https://www.farnell.com/datasheets/2353855.pdf I bought a few of these. I don't know what is intended by "digital transistor" since the charge storage time defeats any attempt to use them at a high frequency, and there's no way to add a Schottky diode to prevent saturation, or a speed-up capacitor across the base resistor. I'm using this and the corresponding PNP type to do level shifting of both high and low levels (*). At 25 kHz it's OK, albeit with a significant change in duty cycle, and is adequate for my purpose, but I wouldn't want to push it much higher. I suppose the lack of description of the switching characteristics should have been a give-away. Sylvia. (*) Whether I needed to do that is a different discussion.
Digital transistor?
Started by ●March 25, 2022
Reply by ●March 25, 20222022-03-25
lørdag den 26. marts 2022 kl. 00.52.43 UTC+1 skrev Sylvia Else:> https://www.farnell.com/datasheets/2353855.pdf > > I bought a few of these. I don't know what is intended by "digital > transistor" since the charge storage time defeats any attempt to use > them at a high frequency, and there's no way to add a Schottky diode to > prevent saturation, or a speed-up capacitor across the base resistor. > > I'm using this and the corresponding PNP type to do level shifting of > both high and low levels (*). At 25 kHz it's OK, albeit with a > significant change in duty cycle, and is adequate for my purpose, but I > wouldn't want to push it much higher. > > I suppose the lack of description of the switching characteristics > should have been a give-away. > > Sylvia. > > (*) Whether I needed to do that is a different discussion.it saves a couple of resistors, I don't think it ever promised anything else
Reply by ●March 25, 20222022-03-25
On Sat, 26 Mar 2022 10:52:34 +1100, Sylvia Else <sylvia@email.invalid> wrote:>https://www.farnell.com/datasheets/2353855.pdf > >I bought a few of these. I don't know what is intended by "digital >transistor" since the charge storage time defeats any attempt to use >them at a high frequency, and there's no way to add a Schottky diode to >prevent saturation, or a speed-up capacitor across the base resistor. > >I'm using this and the corresponding PNP type to do level shifting of >both high and low levels (*). At 25 kHz it's OK, albeit with a >significant change in duty cycle, and is adequate for my purpose, but I >wouldn't want to push it much higher. > >I suppose the lack of description of the switching characteristics >should have been a give-away. > >Sylvia. > >(*) Whether I needed to do that is a different discussion.The idea was to drive them from TTL without a base resistor. I suppose some people liked not having to do the required higher-level math. Logic-level mosfets made these mostly obsolete. FDV301 costs us 3 cents. I bet that base resistor has a radical tempco. -- If a man will begin with certainties, he shall end with doubts, but if he will be content to begin with doubts he shall end in certainties. Francis Bacon
Reply by ●March 25, 20222022-03-25
