Who has experience with lateral mosfets ?

[larryB]

Hi all you audio fellas .

So from what I can read , these devices offer significant advantages over BJT .

But I hear they are difficult to drive without oscillation .

Solutions to get clean signal is apparently the following , resistor close to the gate , good low inductance layout , operating @ high voltages .

I understand the 2 first ones , but the operating @ higher voltages is a mystrery to me .

When I called profusion the engineer there said to me they were linear @ all voltages and would work in the 15-20 volt range

But then I read these from the now deceased honorable mister Randy Sloan I wonder what is the for these mosfets . Me and this man have the same views on thing and share the same opinions , obviously he pwn my azz . I am also a student in industrial electronics , the big stuff you know ...

Quote: from mister sloan

http://www.zusaudio.com/technology/whitepapers.htm

Secondly, they provide a moderately-high voltage amplifier stage output impedance, which is preferable for driving the Lateral MOSFET output devices, since this is the primary means of eliminating gate oscillation problems


Peronally my opinion is that the research on these devices has been lagging and they are not as quality controlled as they should be , they vary alot ...

But these new alfet ones here look very good and advanced .

Magnatec. ALFET Lateral MOSFETs


Do post and share you opinions , thoughts , etc ...

cheers-LarryB

[jechentau]

ah i have only read you post, but, don't they mean a moderately high output impedance, of the voltage amplifier stage? (driving the fets) (which is what grid stoppers do in effect as in being applied)

[AndrewT]

Originally Posted by larryB

Secondly, they provide a moderately-high voltage amplifier stage output impedance, which is preferable for driving the Lateral MOSFET output devices, since this is the primary means of eliminating gate oscillation problems

the "moderately high" applies to the "output impedance" that belongs to the "high voltage amplifier stage". It's the way english grammar works.
The following oscillation avoidance clarifies that moderately high output impedance which is precisely what most writers/builders designers choose to call the gate resistor.

I think Slone/Sloan? was simply trying to make it sound very technical to avoid appearing to copy the simple direction given by all his coworkers.

[larryB]

Omg I typoed the name of the legend mister Slone , so sorry .

So yes my english it not very good , but I can manage .

So are you 100%% certain that what is meant here is high output impedance of the gate driving circuit and not high voltage of the driving signal .

THis is 100% important as I am embarking on a build adventure as of today using about 7-8 amps ( max peak ) @ about 16 volts with a double die device , will probably spend 200$-300$ on this and much time .

thx you .

[nigelwright7557]

Its not all good with MOSFETs'
BJT need 0V7 to switch on, MOSFETS need a few volts and so you cant drive rail to rail with some messing with the power supply to the driver stage.

MOSFETs are also a bit touchy about the bias. Too little bias and they distort, too much and they cook in thermal runaway. I would recommend some thermal feedback although laterals are better than verticals with heat. Some people like a high bias voltage for a cleaner sound and this can put MOSFETs on the edge of thermal runaway.

[analog_guy]

I have designed several amps with the BUZ501/901 lateral MOSFETS and have had good luck. One was a 400W/ch audio monoblock, and the other was a 1 - 5 MHz, 80 vPP piezoelectric transducer driver capable of driving 10 ohm (complex) loads. As mentioned earlier, these devices are prone to oscillation due to parasitic inductances, so the power supply and gate drive circuits need to be laid out carefully. Multilayer PCBs are recommended, with separate planes dedicated for V+, V-, and ground. Use of multiple 0.1 to 1.0 uF SMT bypass capacitors bypassing the V+ and V- planes to the ground plane is also a good idea. With these precautions the devices work well.

In an earlier thread someone mentioned that a Vgs of several volts, as opposed to ~0.6 Vbe for bipolar devices, is required. This should not be a problem when the MOSFETS are used as source followers, although it does imply a slight reduction in Vswing unless the driver stage operates from higher rail voltages, (something I recommend).

[larryB]

I am most satisfied by the feedback , I give you all a good manly handshake .

Analog guy thats a nice project .

Time for some pcbing .

[CBS240]

......although it does imply a slight reduction in Vswing unless the driver stage operates from higher rail voltages, (something I recommend).

or bootstrap. Also with respect to ocilations, placement of the gate stopper as close to the device package as well as a gate Zobel will help. Gate Zobel will help you more acurately set the roll-off freaquency. In addition, a decoupling cap as close to the drain pin as possible is also a good step.