- If this is spam, hope the company deletes the account, bans the URL, and removes this thread ASAP.
The threshold voltage of the FETs can differ. With a common gate resistor the gates are connected and the transistor with the lower threshold voltage will catch all the current.
I have seen that as well, it stems from transistor where you had to add an emitter resistor to balance them. The MOSFETs do that thermally. Resistance in the gate lead causes the switching to slow down leaving the load in the ohmic region longer.
That is why you turn them on quickly and keep them in there SOA (Safe Operating Area). A gate resistor will form an RC with the gate capacitance and slow down on the turn on exasperate the problem.
The gate resistor protects the driver stage against too high currents while switching.
See my #4. This thread starts with a MOSFET killer circuit. Intended to increase the sales of the TO?
so are you saying to not use a gate resistor or
to use a low value 10 ohm gate resistor and one will suffice or
to use low value gate resistors one per MOSFET?
a7
Yes, the resistor tends to keep it from oscillating. Also you will need something to drive it, a logic gate will not source/sink enough current. Here is an example of one.
This shows the capacitance associated with the MOSFET. It also shows the wrong way to drive the gate, The pull down resistor (R1) should be connected to the emitters. The way it connected is forming a voltage divider reducing the Vgs voltage. R2 the source lead represents the resistance in the power source to the drain, this can sometimes be a source of oscillation. .
That would happen for sure. I expect most users on this forum do not know what the SOA (Safe Operating Area) is or how to read the graph.
Isn't this quite usual in load switches where either the gate capacitance or an additional capacitor is used to hold the mosfet in the linear region momentarily for say implementing a "soft start" for something like a boost converter etc.
Two questions for the thread participants...
- Is the original post contrived to make the product look good?
- Is there anything of value in this thread? Anything that might help someone?
1 Like
-
What’s been discussed here has been covered before, except for the possible deception.
-
Fool me once shame on you, fool me twice shame on me.
2 Likes
That's a killer approach. In contrast to a linear amplifier a switch has to avoid the linear region. High (capacitive...) switching currents have to be reduced by resistors or inductors in series to the load.
- Last time I participate in this thread.
For MOSFET switching applications, we must operate in the ohmic region i.e. linear region.
- Please read this . . .
https://www.baldengineer.com/7-mosfet-myths-and-misconceptions-addressed.html
An interesting point of view. However, here is a Load Switch (Texas Instruments tps27081a) where the user can configure the rise time with external components, a capacitor and a resistor, the capacitor being positioned between the PMOS drain and gate: https://www.ti.com/lit/ds/symlink/tps27081a.pdf?ts=1719061361032. There is a table of suggested values where the user can pack in an upto 1uF capacitor for a rise time of nearly 4ms.
Special loads can require special drivers. A general purpose switch has minimal rise and fall times.
Is there a question in there?
You don't define Vin, but in general driving a MosFet via a 10k resistor is not a good idea. The reason is the drain charge will delay the transition from off to on.
Your very pretty waveforms unfortunately do not relate the reality of the situation, they simplify too much and don't consider the Qgd or the define the load or the input power supply.
There's enough technical stuff to warrant keeping the thread. The original poster is gone.
1 Like
I am not happy with the way this thread has developed.
OK. It was started by a representative of a semiconductor manufacturer on some rather contrived scenario, but it could, nevertheless, have developed in a more positive way.
My opinion is that it would be useful to welcome such contributors to the forum to hear at least what their opinions are when dealing with topics which appear here from time to time. For example, it would have been useful to have had the point of view of a Mosfet manufacturer on this recent thread: What is the difference between IRL3705N and IRL3705Z? and this is the reason for me choosing to comment again.
Of course it is clear that non-relevant advertising of the Viagra type should be prohibited and also agressive pushing of products even if relevant should also be stopped but if a manufacturer of relevant products, in a transparent and passive way, offers some free consultancy for a bit of company profiling, then why not? For example, if a user has a question about PCB track width for a specific current in the middle of a multi layer board, would an answer from a JCLCPB representative be met with open hostility ? From what I have already seen, the answer is almost certainly no.
Incidentally, the OP was even offering free samples but was banned before he could be contacted or fully explain him/her self.
I find topics like this the hardest to judge from a moderation point of view, helpful contribution or spam? Hard to say. Moderators will disagree behind the scenes, probably without you ever being aware of the disagreement.
OK. Thanks. I guess that someone reported it as spam and hence the moderator intervention. The subject matter of the thread is complex so could well have been difficult to assess. The same dogmas appear in similar threads, one of which has been referenced, so it would have been good to have more opportunity to quiz an "expert" who even possibly has engineering support in the back ground. Anyway it is water under the bridge.
Two people. But that just gets our attention.
The original poster stopped engaging despite being asked a few direct questions. That was the triggering event: they wandered off.
Based on the original poster's behaviour I can confidently say they are neither of those.
Please try to have more trust in your moderators. Despite outward appearances we are not quick to ban.
In my experience, most chip manufacturers do that or offer inexpensive "starter kits" when politely asked. They seem to especially like when their parts are mentioned in Exhibition / Gallery.
2 Likes
Thanks for responding. I agree that the OP could have done a better job of giving a specific use case to support his argument, which seemed to give rise to most of the direct questions, instead of simply saying that it was a "hypothetical situation". However, some of his responses were met with a degree of hostility which is the most likely explanation for his/her apparent reluctance to engage further.
I don't see enough material to support such a conclusion but we will probably never know.
I can agree that it was a difficult thread to moderate.
2 Likes