lørdag den 26. marts 2022 kl. 01.32.51 UTC+1 skrev John Larkin:> On Sat, 26 Mar 2022 10:52:34 +1100, Sylvia Else <syl...@email.invalid> > wrote: > >https://www.farnell.com/datasheets/2353855.pdf > > > >I bought a few of these. I don't know what is intended by "digital > >transistor" since the charge storage time defeats any attempt to use > >them at a high frequency, and there's no way to add a Schottky diode to > >prevent saturation, or a speed-up capacitor across the base resistor. > > > >I'm using this and the corresponding PNP type to do level shifting of > >both high and low levels (*). At 25 kHz it's OK, albeit with a > >significant change in duty cycle, and is adequate for my purpose, but I > >wouldn't want to push it much higher. > > > >I suppose the lack of description of the switching characteristics > >should have been a give-away. > > > >Sylvia. > > > >(*) Whether I needed to do that is a different discussion. > The idea was to drive them from TTL without a base resistor. I suppose > some people liked not having to do the required higher-level math.more like less parts and less space> Logic-level mosfets made these mostly obsolete. FDV301 costs us 3 > cents.quite few "digital transistors" here for less than half that https://lcsc.com/products/Digital-Transistors_562.html but also for fets https://lcsc.com/products/MOSFETs_381.html
Reply by ●March 26, 20222022-03-26
On 26-Mar-22 11:32 am, John Larkin wrote:> On Sat, 26 Mar 2022 10:52:34 +1100, Sylvia Else <sylvia@email.invalid> > wrote: > >> https://www.farnell.com/datasheets/2353855.pdf >> >> I bought a few of these. I don't know what is intended by "digital >> transistor" since the charge storage time defeats any attempt to use >> them at a high frequency, and there's no way to add a Schottky diode to >> prevent saturation, or a speed-up capacitor across the base resistor. >> >> I'm using this and the corresponding PNP type to do level shifting of >> both high and low levels (*). At 25 kHz it's OK, albeit with a >> significant change in duty cycle, and is adequate for my purpose, but I >> wouldn't want to push it much higher. >> >> I suppose the lack of description of the switching characteristics >> should have been a give-away. >> >> Sylvia. >> >> (*) Whether I needed to do that is a different discussion. > > The idea was to drive them from TTL without a base resistor. I suppose > some people liked not having to do the required higher-level math. > > Logic-level mosfets made these mostly obsolete. FDV301 costs us 3 > cents.LTSpice shows me 0.7V of undershoot when driving it with 3V pulses, 10ns rise time, using Onsemi's model, and 10K load resistor. https://www.onsemi.com/design/resources/design-resources/models?rpn=FDV301N Limiting that with a Schottky diode then gives me 22 mA spikes through the diode. Taming it with a 1K gate resistor helps, at the expense of slowing the output, but now we're back to needing an extra component. Not that such spikes always matter. Sylvia.
Reply by ●March 26, 20222022-03-26
On Sat, 26 Mar 2022 14:16:46 +1100, Sylvia Else <sylvia@email.invalid> wrote:>On 26-Mar-22 11:32 am, John Larkin wrote: >> On Sat, 26 Mar 2022 10:52:34 +1100, Sylvia Else <sylvia@email.invalid> >> wrote: >> >>> https://www.farnell.com/datasheets/2353855.pdf >>> >>> I bought a few of these. I don't know what is intended by "digital >>> transistor" since the charge storage time defeats any attempt to use >>> them at a high frequency, and there's no way to add a Schottky diode to >>> prevent saturation, or a speed-up capacitor across the base resistor. >>> >>> I'm using this and the corresponding PNP type to do level shifting of >>> both high and low levels (*). At 25 kHz it's OK, albeit with a >>> significant change in duty cycle, and is adequate for my purpose, but I >>> wouldn't want to push it much higher. >>> >>> I suppose the lack of description of the switching characteristics >>> should have