Maker Pro
Maker Pro

how can stabilize the mosfet's drain current help (linear amplifier)

tzitzikas

May 9, 2010
16
Joined
May 9, 2010
Messages
16
hi, i have this linear amplifier project http://tzitzikas.webs.com/irf640-1.JPG with 2 irf640 mosfets.
My problem is, when i adjust the both mosfets drain's current at 400mA (current without input signal), after some minutes

when the heatshink is warm (with input signal), the current becomes very high and one of the 2 fet goes off, and i must

replace it. Do you have any idea-schematic (with thermistor or diode) to control this current and to keep it stable at 400mA?
 

55pilot

Feb 23, 2010
434
Joined
Feb 23, 2010
Messages
434
Sorry to say this, but that is a really lame design. The idea that you can control the output current of a MOSFET by just setting its gate voltage open loop is plain stupid and ignorant. As the MOSFETs heat up, their gate threshold voltage is going to decrease, the MOSFETs will turn on more at the fixed input voltage and hence the current through them is going to increase and you will end up with a failure. All quite expected.

It is equally lame idea to have two MOSFETs be driven by a voltage and think that the current through them would be identical. In real life, any two given parts are never than well matched, except by sheer luck or coincidence.

The way this type of oscillators are supposed to work is by tailoring the inductance of the transformer to the drive frequency so that the current through the inductor does not have the time to increase above the desired limit during the time that the MOSFET is on during the cycle.

It is likely that the circuit is supposed to work only when the input is active and rather than providing interlocks to ensure that, they are just hoping that it will always be true.

I suggest you contact the website where you got the circuit from. Maybe someone there has some additional information for you.

---55p
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Wow, push-pull class A amplifier.

This is totally outside of my area, so take this with a grain of salt.

I am surprised that it dissipates 9.6W for an output of 2W. Are you sure the quiescent current should be set so high?

I'm also surprised both MOSFETs turn on at the same voltage. Could one be turning on more than the other? Although with sufficient heatsinking that should not be an issue.

edit: *slaps head* -- Vgs(th) Vs Tj (even *I* should have seen that one)
 
Last edited:

tzitzikas

May 9, 2010
16
Joined
May 9, 2010
Messages
16
Sorry to say this, but that is a really lame design. The idea that you can control the output current of a MOSFET by just setting its gate voltage open loop is plain stupid and ignorant. As the MOSFETs heat up, their gate threshold voltage is going to decrease, the MOSFETs will turn on more at the fixed input voltage and hence the current through them is going to increase and you will end up with a failure. All quite expected.

It is equally lame idea to have two MOSFETs be driven by a voltage and think that the current through them would be identical. In real life, any two given parts are never than well matched, except by sheer luck or coincidence.

The way this type of oscillators are supposed to work is by tailoring the inductance of the transformer to the drive frequency so that the current through the inductor does not have the time to increase above the desired limit during the time that the MOSFET is on during the cycle.

It is likely that the circuit is supposed to work only when the input is active and rather than providing interlocks to ensure that, they are just hoping that it will always be true.

I suggest you contact the website where you got the circuit from. Maybe someone there has some additional information for you.

---55p

here you can see some photos from my linear. sometimes i am using 4 mosfet (2+2)
http://tzitzikas.webs.com/linear2xirf640-1.jpg
http://tzitzikas.webs.com/linear2xirf640-2.jpg
http://tzitzikas.webs.com/linear2xirf640-3.jpg
i don't know if the problem is the heatshink, because i am using a fan for cooler and the heatshink is cold.

do you have any idea to modificate the circuit to keep stable the drain's current? Maybe if i will use 2 lm7805 and 2 different gate voltages with 2 variable 5kohm resistances? But i don't know how to modificate the circuit to achive this.
Any other idea, to keep this circuit but to keep stable the drain's current regardless the temperature?? maybe a thermistor or a diode?
see this. i have construct this also: http://tzitzikas.webs.com/linear500w.jpg (12 mosfets irfp360 6+6) .i have not this problem with this linear. i have test it many times, and i have not destroy any of the mosfets.
(it is using a diode to keep stable the drain's current)
Can i use a 1N4007 diode to the lm7805 like the linear 500w?

(*steve*): The amplifier with 12v voltage with 2 mosfet gives 2watt output r.f power and the whole current is approximately 1,5A, without modulation. I think that the designer have design this linear for linear operation to achive the best quality of the modulation.
The linear's heatsink isn't warm beacause i am also using a fan for extra cooling.
sometimes i adjust the drains current without input signal about 300mA. After some minutes of linear's operation with input signal, and when i stop the input signal this current has been increasedup to 600mA !
i think 400ma is very high current for 2 only mosfets. what value do you think that this current must have?
 
Last edited:

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Your major problem is that Vgs(th) drops with increasing Tj. It results in what is essentially thermal runaway.

You are seeing that in the increasing quiescent current.

The current probably increases slowly at first, then at an ever increasing rate. At some point the device self destructs. It may happen so quickly that the case doesn't have time to get really hot.

Of course that would require that you're running it from a fairly hefty power supply (are you running this from a car battery?). I'd expect your fuse would blow if you have one.

Small resistors in series with the source terminal may well prevent thermal runaway. Maybe 1 ohm with a suitable capacitor in parallel to shunt them at high frequency.

