EWP Controller

[richy_rich]

49 minutes ago, Nick Jones said:

Andy....... Vowell?

The very same.

[Nick Jones]

Haven’t heard from him in a long while......

[richy_rich]

He's still around on the facebook site.  Don't think he's raced much since ~2011ish? I should maybe ping him and see if he's got any spares (or spare cars) going cheap

[Nick Jones]

He lives (or used to) fairly close to me, though I think the only time I ever met him in person was at Castle Coombe. 

[richy_rich]

Last time I saw him was Spa 2010 I think https://richyrichracing.com/2010/07/spa-summer-classic-2010/

Bringing things back on track, I did some experimenting with voltage dividers to slow the pump down.  Works, but you end up shunting a lot of power through the resistors which means they get hot.  I did some sums on a bit of paper and I think you're looking at circa 10-20W through the resistors, regardless of configurations. I think a more reasonable way to do this is using PWM as the expensive controllers do, or if you're on a budget a flasher relay   I will keep playing.

[Nick Jones]

PWM controller makes sense - except you then need to find a way to tell it what voltage to output for a given operating temp. Most PWM controllers (cheap ones anyway) need a continuous analogue input. Maybe you can condition a voltage using a temp sensor?

Suspect the life of a flasher relay might be a bit short when continuously switching a fairly hefty inductive load.......

[oldtuckunder]

A quick search does indicate there are cheap Heavy Duty Flasher Relays out there (within Rich's budget parameters :-) that will handle way above 20W.

[DeTRacted]

2 hours ago, Nick Jones said:

PWM controller makes sense - except you then need to find a way to tell it what voltage to output for a given operating temp. Most PWM controllers (cheap ones anyway) need a continuous analogue input. Maybe you can condition a voltage using a temp sensor?

Suspect the life of a flasher relay might be a bit short when continuously switching a fairly hefty inductive load.......

Proportional control is fairly easy to do with a PRT  or thermocouple but for simplicity you could arrange a PWM unit to switch between two fixed rates using an on/off sensor or  even use a fixed PWM for slow speed and short it out when the temp switch cuts in for full voltage. A module like this is cheap enough to play around with:

https://www.amazon.co.uk/6V-90V-Pulse-Width-Controller-Switch/dp/B00CWSO2IY/ref=pd_sim_23_1?_encoding=UTF8&pd_rd_i=B00CWSO2IY&pd_rd_r=b24f3997-0cf8-11e9-8210-ed08ba772cb1&pd_rd_w=Sit4l&pd_rd_wg=gzvGz&pf_rd_p=1e3b4162-429b-4ea8-80b8-75d978d3d89e&pf_rd_r=984P5MJTXR5KHJ2VHCRT&psc=1&refRID=984P5MJTXR5KHJ2VHCRT

You are right that its not a good idea to pulse the motor current on and off at a low rate. Unless you buffer the output somehow (maybe with a large super-capacitor ) the current surges will soon eat the contacts and the motor won't like it much either.

[richy_rich]

Yep, I'm thinking the same.  Basically low rate (~4-5V) fixed and fully on above a certain threshold.  As you say PWM controllers are cheap and plentiful.

[rogerguzzi]

1 hour ago, DeTRacted said:

Proportional control is fairly easy to do with a PRT  or thermocouple but for simplicity you could arrange a PWM unit to switch between two fixed rates using an on/off sensor or  even use a fixed PWM for slow speed and short it out when the temp switch cuts in for full voltage. A module like this is cheap enough to play around with:

https://www.amazon.co.uk/6V-90V-Pulse-Width-Controller-Switch/dp/B00CWSO2IY/ref=pd_sim_23_1?_encoding=UTF8&pd_rd_i=B00CWSO2IY&pd_rd_r=b24f3997-0cf8-11e9-8210-ed08ba772cb1&pd_rd_w=Sit4l&pd_rd_wg=gzvGz&pf_rd_p=1e3b4162-429b-4ea8-80b8-75d978d3d89e&pf_rd_r=984P5MJTXR5KHJ2VHCRT&psc=1&refRID=984P5MJTXR5KHJ2VHCRT

You are right that its not a good idea to pulse the motor current on and off at a low rate. Unless you buffer the output somehow (maybe with a large super-capacitor ) the current surges will soon eat the contacts and the motor won't like it much either.

Hello All

                  I think that is what I said in a roundabout way

I think I may have a couple of those units as I bought a few direct from China (5 or 6 for the price of one)

Being off grid we use the 24volt battery supply for a lot of things to save running the inverter or generator.

I you want one I will dig one out but I think they are only rated at 2amp with a heat sink

Roger

[spitfire6]

Here is a picture of the large plastic pump controller graph contradicting my erroneous memory.

Hi,

 If it was done without semiconductors; A suitably preset thermostat sat low in the coolant to short-out a suitable power resistor that is used to drop 8 volts would give the two speeds.

In the end, I ran the 2000 Hrs rated pump at system voltage from an alternator switched relay powered by the ignition.

