john larkin <jl@650pot.com> Wrote in message:r> https://www.dropbox.com/scl/fi/weo0og6siaf5ano8thng3/Opto_Shift_Reg.jpg?rlkey=g170fs3jzg6zftzrrvr3phmd4&raw=1A short low pulse shifts a 1 and a long pulse shifts a 0.Dallas one wire interface? I recall some Manchester chips that could do that. Ran some hp optical interfaces, worked ok. Cheers -- ----Android NewsGroup Reader---- https://piaohong.s3-us-west-2.amazonaws.com/usenet/index.html
opto shift register
Started by ●December 18, 2023
Reply by ●December 20, 20232023-12-20
Reply by ●December 20, 20232023-12-20
On Wed, 20 Dec 2023 06:33:50 -0800 (PST), Fred Bloggs <bloggs.fredbloggs.fred@gmail.com> wrote:>On Monday, December 18, 2023 at 1:24:58?PM UTC-5, john larkin wrote: >> https://www.dropbox.com/scl/fi/weo0og6siaf5ano8thng3/Opto_Shift_Reg.jpg?rlkey=g170fs3jzg6zftzrrvr3phmd4&raw=1 >> >> A short low pulse shifts a 1 and a long pulse shifts a 0. > >You can't beat that for simplicity and reliability. The first bit always has to be a 0,Why?>and you probably want to blank the output whenever you put the register into a new state.That wouldn't matter in my case. I'm just configuring current sources and gains, a setup-type operation.
Reply by ●December 20, 20232023-12-20
On Wednesday, December 20, 2023 at 4:21:57 AM UTC-8, John Larkin wrote:> On Wed, 20 Dec 2023 05:18:42 GMT, Jan Panteltje <al...@comet.invalid> > wrote: > > >On a sunny day (Tue, 19 Dec 2023 07:31:03 -0800) it happened John Larkin > ><j...@997PotHill.com> wrote in <r9d3oipchi8i1no29...@4ax.com>: > > > >>On Tue, 19 Dec 2023 06:04:01 GMT, Jan Panteltje <al...@comet.invalid> > >>wrote: > >> > >>>On a sunny day (Mon, 18 Dec 2023 10:24:39 -0800) it happened john larkin> >>>>A short low pulse shifts a 1 and a long pulse shifts a 0. > >>> > >>>What,s the use? > >> > >>One channel of an isolated temperature acquisition box. It will do > >>RTDs, thermocouples, IC sensors, and cryo diodes. I need several > >>isolated logic levels... > >I usually grab a Microchip PIC micro. > > > >It has many outputs that can drive optocouplers, > >both for in - and outputs. > >Then it can do Ethernet to a main computer too with one extra chip. > >I am a bit worried about the timing capacitors in your idea,> ...People sometimes weld thermocouples to whatever, or bolt raw > cryo diodes to metal things, so I need true isolation. > > If I do the long/short pulse thing, the pulse width ratio can be 10:1 > or 100:1 so capacitor tolerance wouldn't matter.It's also possible to cut out the isolation part of the problem by using an Ethernet signalling scheme (one of the old ones) and an Ethernet integrated connector/magnetics with some suitable POE attached. The integrated connector/transformer/PHY front end is cheap. Early Ethernet was basically just a serializer scheme anyhow; ditch TCP/IP, go to a user datagram. As for power isolation, there's no reason you couldn't have a little rechargable battery with a DPDT relay to connect it long enough to poll a sensor, then revert to the float charger.
