How To: Build a Dew Heater with Resistors

[netwolf]

Phill,
Can you advise where you bought your parts from. Would the wire be available at Jaycar? I was thinking for this type of wire it might just be simpler to make a simple regulator with a pot control. I could mange this. This way you dont get any interference, or spikes on the power. Can run it off the same battery pack as the scope without any ill effects on the LX90.

Regards

[acropolite (Phil)]

Runs no problems on the same power source with the LX, the switch rate is slow so interference is not an issue. The 555 and transistor (see what I said about substitutes in the linked thread) will be available from DSE or jaycar, the nichrome wire you might have to substitute as I found that gauge hard to come by in OZ. I didn't build the dewstrap as I bought one from scopestuff with some other bits. Laurie reports that old electric blankets are a good source for the nichrome wire, so if you can find an old blanket that might be the go. Silicon tubing is available as fuel tubing from most hobby suppliers.

[netwolf]

Would this work? Looks to be about the same resistance per meter ~14ohms.
http://www1.jaycar.com.au/productView.asp?ID=WW4040

What did you use for Insulation? I can not find this Teflon stuff.

Regards

[AndrewJ]

Gday Fahim

Are you speaking of NPN vs PNP Transistors?

Yep, but not Transistors, you NEED to use Mosfets
( with really low "on" resistances )
Transistors have a much higher "ON" resistance than mosfets.
As a Dewheater strap is a very low resistance item by itself, there is a lot of power dissipated across a transistor ( hence in the 555 designs you see, the transistor has to be mounted on a heatsink, thats just wasted energy )

No idea, please explain pros and cons of each.

PChannel are harder to drive using logic level supplies ( ie PICs ) and are more costly, but give better load isolation.
The "driving" is not a problem for my type 1 or 2 designs though, as they all operate at full rail voltages, not 5V like the PICs do.
If you look at the circuits i have posted, you will see how the Mosfet acts as the switch.
Nearly all units use an NChannel mosfet. In this configuration, the switch is between the strap and ground. As such, when the switch is fully open
( ie heater turned off ), the heater wires and leads are all at a 12V potential relative to ground. Thus you need to be very sure you have everything insulated to avoid accidentally shorting anything out.
If you use PChannel, the switch goes between 12V and the strap, hence ensures the strap is always at ground potential when its off.

Re the Nichrome wire, the Jaycar stuff is OK, you just calc how much heat you want, then work out what length you need.
remember P = VI = IIR
I made up a heated dewshield using this wire, and just pulled it through some heatshrink for insulation.

As to the type 1 and type 2 units, i have loaded some zips with the circuits and also piccies of the one i have in my fork arm ( a type 2 )
The benefit of the design used in these ( over 555 types ) is the units can be made to run from full on to full off smoothly. A lot of the 555 designs cant go full on or off.
In the type i use, frequency can be easily set using R2 and C1.

Andrew

[snowyskiesau]

Originally Posted by netwolf

Would this work? Looks to be about the same resistance per meter ~14ohms.
http://www1.jaycar.com.au/productView.asp?ID=WW4040

What did you use for Insulation? I can not find this Teflon stuff.

DSE also have nichrome wire, 28B&S gauge, 13 ohm per meter, W-3205 is a 4 meter pack for about $12.
For teflon or silicon tubing, try a model aircraft hobby shop.

[AndrewJ]

Gday Fahim

Can the PIXAXE handle two or three heaters automatically, what are the input output limitations?

Yes and No, ( and all variations inbetween )
The PICAXE is only the brains, not the brawn here.
All the power is handled by the selected Mosfets, the PICAXE just has to drive these ( via a buffer opamp/comparator )
Ie the PICAXE chip could control as many devices as it has inputs and outputs. But is limited from doing this directly by its hardware PWM
I havent looked recently, but one other idea i had a long while ago was to get some of the new digitally settable onewire potentiometers.
Thus, you would build a std sawtooth oscillator ( like i use in my no 1 and no 2 designs ).
The output from this would be fed into a series of comparators/opamps
Each opamp would also be fed by a dedicated digital pot
The output of the opamp would then drive the mosfet
This would give full control of all channels, with limited load on the controlling PICAXE.
Another thing to investigate.

Andrew

[netwolf]

AndrewJ,

Thanks a lot mate, very useful information. I am going to have to crack open and dust of some of my old textbooks. I will probably go with you 1 or 2 design for now. I was talking to some techs at work today, and one of wiser ones told me i could always use a rectifier at the output to reduce RFI caused by the inductive wire loads. What do you think?


