Positioning a current sense resistor in a mains powered dimmer circuit

[ramuna]

KeepItSimpleStupid said

If this thing is a "one of", thing or proof of concept?
Its a one off.

Can you use DC?
Sadly no. The material being heated will decompose in a DC field.
I agree that if I could use DC it would make things much easier to control.

 

[rjenkinsgb]

As the power is only 10W, could you use a small 240V isolation transformer to power the load and connect one side of the secondary to 0V via a simple shunt resistor?

That's surely the simplest and most direct method to sense load current as such low levels?

 

[ramuna]

rjenkinsgb said
As the power is only 10W, could you use a small 240V isolation transformer to power the load and connect one side of the secondary to 0V via a simple shunt resistor?
I'm afraid I cannot visualise what you mean. Please provide a rough schematic.

 

[rjenkinsgb]

Run the whole circuit through an isolation transformer.
Something like this, as an example:

RS PRO 50VA 1 Output Chassis Mounting Transformer, 230V ac, IEC 61558-2-4 | RS

uk.rs-online.com

That completely does away with the "connecting to a live circuit" hazard, as the secondary can safely be connected with the low voltage electronics.

Back to your non-isolated, direct sense circuit here, in other words - but on a 240V isolated supply from such a transformer;
the low voltage side can then be earthed for safety & the sense signal can be taken directly from the shunt resistor:

 

[KeepItSimpleStupid]

Another idea. I'm just throwing them out at you.

A power OP amp from say Apex Microtechnology or build a high voltage stage for a standard OP amp.
If you need multiple isolated outputs, use custom transformers, but use the primary to indirectly measure the secondary current. Initially, you can wind your own on some toroids.

Use an I-V converter or feedback ammeter to measure current. Really it;s good for +-10mA or smaller.
I made one that was 2 or 4-terminal and had a +-100mA output with +-5V bias and +-50mA suppression. There were 4 ranges with +-10V output. A feedback ammeter gives you a voltage drop of a few mV.

You didn't answer the question if it's a:

1) One of (research purposes only)

You could work on other control methods later.

Another idea is to use a current source across each resistor?

Could you even multiplex the measurements to get R?

I'm just throwing out stuff. You can do that initially in the "brainstorming" session. It's actually a good way to start.

 

[ramuna]

rjenkinsgb said:
Run the whole circuit through an isolation transformer.
...
That completely does away with the "connecting to a live circuit" hazard, as the secondary can safely be connected with the low voltage electronics.

I'm sorry, but I still don't get your drift. The whole point of isolation from the mains is to provide operator safety. Hence the opto-triac which isolates the high voltage triac from low voltage control circuitry operated by the user. An isolation transformer reproduces on its secondary winding the same high voltage which is present in it primary (the 240V rms mains voltage). Thus the secondary voltage is as dangerous to the operator as the primary. Nor can this secondary voltage be directly applied to the low voltage control circuitry. It would need a step-down transformer, rectification & voltage stabilisation. So a 'like-for-like' high voltage isolation transformer is of no use in this application.

KeepItSimpleStupid said

You didn't answer the question if it's a:

1) One of (research purposes only)

I did KISS. Please read post #21.
My favourite solution, would have been to use the TMCS1100A4 as suggested by Nigel in post #8. unfortunately, this part is not available from my regular suppliers RS & Rapid. Yes, I can buy it from Mouser or from Digikey at quite a reasonable price (around £4.20, including VAT). But the delivery cost from these US suppliers is £14.40 (including VAT). Both RS & Rapid have much lower delivery costs. And since I have an account with RS, I get free delivery. So, the all-in cost to me for a TMCS1100A4 is £18.60, which I find too expensive compared to the alternative back-to-back transformer solution.

Another idea is to use a current source across each resistor?
Do you mean temporarily remove the mains (say with a relay or with a switched off triac) and instead place a DC voltage source across the load resistance? Please again refer to post #21 (the second part of my reply), where I mention that applying DC across the load will decompose it, and give an erroneous resistance reading

Could you even multiplex the measurements to get R?
Please see my previous answers.

I'm just throwing out stuff. You can do that initially in the "brainstorming" session. It's actually a good way to start.
I agree with your views on brainstorming. But I also subscribe to 'Design Freeze' - that at some point one has to bite the bullet, commit to the best design to date, order the parts and start assembly.

 

[rjenkinsgb]

Nor can this secondary voltage be directly applied to the low voltage control circuitry. It would need a step-down transformer, rectification & voltage stabilisation.

