Changing the sensitivity/resolution of a potentiometer in an SCR Circuit

Thread Starter

Andrew Leigh

Joined Sep 8, 2008
131
I have a commercially purchased SCR controller which I use to control a 220v element. The controller has performed exceedingly well and is capable of an output of 4 000W, my current element is only 2 700W. My only issue it that I would like the output to be controllable to +-5W, to achieve this takes significant back an forth adjustments with the potentiometer to dial in the required wattage.

I though a solution would be to install additional 2W resistors in some form of parallel / series configuration with a selector switch or switches that would allow me to have finer resolution in the ranges that I work. I don't know how to go about this.

This morning I checked the two spare controllers I have. The output power is driven by a pot with markings 500k pot rated for 2W. Oddly enough both are identically marked but in both cases the the pots measure 4.7k when isolated from the circuit.

When sweeping the knob on the pots when connected to the multi-meter I seem to have found the problem. The pots are connect such that fully CCW or eight o clock the reading is 4.6k, at half past three it is 4.0k it then plummets to zero from half past three to 5 o clock. Both are the same and I hope my multimeter is not playing up. Is there something about pots in SCR circuits that I fail to understand?

Pot data sheet WTH118-1A series potentiometer for resistance input SSR or voltage regulator (finglai.com) The pot data sheet does not indicate that a 500K is available.

So apart from the fact that the pots appear to be faulty. Could anyone make any recommendations that will allow me to alter the resolution of the power output.

thanks
 
Last edited:

Irving

Joined Jan 30, 2016
3,845
The link you gave doesn't appear to be the manufacturer's site but a retailer. Other retailers list a wider range of values including 560k and 680k versions as well as double and triple versions. It wouldn't be unusual for a large-scale purchaser to spec a different marking, so 470k could be marked as 500k as its within 10 - 20% which is the usual tolerance for a carbon track pot - this is done so their inventory & stock is always marked 500k irrespective of supplier.

The pots could be a logarithmic taper rather than a linear taper, though a log taper would be as per the table below, so something seems strange with your measurement. Indeed its odd they measure 4k7 out of circuit though - they were completely disconnected? When you did your measurement you measured from pin 2 to pin 3 (shaft towards you)? Try measuring from 1 to 2, should start at 0 and get bigger - is it the reverse of what you measured?

1637585812796.png

1637586963129.png
 

Audioguru again

Joined Oct 21, 2019
6,674
The pots are extremely cheap then probably their resistance is marked wrong. The datasheet shows a photo of one that is marked 470k and available on the second page but is not available on the first page.

If you measured the pot to be actually 4.7k then the motor will probably runs slower with a resistor in series with the pot. Try 47k.
 

Thread Starter

Andrew Leigh

Joined Sep 8, 2008
131
Thanks for the responses.

The circuit connects to pins 2 and 3. Pin 1 is open as is the earth.

The pot connects to the circuit via a small two pin plug in connector, I unplugged the connector so the readings were completely independent of the circuit.

My multi-meter has just died o_OI think I was getting spurious readings as a result, standby, on my way to buy a new battery.

Apologies for feeding you bum data, will correct it in a while.
 

Thread Starter

Andrew Leigh

Joined Sep 8, 2008
131
I am a little lost now. New battery and still cannot get repeatable readings on the pots, the other battery was very low. In auto mode (resistance) my display is stuck on 4.63k, touch the leads and the display reads zero, remove the leads and the display kicks back to 4.63k so the meter is reading. Hold function is off. Somehow I think I have activated the REL (Relative) mode and my manual does not say how to cancel it.

Measured a couple of 220R resistors and they all tested out at 173R.
Measured two 1k0 resistors and they measure 690R

Measurements on the posts are erratic and there is no fluid increase from 0 to 4.7k it is zero, then some arbitrary reading around 100k then 4.3k. Both pots Faulty?

Ok back back to the point in question. Assuming the pots were working how would one uses a network of resistors to achieve better resolution in the output?
 

MisterBill2

Joined Jan 23, 2018
18,176
OK, one suggestion is to try some fixed resistors instead of the pot. Those are available with 5% and 10% tolerance. In fact, if you have any well marked fixed resistors you can try measuring them with your meter and seeing how close the mete readings are to what the markings are. Ten percent tolerance resistors will be within ten percent, but not within 5percent of the marked value.

I am guessing that as the control resistance is reduced that the wattage increases, that is how it usually works. If you can verify that then I can provide a suggestion about how to have better resolution for setting a specific wattage level. It will require a different potentiometer and some fixed resistors. Since I don't know what part of the world you are in I don't know how hard it is to get electronic parts.
 

Irving

Joined Jan 30, 2016
3,845
Somehow I think I have activated the REL (Relative) mode and my manual does not say how to cancel it.
Hold the leads touching each other firmly, then press REL again to get readings RELative to the nominal '0'.

Try measuring between pins 1 and 3 on the pot, that should be the total track resistance. If that's open circuit, or varies as the pot is turned then its junk.

To change the range/sensitivity first we need to know which end of the track is needed, ie smaller R = more heat/speed/etc or larger R = more heat/speed/etc.