Geoff, Thanks looks like the DSE is way overpriced is there any benefit with that wire?

Regards
Fahim

[snowyskiesau]

Originally Posted by netwolf

Geoff, Thanks looks like the DSE is way overpriced is there any benefit with that wire?

My mistake, that price was from an invoice and was for 4 packs. It's around $3 per package (4 meters)

[netwolf]

About the same price then, I went and bought a Camping mat from Bigw today for 13$. I chose the 10mm green one, the blue one was 28mm thick. did have The Velcro they was in strips only about 10mm wide. I cant see how this stuff will hold it together. There sew on, stick on, and iron on, I was thinking of stick on but how well will it stick to the mat material. I will see if Bunnings have got a better sized velcro selection or else maybe Spotlight. I will give this a whirl once the clouds clear and see how it does.

Regards

[bird (Anthony Wesley)]

The PICAXE system is terrific, I use one for the heart of my 4-port temp sensor gear on the scope. Easy to program, and easy to interface with. Hard to go wrong.

cheers, Bird

[netwolf]

I think perhaps the Dewbuster also uses PIC or PICAXE.
This is from there website, and from what is shown it does not look very complex. I guess most of the work is in the code.
http://www.dewbuster.com/images/db4_...-animation.gif

Regards

[netwolf]

AndrewJ,

Ok I am begining to see what you are trying to do here. Your PICAXE is I think a really good design. What if the PWM output from the PICAXE was rectified, would this not reduce the RFI? I looked at the Microzed page, and I found the PICAXE-28 has a few more PWM outputs. What is the diffrence between the A and X series?
Is there a good internet resource for PICAXE and PIC newbies. Is there also someplace to find most used input output circuits. I mean a encyclopedia of circuits to provide input and use ouputs from the PIC's to control motors, sensors, etc etc. Am I correct in assuming the PIC/PICAXE is not a uP but a controller with a uP inside, designed for general applications. I ask because the microprocessor I designed with at Uni one of the motorolla's, had an address bus, data bus, IRQ's etc etc and you provided the program from a ROM chip. Am I correct to assume then the PIC/PICAXE has all of these inside?


Regards
Fahim

[AndrewJ]

Gday Fahim

one of wiser ones told me i could always use a rectifier at the output to reduce RFI caused by the inductive wire loads. What do you think?

In the words of the castle, he's dreamin.
The RFI from the PIC/Controller is easily covered
However, as the system uses PWM as its output, ALL wires going to and from the heater wires have up to 2Amp / 12V squarewaves pulsing along them.
( Ref my earlier post ) As such, the RFI comes from the leads and the heater wires, not the controller.
You could apply massive LC output filters, but thats starting to get bulky and costly. In most cases its not warranted.
All the Dickies controllers ( and my type 1 and 2 ) put out RFI from the wires, its just something to be wary of if you see interference in other things ( like wireless video feeds etc ). I havent had any probs with my scope so far,
but thats not an absolutely conclusive test.

Andrew

[netwolf]

So how to the Dewbuster and other commercial controllers mange to keep this down? Or at least advertise that it is reduced.

Regards

[AndrewJ]

Gday Fahim

What if the PWM output from the PICAXE was rectified, would this not reduce the RFI?

Yep, but it would kill the Mosfet
To use PWM, you want the mosfet to toggle between hard on or hard off. This gives the minimal resistance drop across the mosfet.
To try and use it like a transistor, ( where output voltage is proportional to base input voltage ), it will overheat ( and die ) quickly
Thus, the major RFI comes from the leads, not the controller.

I looked at the Microzed page, and I found the PICAXE-28 has a few more PWM outputs.

It only has two, hence, for yr 4 channels, its probably better to look at a distributed system.

What is the diffrence between the A and X series?

Different underlying PIC
PICAXE is just a std PIC with a proprietary "basic" interpreter loaded ( hence its slower running ). The A vs the X just signifies what type of PIC chip its based on, and hence its extra capabilities.

Is there a good internet resource for PICAXE and PIC newbies.

I wouldnt suggest PIC straight up, but for PICAXE, there is a yahoo group
http://groups.yahoo.com/group/picaxe/
and also a REvEd forum group
http://www.rev-ed.co.uk/picaxe/forum/
lots of good resources and links there.

Am I correct in assuming the PIC/PICAXE is not a uP but a controller with a uP inside, designed for general applications.

