Controller Build Advice Request

[Longhairedcountryboy]

I have outgrown my old, thrown together controller. I started compiling components for a new controller. I was planning on using this meter... https://www.amazon.com/gp/product/B00YY1KOHA

and then Oldvine Zin posted his new controller build... viewtopic.php?f=85&t=72738. In his thread, there is discussion about digital meters not working correctly in this kind of circuit and using analog meters instead.

So I ordered a couple analog meters. I was just going to use an ammeter, but then I thought I should include a voltmeter as well so I can more accurately calculate power.

Now that I have all the parts in front of me, I realize it will be difficult to fit both the analog meters in the enclosure I have for this project. It is doable, but its going to be really tight which will make it very difficult to wire up.

My question to the community is, if you had a choice of the three meters I have, what would you use? Would you just put in the ammeter alone or try to cram the voltmeter in as well? Should I scrap the analog meters all together and just use the digital and deal with the potential problems caused by the SSR?

[zed255]

Analog ammeter. You do not need to know the voltage to figure out power since heating elements are just big resistors. Analog will do a decent job of reading average values (they do not read RMS for anything non-sinusoidal) but cheap digital can be a crap shoot on what it reads unless you get a tru RMS unit ($$$).

[Longhairedcountryboy]

Analog ammeter. You do not need to know the voltage to figure out power since heating elements are just big resistors. Analog will do a decent job of reading average values (they do not read RMS for anything non-sinusoidal) but cheap digital can be a crap shoot on what it reads unless you get a tru RMS unit ($$$).

Thank you, zed.

That is exactly what I was thinking. The voltmeter idea was just to monitor any fluctuation in the line voltage so as to be able to calculate the power more precisely . My hunch is that the voltage is probably consistent enough to get repeatable results with the ammeter alone. I just wanted to double check with people that have more experience before committing to the design.

[Expat]

For reference, I'm using the same larger analog ammeter, no issues. The larger display surface makes quick reads very easy.

[zed255]

Line voltage fluctuations will show as a fluctuation in current drawn. Ohm's law.

Just measure the resistance of your element(s) or calculate based on rated power and rated voltage. R=V^2÷P, so a 5500W element rated at 240V would have a resistance of: 240^2÷5500=10.5 Ohm, rounded.

For power from current and resistance:

P=I^2×R

I just wrote down a table to correlate amperage readings to power output. Want x watts, set it for y amperes.

[Yummyrum]

+1 what Zed wrote .

Work out your element resistance ( one time calculation ) then calculate power from that and current .

The graph or chart is a great way to do it .

[Oldvine Zin]

I agree with the comments above , the ammeter will be the most useful meter. I included a volt meter in mine because I love bling and I had the space.

Good luck
OVZ

[shadylane]

I prefer just a cheap voltmeter.
Cheap analog amp meters are more prone to going up in smoke.
I don't worry about accurately calculating power
Especially since most controllers produce a non-sinusoidal wave form
That makes it difficult to accurately measure voltages or amps.
And therefore power.

[Oldvine Zin]

And hell - my last controller just had a control pot and I was able to run just fine- but the added bling will make a finer product

still trying to get it right
OVZ

[sambedded]

Analog ammeter. You do not need to know the voltage to figure out power since heating elements are just big resistors. Analog will do a decent job of reading average values (they do not read RMS for anything non-sinusoidal) but cheap digital can be a crap shoot on what it reads unless you get a tru RMS unit ($$$).

Thank you, zed.

That is exactly what I was thinking. The voltmeter idea was just to monitor any fluctuation in the line voltage so as to be able to calculate the power more precisely . My hunch is that the voltage is probably consistent enough to get repeatable results with the ammeter alone. I just wanted to double check with people that have more experience before committing to the design.

