Re-reading the original post, I realize that I don't think most respondents understand the actual problem.
I'm going to restate; @Technological, it would be great if you can confirm if this is correct.
My understanding is that you have some sort of remote input. This input generates a short 'startup' voltage pulse, but does not represent a reliable power supply. You also have an input 'X' that needs to be fed +5V for at least 2s when the 'startup' pulse is seen. The only 'onboard' power supply you have is the battery bank itself. You can feed 'X' its +5V indefinitely, but doing so from the battery bank is strongly undesirable.
If that's the case, I think you have only three choices:
- Have the 'startup' pulse deliver the input signal you need. That is, the pulse needs to last the full 2s.
- Convert the battery 80V to a more usable voltage.
- Add a secondary power supply that can power 'X'.
I'm guessing (1) is a problematic solution. You may be stuck with (2).
That said, I came up with
this possibility for (2) or (3). It needs quite a few parts, but has the advantage that the power supply is disconnected (physically, if you use EMRs rather than SSRs). The 'startup' pulse does need to be long enough for
both relays to engage. This might make (3) feasible, but at least it achieves the objective of disconnecting 'X' after the required interval.
This works as follows... The top voltage represents an always-available supply, e.g. the batteries stepped down to 5V. The bottom voltage plus momentary switch is the 'startup' pulse. Note that this needs to be isolated from everything else. The bottom relay momentarily energizes the circuit. The 555 is set up as a monostable vibrator; as soon as power is supplied, it turns on, and stays on for 2s (adjust the 200k resistor to change the timing). The output is connected to a
separate relay (top) that connects the supply to the circuit. This acts like a start-latch; powering it either via the 555 output or the lower relay turns it on, and since it is back-fed from the 555, it stays on until the 555 turns off, allowing power to the lower relay to be interrupted.
The caps drain through the 555 and the whole thing resets relatively quickly (<0.2s). Because it just connects power that is already connected, repeated 'startup' signals during the 2s do nothing.
To be clear:
something has to power the timing circuit; I don't know of any
mechanical way to accomplish what you're asking. It seems, however, that your problem with the manufacturer's suggestion "connect a Li-ion battery briefly to this input" is that you need an unattended way to
disconnect said battery again. IOW, you need a magic switch that can connect the supplemental power supply for 2s when it receives an input signal, then disconnect it again... and doesn't draw power when it's 'off'. This is exactly option (3) with the circuit shown. The 555 is powered by the same supplemental supply that powers 'X'. It draws no power when off (everything is physically disconnected by EMRs), and uses an electrically isolated signal to turn 'on' that can be a fairly short pulse. While it does draw a modest amount of power while 'on' that is in addition to your load, it seems like that may be acceptable for your use case?
...and if I'm totally misunderstanding and 'startup' is just a constant voltage until the whole thing turns off, then EPAIII's answer is probably the way to go. You can implement that with a 555 by ripping out the relays and the output-to-relay connection and just feeding voltage to the resistors/Vin/rst.