But the current sense resistor CAN be, as you can connect the floating 240V to the electronics ground, which you cannot do with direct mains.

It does away with any requirement to try and isolate and couple the current sense signal, which seemed to be the main problem earlier in the thread.

 

[Nigel Goodwin]

rjenkinsgb said:
Run the whole circuit through an isolation transformer.
...
That completely does away with the "connecting to a live circuit" hazard, as the secondary can safely be connected with the low voltage electronics.

I'm sorry, but I still don't get your drift. The whole point of isolation from the mains is to provide operator safety. Hence the opto-triac which isolates the high voltage triac from low voltage control circuitry operated by the user. An isolation transformer reproduces on its secondary winding the same high voltage which is present in it primary (the 240V rms mains voltage). Thus the secondary voltage is as dangerous to the operator as the primary. Nor can this secondary voltage be directly applied to the low voltage control circuitry. It would need a step-down transformer, rectification & voltage stabilisation. So a 'like-for-like' high voltage isolation transformer is of no use in this application.

Your entire premise is wrong - the whole point of 'isolation' is to break the earth connection from the mains (nothing to do with lower voltages).

The incoming neutral is earthed at the sub-station, thus you can get a shock from touching just the live wire, through your body, down to ground, through the ground, and back to neutral at the substation.

Fitting a mains isolation transformer on the incoming supply breaks this earth connection, you don't have live and neutral any more, just two wires with mains between them. Obviously you can stil get a full mains shock by touching both mains wires, but you can on your opto-isolated design just the same.

In either case, the point is you can't get a live mains shock from a single body connection.

The rest of the circuit can stay pretty well the same, an opto-isolator (particularly a tric one like you're using) is still a good idea, just not for it's isolation purposes.

This is really basic electrics 101 - and it worries me somewhat that you're playing with mains without knowing the absolute basics.

 

[ramuna]

Nigel Goodwin said:
Your entire premise is wrong - the whole point of 'isolation' is to break the earth connection from the mains (nothing to do with lower voltages).

The incoming neutral is earthed at the sub-station, thus you can get a shock from touching just the live wire, through your body, down to ground, through the ground, and back to neutral at the substation.

Fitting a mains isolation transformer on the incoming supply breaks this earth connection, you don't have live and neutral any more, just two wires with mains between them. Obviously you can stil get a full mains shock by touching both mains wires, but you can on your opto-isolated design just the same.
...
This is really basic electrics 101 - and it worries me somewhat that you're playing with mains without knowing the absolute basics.

I admit that my knowledge of and experience with electronics to date, has been concerned with RF, SMPS, digital & analogue circuits all of which are powered by lower voltages than the mains.

This is my first mains project. But it has to be done. There has to be a beginning. If not now, then when?

I have designed & built Tesla coils & voltage multipliers which produce upto 15 kV, all of which incorporate the required safeguards, BUT these too are powered with 24V sources. So I am not a stranger to EHT and its precautions/dangers. But this project is my first direct mains powered one.

Obviously you can stil get a full mains shock by touching both mains wires...

This is the main danger in this project, which is why I am housing the load resistance in an stainless steel mesh enclosure whose door has a failsafe interlock switch. The electrodes connected to the resistance being the main point of high voltage electrical contact with the operator.
The control circuitry is housed in an earthed metal enclosure located outside the load resistor housing.

I had forgotten about the neutral being earthed at the substation - as knowledge not used is knowledge lost. Thank you for reminding me.

 

[Nigel Goodwin]

Nigel Goodwin said:
Your entire premise is wrong - the whole point of 'isolation' is to break the earth connection from the mains (nothing to do with lower voltages).

The incoming neutral is earthed at the sub-station, thus you can get a shock from touching just the live wire, through your body, down to ground, through the ground, and back to neutral at the substation.

Fitting a mains isolation transformer on the incoming supply breaks this earth connection, you don't have live and neutral any more, just two wires with mains between them. Obviously you can stil get a full mains shock by touching both mains wires, but you can on your opto-isolated design just the same.
...
This is really basic electrics 101 - and it worries me somewhat that you're playing with mains without knowing the absolute basics.

I admit that my knowledge of and experience with electronics to date, has been concerned with RF, SMPS, digital & analogue circuits all of which are powered by lower voltages than the mains.

This is my first mains project. But it has to be done. There has to be a beginning. If not now, then when?

