Trying to design a super efficiant heater

[dimwatt]

Here is what people on ebay are passing off as a electronic incubator thermostat. **broken link removed**

Anybody care to guess what brand of hotwater heater these come on.. You can pick one up at Lowes for $9 and save the shipping.

and the one Cbiblis currently sales on ebay.
**broken link removed**

And just for comparison a very expensive high end incubator at $3995
**broken link removed**

I dont know, who's ripping off who. Which would you rather have and which are you willing to payfor.

 

[cbiblis]

Here is what people on ebay are passing off as a electronic incubator thermostat. **broken link removed**

Anybody care to guess what brand of hotwater heater these come on.. You can pick one up at Lowes for $9 and save the shipping.

and the one Cbiblis currently sales on ebay.
**broken link removed**

And just for comparison a very expensive high end incubator at $3995
**broken link removed**

I dont know, who's ripping off who. Which would you rather have and which are you willing to payfor.

Wow, you do your homework. What interest do you have here?
As for the thermostat on e-bay anyone see a similarity to the schematic i posted on this forum?

 

[cbiblis]

Cbiblis, that's a very impressive cabinet! You need a nice controller to go with it.

I recently created a nice open-source temperature controller, complete with LCD that displays the actual temperature and the setpoint temp. It was designed to use the very cheap PIC 12F675, and my little cheap Shift1 PCBs. And 2 buttons to turn the temp setpoint up or down.

Have a look at controller here;
SH1_Temp2 Temperature Controller

You don't need much more than a transistor and relay to operate your heater element, and the temperature sensor I chose was LM335 again because its a cheap and fairly accurate sensor. This controller is fully designed, tested and proven, it's perfect for low temp dryers and incubators and you can assemble them very quicky as there is only a few parts on the PCB.

If you are intertested in using this controller I can sell you a small quantity of the little Shift1 PCBs very cheap and you can source relays and LCDs at your end.

Thanks for the link it looks good. I really, really don't want a lcd screen, and my thermostat works great as it is. I am getting 1- .5 degree variants between on/off. Just could really use the help with the leds.
I will look through your page more and see if i can canalize the PCB board. I to use the lm335's with my controllers. I just don't know ditty about how to incorporate the leds into my design. I did however offer to buy a schematic to do this in my other post if you interested in helping me.

 

[cbiblis]

Some of the best insulation I have found is the simple double foil backed bubble wrap stuff. Around here the more common brand is Reflectix.

I use it on all of my hot water lines for my boiler and three layers of it, about an inch thick total, is enough to keep snow from melting on 200 F water lines all winter.

Its R value is about 7 per layer but the ability to reflect heat gives it a much higher equivalent R number.

For lower cost high heat retention I dont think there are too many other types that can compete!

I pick up a roll of this stuff today. I can spray adhere right to the interior of the box. I really can't believe that this thin stuff is going to give a r value twice as high as 1" foam. That's amazing and well worth the 22 dollars for a 2'x25' roll of the stuff.

 

[cbiblis]

Cbiblis, that's a very impressive cabinet! You need a nice controller to go with it.

I recently created a nice open-source temperature controller, complete with LCD that displays the actual temperature and the setpoint temp. It was designed to use the very cheap PIC 12F675, and my little cheap Shift1 PCBs. And 2 buttons to turn the temp setpoint up or down.

Have a look at controller here;
SH1_Temp2 Temperature Controller

You don't need much more than a transistor and relay to operate your heater element, and the temperature sensor I chose was LM335 again because its a cheap and fairly accurate sensor. This controller is fully designed, tested and proven, it's perfect for low temp dryers and incubators and you can assemble them very quicky as there is only a few parts on the PCB.

If you are intertested in using this controller I can sell you a small quantity of the little Shift1 PCBs very cheap and you can source relays and LCDs at your end.

