This is going to sound crazy, but what about a logic level MOSFET? A 1V signal won't reach the official threshold voltage, but many of them will pass some small amount of current all the way down to Vgs values well below their threshold.

Not crazy but not workable.
The MOSFET Vgs(th) is the point at at where it just starts to conduct a tenths of a mA of current.
Slightly below the Vgs threshold voltage essentially no current flows.

 

Not crazy but not workable.
The MOSFET Vgs(th) is the point at at where it just starts to conduct a tenths of a mA of current.
Slightly below the Vgs threshold voltage essentially no current flows.

I must've misread (or perhaps just mis-remembered) some datasheets. Thanks for the clarification on that. That certainly explains the complete and utter failures I had with at least a dozen versions of my idea in LTspice!

 

I like the trim pot idea, and I also don't like it. The flexibility in the current / voltage coming from the CT is good, but can that be achieved within the circuit without an adjustable component that requires user input? Something perhaps that just has a wider operational range and can work with the lower amount of power and still work with a slightly higher amount of power?

 

I like the trim pot idea, and I also don't like it. The flexibility in the current / voltage coming from the CT is good, but can that be achieved within the circuit without an adjustable component that requires user input? Something perhaps that just has a wider operational range and can work with the lower amount of power and still work with a slightly higher amount of power?

Of course - it's just hard to do that up front without some real world measurements.

The really easy way (in terms of being lazy about math) is to build a temporary version of the circuit with a trim pot, find the level that works, and then replace the trim pot with two fixed resistors.

The other approach would be to install the transformer and probably also the filter portion, then measure voltages. You'd want to see what the output looks like when the amp is off (how much noise is on the signal.) The threshold obviously has to be safely above this point. You'd also want to get voltage readings with the amp on, trying a variety of conditions. See how much the output varies with the amp on, but idling, vs. working hard cranked up to 11! You would then choose your two fixed resistors to provide a reference voltage in the middle of your highest off voltage and your lowest on voltage.

 

Hmph. I'm starting to doubt myself, even more than usual. I think I may have created one of those simulations of ideal parts that can't exist in the real world.

It occurred to me when looking back at my circuit that if the signal from the transformer is 200mV, it shouldn't overcome the forward voltage of the diode. Digging a little deeper, I realized that I tricked myself with the nature of my transformer simulation. If I disconnect my filter parts and just read the voltage across R2, it's peaking around 800mV. That's cause I have no idea how to simulate transformers accurately, so I just fiddled with numbers until I got the behavior I expected, ignoring the effect of my filter circuit loading the transformer.

So here's a million dollar question that gets to the heart of how I should simulate this circuit, and also what the possible fixes are for this circuit: What is the nature of the output from this current transformer? Does it behave like a voltage source, a current source, a power source? If I dramatically increase the resistance of R2, should I expect a proportional increase in voltage across it, a proportional decrease in current through it, or some complicated combination of both?

I'm going to go off into Google and the rest of the inter-webs and see if I can get my head around some of this. Until then, I think it's safe to assume my last circuit was built on lies and bad assumptions. Sorry about that!

 

Hmmm... if I'm reading things right, it looks like the output of this transformer can be modeled as a current source whose maximum voltage output is limited by zener diodes built into the output for protection purposes (the real transformer's output voltage is also limited by magnetic properties beyond my comprehension, but for the sake of simulation I think I can just use zeners to limit it?)

I'll do a little more reading and then try a new circuit model sometime this weekend.

Sorry I've gone off the rails here on my voyage of discovery. Someone else who knows what they're doing should save the OP from me and my ramblings!

 

ebeowulf17, are you available for a brief phone conversation? I'd love to ask you a few questions!

Mark
305-775-3020

 

ebeowulf17, are you available for a brief phone conversation? I'd love to ask you a few questions!

Mark
305-775-3020

Sorry, it's not a great time right now, plus I have profound hearing loss which makes phone calls excruciatingly difficult, so in a way it's never a great time. I think this forum has a messaging system though, if there are questions you want to keep semi-private for whatever reason. I haven't used it before, but I believe if you click on my profile and then click "start a conversation" we can exchange messages that way.

Having said that, if they're electronics questions, they should stay on the open forum so that others can learn too... even if it's just learning from our bone-headed mistakes so they don't have to repeat them!

 

Indeed! fair point.

I wonder, if I were to purchase 2 or 3 different Current Transformers, and take some readings with a multimeter, would that be helpful real world data to help these ideas move along a little?

