[moved] Ultra Low current AC switch

[YdaTron]

Hello Guys,
This is my first ever post in this forum ( If I'm in the wrong forum, please let me know)

I'm building a Test rig, switching couple of voltages both AC and DC and applying them to the DUT.

for the AC voltage application 100VaC/200VaC with an On state current of 1uA - 3uA tops. I'm evaluating different solutions that would be both functional and simple.
At first it was as simple as relay, then figured out that relays have a minimum load switching current in the order of milliamps. Then went for the solid state relay, based on Back to back MOSFETs, but the solution would be too complicated for such a simple application, and the used MOSFETs should have very low IDss compared to the actual load...

Couple of other ideas down the road, than it hit me why not use an actual OptoTriac, knowing that no significant load will ever be driven by it. and as long as it has a small Off current (10nA or less) and a voltage rating of 400V should be fine. I can also use one with Zero cross detection.

what do you guys think ? Any feedback would much appreciate.

 

[KlausST]

Hi,

I don´t think Optotriacs are a good solution because they need some holding_current. More than your couple of mA.

I think about
* solid state MOSFET AC relays. PhotoMOS, like AQY214 ... check specification.
* REED relays for low currents high voltage.

You should specify
* max acceptable leakage current on OPEN contacts
* max acceptable isolation leakage current
* switching time and bouncing time
* max acceptable ON sate resistance
* control voltage
* max acceptable control current
...

Klaus

 

[YdaTron]

Hi,

I don´t think Optotriacs are a good solution because they need some holding_current. More than your couple of mA.

I think about
* solid state MOSFET AC relays. PhotoMOS, like AQY214 ... check specification.
* REED relays for low currents high voltage.

You should specify
* max acceptable leakage current on OPEN contacts
* max acceptable isolation leakage current
* switching time and bouncing time
* max acceptable ON sate resistance
* control voltage
* max acceptable control current
...

Klaus

Thank you Klaus for your feedback.
It is a calibration rig as well, so the leakage current when the *Switch* is of will disturb the other cases ( for example case one : 200Vac, and case 2 : 15VDC. So any possible leakage current shouldnt be significant to disturb the calibration results).

I tried to simulate something like Vishay VO302 (with a leakage off stage current of 10nA) it did the job when used alone. but when trying to simulate the other cases. not so much.

As for the PhotoMos SSR, they should work, but the minimum leaked current that I could find was 1uA ( which should be my on value and not Off)

The REED relays, from I've seen doesn't specify the leakage current, not a minimum load current. but they are a bit expensive (especially if I'm using 10 of them in the same board)

For the specs :
- What is the difference between the first two ? isolation leakage current and Open contact leakage current ? The leakage should insignificant compared to the operational current of 1uA. if not I don't think there would be a difference between Open and close at this point.
- The timing is not an issue, as I can adapt to it and take into account the Turn On/Off durations in my different case.
- Control voltage shouldn't be a problem, as I have 3.6/5/10/12/15.
- Control current would be taken into account in the power tree of the control voltage source.

 

[KlausST]

What is the difference between the first two ? isolation leakage current and Open contact leakage current ?

The one is between open contacts
The other beween control and contacts

So any possible leakage current shouldnt be significant to disturb the calibration results

Then your job is to specify it. ;-)

but when trying to simulate the other cases. not so much.

Don't know what this means.

As for the PhotoMos SSR, they should work, but the minimum leaked current that I could find was 1uA ( which should be my on value and not Off)

As said the AQY.. was just an example. I guess there are photomos that suit better to your application.

***
Good REED relais may be not cheap, but some parameters are outstanding, like the leakage currents.
Thus they are still often used in ATE.

Klaus

 

[Warpspeed]

Off state leakage (and capacitance) will be your biggest problem, especially at higher frequencies.

The trick is to use more than one switch, which can be of any suitable type. Use two switches in series on the hot side, and connect a third shunting switch between those, down to ground.
The shunting switch diverts any potential leakage straight down to ground when both the hot side switches are open.