I might use an LED+mosfet type SSR to switch a voltage divider at the input of an isolated ADC. The ON resistance of the switch is an error contributor, and a super-low Ron is expensive. 1 or 2 ohms is affordable. I'm thinking that driving the LED side from the proper Thevenin source might temperature compensate the ON resistance of the mosfets. Or even program the LED supply with a DAC. LT Spice dosn't seem to have any led-mosfet SSR models. I guess I'll have to try it.
SSR tempco
Started by ●December 13, 2023
Reply by ●December 13, 20232023-12-13
onsdag den 13. december 2023 kl. 21.23.40 UTC+1 skrev John Larkin:> I might use an LED+mosfet type SSR to switch a voltage divider at the > input of an isolated ADC. > > The ON resistance of the switch is an error contributor, and a > super-low Ron is expensive. 1 or 2 ohms is affordable. > > I'm thinking that driving the LED side from the proper Thevenin source > might temperature compensate the ON resistance of the mosfets. Or even > program the LED supply with a DAC. > > LT Spice dosn't seem to have any led-mosfet SSR models. I guess I'll > have to try it.TLP3905 or similar and two low Ron fets ?
Reply by ●December 13, 20232023-12-13
On Wed, 13 Dec 2023 12:39:30 -0800 (PST), Lasse Langwadt Christensen <langwadt@fonz.dk> wrote:>onsdag den 13. december 2023 kl. 21.23.40 UTC+1 skrev John Larkin: >> I might use an LED+mosfet type SSR to switch a voltage divider at the >> input of an isolated ADC. >> >> The ON resistance of the switch is an error contributor, and a >> super-low Ron is expensive. 1 or 2 ohms is affordable. >> >> I'm thinking that driving the LED side from the proper Thevenin source >> might temperature compensate the ON resistance of the mosfets. Or even >> program the LED supply with a DAC. >> >> LT Spice dosn't seem to have any led-mosfet SSR models. I guess I'll >> have to try it. > >TLP3905 or similar and two low Ron fets ?That would work; we have some PV optocouplers in stock. But I'd prefer a simpler cheaper smaller way to switch gains. I could use an analog mux with a photocoupler of some sort, and have zero error. But that has its own problems. https://www.dropbox.com/scl/fi/7q6kjklsjfobfo4s0fv8y/B450_Amp_1.jpg?rlkey=o1uspw3bs78kqr09y9g78szxo&raw=1 The four ADC ranges might be +-100 mV to +-3 volts, something like that.
Reply by ●December 13, 20232023-12-13
On Wed, 13 Dec 2023 12:22:43 -0800, John Larkin <jl@997PotHill.com> wrote:>I might use an LED+mosfet type SSR to switch a voltage divider at the >input of an isolated ADC. > >The ON resistance of the switch is an error contributor, and a >super-low Ron is expensive. 1 or 2 ohms is affordable. > >I'm thinking that driving the LED side from the proper Thevenin source >might temperature compensate the ON resistance of the mosfets. Or even >program the LED supply with a DAC. > >LT Spice dosn't seem to have any led-mosfet SSR models. I guess I'll >have to try it.https://www.renesas.com/us/en/products/interface/optoelectronics/solid-state-relay-optical-coupled-mosfet-structure-and-features Looks like four chips inside: LED, PV, and two mosfets. For about 70 cents! I'm thinking there would be no useful relation between LED current and Ron. It looks like on resistance is about PTAT. 1 ohm at 0c might become 1.5 at 100c. Maybe I can live with that. Or buy a lower resistance part.