 

Indeed! fair point.

I wonder, if I were to purchase 2 or 3 different Current Transformers, and take some readings with a multimeter, would that be helpful real world data to help these ideas move along a little?

Definitely! You'd also need to get burden resistor, or a variety of them to work with. In order to get the most signal with the least effort, I think you want a current transformer that's optimized for lower amperage ranges. If your amp only draws around 3A, then we're using a tiny fraction of the potential of a 120A transformer. If there was one optimized for maybe 5A max range, we'd probably get a more suitable output.

On the other hand, I think even with a transformer optimized for 120A, if we just use a larger burden resistor, we can get a higher voltage output from the same transformer output. As you can tell, I'm still learning on all this.

Regardless of all my speculation, getting some readings from one or more real transformers in the real world is a great first step - just be sure to grab suitable resistors too.

 

I did an hour meter for a light source. I found a battery powered hour meter that accepted an Open collector or contact closure input. So, a photo-transistor works fine. I don't remember the model of the hour meter.
Most of them are so cheap, that you have to replace them when the battery dies with a typical 10 year life. It took some searching. Some had reset options available.

A quick search revealed this http://www.crmagnetics.com/remote/cr2550

Unfortunately, high efficiency LED doesn't tell you much, But it could be, say 2 mA.

I suspect, YOU SHOULD, be able to drive an OPTOMOS relay, optocoupler or FET optocoupler. 2 mA is feasible for the optomos. The OPTOMOS "relay" is an IC.

You can increase the sensitivity by adding more turns around the core. Remember, it has to be one of the wires, not both that goes through the core.

 

You can increase the sensitivity by adding more turns around the core.

I spaced out and forgot about that. Regardless of which circuit or method is used, that could be a useful technique.

 

On that website, there's a couple of other "more expansive" parts on that website.

e.g. http://www.crmagnetics.com/switches/ac-current/ac-voltage/cr9321

and

http://www.crmagnetics.com/relays/cr4395

Don;t forget to look at the catalog pages.

Sometimes you have to add a bit of labor or say a small PCB for the optomos relay or even DIP to terminals, so the effective cost goes up.

Not sure if, your going to build a box that has power in/out, wire separation etc or are going to try to "build it in" to the amp.

I might look for an hour meter.

 

Something like this http://www.redlion.net/sites/defaul...ference only - see CUB7T for new designs).pdf one, but it appears to obsolete, but available from digikey.

 

Replacement: http://www.redlion.net/product/cub7t-8-digit-timer-low-voltage-input-reflective-display

 

Thank you KeepIsSimpleStupid,

I have indeed looked at Red Lion Controls. As I am trying to keep the total price down, their controls are a little on the pricey side. However, you inspire another way of looking for a solution. If I can find the right CT and the right hour counter for the job, perhaps the correct combo will allow for a simpler circuit design and more flexibility in the types of amplifiers that I can connect to and use to trigger the device.

The problem I see with the red lion item that you sent a link to is the life span, Just 7 years on an internal battery. I am trying to create a device that will last longer than that - so perhaps a replaceable battery in the timer or an ultra low power requirement from the timer so I can run the entire device on a set of 4xAA batteries for a good long time.

I think my next step is to research the timers more, and compare the different trigger voltages and also the operating power consumptions. I also have to compare the CTs that are on the table for this project. I will assemble a spreadsheet to compare the specs, and also buy a few CTs and test them.

Can I ask for some advice? I will measure CTs with the multimeter. You mentioned some components to add to that test. to keep it simple, can I test these with out a PCB? Can I just strip the wires on the CT, and connect to a component that you mentioned above, and then measure that with the Multi meter?

At a minimum, which component(s) should I attach to the CT for these measurements?

thanks!!

Mark

 

here's https://www.digikey.com/product-det...rcuits-division/LBA110STR/CLA240CT-ND/1212851 the device I didn't link to. Note, the LED only takes about 2 mA making it a high efficiency LED.

The general problem you have, is that most multimeter's measure the voltage across a fixed resistor. You need to look at the specs of your multimeter and see what kind of voltage drop you would get when currents are on the order of 1-10 mA. So with that unknown cheap part, I suggested, we have idea what I is or what Vf is. Vf is the easy parameter to measure. There has to be a resistor somewhere, I would think. That's unknown too.

Now, you can measure Vf.