Reply by ●December 20, 20232023-12-20
On Wed, 20 Dec 2023 13:04:58 GMT, Jan Panteltje <alien@comet.invalid> wrote:>On a sunny day (Wed, 20 Dec 2023 04:20:54 -0800) it happened John Larkin ><jl@997PotHill.com> wrote in <lim5oi5afhvee38tb5can9id36k1bgcic9@4ax.com>: > >>On Wed, 20 Dec 2023 05:18:42 GMT, Jan Panteltje <alien@comet.invalid> >>wrote: >> >>>On a sunny day (Tue, 19 Dec 2023 07:31:03 -0800) it happened John Larkin >>><jl@997PotHill.com> wrote in <r9d3oipchi8i1no29gu1rlccg0mlnn3llc@4ax.com>: >>> >>>>On Tue, 19 Dec 2023 06:04:01 GMT, Jan Panteltje <alien@comet.invalid> >>>>wrote: >>>> >>>>>On a sunny day (Mon, 18 Dec 2023 10:24:39 -0800) it happened john larkin >>>>><jl@650pot.com> wrote in <re31oid9dqb1136cj5rctmnruvvhrm47nl@4ax.com>: >>>>> >>>>>>https://www.dropbox.com/scl/fi/weo0og6siaf5ano8thng3/Opto_Shift_Reg.jpg?rlkey=g170fs3jzg6zftzrrvr3phmd4&raw=1 >>>>>> >>>>>>A short low pulse shifts a 1 and a long pulse shifts a 0. >>>>> >>>>>What,s the use? >>>> >>>>One channel of an isolated temperature acquisition box. It will do >>>>RTDs, thermocouples, IC sensors, and cryo diodes. I need several >>>>isolated logic levels to switch ranges and such, and a multi-channel >>>>logic isolator is big and expensive. I don't need speed. >>>> >>>>The alternative is to use a bunch of single-lane SSRs. >>>> >>>>Or maybe float a uP per channel and talk to it with one bidirectional >>>>isolator. >>>> >>>>Early stages of design, many possibilities. >>> >>>For that, if I understand your description right, >>>I usually grab a Microchip PIC micro. >>> >>>It has many outputs that can drive optocouplers, >>>both for in - and outputs. >>>Then it can do Ethernet to a main computer too with one extra chip. >>>I am a bit worried about the timing capacitors in your idea, >>>caps may drift over time and temperature? >> >>I plan to use one RP2040 in the box, as the top level manager. The >>question is then how to talk to four isolated temprature measurement >>channels. People sometimes weld thermocouples to whatever, or bolt raw >>cryo diodes to metal things, so I need true isolation. >> >>A uP per channel is probably overkill. I've done that before but it's >>a nuisance. > > >Well the thing will have to understand commands, and sent data.. >or is it just one-way traffic? sensor to raspberry? >Using logic takes more chips than a micro. > > >>If I do the long/short pulse thing, the pulse width ratio can be 10:1 >>or 100:1 so capacitor tolerance wouldn't matter. > >Looks like a simple serial protocol like for example used via PC serial port would be OK. >2 photo couplers per channel, one for each direction, >You need supply insulation too, so one potcore at RF with a few flat cable thick windings, one for each sensor unit > https://www.panteltje.online/pub/1_to_1_transformer_in_1_minute.jpgIt will take more than a minute to mount and connect that on a PCB.>1 transistor sine wave oscillator to power all units: > https://www.panteltje.online/pub/tuned_eprom_programmer_voltage_generator.gifA stock tiny surface-mount isolated dc/dc converter costs about $2.>Raspi can mimic several serial ports via the GPIO pins. >Depends a bit how many task interups happen via the software. > >On the Pi4 there seem to be 4 hardware serial ports? > https://forums.raspberrypi.com/viewtopic.php?t=244528 >I only ever used ttyAMA0 (GPS module connected to it).I could bit-bang it too. But I'll use the Pico processor, the RP2040 chip. The Pi4 is about as big as my entire product! One might use Ethernet magnetics, two transformers, as a logic isolator too. Clock and data.
Reply by ●December 20, 20232023-12-20
On Tue, 19 Dec 2023 07:31:03 -0800, John Larkin <jl@997PotHill.com> wrote:>On Tue, 19 Dec 2023 06:04:01 GMT, Jan Panteltje <alien@comet.invalid> >wrote: > >>On a sunny day (Mon, 18 Dec 2023 10:24:39 -0800) it happened john larkin >><jl@650pot.com> wrote in <re31oid9dqb1136cj5rctmnruvvhrm47nl@4ax.com>: >> >>>https://www.dropbox.com/scl/fi/weo0og6siaf5ano8thng3/Opto_Shift_Reg.jpg?rlkey=g170fs3jzg6zftzrrvr3phmd4&raw=1 >>> >>>A short low pulse shifts a 1 and a long pulse shifts a 0. >> >>What,s the use? > >One channel of an isolated temperature acquisition box. It will do >RTDs, thermocouples, IC sensors, and cryo diodes. I need several >isolated logic levels to switch ranges and such, and a multi-channel >logic isolator is big and expensive. I don't need speed. > >The alternative is to use a bunch of single-lane SSRs. > >Or maybe float a uP per channel and talk to it with one bidirectional >isolator. > >Early stages of design, many possibilities.Incidentally, cryo diodes are interesting parts. They are just well-charactized small diodes, or maybe diode-connected transistors. They act like normal diodes down to around 20K, and then "carrier freeze-out" kicks in and voltage drop goes way up as temp goes down. I think they get ohmic too. Lakeshore makes good ones. They are specified at 10 uA forward current, probably to minimize self-heating. https://www.lakeshore.com/products/categories/overview/temperature-products/cryogenic-temperature-sensors/dt-670-silicon-diodes I think transistor action stops below 20K too. I recall that jfets still work cold.