been a give-away. >>> >>> Sylvia. >>> >>> (*) Whether I needed to do that is a different discussion. >> >> The idea was to drive them from TTL without a base resistor. I suppose >> some people liked not having to do the required higher-level math. >> >> Logic-level mosfets made these mostly obsolete. FDV301 costs us 3 >> cents. > >LTSpice shows me 0.7V of undershoot when driving it with 3V pulses, 10ns >rise time, using Onsemi's model, and 10K load resistor. > >https://www.onsemi.com/design/resources/design-resources/models?rpn=FDV301N > >Limiting that with a Schottky diode then gives me 22 mA spikes through >the diode. > >Taming it with a 1K gate resistor helps, at the expense of slowing the >output, but now we're back to needing an extra component. > >Not that such spikes always matter. > >Sylvia.10K load implies very low current, so if you want less spikes you'd need a part witn low Cg-d, namely a wimpier smaller-geometry device. An open-drain cmos gate might work, for whatever you are doing. Got a schematic? -- I yam what I yam - Popeye
Reply by ●March 26, 20222022-03-26
On 26-Mar-22 2:40 pm, jlarkin@highlandsniptechnology.com wrote:> On Sat, 26 Mar 2022 14:16:46 +1100, Sylvia Else <sylvia@email.invalid> > wrote: > >> On 26-Mar-22 11:32 am, John Larkin wrote: >>> On Sat, 26 Mar 2022 10:52:34 +1100, Sylvia Else <sylvia@email.invalid> >>> wrote: >>> >>>> https://www.farnell.com/datasheets/2353855.pdf >>>> >>>> I bought a few of these. I don't know what is intended by "digital >>>> transistor" since the charge storage time defeats any attempt to use >>>> them at a high frequency, and there's no way to add a Schottky diode to >>>> prevent saturation, or a speed-up capacitor across the base resistor. >>>> >>>> I'm using this and the corresponding PNP type to do level shifting of >>>> both high and low levels (*). At 25 kHz it's OK, albeit with a >>>> significant change in duty cycle, and is adequate for my purpose, but I >>>> wouldn't want to push it much higher. >>>> >>>> I suppose the lack of description of the switching characteristics >>>> should have been a give-away. >>>> >>>> Sylvia. >>>> >>>> (*) Whether I needed to do that is a different discussion. >>> >>> The idea was to drive them from TTL without a base resistor. I suppose >>> some people liked not having to do the required higher-level math. >>> >>> Logic-level mosfets made these mostly obsolete. FDV301 costs us 3 >>> cents. >> >> LTSpice shows me 0.7V of undershoot when driving it with 3V pulses, 10ns >> rise time, using Onsemi's model, and 10K load resistor. >> >> https://www.onsemi.com/design/resources/design-resources/models?rpn=FDV301N >> >> Limiting that with a Schottky diode then gives me 22 mA spikes through >> the diode. >> >> Taming it with a 1K gate resistor helps, at the expense of slowing the >> output, but now we're back to needing an extra component. >> >> Not that such spikes always matter. >> >> Sylvia. > > 10K load implies very low current, so if you want less spikes you'd > need a part witn low Cg-d, namely a wimpier smaller-geometry device. > > An open-drain cmos gate might work, for whatever you are doing. > > Got a schematic? > > >Below is my LTSpice version of the bit of the circuit concerned, to change levels up and down. The actual version implemented using the digital transistors performs fairly closely to the simulation. As I've said, it's sufficient for my requirements. I originally used a digital isolator but I've already destroyed