Naked CT's are wierd. They generate high voltages when they don't have a load, thus they usually have a parallel resistance across them. Then they might have a series resistor with the LED. A call to the MFR might help here.

You can build yourself a little zero resistance ammeter, but 20 mA is near the limit of a standard OP-amp.
I've built a few. one used two 9V batteries, 2 bypass caps and a resistor to scale the current. e.g. V=-I*Rf
And a polyester capacitor to limit bandwidth. This device would essentially have a drop of a few mV.

The classic I-V converter is shown here: https://www.passdiy.com/project/other/zen-i-v-converter You have to sprinkle a few other parts and choose an OP amp with a low Vos. Single supply is also possible. Dual 9v batteries work.

I have a picoammeter that goes to 2 mA full-scale and I built a $5000.00 USD instrument for in-house work use that could go to 1 Amp. It had lots more capabilities and was 4-terminal and biasable.

Vf of a LED has a range and a max value depending on color.

Once you know the real current and the voltage drop of the OEM LED and the current needed for the OPTOMOS relay and Vf of the OPTOMOS relay you can size an additional resistor, if needed.

it's possible that the resistor is at the LED under a piece of heat shrink.

You could also try a 1 ohm resistor, so V = 1e-3*1; so you get 1 mV per mA. Not great, but likely better than the DVM.

Generally, you can use you DVM in diode mode to measure the Vf of any LED.

In terms of the OPTOMOS relay, the one I picked has an NC/NO contact AND the datasheet for the timer suggest that longer battery life can be achieved if the input is in one state vs the other most of the time.

Now, it doesn't mean you can;t replace the battery in the timer. It just might be very very difficult. It might also be easier to clip wires and make the battery external.

I have toothbrushes, a Phillips Airfloss that you pay >$50.00 for and they have non-replaceable batteries.
I had a cordless shaver that really did make it impossible to replace the battery.

I do have an old hour meter at home, bit I've never taken it apart to see what the battery is like. It's like 37 years old and dead.

I have an insulin pen with a non-replaceable battery with a 5 year life supposedly.

I have a PERS (Personal Medical Response Transmitter) with two CR2032 batteries inside a case that's totaly sealed. it's the only way to male it totally water tight.

Your AA cells is probably NOT a good idea. Lithium cells have a flat discharge curve, so they maintain 3V nearly to the end of it's life.

I get 20 CR2032 (Sony or Maxell) batteries for around $20.00 postage paid, not at the $5.00 to $6.00 local price.
Digikey sells them for less if i want to pay postage. If I order directly from a local www.mscdirect.com, I can avoid postage too. For that convenience, I may have to wait a week. The local office stocks certain items for local businesses. So, currently, I stock CR2032's and LR44;s and SR1120's?.

I don't know exactly what your up to.

What is it about $10/year?

The OTOMOS relay is bit expensive, but does have a high efficiency LED.

 

Back around post #17 I suggested the same link which KISS later suggested. That being the small Current Transformer from CR Magnetics. I have used literally hundreds of them in industrial designs and they work very well. They also allow you to use multiple turns through the doughnut for increased sensitivity. In keeping it simple you could either cut the LED off or use the LED by placing it in a drinking straw maybe 1/4" with black heat shrink over it and place a simple photo transistor on the other end making your own Opto-Coupler. Use it to drive any battery powered hour meter like what KISS mentioned. There isn't much to it. The CR Magnetics units run as little as $10 from some distributors. This way your have your Current Transformer in one neat little package.

As mentioned, a Current Transformer works with a burden resistor and there are formulas for calculating the end result. Also as mentioned a current transformer, open circuit, can have a very high voltage at the output. You may want to keep things simple.

Ron

 

I know I said I'd have some updated circuit ideas this weekend, but I got distracted - we took a road trip to see the total eclipse:

It was an amazing experience!

Anyway, back to work tomorrow, but I'll give this a little more thought. I'm wondering if a few extra turns and a higher resistance burden resistor might not get us back to a simple, battery free solution with the CT directly providing the signal. At least one of the counters you linked would accept AC in at low voltage with essentially no current demand. I think we might be able to do that directly after all, but I may be getting overly optimistic.

Both of the other two responding to this thread have loads more experience than me, so you're probably better off with their suggestion anyway, but I'm still curious to see what I can figure out.

 

There's always mechanical. https://www.zoro.com/honeywell-ac-hour-meter-2-screw-24vac-28060/i/G5151151/ 24 VAC input.