2
\$\begingroup\$

I need a way to isolate several voltage sources around a large vehicle that feed a single voltage meter some distance away on a panel. Each source is selected by a momentary-spst (at the source).

Here is a block diagram, with 2 discrete options:

enter image description here

All current is DC. Voltage sources run 4.6 VDC to 54.0 VDC. Desired display accuracy is 0.1 Volt. All the power supplies share a common ground.

The (arbitrarily) colored areas display the core differences in the current two options. Option one (green) uses 'ultra-low loss diodes' (ideally an IC array), Option 2 (blue) uses typical signal diodes (1N4148) and a post diode voltage adjustment.

For option 1, the MAX4020 came to mind, but the voltages are out of range. Is their an IC / IC array designed for this?

For option 2, what is the best way to adjust the voltage up 0.7 V?

We're somewhat stuck with the basic layout of using multiple momentary-spst feeding a single voltage meter, so a rotary selector switch is out.


edits:

  • updated 'mom' body text to 'momentary' for clarity.
\$\endgroup\$
  • \$\begingroup\$ Not an answer to your question, but have you considered a voltage divider on the measurement signals to bring them all into a usable range, and then use a multiplier on the display? See MAX4590 for example \$\endgroup\$ – Peter M Nov 8 '18 at 12:48
  • \$\begingroup\$ There's an op amp circuit you may want to look up, called an "ideal diode" or "super diode". \$\endgroup\$ – Hearth Nov 8 '18 at 12:52
  • \$\begingroup\$ Another 'not an answer to your question': alternative is to replace the MOM-spst with a relay + circuit which refuses to switch of there is a voltage at the 'other' side. \$\endgroup\$ – Oldfart Nov 8 '18 at 12:52
  • \$\begingroup\$ There are 'ideal diode' ICs. Even with 'enable' input so they can act as a switch. e.g. MAX40200. \$\endgroup\$ – Oldfart Nov 8 '18 at 13:03
  • \$\begingroup\$ @Oldfart Those IC's typically won't handle up to 48V \$\endgroup\$ – Peter M Nov 8 '18 at 13:05
3
\$\begingroup\$

Since the switch is momentary contact, you don't have to have protection if only one switch can be pressed at any given time.

However, to guard against two or more switches closing, put a 5000 ohm, 1 watt resistor in series with each switch.

Also, even with the resistors included, you can tell if a switch hasn't released if the voltmeter shows voltage when none of the switches should be closed.

Unless the voltmeter is strange low input resistance, you won't even notice voltage drop across resistor.

\$\endgroup\$
  • \$\begingroup\$ Given the additional information that appeared in comments, this is the correct answer. However, 2200 ohms @ 0.5 W would be more than enough -- with the worst-case voltage difference between any two sources being less than 50V, two resistors in series (4400 ohms) would dissipate at most around 0.25W each. \$\endgroup\$ – Dave Tweed Nov 8 '18 at 21:35
  • \$\begingroup\$ @DaveTweed . . I did put in a lot of over design. Sounds like an outdoor vehicle. In case of build up on resistors. \$\endgroup\$ – Marla Nov 8 '18 at 21:50
  • \$\begingroup\$ Fair enough, but I think we can assume that the resistors are protected at least as well as the switches and the power supplies are. \$\endgroup\$ – Dave Tweed Nov 8 '18 at 21:52
  • \$\begingroup\$ @Marla, you hit the root of the issue: more than one switch being closed at a time with higher voltage feeding back to a lower voltage area (and blowing the attached components as tests are done when circuits and systems are live and power supplies are under load). Thanks! -- with no load being carried on these testing circuits, this is a great solution. Thanks also for your comment Dave, much appreciated. \$\endgroup\$ – Spere Nov 8 '18 at 21:59
0
\$\begingroup\$

It sounds like what you want is a super diode. This is an op amp circuit that provides an effectively ideal diode, by using negative feedback to adjust up its output voltage to compensate for the diode drop. It's just like an op amp voltage follower, but with a diode added:

schematic

simulate this circuit – Schematic created using CircuitLab

The downside to this is that you'll need to find an op amp that can take 48V input and also output 48V. This may not be cheap.

Another possibility is using a MOSFET-based super diode. There are devices that can control a FET to act as an ideal diode, such as the LTC4359, and while I know that fully-integrated FET-based super diodes exist (such as the SM74611), I'm not aware of any that can withstand the voltages you need for this particular use case.

\$\endgroup\$
0
\$\begingroup\$

What's the point of having the switches at the sources if the meter is some distance away on a panel?

Instead, you should have the switches at the panel, and wire them like this, so that only one source can be connected to the meter at a time:

schematic

simulate this circuit – Schematic created using CircuitLab

\$\endgroup\$
  • \$\begingroup\$ The question specifies that unfortunately they have to use SPST switches, so this wouldn't solve the problem. It would be a convenient way to solve it otherwise, though. \$\endgroup\$ – Hearth Nov 8 '18 at 13:34
  • \$\begingroup\$ @Felthry: Not really. It says that they're currently using SPST. I'm saying that the switches are in the wrong place to begin with, so while you're moving them, you might as well change them out for SPDT. \$\endgroup\$ – Dave Tweed Nov 8 '18 at 13:37
  • \$\begingroup\$ @DaveTweed, thanks for the suggestion. We are currently using momentary spst, they are at the power supplies and thus removed from the panel as we adjust the voltages at the power supplies, and can view the panel output from across the vehicle. That said, your idea is a sound one as it is generally a 2 person job anyway. \$\endgroup\$ – Spere Nov 8 '18 at 20:53

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.