two, one through carelessness, and one possibly through static discharge while trying to understand why one channel was behaving oddly. They're rather expensive to buy in excess of what is required in case they break. Version 4 SHEET 1 1268 680 WIRE 432 -752 432 -784 WIRE 784 -752 432 -752 WIRE 192 -704 192 -784 WIRE 784 -672 784 -752 WIRE 432 -656 432 -752 WIRE 528 -656 432 -656 WIRE 528 -608 528 -656 WIRE -112 -592 -160 -592 WIRE 192 -592 192 -624 WIRE 192 -592 -112 -592 WIRE 192 -544 192 -592 WIRE 432 -544 432 -656 WIRE 736 -512 640 -512 WIRE 784 -512 784 -592 WIRE 784 -512 736 -512 WIRE -32 -496 -80 -496 WIRE 80 -496 48 -496 WIRE 128 -496 80 -496 WIRE 528 -496 528 -528 WIRE 528 -496 496 -496 WIRE 560 -496 528 -496 WIRE 640 -496 640 -512 WIRE 80 -448 80 -496 WIRE 432 -448 416 -448 WIRE 80 -320 80 -368 WIRE 80 -320 -144 -320 WIRE 192 -320 192 -448 WIRE 192 -320 80 -320 WIRE -80 -272 -80 -496 WIRE 416 -272 416 -448 WIRE 416 -272 -80 -272 WIRE 416 -128 352 -128 WIRE 416 -80 416 -128 WIRE 176 -48 176 -96 WIRE 176 -48 -256 -48 WIRE 352 -48 352 -128 WIRE 352 -48 176 -48 WIRE -256 16 -256 -48 WIRE 416 32 416 0 WIRE 416 32 -96 32 WIRE 176 48 176 -48 WIRE 176 48 64 48 WIRE 64 80 64 48 WIRE -240 96 -256 96 WIRE 176 144 176 48 WIRE 416 144 416 32 WIRE 672 176 624 176 WIRE 768 176 672 176 WIRE -96 192 -96 32 WIRE -48 192 -96 192 WIRE 64 192 64 160 WIRE 64 192 32 192 WIRE 112 192 64 192 WIRE 512 192 480 192 WIRE 544 192 512 192 WIRE 624 192 624 176 WIRE 512 224 512 192 WIRE 416 240 400 240 WIRE -336 256 -384 256 WIRE 176 256 176 240 WIRE 176 256 -336 256 WIRE 400 320 400 240 WIRE 512 320 512 304 WIRE 512 320 400 320 WIRE -256 368 -256 96 WIRE 176 368 176 336 WIRE 176 368 -256 368 FLAG 176 -176 0 FLAG 400 320 0 FLAG 768 256 0 FLAG 192 -864 0 FLAG -144 -320 0 FLAG 432 -864 0 FLAG 416 -192 0 FLAG -112 -592 OUT1 FLAG 736 -512 IN1 FLAG 672 176 IN2 FLAG -336 256 OUT2 SYMBOL npn 480 144 M0 SYMATTR InstName Q1 SYMATTR Value 2N3904 SYMBOL pnp 112 240 M180 SYMATTR InstName Q2 SYMATTR Value 2N3906 SYMBOL res 496 208 R0 SYMATTR InstName R1 SYMATTR Value 47K SYMBOL res 48 64 R0 SYMATTR InstName R2 SYMATTR Value 47K SYMBOL res 48 176 R90 WINDOW 0 0 56 VBottom 2 WINDOW 3 32 56 VTop 2 SYMATTR InstName R3 SYMATTR Value 10K SYMBOL res 640 176 R90 WINDOW 0 0 56 VBottom 2 WINDOW 3 32 56 VTop 2 SYMATTR InstName R4 SYMATTR Value 10K SYMBOL res 160 240 R0 SYMATTR InstName R5 SYMATTR Value 10K SYMBOL voltage 176 -80 R180 WINDOW 0 24 96 Left 2 WINDOW 3 24 16 Left 2 WINDOW 123 0 0 Left 0 WINDOW 39 0 0 Left 0 SYMATTR InstName V1 SYMATTR Value 14 SYMBOL voltage 768 160 R0 WINDOW 123 0 0 Left 0 WINDOW 39 0 0 Left 0 SYMATTR InstName V3 SYMATTR Value PULSE(0 3.3 0 1e-8 1e-8 2e-5 4e-5 10) SYMBOL voltage -256 0 R0 WINDOW 123 0 0 Left 0 WINDOW 39 0 0 Left 0 SYMATTR InstName V4 SYMATTR Value 4 SYMBOL npn 128 -544 R0 SYMATTR InstName Q3 SYMATTR Value 2N3904 SYMBOL pnp 496 -448 R180 SYMATTR InstName Q4 SYMATTR Value 2N3906 SYMBOL res 512 -624 R0 SYMATTR InstName R6 SYMATTR Value 47K SYMBOL res 64 -464 R0 SYMATTR InstName R7 SYMATTR Value 47K SYMBOL res 64 -512 R90 WINDOW 0 0 56 VBottom 2 WINDOW 3 32 56 VTop 2 SYMATTR InstName 10K SYMATTR Value 10K SYMBOL res 656 -512 R90 WINDOW 0 0 56 VBottom 2 WINDOW 3 32 56 VTop 2 SYMATTR InstName R9 SYMATTR Value 10K SYMBOL voltage 192 -768 R180 WINDOW 0 24 96 Left 2 WINDOW 3 24 16 Left 2 WINDOW 123 0 0 Left 0 WINDOW 39 0 0 Left 0 SYMATTR InstName V2 SYMATTR Value 3.3 SYMBOL voltage 784 -688 R0 WINDOW 123 0 0 Left 0 WINDOW 39 0 0 Left 0 SYMATTR InstName V5 SYMATTR Value PULSE(0 4 0 1e-8 1e-8 25e-6 50e-6 10) SYMBOL res 176 -720 R0 SYMATTR InstName R10 SYMATTR Value 10K SYMBOL voltage 432 -768 R180 WINDOW 0 24 96 Left 2 WINDOW 3 24 16 Left 2 WINDOW 123 0 0 Left 0 WINDOW 39 0 0 Left 0 SYMATTR InstName V6 SYMATTR Value 14 SYMBOL res 432 -176 R180 WINDOW 0 36 76 Left 2 WINDOW 3 36 40 Left 2 SYMATTR InstName R8 SYMATTR Value 10K SYMBOL res 400 -96 R0 SYMATTR InstName R11 SYMATTR Value 10K TEXT -288 376 Left 2 !.tran 2E-4
