StillSmart

Hi Ant

Here's the info on my controller, it does indeed use a Triac

http://www.united-automation.com/avr-25 ... scia-dial/

Maybe not all of the scope pics are clear but if you look for the one with the multimeter reads 211 volt, you can see, (reading the scope from left to right) that at the start of each pulse it starts with a flat horizontal line that suddenly becomes vertical, this is the off time before the Triac is triggered, once triggered you see the peak and the back end of the sign wave as it returns to zero, the same is then repeated on the negative half cycle.

Ah yes. Just me trying to expand the thumbnails on my phone. I had stopped trying to open pics at a time when it would cause ads to open up in new windows which was a real problem on the phone. I see that has stopped now and the pics are viewable without being bombarded.

So mia culpa. Standard triac contrloller. Just my old eyes trying to view TNs on a phone. Wrong once more. Won't be the last time. Silver lining is that all this taxing speculation means my memory may linger longer. I may carry how triac controllers work for years to come.

Also it seems that if you did use mosfets instead of a triac you could clip the peak instead of the leading edge? More complicated and expensive I expect but doable if you wanted a lower peak voltage to match your lower average voltage.

ant wrote:

So mia culpa. Standard triac contrloller. Just my old eyes trying to view TNs on a phone. Wrong once more. Won't be the last time. Silver lining is that all this taxing speculation means my memory may linger longer. I may carry how triac controllers work for years to come.

Don't count on it!! I was teaching semiconductor theory and practical fault diagnosis on 4 stage amplifiers just 15 years ago.

I suspect I burned out a few brain cells when I started with this hobby. Bad washes, some really dubious equipment, and drinking product at too high an ABV.

However, I hope I can still remember the important things. At least short term memory is Ok, not a goldfish quite yet

Castle? What castle? Ooh a castle! Lol.

Fifteen years? That's ambitious. If i can still remember this stuff in five years time I'll call that a win. But the fine detail of comprehension does get blurred by time. I remember the basics but can't quote precise figures from memory. Just the general form of things which acts as a map or directory of what I need to look up. We should be comforted by the fact that Einstein did the same thing.

Ok. Did testing of controller and made pot better. Much more resolution and it will cost you less than a quid to do it to yours if you want to.

In the interests of full disclosure I did swap the original 500k pot for a 470k pot with a long shaft so that I could thread the shaft for a brass wingnut to hold my big red control in place. This makes negligable difference and the two graphs can be compared meaningfully. What did make a difference was a 39p fixed resistor between the wiper and the track leg you are wired to. This reduces the pot value to almost half and gives a slight semi anti log curve to the response. Overall I believe a big improvement.

Here is the raw data used to generate the graphs. You will note more data points on the data from the modded pot.


Deg=Watts. Change


0 = 1.3
120 = 1.5 0.2
150=45 43.5
155=95 50
160=175 80
165=245 70
170=310 65
175=420 110
180=505 85
185=594 89
190=677 83
195=785 108
200=890 105
205=960 70
210=1064 104
215=1114 50
220=1278 164
225=1395 117
230=1480 85
235=1550 70
240=1644 94
245=1772. 128
250=1897 119
255=1905 8
260=1976 71
265=2044 68
270=2113 69
275=2158 45
280=2177 19
285=2179 2



Deg=Watts. Change.

0=49
30=59. 10
35=67. 08
40=84. 17
45=100. 16
50=120. 20
55=144. 24
60=166. 22
65=180. 14
70=220. 40
75=230. 10
80=260. 30
85=284. 24
90=309. 25
95=348. 39
100=363. 15
105=393. 30
110=420. 27
115=490. 70
120=524. 34
125=563. 39
130=614. 51
135=659. 45
140=694. 35
145=732. 38
150=774. 42
155=837. 63
160=885. 48
165=940. 55
170=1021. 81
175=1077. 56
180=1158. 91
185=1213. 55
190=1274. 61
195=1332. 58
200=1386. 54
205=1425. 39
210=1500. 75
215=1552. 52
220=1647. 95
225=1704. 57
230=1756. 42
235=1828. 72
240=1899. 71
245=1973. 74
250=2046. 73
255=2121. 75
260=2187. 66
265=2234. 47
270=2270. 36
275=2288. 18
280=2309. 20
285= 2309. 0
300=2309. 0

