Incubator temperature fluctuations

[frazzledfozzle 7,483]

So I need some advice on incubator temperature fluctuations .

is a temperature range of 32.5 to 33.5 to wide a variation?

 

initially our incubator was set at 32 with a .1 differential so basically keeping it at 32 with many cycles of on and off to keep it at that temperature. 

 

If we set it with a .5  it will require less turning off and on but the temperature will fluctuate .5 above and below the desired temperature which is 33 

Edited September 8, 2018 by Trevor Gillbanks
spelling

[Otto 780]

I assume this is for cells?

I think your cells would be absolutely fine with this fluctuation. They can handle a bit.

If you look at where queen cells are built in a hive they are often on the edges or along bottom bars (because this is where there is space for them). In these positions you cannot help but get some temperature fluctuations with environmental up and downs.

Edited September 6, 2018 by Otto

[frazzledfozzle 7,483]

Cheers thanks Otto  

[Alastair 8,627]

As part of the experimenting when I set up my own incubator I ran a temperature probe in a broodnest to find out for myself what the temperatures are.

 

The temperature did fluctuate by more than a degree, even mid broodnest. But the thing was, it would be over the course of several hours, not minutes such as would be typical in an incubator.

 

So end of day I didn't know if a faster fluctuation would be a good thing or not, so I set a very finely controlled PID controller to hold the temperature within 0.1 of a degree, and the queens turned out excellent.

 

However I do know there's people with set ups that allow the temperature to constantly fluctuate up and down by over a degree, and still get good results.

 

The only comment I would have on your post Frazz is that 32.5 is very lower end of required temperature. I experimented with this temperature and found queens took longer to develop (hatch time was later), plus they did not seem as vigorous when first hatched. I found the ideal temperature was 34.5. Development was again affected adversely at 36 degrees, if the set up is such that a fluctuation is going to happen, the maximum high should be 35.5, and the average temperature is best at 34.5.

 

And oh, the other thing I learned is how much out some temperature readers can be. In my view it is best to calibrate any equipment in an actual hive, and use that as your guide, if your gauge reads say, 37, in a hive brood nest, it means your gauge is out so set it to 37 in the incubator.

Edited September 7, 2018 by Alastair

[frazzledfozzle 7,483]

Cheers Alastair that’s good info

[Otto 780]

Thanks for the information Alastair.

I also thought Frazz's set temp was a little low. I have mine at 34 degrees. Haven't really tested how much it fluctuates but will sometime when I get my hands on a datalogger.

In terms of calibrating - I tested my incubator with two different independent thermometers to get an indication of how accurate it was. I thought it unlikely all three would be out by a consistent amount so trust that it is pretty good. 

[David Yanke 470]

4 hours ago, Otto said:

Thanks for the information Alastair.

I also thought Frazz's set temp was a little low. I have mine at 34 degrees. Haven't really tested how much it fluctuates but will sometime when I get my hands on a datalogger.

In terms of calibrating - I tested my incubator with two different independent thermometers to get an indication of how accurate it was. I thought it unlikely all three would be out by a consistent amount so trust that it is pretty good. 

I am in the '34 is best' club as well.

[Philbee 4,675]

6 hours ago, Alastair said:

The temperature did fluctuate by more than a degree, even mid broodnest. But the thing was, it would be over the course of several hours, not minutes such as would be typical in an incubator.

This is exactly my concern

"Frequency"

What are the effects of high frequency temp fluctuations compared to a Hive's low frequency fluctuations? 

I run at 34 deg C  non pid so see a 2 Deg C fluctuation.

Concerns me a little 

[Alastair 8,627]

As in 32 to 36? If so, at what stage do you put the cells in, and how do they turn out?

Edited September 7, 2018 by Alastair

[Philbee 4,675]

39 minutes ago, Alastair said:

As in 32 to 36? If so, at what stage do you put the cells in, and how do they turn out?

I set the controller at 34 high cut out and 34 low cut in
with the minimum 1 deg differential
So from memory it hits 34, cuts out and drifts past to something like 34.8 then falls back through 34 and starts again at 33.
It continues to fall past 33 for a few deg then heads back up through 34 for a repeat.
This cycle might take 15-20 mins sometimes longer depending on ambient temps
Cells ok but realistically who really knows.

 

[Otto 780]

11 hours ago, Philbee said:

Cells ok but realistically who really knows.

 

If you get a very high emergence percentage and your queens are good then you know

 

My incubator is an egg incubator and I don't really know how tightly the temperature is controlled. I'll be switching it on today as I have my first batch of cells coming out so will stick a min/max thermometer in there for a few days to see what sort of fluctuation there is.

