While we are waiting for a final design of a new temperature sensor board we can make use of the old LIGO Temperature Sensor interface board (D980400) that was used to sense temperature before. I had put it aside because I didn't understand the modifications (chips glued upside down with dead bug wiring) and it was noisy (ringing) when plugged into acromag. But we can debug this to get a temp sensor up and running quickly while the Kira and Kevin build something better.
When we plugged in the AD590s to the board and then attached this to the acromag slow inputs the circuits went into ringing mode. Rana has suggested that we look into adding more (100 Ω) series resistance on the output so that the output op amps are not overloaded (pulling more current than they can handle). I understand we need to choose low noise thin metal film surface mount type. I seem to recall a comment from Frank in the elog that he didn't have any no precision resistors at the time and that he only wanted to use it for monitoring anyway: we can fix this when we debug. We need to work out what the true schematic is and why, with a ~ 100 kΩ load connected, the circuit is in an unstable state. As you can see in the picture attached below the dead bug job uses axial resistors (non-surface mount) on a mystery chip.
I think post PSL:108 summarizes the main modifications that Frank made to the board. There are other tit-bits of information that can be found with a search of AD590 back throught the elog
Also, I found the box and lid, which were being used as storage for screws and bits: we'll put foam inside the reduce thermal fluctuations of electronics.
Kevin and Kira should come over and we'll look at it with a scope. |