Path: ...!local-3.nntp.ord.giganews.com!Xl.tags.giganews.com!local-1.nntp.ord.giganews.com!nntp.supernews.com!news.supernews.com.POSTED!not-for-mail NNTP-Posting-Date: Wed, 12 Jun 2024 21:21:13 +0000 From: john larkin Newsgroups: sci.electronics.design Subject: Re: another fast one-shot Date: Wed, 12 Jun 2024 14:21:13 -0700 Message-ID: References: <666a089d$0$1943529$882e4bbb@reader.netnews.com> User-Agent: ForteAgent/8.00.32.1272 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 52 X-Trace: sv3-rp0ZdsJdGUBNbGOlKrPTI6iM/wOX4j2Avu7BlHqB9eTC1gmp115IQJV7xLhJm3gBhcK9t+I9d+7csnc!v1ivtUgStW3yrrRv+8HzxKbR6sPe4uwKAUgIxapHsCLWhX3j2eM0rZ8UJHVSWPXR5lD+epfnVaTn!IXQxbQ== X-Complaints-To: www.supernews.com/docs/abuse.html X-DMCA-Complaints-To: www.supernews.com/docs/dmca.html X-Abuse-and-DMCA-Info: Please be sure to forward a copy of ALL headers X-Abuse-and-DMCA-Info: Otherwise we will be unable to process your complaint properly X-Postfilter: 1.3.40 Bytes: 3191 On Wed, 12 Jun 2024 16:44:14 -0400, bitrex wrote: >On 6/12/2024 2:11 PM, john larkin wrote: >> On Wed, 12 Jun 2024 17:22:49 -0000 (UTC), Cursitor Doom >> wrote: >> >>> On Wed, 12 Jun 2024 09:33:17 -0700, john larkin wrote: >>> >>>> Here's another idea for a fast one-shot. This is the just-apply-money >>>> version and should get close to 1 ns pulse width and should easily >>>> trigger at 150 MHz. >>>> >>>> https://www.dropbox.com/scl/fi/1okfo8964bw5m53mjmxs6/ >>> Pecl_One_Shot_1.jpg?rlkey=dwus6qjyqsynl2rv6uje2xxtj&raw=1 >>>> >>>> I can use the other half of the dual comparator elsewhere. This one has >>>> programmable hysteresis too. >>> >>> Exposing my contemptworthy ignorance here no doubt, but what exactly is it >>> that is fast about this design, John? And what would its application be? >> >> It would have a leading-edge prop delay of 350 ps, a minimum pulse >> width just over 1 ns, and would easily trigger at 150 MHz. Integrated >> one-shots, xxx123 types, are always terrible, even the obsolete ECL >> ones. >> >> It would be a programmable delay in a number of possible products, >> including some optical modulators. Just playing right now. >> >> More dollars could certainly do something faster, but as long as >> customers connect boxes with cables, faster isn't worth a lot. Light >> just moves too slow. >> >> Any suggestions for a faster one-shot? >> > >What about connecting the + of the RC to Q to speed up the discharge Pecl only swings about 0.8 volts, up close to Vcc, so it wouldn't help much. That phemt turns on pretty hard, maybe 100 mA with Rds-on a couple of ohms. Speeding up the discharge would improve rep-rate but wouldn't affect minimum pulse width. I'd need a faster comparator for that, or some different architecture. The phemt gate driver is a little scary but will work if done carefully. I could give up a bit of speed and do this in CMOS. There's a 1 ns TinyLogic flop that's 36 cents.