---+ List of modifications for Rev 4 SQUID Controller

These are changes for the boards that will be built for SPTpol.

  1. footprint for U18 -5V regulator is wrong, the pad should be connected to VIN, not GND. This part is changing anyway, see below. (done)
  2. Change U19 (beocmes U810) +5VA regulator (3 pin TO-220) to an adjustable (MIC2941AWT, 5 pin TO-220), and set it to 4.5V (done - verified R1/R2=265K/100K --> 4.5V).
  3. Change U20 (becomes U809) +5VD regulator (3 pin TO-220) to an adjustable (MIC2941AWT, 5 pin TO-220), and set it to 4.75V (done - verified R1/R2=285K/100K --> 4.75V, this is the rec'd minimum for the AD5532 DAC).
  4. change U18 (becomes U811) -5V regulator (3 pin TO-220) to an adjustable one and set it to -4.5V (verified R1=2.4K, R2=6.53K)
    • use LM2991S, TO-263 5 pin, http://www.national.com/pf/LM/LM2991.html (done)
    • Double check that the power-pad is connected to vin, not ground.
    • we're reducing the voltage of the regulators because the dfmux boards have an max rec'd input of 5.5V, and there can be substantial voltage drop across the cable to the squid control. This can leave not enough drop-out voltage. The MIC2941 needs 0.2-0.3V of dropout.
  5. Add a resistor in series with capacitor Cx04 connecting SQUID_Vout to ground for each of the 8 channels. (this is the lead-lag).
    • Cx04 is on the top of the board and a via connects to ground. Simply make this via floating, and place an 0603 10 ohm resistor on the bottom of the board.
    • for channel 1, you've called this resistor R8008, which is confusing since all components on a given channel have names like Cx04, where x is the channel number. Could you name this resistor something like Rx99, where x is the channel number, so that it fits the naming scheme and is easier to document the lead-lag?
    • The new resistor, e.g. R8008, sits very close to bypass cap C105. Probably this is ok- but I worry a little about this, since the resistor carries the most important signal on the board, before it is amplifier. Suggest rotating the resistor 90 degrees and moving it slightly out away from the cap, as shown by the red rectange in the image attached to the email.

  • Please change text on schematic: (pad was connected to +7V by mistake) --> (pad was connected to +7V by mistake in previous revision)
  • Please change text on schematic in two places: (pad was floating) --> (pad was floating in previous revision)
Assembly
  1. adjust values to increase Vzero range of op-amp. Martin to provide values in consultation with matt. (this is a parameter change and does not affect layout)
    • RX31 has changed from 8.06 k to 3.6 k. (Here X is the channel number)
  2. for all analog resistors, use low temperature coefficient 25ppm 0.5% parts, such as the SUSUMU RR0816P-101-D, available from newark for a small premium. Watch out, lead times can be long.
  • NOTE: C41 is a 10 uF ceramic cap. This produces a 10 kHz oscillation and will need to be replaced. The replacement component is a 10 uF >= 16 V, tantulum cap. size: 1206. The component part number we've used is: p/n 293D106X9016A2TE3
Lead_lag_modifications - Testing modifications to the rev 4 lead lag filter

-- MattDobbs - 28 Aug 2009

This topic: AnalogFMux > WebHome > SquidControllerRev4Mods Topic revision: r10 - 2012-10-15 - JamesKennedy
© 2020 Winterland Cosmology Lab, McGill University, Montréal, Québec, Canada