I would like to isolate my thermocouple amplifier using a adum1200. Given the following illustration of the chip, does this chip allow me bidirectional comms on the same pin?

enter image description here

I hooked up power on vdd1 and gnd1 to 5v rail on my uno.

then provided another sperate 5v (plug pack and 7805) to vdd2 and gnd2.

ViA ----> to my digital pin 2 then voA to my one wire amplifier.

The code says it can't find the amplified but it works pefectly fine when I don't use the amplifier.

I'm guessing I need a chip that has an open drain output

I fell into the trap of believing the banner specs on the datasheet if that's the case.

out of interest how are they claiming it's bi directional then. iS it because the adum1201 has 1/1 directionality

  • \$\begingroup\$ I'd guess you need the adum1201 ...analog.com/media/en/technical-documentation/data-sheets/… but even then more digging is needed \$\endgroup\$ – Spoon Feb 21 '17 at 13:12
  • \$\begingroup\$ ... and that chip does not offer bidirectional coms if the figure is correct-- either on the same pin or using different pins. Both channels go in the same direction. \$\endgroup\$ – Scott Seidman Feb 21 '17 at 13:31
  • \$\begingroup\$ maximintegrated.com/en/app-notes/index.mvp/id/5134 \$\endgroup\$ – Scott Seidman Feb 21 '17 at 13:34
  • \$\begingroup\$ TBH i ASSUMED (like an idiot) the square s symbol means it's bidirectional. I now know it stands for schmitt trigger. WHEN I FLAMING LEARN \$\endgroup\$ – Ageis Feb 21 '17 at 15:52

Use directions in Dallas chip directly and properly, instead of isolation and get that working first. Then show waveforms if flawed.

STRONG PULLUP To provide energy for a temperature conversion or for copying data from the scratchpad to the EEPROM,a low-impedance pullup of the 1-Wire bus to 5V is required just after the corresponding command has been sent by the master. During temperature conversion or copying the scratchpad, the bus master controls the transition from a state where the data line is idling high via the pullup resistor to a state where the data line is actively driven to 5V, providing a minimum of 1mA of current for each DS1920 doing temperature conversion. This low impedance pullup should be active for 0.75 seconds for temperature conversion or at least 10ms for copying to the scratchpad. After that, the data line returns to an idle high state controlled by the pullup resistor.

read the following


in particular ... impedance matching and active pullup

Consider for EMI reduction, and use twisted pair with CM choke for improved noise rejection.

If you need isolation for hiV or CM rejection give reasons and consider above instead.


Assuming your previous thermal questions are related, do you have the negative side of the Thermocouple wire grounded to aluminum block? The tip must be electrically insulated yet thermally coupled, perhaps with oil or thermal grease.

enter image description here

extra info

Thermal oils are available in a wide range of specifications. Different manufacturers produce different oils under different brand names. Different specifications suit various needs of the process and applications. Some of the commonly used brands are

  • Therminol from Solutia Inc.
  • Dowtherm from Dow Chemical Co.
  • Exceltherm from Radco Industries Inc.
  • Paratherm from Paratherm Corp.

The most important characteristic is the maximum temperature of service. The oil is thermally stable till this temperature. It should be borne in mind that the cost increases exponentially with increase in the maximum temperature limit. Currently the thermal oils are available up to 400 °C.

  • \$\begingroup\$ hi tony can you explain what you mean by directions? the dallas temperature library am using doesn't mention this. It does work without the isolator. The reason I need isolation is to have a floating front end for my thermocouple as suggested by another EE on this site. \$\endgroup\$ – Ageis Feb 21 '17 at 16:09
  • \$\begingroup\$ Sorry Temp only is unidirectional, but commands require bi-directional communication. I think a balun or CM choke is better. It gives low differential impedance and high CM impedance over the frequency of interest, which is low here so large CM choke is needed, larger than ethernet PHY balun \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Feb 21 '17 at 16:13
  • \$\begingroup\$ Sorry tony so you;re saying you think I should just add a common mode choke and forget about the digital isolator? if so can you advise I'm just a average hobbiest wanting a working 3d printer. \$\endgroup\$ – Ageis Feb 21 '17 at 18:12
  • \$\begingroup\$ something of this size you mean uk.rs-online.com/web/p/leaded-inductors/7926173 \$\endgroup\$ – Ageis Feb 21 '17 at 18:33
  • \$\begingroup\$ no this is around same inductance but higher current. First scope your signals in lab to simulate environment then see what problems exist (vs datasheet) and implement active pullup then twisted pair for sensor and ferrite sleeve over printer motor wire pairs if not already, then last resort CM choke for sensor. \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Feb 21 '17 at 19:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.