I've read that MOSFET need no thermal compensation since when their temperature rise, the current decrease. Does it means I can do without ??? (I use IRFP240 & 9240)
Thanks for any comment.
David
Thanks for any comment.
David
temp comp MOSFETS
MosFets like the IRF 240 and equiv. do need some kind of thermal compensation. Only Power Mosfets Like the BUZ900 / 2Sk1058 / 2SK135 do not need this. These devices do have a negative temperature coeficiente, and this will avoid thermal run-a-way.
MosFets like the IRF 240 and equiv. do need some kind of thermal compensation. Only Power Mosfets Like the BUZ900 / 2Sk1058 / 2SK135 do not need this. These devices do have a negative temperature coeficiente, and this will avoid thermal run-a-way.
thanks for your input,
but it's irf"P"240 and IRF"P"9240, I think they are power devices ...
David
but it's irf"P"240 and IRF"P"9240, I think they are power devices ...
David
Re: temp comp MOSFETS
Is there a generic name or classification of FET's with -ve temp coefficient or does one go through the datasheets of each device....
wimdehaan said:
Is there a generic name or classification of FET's with -ve temp coefficient or does one go through the datasheets of each device....
Indeed. IRF (any type) are vertical devices intended for switching and the -ve tempco only happens at quite high currents. Lateral MOSFETs are specifically for audio -- Hitachi, Exicon and Magnatec are the most prominent -- and these have -ve tempco starting in the milliamps range.
Mr Pass stated in one of his articles that he has a "dumpster" load of them, the main reason for his choice.
I think he referred to it as a stategic reserve, good thing he is not in the Silver business.
I have the impresion it was Mr Pass who made the vertical IRF's more accepted.
The main reason for lateral Hitachi Mosfet in the 80s and 90s was their availability, the ruggid TO3 package, low Ciss, and most of all their 'good' sound.
In those days verticals were generally considered a NoNo., even diy mags advised not to go for vertical mosfets.
Overhere companies as Silltech pushed the popularity of laterals, in Oz a company as Perreaux, in the US Conrad Johnson and some others.
I built a few amps designed for lateral mosfets with vertical ones, sounded ok to me (Class A, Naturally !)
Temperature with class A amplifiers is much more stable, dissipation actually decreases if a load is driven.
btw: does anyone have some experience with Magnatec double Buz's in a single TO3 ?
I think he referred to it as a stategic reserve, good thing he is not in the Silver business.
I have the impresion it was Mr Pass who made the vertical IRF's more accepted.
The main reason for lateral Hitachi Mosfet in the 80s and 90s was their availability, the ruggid TO3 package, low Ciss, and most of all their 'good' sound.
In those days verticals were generally considered a NoNo., even diy mags advised not to go for vertical mosfets.
Overhere companies as Silltech pushed the popularity of laterals, in Oz a company as Perreaux, in the US Conrad Johnson and some others.
I built a few amps designed for lateral mosfets with vertical ones, sounded ok to me (Class A, Naturally !)
Temperature with class A amplifiers is much more stable, dissipation actually decreases if a load is driven.
btw: does anyone have some experience with Magnatec double Buz's in a single TO3 ?
"Is there a generic name or classification of FET's with -ve temp coefficient or does one go through the datasheets of each device...."
In general, lateral mos are good for linear operation, vertical are for switching.
If the DRAIN is hooked to the case, is good for switching.
If the SOURCE is hooked to the case, is good for linear.
In general, lateral mos are good for linear operation, vertical are for switching.
If the DRAIN is hooked to the case, is good for switching.
If the SOURCE is hooked to the case, is good for linear.
djk said:
I'm not trying to argue, but I don't understand this exactly... what does the case have to do with the construction of the FET?
What I do know, the naming of a DRAIN and SOURCE of a FET is sort of arbitrary, they simply refer to which terminal is at higher or lower potential. FETS would work exactly the same if you reversed the pins (not true of BJTs). However, that is only true if the substrate is independantly controlled. Usually the FETs we buy have the substrate internally tied to one of the pins and that generally forces the designer to use that pin as the SOURCE.
--
Danny
You are right, it has nothing to do with it whether source or drain is connected to case. I think what djk was trying to say was that vertical MOSFETs can reliably be identified by having a drain-case connection, and lateral MOSFETs by having a source-case connection. Handy I suppose if you have no datasheet or part number on the device. Although vertical are usually pinned GDS and lateral GSD.
azira said:
This is true of most, but not all, JUNCTION FETs (small signal). It is not true of the power MOSFETs that are under discussion. DJK is correct, the pinout is one of the easiest ways to identify the lateral parts versus the vertical parts.
