Forums

High voltage pnp's (250 volts)

Started by Hul Tytus April 14, 2020
I'll take a look. Thanks

Hul

bitrex <user@example.net> wrote:
> On 4/14/2020 10:58 AM, Winfield Hill wrote: > > Hul Tytus wrote... > >> > >> sci.electronics.design, comp.arch.embedded > >> high voltage pnp's (250v) > >> > >> Anyone know of a good pnp transistor spec'd up to 250 volts > > > > Suggest using a MOSFET, good choices all the way to 600V. > > > >
> What about the 2N5415 apparently you can still get 'em:
> <https://www.mouser.com/ProductDetail/Central-Semiconductor/2N5415-PBFREE?qs=sGAEpiMZZMve4%2FbfQkoj%252BLZHvMXp4Si4f9fQJjExGMA%3D>
If you look at a power stage with less than 0.5A or so, a BJT has lower conduction loss for the same price vs a FET

Problem is they are a lot more sensitive to overvoltage, secondary breakdown and are a pain to drive (or so one would think)

Even though beta is small and base currents seems excessive, they are only multiplied with base voltage, so ends up with low base drive, conduction loss wise

Klaus - what's the "baker clamp" method? I've been woking along the line you suggest: first driving the transistor from 3 pal'd AT13 
outputs through 3300 pf parrallel with 3300pf in series w/580 ohms along with a diode & 10k resistor to maintain bias. On the scope that 
appears pretty good with just a hint of curvature in fall time when measured at the AT13 output. With just one output, the curvature is 
quite noticable.

Hul


klaus.kragelund@gmail.com wrote:
> This is dirt cheap :
> https://www.mouser.dk/datasheet/2/389/dm00040355-1797494.pdf
> You need to drive the base either hard negative to shut it off fast or use a baker clamp type method
> Cheers
> Klaus
Hi

I do not know the AT13, do you have a link?

About baker clamp:

https://en.m.wikipedia.org/wiki/Baker_clamp

It&rsquo;s used to make sure the npn is never hard on which would cause long storage time when turning off

Cheers 

Klaus
By the way the baker clamp is difficult to tune robustly for low VCE

If you know your load you can do intelligent trimming, but that is even more flaky

Last I looked at it there is tons of papers and patents on these techniques 
On 2020-04-14 15:01, Hul Tytus wrote:
> Phil - you might read the message again.
I see a reference to a 500 ns pulse width--is that a design requirement? Still no schematic. I'd be interested to see it, whether or not I could contribute. Your call. Cheers Phil Hobbs
On 2020-04-14 18:06, Klaus Kragelund wrote:
> By the way the baker clamp is difficult to tune robustly for low VCE > > If you know your load you can do intelligent trimming, but that is even more flaky > > Last I looked at it there is tons of papers and patents on these techniques >
Any favourite references? Cheers Phil Hobbs -- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net http://hobbs-eo.com
On Wednesday, April 15, 2020 at 2:00:03 AM UTC+2, Phil Hobbs wrote:
> On 2020-04-14 18:06, Klaus Kragelund wrote: > > By the way the baker clamp is difficult to tune robustly for low VCE > > > > If you know your load you can do intelligent trimming, but that is even more flaky > > > > Last I looked at it there is tons of papers and patents on these techniques > > > > Any favourite references? >
There is a lot of IEEE papers, but you need access to get those, and the wording is hidden, so cannot just search for "baker" clamp This app note is ok: http://notes-application.abcelectronique.com/040/40-33424.pdf Some of the better patents: https://patentimages.storage.googleapis.com/9b/13/fa/0bbacd5941291d/US20030062931A1.pdf https://patentimages.storage.googleapis.com/f4/fc/c9/539a73945189d6/US4013903.pdf
On 14/04/2020 2:44 pm, Hul Tytus wrote:
> sci.electronics.design, comp.arch.embedded > high voltage pnp's (250v) > > Anyone know of a good pnp transistor spec'd up to 250 volts, a probable > current limit of 1 amp and designed as a switching device. It's to serve as > the switcher in a variable frequency (pulse width held constant) power supply. > So far a 2n2907 has been tried, with a 100ma 50 volt limit, with > impressive results. Zetex (now Diodes Inc) fzt957 & fzt958 with a 1 amp 300 > volt limit were tried but they were inefective. The 2n2907 showed 47% > efficiency with a 500ns pulse width; the fzt957 showed 38%. With 4 micro > second widths, 2n2907 showed 61% & fxt957 showed 32% (). > Apparantly, the higher collector/emitter voltage implies a greater > effective restance and higher saturation voltage, which might explain why n > channel fets with convoluted "high side" drive circuitry are used. I'm hoping > there are some pnp devices in the relatively low 1 amp and 2 amp range that > are more useful than the fzt957 units. Any suggestions? > > Hul >
My go-to jelly bean HV PNP is the MPSA92 but that might struggle above 500mA. Can you spare some Vce and use one of those in a Sziklai with a beefier NPN? Have you checked if the MJE5731 fits? But a p-channel MOSFET may work better. These may help: IRFR9214; FQT2P25; FQP9P25. More details of your circuit should help us find an answer. piglet
On Wednesday, April 15, 2020 at 12:22:32 PM UTC+2, piglet wrote:
> On 14/04/2020 2:44 pm, Hul Tytus wrote: > > sci.electronics.design, comp.arch.embedded > > high voltage pnp's (250v) > > > > Anyone know of a good pnp transistor spec'd up to 250 volts, a probable > > current limit of 1 amp and designed as a switching device. It's to serve as > > the switcher in a variable frequency (pulse width held constant) power supply. > > So far a 2n2907 has been tried, with a 100ma 50 volt limit, with > > impressive results. Zetex (now Diodes Inc) fzt957 & fzt958 with a 1 amp 300 > > volt limit were tried but they were inefective. The 2n2907 showed 47% > > efficiency with a 500ns pulse width; the fzt957 showed 38%. With 4 micro > > second widths, 2n2907 showed 61% & fxt957 showed 32% (). > > Apparantly, the higher collector/emitter voltage implies a greater > > effective restance and higher saturation voltage, which might explain why n > > channel fets with convoluted "high side" drive circuitry are used. I'm hoping > > there are some pnp devices in the relatively low 1 amp and 2 amp range that > > are more useful than the fzt957 units. Any suggestions? > > > > Hul > > > > My go-to jelly bean HV PNP is the MPSA92 but that might struggle above > 500mA. Can you spare some Vce and use one of those in a Sziklai with a > beefier NPN? > > Have you checked if the MJE5731 fits? But a p-channel MOSFET may work > better. These may help: IRFR9214; FQT2P25; FQP9P25. > > More details of your circuit should help us find an answer. >
APT27 is also extremely cheap, 0.5A: https://www.diodes.com/assets/Datasheets/APT27.pdf Beta is ok up to 200mA