BAV199 fw recovery

Probably far fetch but anyway...

I need to use some of these BAV199 for a low leakage 2nd level ESD clamp (after a nice Diode's D20V0L1B2WS low leakage but circa 30V clamp TVS. The ESD events I'm concerned with are the low impedance machine ones.

Having quite a bunch of those ESD cells on board I'm trying to simplify the cell and need some info on forward recovery. ATM I don't have samples at hand to do some measurements and my google-fu gave nothing either, except that Trr is rather bad at 3us (but I don't care) which suggests that forward recovery voltage and time aren't great either.

Anyone knowing either some figures from measurements or manufacturer of BAV199 with fw recovery figures, or even some other low leakage diode which have that fw beahaviour specfied as well (looking for sub nA

--
Thanks, 
Fred.
Reply to
Fred Bartoli
Loading thread data ...

Here are some forward recovery tests on a 1N914.

formatting link

formatting link

formatting link

formatting link

I've run power diodes at 50 volts forward bias, but that's a whole nother story.

A BFT25 c-b junction is a fast femtoamp-leakage diode.

There are fast diode arrays specifically for ESD protection, like CM1213, but you'd have to qualify them for leakage.

--
John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    
 Click to see the full signature
Reply to
John Larkin

how does 1N914 compare to 1N4148?

Reply to
RobertMacy

Grekhov's en even bigger wildman than you. ;)

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
 Click to see the full signature
Reply to
Phil Hobbs

There's a zero-TC region (in theory anyway) where the resistive postive tempco cancels the junction negative tempco. Something like a couple volts forward bias?

Reply to
Spehro Pefhany

Don't know. They are really the same part, and I bet lots of them come off the same wafers. I'd also expect a lot of variation in recovery times from different manufacturers; I've seen that in 1N4005 types.

--
John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    
 Click to see the full signature
Reply to
John Larkin

Some small-signal schottky diodes have their zero-TC point at useful currents, like 10 mA, 0.5 v or something like that. PN diodes have zero TC at (very roughly) 1.1 volts. Some power diode data sheets have the curves to locate that point.

--
John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    
 Click to see the full signature
Reply to
John Larkin

It's thought of as the same part, it is not. It *is* a similar part. Jedec numbers are assigned based upon some performance characteristic differences. From memory, I found the 1N4148 to be preferential to the

1N914, may have been voltage breakdown. One is 50V the other 100V ?

When it comes to semiconductors, even LOT variations are wild. So wild that it's usually solved by testing, which means that when you get a part, and you purposely bought a 'cheap' part, you often don't get a gaussian distribution of performance; you get a HOLE with edges distribution of performance, because all the 'good' ones were culled out!

Reply to
RobertMacy

My great (but profitless) insight was that the c-b junction of certain high-voltage horizontal-output transistors make great drift-step-recovery diodes. Of course, nobody makes those HOTs any more.

Got nice 2KV, 1.5 ns wide pulses at 100 KHz, with a little water cooling.

formatting link

formatting link

formatting link

--
John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    
 Click to see the full signature
Reply to
John Larkin

The turn-on delay is mostly due to diffusion in undepleted silicon, so it would depend a lot on doping profiles and so on.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
 Click to see the full signature
Reply to
Phil Hobbs

Gee, I'd think that could only be done for a short instant. Some time ago, I was trying to figure out why a full-bridge PWM servo amp was blowing up. I got a digital scope on it and tesed it right up until it was at the edge of catastrophic failure. The was when the half bridge section was sourcing current to the output filter inductor at about 20 A, and then needed switch to the low-side transistor. During the deadtime, the body diode of the lower FET could develop

-12 V across it, and this would last for several microseconds before the low-side transistor was turned on. That fried the FET driver chip. The fix was to put an ultra-fast power diode across the low-side FET to do what the FET's own body diode wouldn't do.

So, the body diode could develop 12 V forward bias for several us before it started any serious conduction. But, of course, FET body diodes are known to have pretty awful characteristics. I used something from Motorola for a while, and now use the ES3D surface mount diode. I don't know what the turn-on time actually is, but it is fast enough to solve this trouble.

Jon

Reply to
Jon Elson

The 4148 I think has less cap on it..

I can't remember the exact value but I think it was half as much as the 914.

Jamie

Reply to
Maynard A. Philbrook Jr.

Back when we were playing with this stuff, it seemed to us that people were fabbing maybe two different wafers to span 1N4001 through 1N4007, with the higher voltage parts being a PIN-like structure.

--
John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    
 Click to see the full signature
Reply to
John Larkin

Yup. At +50 volts, the current would ramp up linearly to maybe 100 amps in a few hundred ns. Then we reverse-biased it from a -400 volt supply, and it would conduct a couple hundred amps for maybe 50 ns, then suddenly snap off.

Body diodes can act like step-recovery diodes, too. The resulting EMI can be horrific, and can blow out gate oxides.

One of my favorite waveforms,

formatting link

the output of an LM3102 synchronous switcher. In the first 40 ns, you can see four distinct regimes: lower fet on, substrate diode conducting, upper fet fighting substrate diode, snap. Useless damned part.

--
John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    
 Click to see the full signature
Reply to
John Larkin

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.