I happen to agree with him that the lateral parts are better suited to linear applications such as audio amplifiers. Nelson Pass does not.
If you use a lateral mosfet in a transnova configuration with the source tied to ground, you don't need to isolate the fet from the heatsink, and this is a lot convenient especially for the case to sink temperature.
In one of my amplifiers I used CFP of bipolar and MOSFET transistors. Mosfets were IRF540 and IRF9540. Transistor of Vbe multiplier bias circiut (the same as drivers) was placed on the same heatsink as driver transistor. I was surprised that this way temperature compensation was working perfect and quiescent current was not changing much. Output MOSFETs were placed on separate heatsink.
I thought that when the bias circut's transistor isn't mounted on the output devices heatsink, the idle current will rise when the temperature of heatsink rises.
I experienced it too with a lot's of amps.
I experienced it too with a lot's of amps.
It's different in the case of a CFP because there the driver transistor mostly determines bias current, not the output devices.
If the output devices and the drivers are in a CFB configuration a tight local feedback loop results so that the output devices don't need to be directly monitored. The drivers become the point that the Vbe multiplier needs to monitor. At least in priciple -- I've also seen examples of ouput devices, drivers and Vbe multipliers all attached to the same heatsink and everything works well.
EF or SF is you are taking FETs is another mater.
Reply
Hi everybody ,
I think there is need for an input from us too..😀
IRF and APT vertical mosfets are mean't for switching, but there implementation in linear circuits has several benefits over lateral mosfets. They also require thermal tracking and compensation but when driving heavy loads they exhibit +ve temp. coefficient[i.e. large current draw], thus balance the current sharing more appropriately .
regards,
kanwar
Hi everybody ,
I think there is need for an input from us too..😀
IRF and APT vertical mosfets are mean't for switching, but there implementation in linear circuits has several benefits over lateral mosfets. They also require thermal tracking and compensation but when driving heavy loads they exhibit +ve temp. coefficient[i.e. large current draw], thus balance the current sharing more appropriately .
regards,
kanwar
Reply
WHY THE SWITCHING [VERTICAL] MOSFET IS SUPERIOR THAN LATERAL MOSFET?
Lower internal resistance
Cooler running of amplifier
Better efficiency
The internal resistance is the main difference between the switching MOSFETs (0.18ohms ) and the Lateral MOSFETs (2.5ohms ) and the reason of the better efficiency of the former. The consequence of all that is an operating temperature of 10-15º C lower for the switching Mosfets when compared to the Lateral Mosfets. This fact is of great importance because a lower thermal stress can double the life of an amplifier.
Thus it is :
Faster
Better slew-rate
Another difference is
the gain-bandwidth product:
8MHz for the Lateral MOSFETs.
15MHz for the switching MOSFETs.
Moreover, the switching MOSFET is more linear with less harmonic components. The result is a very fast Slew rate and in consequence high and low frequencies response improvement achieving very high instantaneous output voltage increments. Loudspeakers are tightly driven thus account for Good damping factor . Music reproduction with hard transients has an amazing reality , when listened.
hope this helps.
kanwar
WHY THE SWITCHING [VERTICAL] MOSFET IS SUPERIOR THAN LATERAL MOSFET?
Lower internal resistance
Cooler running of amplifier
Better efficiency
The internal resistance is the main difference between the switching MOSFETs (0.18ohms ) and the Lateral MOSFETs (2.5ohms ) and the reason of the better efficiency of the former. The consequence of all that is an operating temperature of 10-15º C lower for the switching Mosfets when compared to the Lateral Mosfets. This fact is of great importance because a lower thermal stress can double the life of an amplifier.
Thus it is :
Faster
Better slew-rate
Another difference is
the gain-bandwidth product:
8MHz for the Lateral MOSFETs.
15MHz for the switching MOSFETs.
Moreover, the switching MOSFET is more linear with less harmonic components. The result is a very fast Slew rate and in consequence high and low frequencies response improvement achieving very high instantaneous output voltage increments. Loudspeakers are tightly driven thus account for Good damping factor . Music reproduction with hard transients has an amazing reality , when listened.
hope this helps.
kanwar
- Status
- Not open for further replies.