On 16 Mar 2018 11:30:59 +0000 (GMT), Theo
wrote:

T i m wrote:
On 15 Mar 2018 19:37:12 +0000 (GMT), Theo
wrote:
I'd spend £ 2.19:
https://www.ebay.co.uk/itm/12V-Batte.../232102823975?

That looks interesting. Any idea if there are any instructions that
cover the two pots Theo?

No idea, but I think I saw LM358 on a picture somewhere. That suggests it
is just a comparator circuit and you can work out what it does from the
layout.

Voltage thresholds and (therefore) any hysteresis maybe?

I'd ignore the rather complicated schematic about charging they
show,

I had. ;-)

and just understand that it has a power supply port, a comparator
input port, and a set of relay contacts.

So that layout is what I was suggesting re the LVD I mentioned, the
load and sense paths managed separately?

It wouldn't take very long to
experiment with what the comparator inputs and pots do.

It depends if they are raw or 'damped' like the one I was playing
with. It made it quite difficult to accurately bench-set the trigger
threshold (even ignoring the voltage variation because of different
loads on the measure and the battery etc). Nothing stopping one
pulling the cap they probably use for that whilst calibrating it
though?

Not to stop anyone doing a project of course, but buying one of these and
repurposing it is cheaper than buying the bits.

Quite! I would often rather tweak something that already exists (or
even assembling a kit) than to start from scratch.

The other thing I'm thinking of doing is powering the relay (or the
voltage sense input) from a separate supply, rather than the generic
input connections. That's because if you have a reasonable load even
though the heaviest cables you can fit into those connectors you still
get some voltage drop variance, if you are running different loads (as
I was when doing battery capacity discharge tests).

Alternatively, you could use the onboard relay to drive larger
external relay(s) and so 'offload' most of the load current
completely.

It also depends how you want to handle the load of the relay. On my digital
widget (VAC-1030A) the 30A relay takes just under 100mA, so energy
consumption is about 1W continuous.

Whilst not much when say discharging a battery at 5A, it might if
discharging a smaller battery or over a longer time.

You could do something with bistable
relays, but would need to ensure there's enough energy available for the
final disconnect.

If you wanted to ensure the final load (inc the HD relays, if powered
by the load and included in the calculations etc) was constant then
you could power two, having one in both the output states (you could
use two single pole c/o relays as a two pole switch or switch the +ve
of a battery between charge or discharge paths)?

Cheers, T i m