Here is the online calculator that will tell you what size resistor will mod your pot

http://www.electronics2000.co.uk/calc/series-parallel-resistor-calculator.php

You want the third calculator. Put in the current value of your pot. Put in the value you want it to be. It will tell you which resistor to use. Err on the side of making the pot value a little smaller than you think you need to eliminate the dead zone. If you leave a bit of dead zone resistance on the pot it will magically become a bigger new dead zone than you thought it would. As you are reducing the effective value of your pot you are increasing the turn it takes to change the value a given amount. So just a few percent of missed dead zone can become a big dead zone all over again. Cut it all off in one go. As you can see from my first graph my deadzone was almost half the 300deg turn range. Some mucking around with half watt fixed resistors got my pot down from 470k to 263k. That was my sweet spot. Yours may vary. If you have an ohmmeter you can measure the value of your pot just as it leaves the dead zone. Then use the calculator to see what resistor you need to put the whole pot in that sweet zone. It will be the best 39p you spend that day (Maplins price). Or you can print out the dial and see what turn percentage is dead zone and work it out from there. If you accidently make your pot too small it will just start at a higher power. You can't hurt anything and if it takes a few goes to zero in on your sweet spot then it's only 39p a go. All the fun of the fair.

That's about it really. The meter has an annoying lag before it updates and is a bit flakey on very low power. Made worse if you have flakey line voltage. Mine fluctuates a bit anyway. Larger loads also produce larger voltage drops so there is a bit of a load dependant response going on as well. Ungrateful of me to complain considering what I have for what I paid. But now you know in case you are thinking of one. The resistor is the real killer mod though.

I tested the passive heatsink with over 4kw for 50 mins. Brought 25-30lt of water to boil. Heatsink was a little hot at the base but comfortable to the touch using various bits of my hands. Just warm for a lot of it. Think it will do until further data comes in.

The only thing this still needs a handle to pick it up and carry it around. It is heavy and the control knob is a tempting unconscious mistake waiting to happen until I get something sorted.

Ok, it's done now, for now.

Fantastic Ant you are not a prop maker for Dr Who are you?

Is there a slot for the 50p ?

Impressed, I bought the StillDragon sticker with the scale on it for my power controller, I'm tempted to plot a similar curve with mine to see how it compares and more importantly if it could be improved

I'm glad that you got there.and have enjoyed trying to keep up with you.
I don't expect to copy you,but it was interesting to follow you,

Robert.

Dr Who? I wish. Thank you for the kind words. I do aim for a style as well as functionality. The slot takes shillings which used to be valuable until the government and the banks manipulated inflation to take back every wage rise we ever won.

Yeah. It would be great to see your own graph here. I feel sure it can be improved. Those stickers kinda entrench the assumption that it's just something that can't be helped. A function of tbe device as Myles believed. They take an ill thought through design and label it neatly to make it look good. Then it seems proper. It happens a lot in history. A paradigm is established and held onto. I do wonder why the pots are so massively overestimated by the manufacturers though. The designer must have some reason. Different load and/or voltage conditions I guess. Dunno really. The good news is that it's an easy fix. Especially for you.

Thank you Robert for sticking with me and being someone to teach to. You have helped me with my own thoughts and learning. If you did want to do it then it's only one resistor. Two joints. Just sayin. All you would have to do is use the dial to establish the full turn angle range of your pot. Then the degree angle which passes the dead zone and turns on your controller. That and the current value of your pot are all the information you need to gather for a go. I'd be happy to talk you through it and help if you want to try.