We noticed that the Setpoint contact of several (so far 8 out of 14) IPCMini opens for some minutes every 32 days, although the current is well below the Setpoint. Is this a software bug?

I do not know, but I will inquire.

I just received confirmation that you may be correct, and a revision is in progress.

Thank you. Recently, we noticed something like the "next level".

Two of our IPCMini were delivered approximately 2 years ago.They have a higher number of opertional hours.

Serial IT1841C146 and IT1835C518.

Since this monday 10th May, they show an open SetPoint although the current is well below the set value of 10uA.

Now it is wednesday 12th May and still the contact is open.

So the behavior seems to depend on the total operating hours.

We see the effect in pairs, because we use a set of 2 IPCMini for each Thales Klystron, and operate them manually at nearly the same time.

Thank you, I forwarded your additional information along to the appropriate channels.

Thank you. While we're at it, there is another issue.

The scaling of the IPCMini analog output can be set from menu.

If the 230V power is cut and restored, the IPCMini will remember these 2 set values of the output scaling. Good.

However, if somebody checks these values anytime after boot and presses "confirm", then these 2 values are set to zero.

Thus, a second checking of the values reveals this.

Here is a more detailed information as "picture story" combined with measurements as PDF from Klystron #8 from Thales.

Best regards,

Dr. Alexander Schnase

Dear George, thank you.

Some time has passed, and our measurements have continued.

Now with 12 out of 14 IPCMini, we recorded the issues with the unwanted behaviour of the setpoint contact.

Recently, we found the clue.

The event "setpoint contact opens" seems to happen every 49,7 days. So, for a new IPCMini one has to wait quite a long time.

This event depends on the time, an IPCMini was connected to the 230 V mains.

This number tells me, that there might be a 32-bit timer that counts the milliseconds. If this counter hits 2^32, then it overflows.

Well, 2^32 milliseconds is 49,707 days.

There should be a second counter that counts the time, that the 5 kV high voltage is on.

Probably, this is also a 32 timer or a 32 bit variable. This also overflows, when reaching 2^32.

The overflow of the timer for "230 V" makes the setpoint going to the wrong state.

The overflow of the other timer "5 kV on", makes the setpoint to go back to the correct state.

Thus, if a fresh IPCMini is taken out of the box, and connected to 230V, the timer for "230 V" is counting.

Once the 5 kV is started, the timer for "5 kV" is counting.

If the time between "230 V" and "5 kV" is short, then the wrong state of the setpoint only appears for a short time (we habe seen minutes and hours).

If there is a significant time, when the 230 V is on, and the 5 kV is off, then the difference between both timers is larger and the wrong state of the setpoint appears (we have seen days and weeks).

Finally, each 49,7 days, the situation repeats.

Now you know, where to look up in the controller software.

Best wishes,

Alexander Schnase