I said the minimum voltage was 7 volts & that not to rely on convection alone at ambient and below temperatures. The EWP data sheet says differently. Sorry.

A restriction out off head is needed to prevent boiling, more so with the EWP flow being less at high RPM. I used a coin with a few holes to obtain this in place of OEM stat and fitted a three port modern thermostat. It looked kak with all the hoses. One of the many problems was that the many many hose connections were all under stress. 

 Crap install from my end was the one problem despite the money thrown at it. My expansion tank is too small to allow the 5% expansion of the coolant being the main problem.

I was going to control the EWP from my ECU via a power FET with PWM. I purchased loads of interface chips & IRF's to simply achieve this but lost interest and went back to stock. Cooling fine now if a little bit too cool @ motorway speeds. I suspect I have drilled too many holes in the OEM stat, or that the stat is not really 88c as marked.

The small buck-boost single IC PCB's would work for a lower preheat speed as the current is a fraction of the 7? Amps required a full RPM.

Cheers,

Iain

 

 

 

 

[richy_rich]

Thank Iain, interesting stuff.  So they basically have no flow at lower temperatures, which I guess would explain why heaters etc didn't work if you use those controllers.

This is my setup, no thermostat at all currently (just a sender for the temperature gauge).  It's about as simple as you can get and works well - flat out, on a hot day and a bit too well at other times.

 

[Sprint95m]

Your heater doesn’t work because there is no return. Said return needs to be plumbed into the bottom hose before the pump.

The Davis Craig pumps have a drain hole which obviously needs to be at the lowest. From your photo I cannot determine if that is so. It is disconcerting to have a pool of water under the car when up park up but apparently this is the way to do things?

Whilst your setup may work it is not the best option. You really do need to fit a (remote) thermostat to get the engine up to temperature efficiently.  From experience I know the graph in the previous post  is just mumbo jumbo, having no coolant movement is undesirable.

 

 Thanks,

Ian

[richy_rich]

Who's heater? Mine?  I don't have one..  This is for a race car, there's no thermostat either.. The idea of the controller is to better simulate there being one, mainly to make warm-up easier/quicker.

Like you, I don't agree that having no coolant flow during warm-up is ideal, but with only an on/off switch currently that is my only option.

I think on the basis of reading everyone's posts and my own experiments the simplest/cheapest solution is a 2 speed system (3-4V slow, 12-14V full) controlled by a thermostat that kicks in somewhere around 85-90 controlled by a three way switch (off/auto/on) on the dash.

 

[spitfire6]

Hi Richy_Rich,

 I recollect that with the EWP a thermostat was not required. I know a restriction was required to the flow out of the head and that a thermostat with its gut removed didn't restrict the flow enough at high RPM. load.

Warmup was much improved with a stat + controller +EWP. 

Warmup + heat capacity was maximum with the pump at 14.7 Volts from alternator AND a three port thermostat. Removing the need for the water pump/stat housing makes the 3 port housing relatively simple. Win,Win. The OEM stat will just try to control temperature exiting the head. The three port will regulate according to rad DeltaT; If the dynamic cooling ability of the rad changes, the stat will modulate quicker.

Cheers,

Iain.

 

Edited January 3, 2019 by spitfire6
Added text.

[Sprint95m]

18 hours ago, spitfire6 said:

I know a restriction was required to the flow out of the head and that a thermostat with its gut removed didn't restrict the flow enough at high RPM. load.

Sorry Iain, I have seen this written but I very sceptical. Drill two tiny holes in the thermostat, indeed.
This is a DC method for cars without a header tank....

...during the warmup the EWP just sits and spins, it cannot pump any coolant because there is none available at its input side due to the blocked bypass.

If this is tried on a header tank equipped car it will stall the car because of the load the EWP places on the electrics,
the pump is able to draw in coolant (from the header tank) but can't expel said coolant because the outlet in the head is blocked off (by the thermostat) and
the radiator bypass is blocked off too.
Mind the EWP is trying to pump 80 litres/minute through two small diameter holes.
Putting a clamp on the header tank outlet allows the pump to just sit and spin. Little wonder DC EWPs are so short lived.

 

 

I don't understand why all this faffing around is necessary?
Looking at V8 racing cars in the USA, they just go for a pump and thermostat
which works.

 

 

Ian.

[spitfire6]

Hi,

 I was impressed with the 80L/s flow rate but no where near that fitted to an engine. lol.

The EWP shouldn't be stalling the engine under full load as its only around 80 Watts?

I think my present stat has too many holes, takes forever to warm up.

Cheers,

Iain

PS. On my car, there is/was always coolant flow with a closed OEM stat as I have a fitted heater for warmth & temperature measurement. I have not had a fitted heater control valve for over 20 years. No hot air and I'm low on coolant and overheating; gauge would probably say OK as no coolant. LOL. I do have a gauge to fit and will do.

Edited January 4, 2019 by spitfire6
Added text.