Reply by ●December 21, 20232023-12-21
On a sunny day (Wed, 20 Dec 2023 11:19:43 -0800) it happened john larkin <jl@650pot.com> wrote in <42f6oitf4inj4un46nu3nkgj5n9as2mt62@4ax.com>:>On Wed, 20 Dec 2023 13:04:58 GMT, Jan Panteltje <alien@comet.invalid> >wrote: > >>On a sunny day (Wed, 20 Dec 2023 04:20:54 -0800) it happened John Larkin >><jl@997PotHill.com> wrote in <lim5oi5afhvee38tb5can9id36k1bgcic9@4ax.com>: >> >>>On Wed, 20 Dec 2023 05:18:42 GMT, Jan Panteltje <alien@comet.invalid> >>>wrote: >>> >>>>On a sunny day (Tue, 19 Dec 2023 07:31:03 -0800) it happened John Larkin >>>><jl@997PotHill.com> wrote in <r9d3oipchi8i1no29gu1rlccg0mlnn3llc@4ax.com>: >>>> >>>>>On Tue, 19 Dec 2023 06:04:01 GMT, Jan Panteltje <alien@comet.invalid> >>>>>wrote: >>>>> >>>>>>On a sunny day (Mon, 18 Dec 2023 10:24:39 -0800) it happened john larkin >>>>>><jl@650pot.com> wrote in <re31oid9dqb1136cj5rctmnruvvhrm47nl@4ax.com>: >>>>>> >>>>>>>https://www.dropbox.com/scl/fi/weo0og6siaf5ano8thng3/Opto_Shift_Reg.jpg?rlkey=g170fs3jzg6zftzrrvr3phmd4&raw=1 >>>>>>> >>>>>>>A short low pulse shifts a 1 and a long pulse shifts a 0. >>>>>> >>>>>>What,s the use? >>>>> >>>>>One channel of an isolated temperature acquisition box. It will do >>>>>RTDs, thermocouples, IC sensors, and cryo diodes. I need several >>>>>isolated logic levels to switch ranges and such, and a multi-channel >>>>>logic isolator is big and expensive. I don't need speed. >>>>> >>>>>The alternative is to use a bunch of single-lane SSRs. >>>>> >>>>>Or maybe float a uP per channel and talk to it with one bidirectional >>>>>isolator. >>>>> >>>>>Early stages of design, many possibilities. >>>> >>>>For that, if I understand your description right, >>>>I usually grab a Microchip PIC micro. >>>> >>>>It has many outputs that can drive optocouplers, >>>>both for in - and outputs. >>>>Then it can do Ethernet to a main computer too with one extra chip. >>>>I am a bit worried about the timing capacitors in your idea, >>>>caps may drift over time and temperature? >>> >>>I plan to use one RP2040 in the box, as the top level manager. The >>>question is then how to talk to four isolated temprature measurement >>>channels. People sometimes weld thermocouples to whatever, or bolt raw >>>cryo diodes to metal things, so I need true isolation. >>> >>>A uP per channel is probably overkill. I've done that before but it's >>>a nuisance. >> >> >>Well the thing will have to understand commands, and sent data.. >>or is it just one-way traffic? sensor to raspberry? >>Using logic takes more chips than a micro. >> >> >>>If I do the long/short pulse thing, the pulse width ratio can be 10:1 >>>or 100:1 so capacitor tolerance wouldn't matter. >> >>Looks like a simple serial protocol like for example used via PC serial port would be OK. >>2 photo couplers per channel, one for each direction, >>You need supply insulation too, so one potcore at RF with a few flat cable thick windings, one for each sensor unit >> https://www.panteltje.online/pub/1_to_1_transformer_in_1_minute.jpg > >It will take more than a minute to mount and connect that on a PCB.OK, example, take a 8 strand piece of flat cable mark both ends at the same side with a black dot. fold the cable and put it 3 times through the ringcore. Now press the 16 strands dot up in one of those flat cable connectors. Takes 1 minute for somebody who plays a lot with electronics. Now you have 8 windings of 3 turns! Now here