I have designed & built Tesla coils & voltage multipliers which produce upto 15 kV, all of which incorporate the required safeguards, BUT these too are powered with 24V sources. So I am not a stranger to EHT and its precautions/dangers. But this project is my first direct mains powered one.

Obviously you can stil get a full mains shock by touching both mains wires...

This is the main danger in this project, which is why I am housing the load resistance in an stainless steel mesh enclosure whose door has a failsafe interlock switch. The electrodes connected to the resistance being the main point of high voltage electrical contact with the operator.
The control circuitry is housed in an earthed metal enclosure located outside the load resistor housing.

I had forgotten about the neutral being earthed at the substation - as knowledge not used is knowledge lost. Thank you for reminding me.

Bear in mind that earthed metal enclosures are a further danger point, NOT a safety point, under many circumstances - you need to understand why earthing 'can' be a good idea, and why it can be a very bad idea. With a mains isolation transformer it's generally a very bad idea, as the whole point is to remove dangerous earths.

 

[ramuna]

Nigel Goodwin said:
...you need to understand why earthing 'can' be a good idea, and why it can be a very bad idea. With a mains isolation transformer it's generally a very bad idea, as the whole point is to remove dangerous earths.

Nigel,

a) I will not be using an isolation transformer. Rather I will be using the opto-triac and the back to back transformer design which we discussed earlier.

b) The mains outlet which I will be using for this application, is controlled by a RCD (Residual Current Detector), aka Earth Leakage Circuit Breaker.

c) correct me if I am wrong, but would not 'Earth' be 'floating' with respect to either of the two terminals of the secondary of the isolation transformer and consequently there would be no advantage in earthing the secondary voltage powered apparatus? Furthermore, is it not the whole point of the isolation transformer, as you reminded me, that you can hold an earthed wire in one hand and a wire connected to either of the two isolation transformer secondary terminals in the other hand and no current would flow since Earth is floating w.r.t. the secondary terminals?

 

[Nigel Goodwin]

Nigel Goodwin said:
...you need to understand why earthing 'can' be a good idea, and why it can be a very bad idea. With a mains isolation transformer it's generally a very bad idea, as the whole point is to remove dangerous earths.

Nigel,

a) I will not be using an isolation transformer. Rather I will be using the opto-triac and the back to back transformer design which we discussed earlier.

b) The mains outlet which I will be using for this application, is controlled by a RCD (Residual Current Detector), aka Earth Leakage Circuit Breaker.

c) correct me if I am wrong, but would not 'Earth' be 'floating' with respect to either of the two terminals of the secondary of the isolation transformer and consequently there would be no advantage in earthing the secondary voltage powered apparatus? Furthermore, is it not the whole point of the isolation transformer, as you reminded me, that you can hold an earthed wire in one hand and a wire connected to either of the two isolation transformer secondary terminals in the other hand and no current would flow since Earth is floating w.r.t. the secondary terminals?

Yes, but if you have an earthed metal case, then there's a risk of a short between that case and one of the mains wires, which brings you back to the normal live/neutral thing and shock hazard to ground.

It's just an added risk, which is one reason why most type II (non-earthed) appliances don't have metal bodies, and certainly not earthed metal bodies.

By all means, try the back to back transformers, and see how it works - essentially it's a simple, cheap and easy, method to give a go.

 

[ramuna]

Nigel Goodwin said:
Yes, but if you have an earthed metal case, then there's a risk of a short between that case and one of the mains wires, which brings you back to the normal live/neutral thing and shock hazard to ground.

Yes, I see your point - you saw a couple of moves ahead of me, touche. I am now torn between the Jenkins isolation transformer route and the Goodwin back-to-back transformer route.

 

[whit3rd]

Good point Schmitt.

I've been thinking about this and have come up with a hypothetical solution, which I want to run past you.

Please take a look at the diagram attached to this post. On the left of the diagram is the original back-to-back transformer proposal (split secondaries, only one of which is being used).

You seem to be circling the 'current transformer' concept, in which a sense resistor is
transformer-coupled in series with the high voltage source. If the impedance of the
resistor divided by the turns ratio is low (which is easy to accomplish) it acts as
an accurate AC current-to-isolated voltage element.

Commercial current transformers are calibrated for a particular load resistor value, typically
with single-turn primaries (i.e. thread one strand of the AC wire through the hole).
Clamp-on meters work this way (but you probably want to run the waveform to
an oscilloscope rather than an average-reading meter).

Here's a typical one
clamp-on current transformer