Regular analog ammeter voltmeter will not give you correct numbers on non-sinus wave same way as cheap digital voltmeter. The readings are actually pretty close on cheap digital and regular analog meters. You need a True RMS voltmeter (digital or analog) to calculate power correctly.
Let say you have set 40% of power by dial. and your mains voltage is 240V and you have a 5500Watt element

Voltmeter's reading
Regular - 108.1V True-RMS - 158.7V

Actual power 2862 Watt. Calculated by analog voltmeter reading - 1327Watt.

But now you can buy relatively inexpensive True RMS meter like this one - https://www.ebay.com/itm/AC-110V-220V-D ... 1615977127" onclick="window.open(this.href);return false;" rel="nofollow

[zed255]

Seen lots of meters like that. It did not say anywhere in my perusal about it being 'True RMS' and most others don't either. I'd highly doubt they really are.

You also can't install it after your control element, it must get its power from the mains before the phase controller. This means power values displayed are based on the input voltage and the output current. If you did install it after the meter likely wouldn't turn on or operate properly at a low power setting of the controller. I have even read accounts of these meters failing if fed power from the output of the controller.

I compared my analog ammeter against a respectable industrial true RMS clamp meter and the error wasn't gross in the reasonably usable part of the controllers range, remembering that the average reading of the analog is NOT the same as the true RMS value on the test meter. Most of the error is in the lower range; as the waveform get closer to a complete sinus the error diminishes. The analog meter has a nice smooth movement and is pleasant to read. They are also essentially bullet proof, so to speak.

We really we only need a reference and not necessarily a truly accurate value so frankly I wouldn't stress it too much.

[sambedded]

Its a PZEM-004 meter based on the SD3004 chip - an electric energy measurement SOC with built in MCU http://www.sdicmicro.com/DataSheet/SD30 ... 0v0.2c.pdf" onclick="window.open(this.href);return false;" rel="nofollow

Yes the voltmeter part should be connected "before" SSR. And it will always show mains voltage. But it will show True RMS amps and calculate power correctly because it uses not an average but instant values of Voltage and current measured several thousand times per second.When triac inside an SSR is closed instant current equal 0. So multiplication U*I will give you a zero power even if measured voltage is not zero due to "unproper" connection.

[zed255]

OK, I wasn't thinking its operation fully through. Thanks for pointing it out that power will still be calculated properly.

I had it stuck in my head it had to sample the voltage after the controller to do that, my misunderstanding.

I still have a soft spot for analog meters for some applications. Anyhow, I appreciate the enlightenment.

[Yummyrum]

OK, I wasn't thinking its operation fully through. Thanks for pointing it out that power will still be calculated properly.

I had it stuck in my head it had to sample the voltage after the controller to do that, my misunderstanding.

I still have a soft spot for analog meters for some applications. Anyhow, I appreciate the enlightenment.

Naa . Zed you are right first time .

Yes the voltmeter part should be connected "before" SSR. And it will always show mains voltage. But it will show True RMS amps and calculate power correctly because it uses not an average but instant values of Voltage and current measured several thousand times per second.When triac inside an SSR is closed instant current equal 0. So multiplication U*I will give you a zero power even if measured voltage is not zero due to "unproper" connection.

Sambedded . Thats not right . If I’m understanding what you are trying to say .
You must measure the voltage on the element side of the Triac and multiply it by the current through the element .

You will get a wrong power reading if the voltage is measured before the Triac .

[still_stirrin]

We really we only need a reference and not necessarily a truly accurate value so frankly I wouldn't stress it too much.

+1.

The analog ammeter is perfect for this duty. It’ll give you a quick (integrated) visual of the (relative) power to the heat element. This visual is more than adequate to know roughly the rate of vapor produced throughout the run.

And you don’t need to make calculations to drive the boiler either.
ss

[Longhairedcountryboy]

My old controller had a digital ammeter and worked just fine. It was sloppily thrown together in a hurry with the intention of rebuilding it someday. I think it's been almost 2 years now and i finally am getting around to it.

I planned it out and ordered some parts, but then Oldvine's thread started. I got carried away with wanting something unnecessarily complicated. Hey, it happens sometimes.