comes your PCB experience: you can make 8 windings of 3 turns or 1 winding of 6 and 6 of 3 or any other combination by PCB layout to those header pins the connector fits in, So 7 insulated power output windings, 1 input maybe 6 outputs and one feedback using the ringcore thing as oscillator. whatever combination you want. I should have patented that, but 4 sure somebody would have found it earlier and I do not like patents anyways, open source idea!... Ringcore oscillators I use up to very high power: https://panteltje.online/pub/h501s_drone_remote_power_lab_test_IMG_6271.JPG Ringcores everywhere, in the junk box, many from old computer supplies.>>1 transistor sine wave oscillator to power all units: >> https://www.panteltje.online/pub/tuned_eprom_programmer_voltage_generator.gif > >A stock tiny surface-mount isolated dc/dc converter costs about $2. > > >>Raspi can mimic several serial ports via the GPIO pins. >>Depends a bit how many task interups happen via the software. >> >>On the Pi4 there seem to be 4 hardware serial ports? >> https://forums.raspberrypi.com/viewtopic.php?t=244528 >>I only ever used ttyAMA0 (GPS module connected to it). > >I could bit-bang it too. But I'll use the Pico processor, the RP2040 >chip. The Pi4 is about as big as my entire product! > >One might use Ethernet magnetics, two transformers, as a logic >isolator too. Clock and data. > >
Reply by ●December 21, 20232023-12-21
On a sunny day (Wed, 20 Dec 2023 17:53:13 -0800) it happened John Larkin <jl@997PotHill.com> wrote in <eo57oi5oooqal8g9hc60c0iq2i87mmbs4f@4ax.com>:>On Tue, 19 Dec 2023 07:31:03 -0800, John Larkin <jl@997PotHill.com> >wrote: > >>On Tue, 19 Dec 2023 06:04:01 GMT, Jan Panteltje <alien@comet.invalid> >>wrote: >> >>>On a sunny day (Mon, 18 Dec 2023 10:24:39 -0800) it happened john larkin >>><jl@650pot.com> wrote in <re31oid9dqb1136cj5rctmnruvvhrm47nl@4ax.com>: >>> >>>>https://www.dropbox.com/scl/fi/weo0og6siaf5ano8thng3/Opto_Shift_Reg.jpg?rlkey=g170fs3jzg6zftzrrvr3phmd4&raw=1 >>>> >>>>A short low pulse shifts a 1 and a long pulse shifts a 0. >>> >>>What,s the use? >> >>One channel of an isolated temperature acquisition box. It will do >>RTDs, thermocouples, IC sensors, and cryo diodes. I need several >>isolated logic levels to switch ranges and such, and a multi-channel >>logic isolator is big and expensive. I don't need speed. >> >>The alternative is to use a bunch of single-lane SSRs. >> >>Or maybe float a uP per channel and talk to it with one bidirectional >>isolator. >> >>Early stages of design, many possibilities. > >Incidentally, cryo diodes are interesting parts. They are just >well-charactized small diodes, or maybe diode-connected transistors. >They act like normal diodes down to around 20K, and then "carrier >freeze-out" kicks in and voltage drop goes way up as temp goes down. I >think they get ohmic too. > >Lakeshore makes good ones. They are specified at 10 uA forward >current, probably to minimize self-heating. > >https://www.lakeshore.com/products/categories/overview/temperature-products/cryogenic-temperature-sensors/dt-670-silicon-diodes > >I think transistor action stops below 20K too. I recall that jfets >still work cold.Thermocouples work fine too: https://panteltje.online/pub/cryo/ you can make your own, and the electronics that comes with it: https://panteltje.online/panteltje/pic/th_pic/ Exploring things is fun. Curiosity :-) and then apply it when needed!
