phil99 said  

As the coil current takes time to rise the 50% PWM period isn't needed.
The short duration of the initial pulse keeps the coil current within safe limits.

Have used 5V relays on 24V this way.

Surely the Kilovac has a pull in time? So we can interrupt the current flow a microsecond after it reaches 50% on its way to exceed coil max voltage of 30V - I think not.
My method is a working example with circuit and tested.
Edited 2026-05-01 12:10 by wiseguy

Pull time is between 15ms and 20ms depending on unit and manufacture.

The stated voltage limit is the continuous voltage that produces the maximum safe temperature rise.
The coil is just wire, and like motor windings can tolerate a much higher start current than it's continuous rating due to the thermal mass slowing the temperature rise.

The 5V relays I used were undamaged by a 1 second pulse of 24V which is far longer than the 10mS they needed to pull in. That is a big safety margin, despite the over-voltage ratio being almost 5:1.
The thermal mass of a contactor coil would be much greater than for those tiny relays so the safety margin would also be much greater, especially with an over-voltage ratio of only 2:1

Edit.
P = V^2 / R so for 2:1 Pmax is 4 times normal.
For 24:5 Pmax is 23 times normal, yet no damage as it never gets time to reach that.
Edited 2026-05-01 13:17 by phil99

  phil99 said  

The stated voltage limit is the continuous voltage that produces the maximum safe temperature rise.
The coil is just wire, and like motor windings can tolerate a much higher start current than it's continuous rating due to the thermal mass slowing the temperature rise.

The 5V relays I used were undamaged by a 1 second pulse of 24V which is far longer than the 10mS they needed to pull in. That is a big safety margin, despite the over-voltage ratio being almost 5:1.
The thermal mass of a contactor coil would be much greater than for those tiny relays so the safety margin would also be much greater, especially with an over-voltage ratio of only 2:1

Edit.
P = V^2 / R so for 2:1 Pmax is 4 times normal.
For 24:5 Pmax is 23 times normal, yet no damage as it never gets time to reach that.

Do you actually have a solution or schematic etc ? Maybe purchase a unit at ~$100 and play with it using your ideas. I was not after an argument I simply presented a working system that provides the right current during pull in time period and the right current for hold time and did not exceed the manufacturers stated maximums. What exactly did I get wrong that needed the exta posts?
Edited 2026-05-01 13:35 by wiseguy

Something that needs to be considered is this: Non technical builders are already feeling a little, or even very overwhelmed. The last thing they need is the thought of somehow trying to makes something themselves, without a PCB, and to work with something they likely don't fully understand in a high power Inverter, introducing another unknown and doubt into the build is not wise IMHO.

The schematic was simply an Arduino Nano, resistor, transistor, diode and relay.

The experimental solution was start with a 1S pulse and keep reducing until the relay didn't pull in reliably, then go back a step.

A more conservative approach would be to start with a 1mS pulse and increase until it pulls in reliably at the minimum expected battery voltage.

The point is, if you want to make one and test it with a 24V solenoid, most come with an economiser, and design a PCB that someone who has learnt to solder can build, and give detailed instructions on how to connect it into the inverter, supply the code to program it, then someone new to this might be able to do it if needed.
.
Edited 2026-05-01 14:08 by KeepIS

Ok my inverter isn't one of Mikes designs and for the pre start that resistor bank and 450amp switch is a true KISS concept I used for my inverter.

For startup just turn on a circuit breaker thats connected to VBat and watch on the LCD screen the voltage rise, when the voltage is stable turn on the 450 amp switch and turn off the circuit breaker then hit the start switch on the inverter.

With the wiseguy system you dont have to be there to toggle switches. If for instance the inverter shuts down due to low battery (not enough solar etc), the inverter will resume in the correct fashion with auto precharge etc automatically when solar/generator charges the batteries again. You can be away for a week and this can all happen in the background without user intervention.

I have no problem with pre-charge switches etc but an on off switch is also easy to implement and you can't accidentally have a splat event.
Edited 2026-05-01 14:35 by wiseguy

@Bryan1, the major reason I / we use a Kilovac, is not just for auto fail safe pre-charge, it is for fault and over current protection, no burnt boards, very limited damage, and more technical reasons, like over-voltage protection for the Inverter, DC input over current protection, most of this protection under full supervision of the controller and Micro-controller, this is by my definition, a real Inverter.

Looking at the circuit for the Nano Control Board all the hardware to drive the solenoid with either a Pulse/PWM or PWM/PWM is already there, no changes needed.

If that Nano pin can be a PWM output it could be done in the firmware.

Footnote added 2026-05-01 18:33 by phil99
A rhetorical question.
Why does trying to save $40 or so for new constructors produce such an angry response ?

I can understand electrics but people will forever be a mystery.

Not going to happen EOS!

  phil99 said  

Looking at the circuit for the Nano Control Board all the hardware to drive the solenoid with either a Pulse/PWM or PWM/PWM is already there, no changes needed.

If that Nano pin can be a PWM output it could be done in the firmware.

That is true with some caveats. When the Nano was first proposed to be used we only wanted it to run the inverter and not take its eye off the ball lest there be grief.
Since then Poida & KeepIS have included bells and whistles galore. It is a credit to their programming skills that grief has not visited the Nano inverter so far.