Nope. As per above, the PICAXE is just a raw PIC, preprogrammed with a basic interpreter that knows how to do lots of things ( but not all )
Ie, it has a single command to read a dallas onewire digital temp sensor ( but only allows one sensor per input )
It also has a command to output PWM, etc etc
download the software and have a look through all the examples for whats available.
What isnt available can sometimes be bit banged if necessary.

I designed with at Uni one of the motorolla's, had an address bus, data bus, IRQ's etc etc and you provided the program from a ROM chip. Am I correct to assume then the PIC/PICAXE has all of these inside?

Yep. Its basically the same process, but now you have a human readable code generator between you and all the guts of the PIC
As such, you program at a high level, and it takes care of the rest ( within limits ).
As i mentioned earlier, download the software and read the manuals / tutes on what it can and cant do.

Ie for my basic controller, the program is basically
read temp sensor air
read temp sensor ota
if temp air - temp OTA > 2 set pwm to 100
if temp air - temp OTA < 0 set pwm off
if between, set PWM to a proportional value
This provides a ramping application of power that avoids having the system hunt or thump on/off
( my no 2 circuit also works this way )
ie i can programatically ramp the power based on temp difference.
Also, as it has an onboard ADC, you can measure yr battery voltage
and warn/shutoff the sytem before it damages yr battery.
Lots of options available.

Andrew

[AndrewJ]

Gday Fahim

So how to the Dewbuster and other commercial controllers mange to keep this down? Or at least advertise that it is reduced.

Dunno, but i dont always believe everything i read ;-)
I have never seen/tested a Dewbuster, but i have seen a new Kendrick
It Has a metal chassis, and if this is earthed properly, then the RFI and noise back up the lines shld be reduced as the alloy chassis will form a shield. Also, the old kendricks i tested ran at about 5 on/offs per minute whereas the new ones were about 52Hz, thus the new ones dont thump on/off as much as prev, hence reduce noise spikes back up the supply lines, but they do produce rfi.
The problem ( with RFI ) comes when you attach a heater strip.
The wires between the controller and the strip radiate away like blazes
Again, i must stress, i know i have RFI coming from my units wires ( and anyone running a PWM device will also have RFI ). Its just a matter of "does it affect the scopes electronics".
To date, mine hasnt, so i dont worry too much.
Unfortunately, RFI reduction falls into the "black arts" category.

Andrew

[netwolf]

Andrew,

I suppose you could build some kind of circuit that would produce a PWM from a standard output..
How would you go about setting up a distributed system?
Would a Basic Stamp be better? OR maybe a Atmel AVR based system. I wonder what Kendrick and Dew buster use?

Regards

[netwolf]

I was checking out the SMD dobs (drooling is a better word), and noticed Peter mentioned the dew-not heaters come standard with his scopes. As far as I know no one else is bringing these in. I sent him an email yesterday, and got a reply this morning, he does have them. But not for the SCT's, though he is happy to order and get them in for me. These i believe us a carbon strip, as opposed to the Kendrick nichrome wire ones. Now I am tempted to get this, but I think i can build a nichrome one for about 10bucks. So I will give that a go first.
So if anyone wants the Dew-not contact Peter at SDM. Arguably the best commercially available heaters.

Regards

[Gama]

Why not just use a LM35 to feed back a voltage to a LM317 to maintain a steady temperature. Use an Opamp to get the exact amplification if required to set the correct temperature if the voltages are too small.
Basically, the LM35 is 10mv per deg. So say you want it at 20 deg = 200mV.
You can feed this value back into the LM317 via a pot for setting your desired Temp. This way your temp stays stable, if the temp varies, the current will either increase or decrease to compensate. Plus its slow DC, so no noise.
This is similar to a design i made for a turbo boost controller about 10 years ago to control boost if temperature increased over a set position.

[netwolf]

Hey Theo,

Have you considered building a local unit for Australia? Or just one for me.
I should really learn to solder, after all I spent 5 years trying. I recall my Signals and systems professor kept telling us how every year we are at uni we are loosing 30K of income. So if you dont want to be here dont. He was a nice guy, old school.

I like the simplicity of PicAXE version, compared to that the no2 circuit is quiet complex. My head hurts looking at all those op amps. As it used to hurt in electronics lectures. No idea how I passed that one. However uP's and coding are more my cup of tea.

But build wise the no2 looks easier to do for starters.
Theo can you draw me a picture of your design.

Regards
Fahim