• State
  • Locked Locked
  • Replies 4 replies
  • Subscribers 16 subscribers
  • Views 914 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TS5A22364 / TS5A22362 negative signal

Steven Ricketts
Prodigy 30 points
Other Parts Discussed in Thread: TS5A22364, TS5A22362

Hello,

I'm using a TS5A22364 in an application where the signal will be below ground (which this component is supposed to support), but I'm getting severe distortion when the voltage drops more than a few mV below ground. The TS5A22364 is basically a solid-state DPCO relay.

The application is for zero calibration of a differential input (to an ADC) and the part is used to disconnect one side from the signal, and short it to the other side - the ADC can then read this level and use it as a zero reference. The resolution is in the region of a few hundred nV, and voltages are likely to range between -1V and +1V.

The problem I have is that when the voltage drops below about -0.05V, I start getting a noticeable voltage difference between C and NC. For positive voltages, and slightly negative voltages, it works fine.

My interpretation of this is that there's a shunt diode to (from?) ground, and this starts conducting when the voltage drops below ground. If I disconnect the power ground to my circuit, and connect my inputs (which connect to C & NO) to ground, the circuit operates, albeit riding 0.33V above ground at 15mA power consumption or 0.93V at 45mA. Admittedly, this does sound like a resistance of about 21 Ohms, rather than a diode, but whatever it is, it's within the TS5A22364.

I'm running the TS5A22364 at 3.0V, and the datasheet (Rev. E) says it can handle inputs down to V+ -5.5V, so I should be okay from -2.5 to +3.0V?

Interestingly, the "Absolute minimum and maximum ratings" states:

II/OK Analog port diode current
VNC,VNO,VCOM <0 or VNC,VNO,VCOM >V+
(min) -50 mA
(max) 50 mA
Excuse the formatting, but I'm reading this to indicate that there's a clamp diode (rated at 50mA) that comes into effect if any of the analogue pins go below 0V. Surely it should be if it goes below "V+ -5.5V"?
If it's relevant, the impedance of the source is very low (<1 Ohm), and the impedance of the load is negligible (>1 MOhm).
I've thought it was something I'm doing wrong, but we've built three boards up now, and they all behave in exactly the way I've described. The TS5A22364 actually replaced a proper relay that we used before. This worked perfectly, but we need to operate it about every 15 seconds, so the clicking was somewhat irritating!
Any advice or suggestions would be greatly appreciated.
Many thanks,
Steven.
  • 0
    TI__Expert 6395 points

    Hello Steven,

    This is interesting behavior.  Can you provide a schematic that I can view to help determine a root cause? 

    Regards,

    Andrew

  • 0 in reply to Andrew_Mason
    Prodigy 30 points

    Hello Andrew,


    I think I may have found the cause of this problem, and it's related to the shunt switch which is to reduce audible click & pop when changing source. This very rough circuit might highlight the issue, where the "relay" is the TS5A22364 (running on 3.0V):

    The usual application would be to switch between two sources, so it's irrelevant what happens to the unselected source (although shorting it to ground seems a bit harsh). In my application, the relay is used to short the inputs of the ADC together to enable a "zero" reading to be taken.

    The TS5A22364 shorts the unselected input (NO or NC) to ground, so while this doesn't really matter during a "zero" reading (C connected to NO), when I make normal readings (C connected to NC), the other side of my differential input gets shorted to ground.

    The symptoms I was seeing didn't really indicate this to be the problem, and it does seem to work okay when the inputs are very close to ground, but the 21 Ohm shunt to ground I was reading does seem to match the 25-50 Ohm shunt that the datasheet mentions is applied to the unused input.

    I've ordered up some TS5A22362 chips (that don't have the shunt), so we'll have to see if that fixes the problem.

    I think the act of documenting the problem to explain it on the forum was probably enough for me to see where I was going wrong - so thanks for the prompt!

    Steven.

  • 0 in reply to Steven Ricketts
    TI__Expert 6395 points

    Hello Steven,

    The shunt switch is very useful for audio applications but your use case I can see it causing issues.  Let me know how the TS5A22362 works in your application. 

    If there are any further issues you discover I would be happy to provide any support I can.

    Regards,

    Andrew

  • 0 in reply to Andrew_Mason
    Prodigy 30 points

    Hello Andrew,

    I've now put the TS5A22362 into the circuit, and I can immediately see that it's not diverting current to ground through the input pins any more, and performance is greatly improved. However, I'm still getting a change in voltage through the part, proportional to how far away it is from ground.

    I can't measure this directly as the levels are beyond the range of any equipment I have here. However, I've done some tests, and deduced the following:

    Passing a signal at -150mV, I reference this as "perfect" behaviour. Then I pass a signal at -450mV, and there is an error of 360uV (i.e. after the TS5A22362, it's -449.64mV). I've measured this by shorting the inputs together externally (see diagram in earlier post), and connecting them (both) to an external voltage that is lower than ground. I then use the ADC to measure the difference - it's a 24-bit ADC, and 360uV works out to about 930 counts after the divider.

    If we assume this error is pure resistance, and my impedance is 1270k to 5V (5.45V total), which equates to 4.29uA, then the resistance change is 84 Ohms.

    According to the datasheet (fig.3 - Ron vs Vcom), there is a slight lowering of resistance with lower Vcom, but no more than 0.05 Ohm decrease from -150mV to -450mV, whereas I'm seeing a 84 Ohm INCREASE over this range.

    I realise I'm treating this as a pure resistance, and other factors may be having an effect, although I can find no data relating to any other cause of this problem in the documentation. Also, it doesn't measure as an 84 Ohm resistance with a multimeter biased to -450mV, so this needs further investigation.

    I shall do some more testing and get back to you.

    Regards,

    Steven.