Like I've mention before i don't know ditty about led or lcd. So my question might be stupid to thoughts that do but I'll ask anyhow. I like what i see with you shift-1. I like the temp2 project. I would want the timer to keep track of the incubation time. It would be really cool if it counted days and hours. I really like the set point to. The back light would help considering i don't like LCDs because you have to get close to read them. It's not that important but if there it room could the lcd also display humidity? And one other thing, can the program be easily changed to read in Fahrenheit instead of centigrade? My customers are all in the states. If you could help me make these adjustments then i'll use your boards in all of my incubators, hatchers, and brooders.

 

[Mr RB]

Interesting. The LCD is actually quite readble with the backlight on, even from about 10 feet.

The farenheight is not a big deal that's just a small calc. And neither is changing the timer to show days:hours:minutes. The clock should be accurate enough over your 24 day incubation cycle.

Humidity though might be a bit difficult. That will require a separate humidity sensor which means you need to access a "standard" humidity sensor that can be relied upon to work ok without calibration. Also converting the humidity sensor output to a linear humidiy in percent will need some calcs and the PIC code is already a bit full doing the temperature stuff. And there is no spare PIC pin for the humidity sensor, although one of the 3 buttons could be removed to free up a PIC pin.

Humidity is not impossible, but it would be MUCH easier to do without the humidiy sensor.

Either way if you are serious about this I can code up the changes you need and can cut you a good deal on bulk Shift1-LCD kits. I get the Shift1 PCBs made on the edge of a commercial PCB (to use up the surplus PCB panel space during manufacturing) so I end up with quite a lot of Shift1 PCBs per year. It's probably time to move to emails if you want to go further with this. I just PM'd my email addy to you.

 

[cbiblis]

Thank for your help Mr RB and i will e-mail you know.
As for the rest of you. Thanks you for your input i have given up on the efficiency issue for now and per your suggestions i have beefed up the r-value. I probably will revisit this issue later down the road but as for now. my mind is at ease with my design. I have got to get these out before i lose my investors and my interests.

 

[Artificer]

Food for thought; Good HVAC thermostats have a (mechanical) anticipatory circuit that follows the rise (or fall) toward set point. It would be analagous to a PI controller, without the Derivative function. A cost differential of several orders.

 

[cbiblis]

What do you call a deer with no eyes?

I have no eye-deer what you just said. You'll have to turn it down a notch for me to make since of it.

 

[dougy83]

You'll have to turn it down a notch for me to make since of it.

Are you from new zealand?

I think what he was talking about was control techniques that can be used. PI (proportional + integral) allows you to minimise the steady state error.

For your application I would think that it's probably not warranted & it's just extra effort.

 

[Artificer]

Sorry, I was responding to a comment that you needed PID control. A little over engineered for an incubator. The proverbial use of a howizer as a flyswatter.

P I D is "Proportional Integral Derivative". It is a type of proportional control, temp, level, speed, whatever, that is performed with an analogue computer. Essentially, a machine doing integral calculus. It is only recently that micro-processors have reached sufficient power they can simulate the functions with any degree of accuracy.

The concept revolves around a process, say temperature control. A set point is given. What temp do you want? The temp is measured for deviation from the set point. (error) direction, amplitude, rate of change, direction of change,... As the error decreases, approaching set point, the heat is reduced, not switched off, just reduced. Did that slow the rate of change? Is the change still in the correct direction? How close is temp to the set point? Do we need more reduction or is the change falling off too fast? Many, many variables that interact. PID controllers are very accurate, difficult to tune and a bit pricey. (read as God-awful, new car expensive for a hobby)

For an application of this nature, PI control would suffice. No Derivative, merely proportional and integral. Think automotive cruise control.....

A primitive form of this is used as a controller for HVAC systems. The thermostat contains a bi-metal strip(adjustable) to anticipate approach to set point. In the case of large heating plants, the heaters are cut off before actual desired temperature is reached. Anticipating reaching set point.... The response will vary from plant to plant, depending on thermal inertia of the heat source, location of the thermostat, ambient(outside) temp, and other factors. Through inertia, the actual temp rise will slow, hopefully ending just as set point is reached.

This is useful in your case because you want to maintain a fairly narrow temp band. PI would help eliminate overshoot.

Sorry I got so heavy handed. Sometimes I forget, we all here have different backgrounds. Mine includes nearly 40 years in industrial automation.