PWM for Dew Heater - design question

[barees63]

I'm about to build the simple Dew Heater controller shown here: http://www.backyard-astro.com/equipment/accessories/dewheater/dewheater.html

but I just had a question about the frequency, for this design they have used a pulse frequency of 1hz and I was wondering if this seems a bit slow to be efficient - I'm wondering if it would allow the element to cool too much between pulses, especially on the lower duty-cycle? They state this value was chosen to avoid spikes but I won't be using the same supply for any computer gear (just fan and dew heaters) so was wondering if any of you electronics experts had an opinion on the optimal frequency? TIA! Bruce

(Oops! I meant to post this in the DIY section - how do I move a thread?)

[RB]

Looking at the design Bruce the 1M pot (P1) resitor varies the duty cycle so you can try different duty cycles.

[barees63]

Yes, but the pulse frequency is always one pulse per second (with variable duty cycle), this means on low setting it will get a tiny blip of power once per second.. I'm probably way off beam but I thought maybe that was too slow and would allow the element to cool too much before the next pulse.. eg rather than a nice low even heat you would a more pronounced heating & cooling cycle which might not be as good?

[RB]

I get yah Bruce.
Well another option is this:
Put in another Pot to vary this base frequency too.
Mind you it's been 20 years since I've played/designed anything with the 555 timer so can't suggest anything off the bat.

Someone else surly will soon.
Meanwhile I'll think about it.

:)

[Robby]

1Hz will be fine. You'll lose no more in the off period with 1Hz than you will do a 1kHz.
PWM systems at low freq are only less efficient with highly inductive loads. There is a trade-off then between inductive losses (at low freq) and switching losses (at high freq).
But for a mostly resistive load (there will be a small amount of inductance) like yours higher freq's are actually slighlty more in-efficient due to switching losses.
1Hz will be fine! :D

[Starkler]

As the heat from the dew heater is meant to be coupled to a load (corrector plate, eyepiece etc) which will have a considerable thermal mass, 1hz will pose no problems.

As for inductive loads, I would think the tiny inductance in dew heater wires wouldnt be an issue until you get to switching speeds in the mhz region.

[barees63]

thanks guys, I'll go with the design as it stands then, initially I'm only going to use an eyepiece heater since this is where I mostly have problems, my finder dews up a lot but I only use it for initial alignment so it doesn't bother me, I've only had the primary dew up once and everything (including me) was so dripping wet by then it was time to call it anyway :rain:

[Vermin]

If you want to reduce the back EMF spikes caused by switching an inductive load, just put a 1N4007 diode across the heating element as shown in the attached image.

To increase the frequency, either:

1) Reduce the value of C2,

and/or

2) Reduce the values of R3 and P1, whilst maintaining about a 1:20 ratio, i.e P1=500k, R3=24K, or P1=100K, R3=4.7K etc...

But as others have stated this is not necessary. I'd still put the extra diode in though as it offers protection for the switching transistor too.

[sheeny]

I have just sent an article to Iceman re: building a PWM controller if you are interested. Just using a kit from DSE. FYI it oscillates at 2kHz not 1 Hz, and it works well.

Let me know your email address I'll send you the files if you like - not sure how long it'll take Mike to massage it into a suitable article for the site...

Al.

[barees63]

Hi Al, sent you a PM

[norm]

[QUOTE]Mind you it's been 20 years since I've played/designed anything with the 555 timer so can't suggest anything off the bat.

I couldn't but help have a :) on my face when the 555 timer was mentioned. Its been nearly 20yrs too since I last mucked around with one of those or the LM3909 to build a flashing LED. They made a great fake car alarms back in those days!!!!

[acropolite]

Bruce, I have built the exact same circuit on veroboard and it works nicely. It's probably the most simple and effective circuit I've seen. An added bonus of the slow switching rate is that you can check the led to see roughly what the setting is (by checking the pulse length) in the dark. I did use a different transistor as I was unable to source the specified one over the counter locally. :D

[barees63]

Hi Phil, I saw your original post and I have made the same substitutions (transistor and cap) as you based on what DSE had in stock..

[barees63]

Here's my dew controller.. using the 555-based PWM, the right-hand switch is for the fan on my GSO dob, the left-hand switch controls the dew heater which has two outlets, I have two eyepiece heaters from www.dew-not.com (http://www.dew-not.com).. The controller itself cost me just over $20 in parts including the case, knob, switches etc.. I will be using a 7ah 12v gel battery.

http://www.zitafarm.com/astro/dew.jpg

[sheeny]

Cool! Nice looking job!

Al.

[barees63]

Thanks Al, it looks much nicer on the outside :D

I used to make up lots of circuits with veroboard when I was younger but I seem to have lost the knack.. it's quite messy :eyepop:

[sheeny]

I know what you mean... my PWM controller was the first bit of electronics I've done since my early twenties building guitar effects boxes, and leads...

Al.

[acropolite]

Yeah mine is pretty untidy inside as well. Funny in my 20's I would have whipped it up in half an hour, took me at least 2 hours to complete; I blame it on poor eyesight...