[john berry 5,636]

I get perfect temperature and humidity control by just leaving the cells in the cell raiser. You don't even need a power supply.

[Alastair 8,627]

Very true John. 

 

But a well set up incubator is really handy, what I was able to do was put grafted cells into cell builders on a particular week day, then after whatever process was used (starter / finishers / whatever ), remove them to the incubator the same week day the next week, where they stayed till the day before hatch day.

 

What this meant was the queen raising cycle went to a 7 day cycle which was easier mentally. I also knew exactly how many cells there were and so how many mating nucs to make, ahead of time. Also, the day I removed the capped cells from the finisher I could put in a new batch, and it did seem the bees did a better job of them, without the presence of the older cells. 

[Philbee 4,675]

20 minutes ago, Otto said:

If you get a very high emergence percentage and your queens are good then you know

 

My incubator is an egg incubator and I don't really know how tightly the temperature is controlled. I'll be switching it on today as I have my first batch of cells coming out so will stick a min/max thermometer in there for a few days to see what sort of fluctuation there is.

My emergence is excellent but any  concerns I have go beyond emergence.
I like to look further down the track. 

Edited September 7, 2018 by Philbee

[Alastair 8,627]

I think an indication is mating percentage, well raised cells get a higher percentage. But there are so many other variables that it is hard to really know sometimes. 

[cBank 801]

19 hours ago, Otto said:

Thanks for the information Alastair.

I also thought Frazz's set temp was a little low. I have mine at 34 degrees. Haven't really tested how much it fluctuates but will sometime when I get my hands on a datalogger.

In terms of calibrating - I tested my incubator with two different independent thermometers to get an indication of how accurate it was. I thought it unlikely all three would be out by a consistent amount so trust that it is pretty good. 

 

If you are keen on messing about with computers and scripts, you’d be able to get something going quite cheaply and with excellent results using a Raspberry Pi. It can be made to record humidity, weight etc as well which may be useful somehow.  Assuming the hive is near a power source, all the parts would be sub $200.  It could be easily repurposed to control the incubator after using it for monitoring.

 

I got part the way there with brood box monitoring and then got distracted and it turned into a retro game console. Nintendo Mario Kart is just so good. I got a second one and it then turned into a network controller.

 

There are masses of projects out there that are easily installed and run, including an entrance bee counter using a camera. It counted, logged activity versus temp and all sorts.

 

Ill have to try a third.

[Alastair 8,627]

Do you understand PID control software CBank? If so, could a rasberry pi be programmed to do that, or are there programs available? I would not be able to write such a program myself.

 

The issue I've had with every PID controller I've had is they work excellent for a while, but then begin to fail due to cheapo hardware / circuit boards.

[cBank 801]

No, but I hack about with various bits of code on a regular basis

What is the bit of software/hardware your using currently and I’ll have a look? Raspberry Pi controllers are good as they have millions of pre made programs (just found a few for controlling home temps), and the language is reasonably easy to follow and test.

 

Pi hardware is very good, and you can buy a massive range of cases and accessories to get what you want. Unlike most computers, you can yank out the power and it doesn’t mind. Or just replace it’s entire brain by switching out the sd card thy contains its smarts.

 

Here is the sort of point I’d start from, but I assume the incubator you have heats and doesn’t cool? 

I’ll have a poke about this afternoon and see where I get.

Edited September 8, 2018 by cBank

[Alastair 8,627]

Thanks CBank. What I have at the moment is a collection of devices with various software installed, most of them are conceptually good but they have all become unreliable after a time and are on the shelf. Some of them have a cooling function also but this is not needed where I am. Right now I have nothing that works properly, luckily this season I am not raising any cells.

 

The device you show sounds good, possibly, but it does not say if it's a PID controller, so let me first explain what a PID controller is, as against a straight on / off controller.

 

A standard on / off controller waits till the temperature reaches a set level, then turns off power to the heating device. It then waits till temperature falls to some level, and then turns the power back on. The problem with this is there has to be temperature swings for it to work. This is also exaggerated, because when it turns on power to the heating element, it waits for the temperature to get to a certain point, then turns off. But by this time the heating element is hot, and the incubator continues getting hotter until the element has lost all it's heat so the swing is even greater.