Reply by ●March 26, 20222022-03-26
Lasse Langwadt Christensen wrote:> it saves a couple of resistors, I don't think it ever promised anything elseAnd they are quite useful, especially the PNP/NPN+resistors pairs in SOT-23-6/SC-70. Save a lot of board space. Best regards, Piotr
Reply by ●March 26, 20222022-03-26
On Sat, 26 Mar 2022 15:35:12 +1100, Sylvia Else <sylvia@email.invalid> wrote:>On 26-Mar-22 2:40 pm, jlarkin@highlandsniptechnology.com wrote: >> On Sat, 26 Mar 2022 14:16:46 +1100, Sylvia Else <sylvia@email.invalid> >> wrote: >> >>> On 26-Mar-22 11:32 am, John Larkin wrote: >>>> On Sat, 26 Mar 2022 10:52:34 +1100, Sylvia Else <sylvia@email.invalid> >>>> wrote: >>>> >>>>> https://www.farnell.com/datasheets/2353855.pdf >>>>> >>>>> I bought a few of these. I don't know what is intended by "digital >>>>> transistor" since the charge storage time defeats any attempt to use >>>>> them at a high frequency, and there's no way to add a Schottky diode to >>>>> prevent saturation, or a speed-up capacitor across the base resistor. >>>>> >>>>> I'm using this and the corresponding PNP type to do level shifting of >>>>> both high and low levels (*). At 25 kHz it's OK, albeit with a >>>>> significant change in duty cycle, and is adequate for my purpose, but I >>>>> wouldn't want to push it much higher. >>>>> >>>>> I suppose the lack of description of the switching characteristics >>>>> should have been a give-away. >>>>> >>>>> Sylvia. >>>>> >>>>> (*) Whether I needed to do that is a different discussion. >>>> >>>> The idea was to drive them from TTL without a base resistor. I suppose >>>> some people liked not having to do the required higher-level math. >>>> >>>> Logic-level mosfets made these mostly obsolete. FDV301 costs us 3 >>>> cents. >>> >>> LTSpice shows me 0.7V of undershoot when driving it with 3V pulses, 10ns >>> rise time, using Onsemi's model, and 10K load resistor. >>> >>> https://www.onsemi.com/design/resources/design-resources/models?rpn=FDV301N >>> >>> Limiting that with a Schottky diode then gives me 22 mA spikes through >>> the diode. >>> >>> Taming it with a 1K gate resistor helps, at the expense of slowing the >>> output, but now we're back to needing an extra component. >>> >>> Not that such spikes always matter. >>> >>> Sylvia. >> >> 10K load implies very low current, so if you want less spikes you'd >> need a part witn low Cg-d, namely a wimpier smaller-geometry device. >> >> An open-drain cmos gate might work, for whatever you are doing. >> >> Got a schematic? >> >> >> > >Below is my LTSpice version of the bit of the circuit concerned, to >change levels up and down. > >The actual version implemented using the digital transistors performs >fairly closely to the simulation. As I've said, it's sufficient for my >requirements. > >I originally used a digital isolator but I've already destroyed two, one >through carelessness, and one possibly through static discharge while >trying to understand why one channel was behaving oddly. They're rather >expensive to buy in excess of what is required in case they break.Interesting schematic style. You might accomplish the same function with a dual optoisolator or a dual opamp. -- I yam what I yam - Popeye