Reply by ●December 21, 20232023-12-21
On Thu, 21 Dec 2023 06:13:42 GMT, Jan Panteltje <alien@comet.invalid> wrote:>On a sunny day (Wed, 20 Dec 2023 17:53:13 -0800) it happened John Larkin ><jl@997PotHill.com> wrote in <eo57oi5oooqal8g9hc60c0iq2i87mmbs4f@4ax.com>: > >>On Tue, 19 Dec 2023 07:31:03 -0800, John Larkin <jl@997PotHill.com> >>wrote: >> >>>On Tue, 19 Dec 2023 06:04:01 GMT, Jan Panteltje <alien@comet.invalid> >>>wrote: >>> >>>>On a sunny day (Mon, 18 Dec 2023 10:24:39 -0800) it happened john larkin >>>><jl@650pot.com> wrote in <re31oid9dqb1136cj5rctmnruvvhrm47nl@4ax.com>: >>>> >>>>>https://www.dropbox.com/scl/fi/weo0og6siaf5ano8thng3/Opto_Shift_Reg.jpg?rlkey=g170fs3jzg6zftzrrvr3phmd4&raw=1 >>>>> >>>>>A short low pulse shifts a 1 and a long pulse shifts a 0. >>>> >>>>What,s the use? >>> >>>One channel of an isolated temperature acquisition box. It will do >>>RTDs, thermocouples, IC sensors, and cryo diodes. I need several >>>isolated logic levels to switch ranges and such, and a multi-channel >>>logic isolator is big and expensive. I don't need speed. >>> >>>The alternative is to use a bunch of single-lane SSRs. >>> >>>Or maybe float a uP per channel and talk to it with one bidirectional >>>isolator. >>> >>>Early stages of design, many possibilities. >> >>Incidentally, cryo diodes are interesting parts. They are just >>well-charactized small diodes, or maybe diode-connected transistors. >>They act like normal diodes down to around 20K, and then "carrier >>freeze-out" kicks in and voltage drop goes way up as temp goes down. I >>think they get ohmic too. >> >>Lakeshore makes good ones. They are specified at 10 uA forward >>current, probably to minimize self-heating. >> >>https://www.lakeshore.com/products/categories/overview/temperature-products/cryogenic-temperature-sensors/dt-670-silicon-diodes >> >>I think transistor action stops below 20K too. I recall that jfets >>still work cold. > >Thermocouples work fine too: > https://panteltje.online/pub/cryo/ >you can make your own, and the electronics that comes with it: > https://panteltje.online/panteltje/pic/th_pic/ >Exploring things is fun.For sure. And the physics boys often need help with electronics. You would think...> >Curiosity :-) > >and then apply it when needed!How low can a thermocouple work? Any error in the reference junction temp measurement is additive (or worse) and that matters near 0K. What did you use for the ref junction measurement? We usually use a platinum RTD working against a Susumu super-resistor. Old carbon comp resistors have been used as cryo temp sensors, but I think they need to be individually calibrated. Lakeshore has a bunch of exotic cryo sensors. Temperature measurement is generally nasty. I could rant about Global Warming instrumentation but I won't.
Reply by ●December 22, 20232023-12-22
On 22/12/2023 12:14 am, John Larkin wrote:> On Thu, 21 Dec 2023 06:13:42 GMT, Jan Panteltje <alien@comet.invalid> > wrote: >> On a sunny day (Wed, 20 Dec 2023 17:53:13 -0800) it happened John Larkin >> <jl@997PotHill.com> wrote in <eo57oi5oooqal8g9hc60c0iq2i87mmbs4f@4ax.com>: >>> On Tue, 19 Dec 2023 07:31:03 -0800, John Larkin <jl@997PotHill.com> >>> wrote: >>>> On Tue, 19 Dec 2023 06:04:01 GMT, Jan Panteltje <alien@comet.invalid> >>>> wrote: >>>>> On a sunny day (Mon, 18 Dec 2023 10:24:39 -0800) it happened john larkin >>>>> <jl@650pot.com> wrote in <re31oid9dqb1136cj5rctmnruvvhrm47nl@4ax.com>:<snip>> Temperature measurement is generally nasty. I could rant about Global > Warming instrumentation but I won't.For once. Your rant is actually Anthony Watts' rant, and climate change denial propaganda, but you don't know enough to realise quite how silly Anthony Watts' obsessions are. -- Bill Sloman, Sydney
Reply by ●December 22, 20232023-12-22
On Thursday, December 21, 2023 at 5:15:27 AM UTC-8, John Larkin wrote:> On Thu, 21 Dec 2023 06:13:42 GMT, Jan Panteltje <al...@comet.invalid> > wrote:> >Thermocouples work fine too: > > https://panteltje.online/pub/cryo/ > >you can make your own, and the electronics that comes with it: > > https://panteltje.online/panteltje/pic/th_pic/> How low can a thermocouple work?It's not so great at the lowest temperatures; I've used gold/ 0.7%Fe versus gold for helium temperatures, but sensitivity is poor. That third law of thermodynamics says it fails at zero.