 

A PID controller is different. PID controllers have software that when it is first turned on, it sees how fast the incubator warms up. Based on this it calculates when the incubator will reach the desired temperature, and turns the heating element off before the temperature gets there, allowing the element to cool. The PID controller then sends pulses to the heating element, bringing the incubator closer to the desired temperature, the pulses get shorter and less frequent as the temperature gets nearer, so once the temperature is reached there is no overshoot. A good PID controller will learn what is needed to maintain the temperature, and send pulses to the heating element to maintain temperature almost exactly. A temperature genuinely within 0.1 of a degree can often be maintained by a good PID controller, but with a normal controller such accuracy is impossible, even if the device claims to ba able to do it.

 

Some of the ones I have had, the software has been excellent and worked well, but eventually the hardware has fallen over. Seems the market is dominated by cheapo Chinese manufactured stuff, I have not been able to figure out what is actually a quality brand.

 

All I need is something to accurately regulate heating, and send power to a humidity generator if needed. I did not need cooling, or humidity reduction. When I had one that was working accurately, the quality of the queens was superb. I would like to be able to set one up again, but want reliable equipment, not something I'm worried about.

 

The device you linked MIGHT be able to do that, but it does not actually state that it is a PID controller, which makes me suspect it probably isn't. My guess is a straight on / off controller would be OK for brewing. On the plus side it does have a solid state relay, this is a must for a PID controller because of the frequent pulsing and on / off cycles.

[yesbut 6,230]

If you had the heater and thermostat inside an air cavity within a lump of some sort of thermal mass which then acted as the heating "element" I would think you could get away with a very unsophisticated controller.  Air is not a good conductor of heat and because it has little mass it heats and cools quickly. A lump of say concrete if heated from within by a cycling heat source would smooth the variation out. Could cost next to nothing. Who hasn't got a concrete block kicking around somewhere ?

Edited September 8, 2018 by yesbut

[Alastair 8,627]

True, could be done. There is an issue though. Which is, that if using a brick, tub of water, or whatever to buffer the heat, that buffer can only go up to the required temperature of the incubator, it cannot go over or the whole incubator will overheat if left closed for a long time. Which means that when the incubator has been opened it will take a long time to get fully back to the needed temperature, if it ever does.

 

A PID controller will bring the temperature up rapidly, just, slow down once the required temperature is nearly there. They are also not much more expensive than non PID controllers, it's just a case of finding a good one. I'm sure they must exist, just, with my limited knowledge I don't know what is quality and what is a junk version.

 

 

[yesbut 6,230]

 

35 minutes ago, Alastair said:

that buffer can only go up to the required temperature of the incubator, 

Yes

 

36 minutes ago, Alastair said:

it cannot go over

Why would it ?

 

37 minutes ago, Alastair said:

the whole incubator will overheat if left closed for a long time.

If the mass stays at 34 , the incubator will not not go over that.

 

38 minutes ago, Alastair said:

Which means that when the incubator has been opened it will take a long time to get fully back to the needed temperature, if it ever does.

No. As I've said, air has very little mass, takes on and loses heat quickly. A 34 deg brick in a lounge won't heat the room up but that same brick in a shoebox

will be very effective in maintaining box temperature.   

 

[Alastair 8,627]

Maybe you right. But I would still like a PID. 

[cBank 801]

4 hours ago, Alastair said:

A PID controller is different.

 

It certanly is, thanks Alastair.

I have worked with PID devices for a long time and had no idea they were doing that.

I cobbled together some bits and pieces this afternoon and got them working together - not the parts that are needed but broadly similar (temp sensor and a behaviour when too high and a different one when too low). However I was doing straight on/off. Not a PID algorithm. How a device does that is beyond me. You would have to have an element that didn’t change with time and a good understanding of the thermal behaviour of all the component parts. As always, it’s more complicated than I thought. 

 

As a side effect of this thread I’ve now got temperature monitoring running nicely, and the phone is notified of any power outages so that’s a nice side effect.

[Alastair 8,627]

35 minutes ago, cBank said:

I have worked with PID devices for a long time and had no idea they were doing that.

 

A lot of them have a little red led that goes on when power is going to the heating element, so you can see them in action, you can watch the temperature go up on the display, then see the red light start going on and off as the target temperature is approached.

 

Also, here is a youtube video that shows it really well as it uses a lightbulb for the heat source, the temperature probe is put on the lightbulb. The set temperature is set at 38.6. The light stays on, but as temperature approaches 38.6 the bulb starts to pulse, and goes of completely at 38.6. Then, to hold the temerature within 0.1 of a degree it pulses, having calculated what is needed even though the gauge may or may not drop or rise.

 

 

Edited September 8, 2018 by Alastair