Reply by ●March 26, 20222022-03-26
On 26/03/2022 2:54 pm, jlarkin@highlandsniptechnology.com wrote:> On Sat, 26 Mar 2022 15:35:12 +1100, Sylvia Else <sylvia@email.invalid> > wrote: > >> On 26-Mar-22 2:40 pm, jlarkin@highlandsniptechnology.com wrote: >>> On Sat, 26 Mar 2022 14:16:46 +1100, Sylvia Else <sylvia@email.invalid> >>> wrote: >>> >>>> On 26-Mar-22 11:32 am, John Larkin wrote: >>>>> On Sat, 26 Mar 2022 10:52:34 +1100, Sylvia Else <sylvia@email.invalid> >>>>> wrote: >>>>> >>>>>> https://www.farnell.com/datasheets/2353855.pdf >>>>>> >>>>>> I bought a few of these. I don't know what is intended by "digital >>>>>> transistor" since the charge storage time defeats any attempt to use >>>>>> them at a high frequency, and there's no way to add a Schottky diode to >>>>>> prevent saturation, or a speed-up capacitor across the base resistor. >>>>>> >>>>>> I'm using this and the corresponding PNP type to do level shifting of >>>>>> both high and low levels (*). At 25 kHz it's OK, albeit with a >>>>>> significant change in duty cycle, and is adequate for my purpose, but I >>>>>> wouldn't want to push it much higher. >>>>>> >>>>>> I suppose the lack of description of the switching characteristics >>>>>> should have been a give-away. >>>>>> >>>>>> Sylvia. >>>>>> >>>>>> (*) Whether I needed to do that is a different discussion. >>>>> >>>>> The idea was to drive them from TTL without a base resistor. I suppose >>>>> some people liked not having to do the required higher-level math. >>>>> >>>>> Logic-level mosfets made these mostly obsolete. FDV301 costs us 3 >>>>> cents. >>>> >>>> LTSpice shows me 0.7V of undershoot when driving it with 3V pulses, 10ns >>>> rise time, using Onsemi's model, and 10K load resistor. >>>> >>>> https://www.onsemi.com/design/resources/design-resources/models?rpn=FDV301N >>>> >>>> Limiting that with a Schottky diode then gives me 22 mA spikes through >>>> the diode. >>>> >>>> Taming it with a 1K gate resistor helps, at the expense of slowing the >>>> output, but now we're back to needing an extra component. >>>> >>>> Not that such spikes always matter. >>>> >>>> Sylvia. >>> >>> 10K load implies very low current, so if you want less spikes you'd >>> need a part witn low Cg-d, namely a wimpier smaller-geometry device. >>> >>> An open-drain cmos gate might work, for whatever you are doing. >>> >>> Got a schematic? >>> >>> >>> >> >> Below is my LTSpice version of the bit of the circuit concerned, to >> change levels up and down. >> >> The actual version implemented using the digital transistors performs >> fairly closely to the simulation. As I've said, it's sufficient for my >> requirements. >> >> I originally used a digital isolator but I've already destroyed two, one >> through carelessness, and one possibly through static discharge while >> trying to understand why one channel was behaving oddly. They're rather >> expensive to buy in excess of what is required in case they break. > > Interesting schematic style. > > You might accomplish the same function with a dual optoisolator or a > dual opamp. > > > >The sim used 50:50 square waves, if the duty cycle is not too far from that then the level shift could be reduced to just two capacitors or even pulse transformers for a 1950s vibe :) piglet