I know enough to be able to follow a schematic and size wire accordingly, but when it comes to things like what is happening to the waveform on the output of the ssr and how that affects the ability to meter the power, I am (was) clueless. I'm going to keep it simple for now and just wire up the analog ammeter and call it good enough for what it is for. Repeatable results is all i am really trying to accomplish.

Thanks for the help, everyone.

[Anyhowe]

Interesting stuff. I am just a smuck noob and have so far come to the realization for me that so far with the variability of wash abv, ambient temperature variations and different chararcteristics of what I am making, anything more than a controller with a dial with marks at full, 1/2, and 1/4 would be superfluous. I have much to learn and will keep an eye on this thread. Thanks.

[sambedded]

Yes the voltmeter part should be connected "before" SSR. And it will always show mains voltage. But it will show True RMS amps and calculate power correctly because it uses not an average but instant values of Voltage and current measured several thousand times per second.When triac inside an SSR is closed instant current equal 0. So multiplication U*I will give you a zero power even if measured voltage is not zero due to "unproper" connection.

Sambedded . Thats not right . If I’m understanding what you are trying to say .
You must measure the voltage on the element side of the Triac and multiply it by the current through the element .

You will get a wrong power reading if the voltage is measured before the Triac .

What do you mean saying "must measure"? Will I be prosecuted if do it other way?
You are right - measuring on element side gives you a most accurate reading. However it's not mandatory. Generally speaking triac has two states On and Off. When triac is in Off state there is a zero voltage on element. When triac is in On state voltage on element equal to voltage “before” triac minus triac voltage drop. Usually triac voltage drop is around 1.5Volts So difference is not too significant. I cause less than 2 percent error in power calculation. And here I’m speaking about “instant” measurement when measurement time is much less than AC period.

[Yummyrum]

Sam , what you are saying is correct at full power
But at anything less that full power , it will not be correct .

We know that power = volts x current. But at anything other than full power , the RMS voltage on the switched side of the triac will be less than the RMS voltage on its input .

Say at 50% power , the triac is switching on at the peak of the waveform , so only half of every half cycle is applied . The RMS voltage will be only half of what the input voltage is .

So we can’t multiply the RMS input voltage ( which will always be maximum) by the RMS current through the element as it is significantly out compared to when we correctly multiply the RMS voltage on the switched side by the RMS current through the element .

Sorry not trying to be a pedantic wanker but it is the way it is . And if we are going to use a true RMS chip to carry out the measurements to calculate power then it needs to be measuring the right thing .

Incidentally thanks for posting up that chips spec sheet . It does seem that it would be an easy modification to provide two
Voltage inputs , one for the measurement side derived from the triac output side and another from the triac input side that purely supplied operating power for the chip . This would probably stop the burning out that seems to occur with some

[sambedded]

Sam , what you are saying is correct at full power
But at anything less that full power , it will not be correct .

We know that power = volts x current. But at anything other than full power , the RMS voltage on the switched side of the triac will be less than the RMS voltage on its input .

If Triac is always ON RMS on both side of triac will be almost the same. If, say, it On only a halve of AC period the RMS on element side will be 1/2 of the RMS on mains side. You are right here. But RMS is stand for a Root Mean Square! I.e. type of averaging.
But I was talking about instant values. When triac is ON (halve of period) we are measuring (very fast) correct non zero current and multiplying it (to get instant power) on measured voltage on mains side (which now is almost equal to voltage on element side). Then next halve of period triac is OFF. Current will be equal zero, voltage on element side is also zero. Since we a measuring voltage on mains side it's not zero but being multiplied on 0 current it gives us zero as well. So when we calculate an average power from set of instant ones it gives as a 50% of original power.

[Yummyrum]

Arrrh .... I get it .
Now I understand how come those meters will work to display correct Power even though the volts is connected before the triac.
OK , I was assuming that those meters carry out a true RMS reading of Voltage and then a true RMS reading of current and multiply the results say once a second and give a power answer .