Depending on the supply voltage if a Kilovac is to hand or used at all with or without economiser. The ensuing software becomes more complex with a need to alter as required.

I chose to not go down that path despite some on the forum who are able to do this I chose the Kiss principle of a dedicated drive to the correct voltage Kilovac with or without economiser.

The system is there and I agree if someone wants to write code for pull in and hold be my guest but obviously all at your accepted risk.

Oops...
seems I've been running a 12 - 24V solenoid (with economiser) straight off 48V these last few months!
I wondered why it ran so warm... Guess I was lucky.

Just solved the problem with a DD7818TA from Aliexpress, the 12V out version. Around $12AU delivered.

https://www.aliexpress.com/w/wholesale-DD7818TA.html

Anyway, the inverter idle power has dropped by 3W...
Edited 2026-05-01 17:38 by FET cemetery

Hi Jakesea,

Good to hear of  another of these great inverters being built in Wales (i'm in Carmarthenshire). Mine is not yet finished: still running on the temporary transformer, but i'll finish it one day.

I have a spare TLS76L1 that Mike generously provided with my boards. I could pop it in the post to you if you like?

  Quote  

Oops...
seems I've been running a 12 - 24V solenoid (with economiser) straight off 48V these last few months!
I wondered why it ran so warm... Guess I was lucky.

I'm surprised that didn't burn out, they run very hot like that. I did the same thing with my 12v units adding a small DC-DC converter and problem solved.

Thanks for the info and update  
Edited 2026-05-01 18:27 by KeepIS

  mab1 said  

Hi Jakesea,

Good to hear of  another of these great inverters being built in Wales (i'm in Carmarthenshire). Mine is not yet finished: still running on the temporary transformer, but i'll finish it one day.

I have a spare TLS76L1 that Mike generously provided with my boards. I could pop it in the post to you if you like?

Mab I expect to be visiting UK again, hopefully within 6 months so I will put half a dozen in my pocket to post to you when I arrive, for you to keep/distribute at your discretion. If anyone I know visits sooner I will ask them to post to you locally.

  KeepIS said  

Something that needs to be considered is this: Non technical builders are already feeling a little, or even very overwhelmed. The last thing they need is the thought of somehow trying to makes something themselves, without a PCB, and to work with something they likely don't fully understand in a high power Inverter, introducing another unknown and doubt into the build is not wise IMHO.

I can only agree.

Personally, I understand my limits: my knowledge is, compared to yours, basic or... shallow? If you wrote to put 10cm long piece of wire somewhere I wouldn't dare to make it 9 or 11cm. I am simply in no position to "try/improve/experiment/add/tinker" with anything (in the inverter). I trust your knowledge and design, I don't trust myself and I am trying to stick rigorously to the design as troubleshooting and fault finding will be more complicated for me than it is for you guys. Especially, if I introduced anything more than there is in the schematic.

On the other note, I wouldn't like you to think that I am not doing anything to gain some knowledge and understanding. I have read all your responses a few times and I appreciate them all. I also educate myself, reserch, read...


@Phill99, I think, I undestand the concept but it still looks to complicated for me to make and implement at this stage.


@Wiseguy, I studied the post and the schematic you mentioned  and it would create more questions from me to you ie. what diode, what transistor, what capacitor... I don't think you really want that 😂.

Once again, thank you for your design and pcb boards and if you come to the UK I wouldn't mind buying you a drink or two... (I hope you don't drink drinks from high glass with straw, parasol and pineapple pieces in them...)


@FET cemetery, the idea of using 24v kilovac and dc/dc converter looks simple enough even for me, thanks.


@Revlac, I still should have it and I will try it but I was not sure if I would be able to calibrate it correctly as I have not seen arduino software and what is or is not possible to set/calibrate there.


@Mab1, great to hear from you. I am in Tredegar. I will try the one I have from Ozinverter built as adviced by Revlac first. If it doesn't work I might beg you for one pretty soon. "Diolch" anyway.


@KeepIS, the link you provided displays 404 page and not available in my country. But I could see similar items recommended underneath so I assume that you meant:

IHV200HDANA

I spent some time looking at them and I have noticed that many have branding removed or covered (photoshoped) on the pictures. To me it means: fakes. Do you think they will still work as intended?

Thanks a lot  
Edited 2026-05-02 07:35 by jakesea

  jakesea said  

Once again, thank you for your design and pcb boards and if you come to the UK I wouldn't mind buying you a drink or two... (I hope you don't drink drinks from high glass with straw, parasol and pineapple pieces in them...)

Aw crap - is there another type of drink ? wait - yes I remember room temperature beer lol.  Ok just a bit of ribbing the local beers are rather palatable.  I will look up where Tredegar is, currently I have no idea how close we would be during our visit.

  jakesea said  

@KeepIS, the link you provided displays 404 page and not available in my country. But I could see similar items recommended underneath so I assume that you meant:

IHV200HDANA

I spent some time looking at them and I have noticed that many have branding removed or covered (photoshoped) on the pictures. To me it means: fakes. Do you think they will still work as intended?
Thanks a lot  

That linked unit appears to be fine.

EDIT:

This looks better
 
Edited 2026-05-02 09:16 by KeepIS