Reply To: RS485/Modbus RTU Transducers

Well my first post on this is to talk about the powering – not as sexy as the protocol, but a base requirement.
Since this is wired, each transducer is going to require power.
Getting powering r

Well my first post on this is to talk about the powering – not as sexy as the protocol, but a base requirement.
Since this is wired, each transducer is going to require power.
Getting powering right is ALWAYS a challenge for a circuit, especially solar powered, low power useage circuits. Its vital to understand the impedance of the power issues.
Wall power supplies are constant impedance and relatively low impedance sources – they can supply say 1 to 2Amps no problem.
Solar panels are variable impedance and relatively high impedance power sources – when its dark there is no power, when the sun is shining it has some power. Batteries and capacitors manage the impedance profile change between a solar panel and the dynamic power used.
In an experiment with a low cost Utopia ULB-I depth gauge – which requires 8-36VDC I recently hit some of the power issues.
It is an example of transducer, and the specification is missing one critical parameter, power used, and inrush current when power is applied.
For a standard wall connected power supply this isn’t a problem.
However I want to take samples every 15minutes and I want it to be off most of the time – so I need an electronic switch in there.
I started by experimenting with a simple boost circuit, and its tripped me up.
The core experiment was taking a “SUNKEE LM2577 DC-DC” I found on amazon – a 5V to 12V boost max@2.5A board (inputrange 4-34,output4-35V), and tying the input to USB +5V switch circuit on the Olimex-H407 board, and the output to the transducer power supply. The Olimex-H407 5V switcher device MP1482DS seemed to be pretty hefty at somewhere around 2A.
I soldered wires to the SUNKEE LM2577 to the H407 USB +5V, and soldered wires to an RJ11 male, and RJ11 female lead to the transducers power/data.
I put a volt meter on the output of the SUNKEE LM2577 adjusted the output to +12V, and left the meter monitoring its voltaage
The power supply supplying H407 was slightly above the minimium of +7V, initially taking about 40mA

Then plugged the transducers RJ11s together.

It failed to work!.
As I poked around with a scope I was seeing some of the issues with power supply impedance.
On plugging in the +12V dropped to +7V, and then USB +5V dropped to +3V.

I repeated, and monitored the wall supply which kept constant voltage but jumped from 40mA to about 160mA, however it fleeting showed +250mA.
Looking at the switching waveforms on the MP1482DS it wasn’t coping for some reason, so the output was dropping to +3V, which was probably causing the output SUNKEE to droop to +7V – but taking a lot of current in the process.It seems like the MP1482/inductor should have been able to supply at least 1A, but it wasn’t.
I modified the powering, connecting the SUNKEE directly to +5V wall supply, with a current meter.
Hey presto it worked. Watching the +5V wall powersupply, it showed +650mA pulse on plugging in.
While I don’t believe the powersupply’s has a reliable max pulse indication, I did guess that the +5V circuit source is a lower impedance- and so as this was a quick and dirty experiment. It took me less time to do it, than write it up here.
So back to the drawing board.
Generally, in the past for this type of boost powering, what I do is turn on the boost, allow it to power up and fill its output capacitor sufficiently, and then have an electronic switch that switches the steady boosted power directly into the transducer.
If I have an inrush current problem – generally a seat of the pants estimation – I can boost the output capacitor until I solve the problem.
So in this case, I’ve gone back to amazon to order “SunFounder 2 Channel DC 5V Relay Module” – something that I hope I can make work with 3.3V H407 outputs, and switch the +12Vs.

For circuit designs I can use lower cost boosters and +12V MOSFET switcher.

Less...