A rare dodgy piece of German engineering in this propagator heat mat

Germany has got itself a reputation for top quality engineering, so I was somewhat surprised to come across a pretty poor example of German engineering in terms of the Bio-Green Sahara heating mat. The mat itself is fine, and continues to do sterling service. It is foil with heating wire run through it, the aim is to stick this under seeds being propagated.

It addresses the fundamental problem that the UK is too far north for most of our vegetables, which were used to the more balmy climates of lower latitudes. So the seeds don’t really get the feeling that it is time to grow until too late in the year, when the temperature reaches what they expected in the Spring, but we get in Summer.

The heat mat comes with a thermostat with a remote probe

which DGF reasonably assumed to be connected with a wire. I had noticed that this damn thing was adding a pretty outrageous extra load to our daily power usage – it is a 65W heater, so if it were on continuously, then it would draw the same amount as the fridge, namely 24×65/1000=1.5kWh.

It was time to break this out again this year, and I took a look at it. The probe reminded me of the sort of thing used on a gas valve to stop the main gas valve opening until the pilot light is lit. These are often thermopiles driving solenoids these days but in the past they were a copper tube with a volatile fluid in it which vaporised on heating to increase the pressure. A capillary tube takes this to the gas valve and opens the valve under pressure.

True enough, this appeared to be the case here, and the device was associated with a disturbing smell of chloroform, which is presumably the active ingredient.

Now in a gas cooker you don’t expect to move the sensor, so having a rigid copper capillary tube is okay. But a heating mat that is described by some retailers as

These all new aluminium encapsulated versatile heatmats easily roll up when not in use.

should not be supplied with a device that is designed to be deployed to a fixed installation. Flex that sensor tube too often and the bugger will crack, releasing the chemical into the atmosphere so that the thermostat will never turn off. The manual really ought to tell you that this is extremely delicate and should not be flexed repeatedly.

The next thing that is painfully wrong is that they tell you to whack the sensor into the soil. Stands to reason, right, that’s what you are trying to control? Not so fast – there is a problem in that the delay between the heat getting to the sensor means there is a large overshoot, as the sensor tells the heater “turn it up, turn it up, turn it UP WHOA THERE turn it DOWN you’ve gone far too much turn it DOWN”.

For the mathematically inclined this is a control system and the lag mucks about with the poles on your Bode plot. I think that’s what I recall from uni. As an engineer it was a lot easier, the mantra was always get your sensor right next to the heater. Which is counter-intuitive because you are measuring not at the point of delivery, but it gets the delay down. You will get a static error due to the thermal resistance from the heater to the soil, but that’s better than ending up with large temperature swings. Industrial control systems use proportional control and may add rate-of-change and integrating loops to go for greater accuracy but these are seeds, they just want to feel they’ve been shifted southwards about 20 degrees of latitude.

The instructions should also have included the practical stuff to make an efficient plant mat, as well as how to avoid knackering the device. The heat only needs to go upto the seeds, rather than downwards, so the heat mat should be placed on an insulating substrate, otherwise energy will be used worthlessly in heating the potting bench. Celotex is probably ideal, but I used a couple of sheets of expanded polystyrene foam covered with aluminium foil. You obviously want to consider what happens under fault conditions with the heater on permanently and dimension accordingly The heat mat then sandwiches the heating cable in two sheets of thick aluminium foil, spreading the heat better.

For those looking to do this on a budget, this can be made using standard heating cable such as used for keeping pipes frost-free, with thick aluminium foil either side. If you are doing that using mains power, you should know the difference between class I and class II insulation and how that pertains to your construction. For UK / Northern Europe you’re looking at about a maximum power input of 150W per square metre, the thermostat will kick that back as required.

The thermostat probe then needs to be placed above the foil, rather than in the soil, and the whole lot covered in a thin layer of builder’s sand, on which the plastic modular trays with the seeds are placed. Since I am using this in a conservatory and don’t want the floor covered in sand I constructed a tray from some plywood and bits of pallets to contain the sand. The disadvantage of using wood from pallets is all the pieces are different widths which makes you look a rotten carpenter if you don’t have access to power tools to trim them to size, or the patience to do it with a jack plane. You can’t argue with the price, however!

tray with heater mat and sand

I was still left with a defective thermostat, so I replaced this sucker with a Dallas DS1820 digital temperature sensor and a 16F628 PIC microcontroller to drive a triac controlling the mains powered heating mat. It ended up looking more like a piece of laboratory equipment than a cuddly Bio-Green growing device, but is a lot more accurate. That system was originally part of a project to propagate sweet potatoes but the development time was a couple of weeks too long so I missed the start time and the tubers rotted My device had a second sensor and serial output because those sweet potatoes are finicky, this was something that originally grew in Mexico and South America. You are seriously taking the mickey trying to persuade them to think about growing in a chilly British March…

Temperature controller. It doesn't have the friendly curves of the Bio-Green devices so it looks like a piece of lab equipment but it is far more accurate.

For a propagator I don’t need 0.5C accuracy, so if we end up needing more of this sort of thing I may just use an analogue system using thermistors. I also had a Sankey propagator base I used to use for tomatoes, this is thermally balanced and could do with temperature control to save power and get more reliable, it can easily drift up beyond 30°C, which isn’t that great and makes the contents a sod to keep watered.

To save anyone who may come across this having to look the germination temperatures up, here are the values I swiped from a Plant Propagation lecture by the Organic Growers Alliance:

28°C Cucumber

25°C Aubergine, Pepper, Tomato

15°C Celery, Celeriac, Calabrese, Early Cabbbage, Brussels Sprouts

12°C Sweet Corn

10°C All others

All a fair amount of rant from a poor piece of engineering, though it says something for German engineering in general that the odd dodgy one stands out so much. Bio Green do make the electronic version of the device I constructed for about £50.