Assuming further CW = smaller and that smaller => more then from your description in post #1 you want a smaller variable range centred around some point roughly 33% up from minimum.

Calling your pot total resistance R and assuming its actually linear then you need a resistance of 0.33R +/- say 5% of R.
So your pot 'swing' is 10% of R and the mid point is 33% of R so a fixed resistor in series of (33-5)% = 28%R.

(On a 500k original pot that would be a pot of 50k and a series resistor of 140k, so use 150k).

This assumes a linear pot. A log pot is not so easy...
 

LesJones

Joined Jan 8, 2017
4,174
I think controlling the output a 2700 watt element to within 5 watts is going to be difficult as this is within 0.185%. It will only require a VERY small change in your mains voltage to cause this change. (Remember the power is proportional to the square of the voltage.) So if your element produces 2700 watts at 220 volts to produce 2705 watts requires 220.203 volts. Your mains voltage will almost certainly vary by much more than this. I think you would need a closed loop control system to compensate for mains voltage variations. You have not given any details of your SCR controller. I suspect it will only be a simple phase angle controller.

Les.
 

Irving

Joined Jan 30, 2016
3,845
Thinking my last post through further and picking up on what @LesJones just said, your controller looks something like this:

1637606281757.png

Smaller R = earlier in the cycle switch-on = more output....
Bigger R = later/less

The CR time constant <= 10mS (1/2 cycle time)

R = 4k7, C is ~ 3uF
R = 500k, C is ~ 30nF

Looking at sample circuits on the web 500k/30nF is quite common set of values. 3uF would draw a lot of current (930ohm @ 50Hz = 350mA at full output and a lot of watts in the resistor at < full output) so we can rule that out. With 30nF, the max dissipation in R is ~ 0.5W so a 2W resistance is right.
 

Thread Starter

Andrew Leigh

Joined Sep 8, 2008
131
Hold the leads touching each other firmly, then press REL again to get readings RELative to the nominal '0'.
Tried that, not working, will try different permutations.

Try measuring between pins 1 and 3 on the pot, that should be the total track resistance. If that's open circuit, or varies as the pot is turned then its junk.
One reads open circuit the other reads 4,7k.

To change the range/sensitivity first we need to know which end of the track is needed, ie smaller R = more heat/speed/etc or larger R = more heat/speed/etc.
As you rotate the pot CW the resistance decreases.

Assuming further CW = smaller and that smaller => more then from your description in post #1 you want a smaller variable range centred around some point roughly 33% up from minimum.

Calling your pot total resistance R and assuming its actually linear then you need a resistance of 0.33R +/- say 5% of R.
So your pot 'swing' is 10% of R and the mid point is 33% of R so a fixed resistor in series of (33-5)% = 28%R.
OK will get replacement pots and a couple of 2W resistors in the range you mention.

(On a 500k original pot that would be a pot of 50k and a series resistor of 140k, so use 150k).

This assumes a linear pot. A log pot is not so easy...
Understood.

Many thanks
 

Thread Starter

Andrew Leigh

Joined Sep 8, 2008
131
@Irving.

For interest sake there are two caps on the circuit, a 100nF and a 0.14uF, the power resistors on the board are 2 each 33R and one 10R. The triac is a BTA41600B.
 

MisterBill2

Joined Jan 23, 2018
18,176
@Irving.

For interest sake there are two caps on the circuit, a 100nF and a 0.14uF, the power resistors on the board are 2 each 33R and one 10R. The triac is a BTA41600B.
It should be possible to trace out the circuit of the board and let us know.
What I do not see are the line and load connections. They would be useful to know.
 

LesJones

Joined Jan 8, 2017
4,174
I do not agree with the resistor values. I think one is 100 ohms (100 x 10^0) and two are 33K (330 x 10^2) The orange capacitor is in series with the 100R resistor and just forms a snubber between A1 and A2 (Mt1 and MT2)of the triac. I thing the variable resistor connects to the white 2 pin connector. I have not yet worked out the function of the bridge rectifier in the circuit.
This is as far as I've got tracing the circuit. I am guessing the the wire with the question mark next to it is the track that goes under the bridge rectifier.
221121.jpg
Les.
 

MisterBill2

Joined Jan 23, 2018
18,176
Close but not there just yet, at least it does not seem like it. And it is still not clear as to where the two terminals,line and load are. And I wonder if the designation of the bridge terminals is correct.
 

Thread Starter

Andrew Leigh

Joined Sep 8, 2008
131
Close but not there just yet, at least it does not seem like it. And it is still not clear as to where the two terminals,line and load are. And I wonder if the designation of the bridge terminals is correct.
For clarification ;

The track with the question mark is correct..
The pot is connected to the white 2 pin connector.

The resistor colour codes are;
Brown Black Black Black Brown
Orange Orange Black Red Brown
Never know which way to read them from.

Neutral goes directly to the load, line comes into A1 and out at A2. So this controller is connected in series with the line. Make sense?

On the matter of the rectifier I found this.
In order to obtain full-wave power control we could connect a single thyristor inside a full-wave bridge rectifier which triggers on each positive half-wave, or to connect two thyristors together in inverse parallel (back-to-back) as shown below but this increases both the complexity and number of components used in the switching circuit.
 
Top