1

Does someone know a digital gates (inverter, buffer, nand, and, or, nor, etc...) that can go from 0V to 5V in about 1ns?

I have check TTL, but they are not so fast.

Ideally, I should found it in Farnell, Digikey, Mouser, etc...

Alain.

ABU
  • 59
  • 3
  • You should check out these questions: [1](https://electronics.stackexchange.com/questions/2925/cheap-solid-equipment-wire-for-breadboard-from-online-shop), [2](https://electronics.stackexchange.com/questions/94841/how-can-a-small-device-produce-two-million-volts) and finally [3](https://electronics.stackexchange.com/search?q=shopping). I got them from [here](https://electronics.stackexchange.com/search?q=shopping) while also doing a ctrl+f for "closed". – Harry Svensson Nov 08 '17 at 21:59
  • 3
    why do you want a gate that fast? – BeB00 Nov 08 '17 at 22:00
  • 1
    If you just want a fast rise time pulse e.g. to characterize an oscilloscope, there are other ways to generate one. – pericynthion Nov 08 '17 at 22:07
  • 1
    @pericynthion No idea whether this would help the OP, but what would you suggest to generate a fast rise time pulse? – user2233709 Nov 08 '17 at 22:22
  • @user2233709 Jim Williams has a famous design - Appendix D of [this app note](http://cds.linear.com/docs/en/application-note/an47fa.pdf) and google "avalanche pulse generator" for various related designs, both DIY and low-cost off-the-shelf. – pericynthion Nov 08 '17 at 22:27
  • What you are looking for isn't exactly a gate, but a very strong output driver. Driving a cable from 0V to 5V within 1ns is much harder than making a GHz logic. – user3528438 Nov 08 '17 at 23:35

2 Answers2

5

Emitter coupled logic is that you want, like the https://www.digikey.co.uk/product-detail/en/MC10EP08DTG/MC10EP08DTGOS-ND/920778?curr=gbp&WT.z_cid=ref_octopart_dkc_buynow&site=us

Why do you need it to go that fast?

If you want your circuit to actually go at GHz speeds, you'll need to take a lot of things into account other than just the chip.

BeB00
  • 5,303
  • 2
  • 18
  • 34
  • 3
    ...things like controlled impedance PCB layout and line termination; and these ECL systems are power-hungry. Bear in mind the speed of light is about 1ns per 20-30cm, so in that realm cables and pcb traces are no longer just ideal wires, they are transmission lines and delay lines. – MarkU Nov 08 '17 at 22:15
  • OP asked for a pulse going from 0 to 5 V. ECL levels are classically from about -2 to -1 V, or +3 to +4 V if using them as "PECL". (Levels are very rough, but in any case the difference between low and high is ~800-1000 mV, not 5 V) – The Photon Nov 08 '17 at 23:45
  • Hello, thank you for your first response. I need such speed in order to generate a very fast dV/dt (>1500V/us). The target is to make some characterize some device in extreme condition. – ABU Nov 09 '17 at 21:56
  • Hello again! Yes I know that for such speed, I should be very strict in the design. – ABU Nov 09 '17 at 21:59
  • @ThePhoton is correct that the vdiff is only 800mV, but it looks like that may work in your design, because it can switch in 130ps, giving a rise time of 6000V/us – BeB00 Nov 09 '17 at 22:03
2

Potato Semiconductor (yes, really) has some 74-family parts with sub-nanosecond rise/fall times.