1. Home
  2. Hobby Electronics Basics
  3. Measuring mosfet gate rise time?

Measuring mosfet gate rise time?

I'm using a FDN359AN (SOT3 PACKAGE) and a TC427 driver. Rather then explain here is my test cct.

formatting link

The FDN359AN specifies a max Qg = 7nC.The TC427 states a max Igp =

1.5A, and Zo= 6 ohms. My Vcc (8V) is high enough for the driver to source 1.3A .I should be getting approximately 5nS rise time, but I'm measuring 37nS. I know the calculation for tr is an approximation, but this far off?

Tr = Qg/Igp = 5nS

My probe is at 10x according to the data sheet at this setting. Bw = 100MHz Tr = 3.5nS Rin = 10Mohm Cin = 17pf BNC probe connections on the scope are 16 pf.

My scope setting are 1Gs/S and 25k points, (max resolution).Would the probe capacitance have this much effect. If so is there a way to do a more accurate measurement. The scope automatically displays rise/fall times I double checked using the cursor from 10% to 90% and get the same 37nS.

The waveform is here.

formatting link

FDN359AN Data sheet here.

formatting link

TC427 driver data sheet

formatting link

Thanks

Reply to
Hammy
Loading thread data ...

The TC427 simply isn't that fast, even with no load; the risetime on the datasheet is 30 ns with a tiny load. You'd need something else, a gaasfet maybe, to get drive like that.

A bit of series inductance will help, but only by 20% roughly.

Why do you want to drive the fet that fast?

John

Reply to
John Larkin

Isn't the data sheet showing a 35nS into a 1000pf load? The FET I'm using only has CISS of 480pf. I know there are other frequency dependent complex impedances, but typically to estimate switching speed don't you use Qg, the driver's output impedance and the drivers Vcc? For example:

Ipk (driver) = Vcc/Zo(driver) A bit of series inductance will help, but only by 20% roughly.

No particular reason the FDN359AN is cheap and available and small, and I have a TC427 driver. I'll be using a MIC4416BM4 to drive it when I do a layout. I'm just wondering why such a large discrepancy between calculated and measured, even when considering the 17pf my probe contributes. The FDN359AN is soldered onto a scrap piece of PCB so I can test it on the breadboard with an inch gate lead. This may also be factor. I think the inch gate lead and I'm also using an inch of 22AWG wire on my probe is whats doing it.Well dont I feel like an idiot.I just realisied the error when writting it down.

Or am I doing something wrong in my estimates?

Thanks for your opinion.

Reply to
Hammy

Hammy wrote in news: snipped-for-privacy@4ax.com:

According to the data sheet, a 480pf load would have rise time closer to 18nS when using the Rise and Fall Times vs Capacitive Load chart. Might not be the 37 you are measuring, but higher the value you expected.

Overstating the obvious, what about the rise times of your input signal and the impedance/resistance of the connection between the driver and the FET?

Reply to
tbell

Yes I just noticed that ,thats also at a Vcc of 18V!

The rise time of the input signal shouldnt mater on the rise time of the driver.

Out of curiosity I hooked the mosfet up in a high side configuration. I left everything else as is the same driver with the same signal = to the source voltage (Vgs differential nowhere near the threshold). This should be equivalent to hooking the driver up to a 480pf cap (or I could be wrong). The driver was charging the FETS 480 pf input capacitance in 8 to 9 nS, this is close to calculated.

Then I put it back to normal low side as it was. I put a light load on the FET 0.1A, no gate resistor and I get 37nS this is the same rise time as a 1A load. I then put a gate resistor in for current sampling this increases the rise time by 10nS to 47nS. The current measurement across the resistor gives me about 0.2A pulses.

Anybody know what's going on?

Here are pics of my schematics.

formatting link

Reply to
Hammy

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.