Prototype liquid-based heater

[ted.lowe]

Here's a scan of the liquid-based heater diagram from Bob Batson of EVA:

[ted.lowe]

Here is the feedback email i sent to Bob Batson at EVA:

Hi Bob,

i wanted to share some observations we've made so far on the liquid-based heater prototype you provided.  Feel free to ask for more detail, share your thoughts and/or push back as you wish...

1.  The thermostat switch is the wrong polarity.  It is open on cold, closed on too hot.  This will not provide a good fail-safe mechanism if something goes wrong with the relay, control wiring, etc.  (the heater could continue to heat forever in this scenario).  The relay should turn the heater on and then off when too hot (or a failure occurs).

2.  The screw-in inlets and outlets don't seem to be adequately attached to the aluminum tank.  The thickness of aluminum used only provides perhaps 1.5 threads.  This seems like a leak or worse waiting to happen.  We're planning on tig welding the inlets/outlets directly to the tank.

3.  What is the extra switch entitled "Heater Switch Added" for ?  A heater-system-disable ?  And how does it interact with the "12V Heater Switch Existing" controlling a relay (not provided) ?

4. The pump seems too large to me (although others didn't think so).  The pump only needs to pump a small amount of anti-freeze a short distance with almost no vertical head.  It seems like a smaller pump would provide the needs of this system for less money and with less weight.

5. We're probably going to wire up the two 2000 watt heaters separately with a high-low switch to give more versatile heating options.  Our resident pump expert (Wayne McGuire) determined the pump will handle no (or restricted flow), so we will also be using the built-in temperature adjustment (which controls the amount of fluid running through the system).

Thanks!

Kind regards,
ted

[ted.lowe]

i talked with Zak Tyler just now over the phone.  Bob had forwarded him my email and he had read it.

1.  He had thought about the "polarity" issue of the thermostat.  He had a very hard time finding the appropriate part (high quality, high temperature, open on rise).  He may of found an alternative for future units.

2.  Again, he couldn't find a barbed aluminum fitting that could be welded in, so he used the brass NPT's for the side and top.  He used NPT tapered compression type fittings which are not that sensitive to the amount of threads.  He hasn't seen any leaking on the inlets/outlets so far.  He will be using a thicker material for the bottom fittings in future units. 

3.  He wasn't aware of this "extra" switch.  He made the original drawing and Bob made the final one.  Perhaps Bob added something that Zake hasn't seen yet.  He will discuss with Bob.

4. He agrees that the pump is on the large size, but again, he couldn't find a pump that was high quality and lighter.  He decided to go with this pump, even though it was twice the price of a plastic bodied pump because of quality concerns.  He did note that he may try the plastic pump after all because the steel pump is pretty loud.

5. He considered a high-low setting but wanted to keep the complexity (and cost) down.  i told him we probably wouldn't actually install a switch, but perhaps try running the system with a single heater as an experiment to see what the system looks like in the steady state (at various outside temperatures).  Personally, i'd prefer a lower draw system that runs continually than a higher draw system that turns on and off frequently.

He noted that MSDA sells a liquid-based heater system that is $800 and takes 1/2 hour to heat up.  He was trying to produce a similar product at lower cost.

[terry.kane]

I was browsing the Austin EV project cars and ran across this profile.

http://www.evalbum.com/17

Worthy of note is how he used a DC/DC converter to "quiet" his 12V water pump by reducing the voltage to just 5V.  Obviously he felt he didn't need the full 12V volume of flow.  Perhaps this method could be employed to quiet Eric's heater water pump?  Anyone know what circuit it would take to drop the 12V down to 5-6V and at what cost?

[robert.funk]

The motor slow-down is a good application of a "PWM" switching controller.  Since you don't need the smooth voltage output of a complete DC-DC converter circuit, you can drive several amps into a DC motor, with an adjustable duty cycle (12V when "on", 0V when "off"), and the motor reacts as if it's a lower voltage.  This kind of circuit is very energy-efficient (90% is attainable) and the components are readily available.  Shouldn't cost more than $15 if you put it together yourself.  A simple knob for adjustment would act like a dimmer-switch speed control, but for DC input you ***can't*** use a regular household dimmer switch, of course.

Let me know if you'd like more detail.  The full-load current of the motor would be a necessary starting point for a reliable and cost-effective design.

Bob Funk

Addendum;

Since we are using resistive electric heat, the efficiency of the speed control is not an issue -- as long as the "waste" heat makes it to the recirculating fluid, it's as useful as if it were energizing the heating element(s).  So, there's no need for anything electronic at all...

Just put a flexible heater element (or foil heater, or heating tape) of the needed resistance in series (electically) with the motor, and attach it thermally to the tubing or the tank.  The motor runs slower, and you put a bit more heat into the liquid.  Once you know the needed voltage for the reduced speed, and the motor current at that speed, just grab a catalog for foil heaters (like they put in some coffee makers) and you have a "100% efficient" motor speed control.

Summary:
For adjustable pump speed, use a simple PWM control.  For a fixed pump speed (with fairly constant 12V supply), just add a heater element in series with the motor.  Either approach requires some knowledge of the motor current for a complete design solution.

Thanks for listening,

Bob

[terry.kane]

Genius, Bob!  Hopefully, Eric will consider one of these options if the loudness of the circulating pump remains an issue on the long-run.  Thanks!