The resistors will effectively decrease the gate voltage as the drain current increases, which *should* counteract the effect of lowering of Vgs(th) with Tj. 1 ohm is only a guess. You'd have to look at how Id varies with Vgs near Vgs(th) and determine what value is appropriate.
 

tzitzikas

May 9, 2010
16
Joined
May 9, 2010
Messages
16
Your major problem is that Vgs(th) drops with increasing Tj. It results in what is essentially thermal runaway.

You are seeing that in the increasing quiescent current.

The current probably increases slowly at first, then at an ever increasing rate. At some point the device self destructs. It may happen so quickly that the case doesn't have time to get really hot.

Of course that would require that you're running it from a fairly hefty power supply (are you running this from a car battery?). I'd expect your fuse would blow if you have one.

Small resistors in series with the source terminal may well prevent thermal runaway. Maybe 1 ohm with a suitable capacitor in parallel to shunt them at high frequency.

The resistors will effectively decrease the gate voltage as the drain current increases, which *should* counteract the effect of lowering of Vgs(th) with Tj. 1 ohm is only a guess. You'd have to look at how Id varies with Vgs near Vgs(th) and determine what value is appropriate.
Hi. Thank you for your answers. The power supply is 12-28vdc/5A with LM338K regulator (the transformer has maximum operation current 6A).
The designer have propose to me, because i have this problem, to increase the gate's series resistors from 4.7ohm to 10ohm and to add at the sources series resistors 0.1ohm. When i did these replacements, the linear was more stable, i don't have problem with the mosfets but the output power was lower. So i kept the 0.1ohm resistances at gates, but i put again 4.7ohm resistors at gates. After this replacement i had problem with the mosfets again.
What value of resistors do you propose to add at gates and at sources? 0.1ohm at sources do you think is so small resistance? 1 ohm resistors at gates do you believe that will decrease more the output power?
What about the capacitors in paraller with source's series resistors? What value of capacitance do you propose to add there.
I must tell you that the linear operates at the frequencies from 800khz up to 1800khz, with ampitude modulation.
About the power supply fuses: i have had a fuse 230v/1A at primary coil of transformer (at 230volt ac), BUT i don't have fuse at the dc voltage. I am thinking to put a 4A or 5A fuse after the power supply capacitors, at the dc voltage. I must tell you, that the linear and the pll have the same power supply and both these devices need 1.5A-2A current at 12vdc, without modulation (only carrier). With modulation the current goes up to 3A. Do you think that a 4Amper fuse at the dc voltage is a good choise? Do you believe that change will save the mosfets from destruction?
 
Last edited:

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Read this. Note that each MOSFET has its own biasing and a thermistor is used to control the thermal runaway (another option is suggested in the text).

Perhaps your designer could incorporate some of these ideas.
 

55pilot

Feb 23, 2010
434
Joined
Feb 23, 2010
Messages
434
Maybe I do not belong in this place. Reading this thread is giving me high blood pressure. If I were to design something like this, I would shoot myself in the head to rectify my incompetence. If were to offer the type of advice that this circuit's designer has offered, I would make sure I had a gun in each hand when shooting myself in the head. Even the fairly well written and reasonable article linked by Steve shows an alarming lack of insight and a lack of understanding and that is a "published" article not a blog (sorry, no hands available for a 3rd gun).

As for the problem at hand, as Steve suggested, adding a resistor from source to ground will add negative feedback on the Vgs and prevent thermal runaway. It will have to be fairly large value, around 0.5 ohms, maybe as much as 1 ohm. But as the OP found out, even an inadequately small resistor kills efficiency. The solution is to start putting multiple MOSFETs in parallel (each with its own gate and source resistor). It is a horribly inefficient solution, but there is only so far you can get by polishing a turd.

---55p
 
Last edited:

tzitzikas

May 9, 2010
16
Joined
May 9, 2010
Messages
16
Read this. Note that each MOSFET has its own biasing and a thermistor is used to control the thermal runaway (another option is suggested in the text).

Perhaps your designer could incorporate some of these ideas.
i read this but i don't know how to modify my circuit to add the thermistor. the schematic is this http://tzitzikas.webs.com/irf640-1.JPG . can you modify this in a photo editor and upload it here?
Also see this linear which i have construct. It uses a 1N4007 diode which is located on heatsink's body, to keep the current constant. Can i use a diode in my case to avoid high current and to protect the mosfets?

The solution is to start putting multiple MOSFETs in parallel (each with its own gate and source resistor). It is a horribly inefficient solution, but there is only so far you can get by polishing a turd.
---55p
Do you propose to add more mosfets? I am thinking to add 2 additional mosfets . like this schematic: http://tzitzikas.webs.com/irf640-4mosfets.JPG What do you think?
 
Last edited:

tzitzikas

May 9, 2010
16
Joined
May 9, 2010
Messages
16
Hi. I tried to remove the capacitors that you said, but the power was very low after these changes.
After that, i keep the capacitors but i change the secondary coil number of turns (of M1) from 3 to 6 turns. after this, the power was 2w at 12volt dc and 8w at 20vdc with very good modulation and only 300mA current (drain's current of 2 mosfets).
My problem now is the next stage this linear http://tzitzikas.webs.com/linear500w.jpg
The manufacturer claims that with only 2w input r.f power it gives 500w at the output at 110vdc (13A current). But with 5,5w input power at 106vdc my linear gives only 220w output power (7A current). I tried to change C4 capacitor's value but the power was low. Do you have to propose any change at transformers or the mosfet's current without driving power (the manufacturer said to adjust this current at 200ma)
I thing to change the secondary coil of T1 from 6 turns to 9 turns. What do you think?
 
Top