But it appears not .
So what you are saying is it multiplies the many thousands of "samples” per second of voltage and current and carries out an average of these .
It would be rather a quirk of how they work that makes them "happen" to give a correct power value even when the Voltage is connected to the input of the triac.......even if the Voltage reading is irrelevant

Just trying to establish that .....thanks for providing the info .

It's just that there is a lot of confusion and miss-information on how things electrical work and a lot of members can be confused.
So it's good to clarify these points .

[cranky]

Personally I never understand these conversations and how complicated they get I really don't see the need to over complicate things, other than it looks cool. My personal goal at the most is repeatable results and all my meter is is a reference number. I really don't need or care to figure out to the nth degree exactly what the electricity is doing to get repeatable results. I have a digital Volt/Amp meter with the voltage wired before the controller, it tells me I have full voltage going into the controller, which I need for troubleshooting in the event something is wrong. If I set it so the amp meter reads the same as it did last time it will run the same, if I have voltage and no amperage I have a problem downstream of my switch. If I have no voltage it is upstream. The reason I have a digital meter is it is much easier to see and read than any analog meter, I can even read it from 10 feet away.

[NineInchNails]

^ ^ ^ I agree.

[Longhairedcountryboy]

I'm finally getting around to putting this together. I have been putting it off for months. I think cutting the box was the biggest mental hurdle I needed to get over to make this happen. I was dreading it for some reason. I kept changing my mind about what I wanted to put in it and how I wanted to lay it out. I had to come to terms with my ability and available tools. I had to realize and accept it isn't going to be as professional and fancy as I had envisioned. Anyway, today I will wire this up. I wanted to get a few pictures on here before I make a mess of it. Hopefully it will work without letting any of the magic smoke out. Any pointers or suggestions are welcome. Wish me luck.

Edit:
What is a good gauge wire to use for the pots? I have some 20 gauge doorbell wire on hand. How much current does it need to carry?

[zed255]

The potentiometers have very little current flowing through them, 24 AWG would be overkill.

[Oldvine Zin]

Looking good LHCB 20 gauge wire should be fine for the pots


OVZ

[MtRainier]

I agree. That's a nice looking box and looks really cleanly laid out.

[Longhairedcountryboy]

Thanks for the complements, guys. I didn't get it finished today. I had to walk away from it a couple times. This is not one of my strong suits. It isn't looking so nice and cleanly laid out at this point. I'm making this thing up as I go along. At one point, about 25% or 30% into the wiring, I realized I was blocking the fans. I tore it all out and started over. Had to run to the home improvement store for more wire.

I'm taking my time, putting one wire in at a time. I understand the circuit and can mentally visualize the schematic, but don't have a clear picture in my mind of the finished product, so the wire runs are just kind of revealing themselves as I look at it and figure out what attaches to where and how. Now that I have about 75% of it done, it is becoming obvious that I could have made more efficient choices with my wire runs. Its kind of a mess at present. I think i can clean it up some, but it is what it is now. I'm going to try to focus and get it all together tomorrow.

One big regret I am having right now is not including bypass switches to send full power past the ssr's straight to the output plugs. I have plenty of room to fit them in later, though. I just need to get this done. I have a couple projects waiting (for months) on this to be finished.

I'll be back tomorrow, hopefully with pictures of the completed project

[Oldvine Zin]

I don't think that bypass switches are necessary, full on on the ssr should get you close to full power. If the ssr fails to an off state during a run (unlikely, usually they fail to an on state) you can always plug through bypassing the controller.

One simple mod to your box is to add some vent holes, to me it looks like you have both fans pointing the same direction so not much air flow doing it that way.

Good luck
OVZ

[NineInchNails]

If one fan is blowing in and the other is blowing out then you should be good to go.

Another option is to just get rid of one of the fans and install a perforated plate in its stead to allow air to flow through.