the resistors?
Is there a similar differential design for N channel fets? Just for future reference.
You have been a great help so far. Much appreciated!!! I'm learning a lot.
 

crutschow

Joined Mar 14, 2008
23,152
I found a zener. It's a small glass one with 5V on it. I would think it has to be.
Do I run it strait to the base from the positive in reverse? Should I still use the resistors?
Yes, you need one resistor to provide a bias current through the Zener to establish it's voltage (below).

upload_2019-10-6_23-54-5.png
Is there a similar differential design for N channel fets? Just for future reference.
upload_2019-10-7_0-2-56.png
 

Thread Starter

Schidtztorm

Joined Oct 1, 2019
44
Yes, you need one resistor to provide a bias current through the Zener to establish it's voltage (below).

View attachment 187462

View attachment 187463
Thank you so much for that!
Yes, you need one resistor to provide a bias current through the Zener to establish it's voltage (below).

View attachment 187462

View attachment 187463
Thank you so much for that. Do I need to redo the math with the use of the L6s? And/or the N channel or other fet?
 

Thread Starter

Schidtztorm

Joined Oct 1, 2019
44
Thank you so much for that!

Thank you so much for that. Do I need to redo the math with the use of the L6s? And/or the N channel or other fet?
Also, do those two resistor values have any tolerance? 4.99k and 49.9k? >Can I use 5k and 50k? Or will it mess up the output or cause other nonsense?
 

Alec_t

Joined Sep 17, 2013
10,267
Here's an alternative simple switch-mode converter, with feedback to stabilise the output at about 5V. The voltage 'reference' is the forward voltage of the base-emitter junction of Q1. Since this drifts with temperature, the converter output drops by about 0.2V when temperature goes from 15C to 30C.
5vReg-simple3.png
Q1 and Q2 aren't critical and could be L6s.

Edit: The downside of 'simple' is that if the feedback fails the load gets the full 12V.
 
Last edited:

crutschow

Joined Mar 14, 2008
23,152
The heat sink I have is about 1 1/4 wide and 1" tall. 1/16 thick. Will that suffice for 7 watts?
Is that just a flat plate?
Sounds like it's not adequate unless you perhaps have a fan blowing on it.
You need a heatsink with < 10°C/W thermal resistance.
 

Thread Starter

Schidtztorm

Joined Oct 1, 2019
44
Here's an alternative simple switch-mode converter, with feedback to stabilise the output at about 5V. The voltage 'reference' is the forward voltage of the base-emitter junction of Q1. Since this drifts with temperature, the converter output drops by about 0.2V when temperature goes from 15C to 30C.
View attachment 187569
Q1 and Q2 aren't critical and could be L6s.

Edit: The downside of 'simple' is that if the feedback fails the load gets the full 12V.
Thank you for that! How could feedback fail? Anything I could do to prevent that?
Is that just a flat plate?
Sounds like it's not adequate unless you perhaps have a fan blowing on it.
You need a heatsink with < 10°C/W thermal resistance.
OK, so this suckers going to get hot for real! The only other one I have is actually a bigger but thinner heat sink that is around 2.5" X 4" that goes over the whole works. As in over top of the components. Will that do? or should I use the smaller with a fan.
I'm going to stuff this thing in a cpap machine which blows air. I could cut the inside of the intake so it sucks heat out also if you think that will do!
Can you e-mail me in private? Schidtztorm@protonmail.com
There's details I don't want addressed in public.
 

crutschow

Joined Mar 14, 2008
23,152
I have is actually a bigger but thinner heat sink that is around 2.5" X 4" that goes over the whole works. As in over top of the components. Will that do? or should I use the smaller with a fan.
It's hard to say which would be better.
If you are size limited and can't use a larger heat-sink, then you may need to use a fan with either of the sinks you have.

Not sure if the CPAP machine would blow enough air for that.
Can you e-mail me in private?
I normally don't do that unless you need to really discuss something in private that's necessary for me to know, to answer your questions.
 

Thread Starter

Schidtztorm

Joined Oct 1, 2019
44
It's hard to say which would be better.
If you are size limited and can't use a larger heat-sink, then you may need to use a fan with either of the sinks you have.

Not sure if the CPAP machine would blow enough air for that.
I normally don't do that unless you need to really discuss something in private that's necessary for me to know, to answer your questions.
No problem, not quite that important.
As far as that machine not blowing enough air; just how hot is this thing going to get? Cherry red? They blow as much as a computer fan if the machine is being used. Much more if it's just blowing air unrestricted.
If it's gonna get that hot. Then this whole project is moot. The plastic housing of the machine would catch fire. I'm gonna put this thing together next week. I want to build it anyways just to see what it's gonna do. I guess we'll see then.. Lol. Wether it can be used inside of something or not is yet to be seen.
Your help in this matter has been nothing less than phenomenal. Thank you!
 

Thread Starter

Schidtztorm

Joined Oct 1, 2019
44
Wire disconnect or M1, Q1 or R2 short-circuit. No simple solution, but R1/R2/Q1 could be duplicated I suppose.
Is that a common problem with these components in this configuration? Do you think Mr Crutschows model would be a better or worse? Also, how hot do you think the FET will get. Read the above quotes to see why I'm asking.
 

Alec_t

Joined Sep 17, 2013
10,267

crutschow

Joined Mar 14, 2008
23,152
how hot is this thing going to get? Cherry red?
No.
The MOSFET will short or open long before it gets cherry red.
The maximum allowed junction temperature of a MOSFET is typically 150°C.
Beyond that the device starts to deteriorate.
 

Thread Starter

Schidtztorm

Joined Oct 1, 2019
44
As the circuit has never been built the answer is no. :)

In what way?

Depends on the FET, its package, the heatsink size and type, the load and environmental conditions.
I meant in terms of not only heat but in risk of the load taking the full 12 v
The heat sinks I have are (1) flat 1 1/4" wide X 1" tall. (2) thinner sheet aluminum 2.5 X 4".
 

Alec_t

Joined Sep 17, 2013
10,267
I meant in terms of not only heat but in risk of the load taking the full 12 v
The linear regulator uses about 12.5W when driving a 5W load at 5V, with >7W in the FET..
The switch-mode one uses about 5.8W when driving that same load, of which about 460mW is dissipated in the FET.
Both regulators rely on feedback to limit the output voltage to 5V, so could let 12V through if the feedback fails. The switch-mode one has a few more components. In general, the more components the greater the chance of a fault.
 

Thread Starter

Schidtztorm

Joined Oct 1, 2019
44
No.
The MOSFET will short or open long before it gets cherry red.
The maximum allowed junction temperature of a MOSFET is typically 150°C.
Beyond that the device starts to deteriorate.
300 deg F is hot but not fire hot. Thank you. I'll look for a formulae for a sink that will work wiht that temp
The linear regulator uses about 12.5W when driving a 5W load at 5V, with >7W in the FET..
The switch-mode one uses about 5.8W when driving that same load, of which about 460mW is dissipated in the FET.
Both regulators rely on feedback to limit the output voltage to 5V, so could let 12V through if the feedback fails. The switch-mode one has a few more components. In general, the more components the greater the chance of a fault.
Thank you. You guys have all been great. I'll let you know how it works when I put them together next week. I have to find some good sinks.
 
Top