Improved DRSSTC Current Monitor Circuit
HV Enthusiast, Fri Jun 09 2006, 05:09PMHere is the first revision release of my improved DRSSTC Current Monitor Circuit.
Features:
- Bar Graph Meter
- Range up to 1000A
- Uses inexpensive commercial current transformer (or wind your own)
- Real time viewing of peak currents within the primary circuit
- Active current limiting
- Maximum current setpoint indicator
The complete document and design can be downloaded here:
http://www.easternvoltageresearch.com/currentmonitor02.pdf (4.6MB PDF)
Dan
Re: Improved DRSSTC Current Monitor Circuit
Reaching, Sat Jun 10 2006, 10:36AM
wow dan, nice circuit and a very good idea with the bargraph display, exactly what i was searching for to give my teslathon drsstc something special.
but why these horrible expensive comparators?
i´ll try to see if i can build up this circuit propely with lm393 cheapo comps or so, for my needs it dont have to be 0,1A exact.
Reaching, Sat Jun 10 2006, 10:36AM
wow dan, nice circuit and a very good idea with the bargraph display, exactly what i was searching for to give my teslathon drsstc something special.
but why these horrible expensive comparators?
i´ll try to see if i can build up this circuit propely with lm393 cheapo comps or so, for my needs it dont have to be 0,1A exact.
Re: Improved DRSSTC Current Monitor Circuit
HV Enthusiast, Sat Jun 10 2006, 12:17PM
Reaching,
The only comparator used in this circuit is the LM311 which is a generic comparator and pretty inexpensive. I assume you are talking about the LT1630 op-amps.
Yes, i forgot these were quite expensive. These were originally chosen for my prototype because the op-amp needed to be a true rail-to-rail input/output since i was dealing with small signals, but for this particular design, a cheapo LM741 will do the trick just fine, although be a bit slower and not work as well when ran near one of the rails.
But, yes, the LM741 or similar generic op-amp that is capable of running single-ended should be fine.
The only thing to really worry about when choosing your op-amp is how far off the supply rails it can operate. Try to choose something that can operate as close to the rails as possible for maximum performance, especially for low currents.
I'll update my datasheet to include the cheaper op-amps later today. Thanks for noting that.
Please note that I am releasing a whole series of downloadable DRSSTC papers and design guides which will be added regularly to the Amateur Design Resource section of our website. Check back often!!!
HV Enthusiast, Sat Jun 10 2006, 12:17PM
Reaching wrote ...
but why these horrible expensive comparators?
but why these horrible expensive comparators?
Reaching,
The only comparator used in this circuit is the LM311 which is a generic comparator and pretty inexpensive. I assume you are talking about the LT1630 op-amps.
Yes, i forgot these were quite expensive. These were originally chosen for my prototype because the op-amp needed to be a true rail-to-rail input/output since i was dealing with small signals, but for this particular design, a cheapo LM741 will do the trick just fine, although be a bit slower and not work as well when ran near one of the rails.
But, yes, the LM741 or similar generic op-amp that is capable of running single-ended should be fine.
The only thing to really worry about when choosing your op-amp is how far off the supply rails it can operate. Try to choose something that can operate as close to the rails as possible for maximum performance, especially for low currents.
I'll update my datasheet to include the cheaper op-amps later today. Thanks for noting that.
Please note that I am releasing a whole series of downloadable DRSSTC papers and design guides which will be added regularly to the Amateur Design Resource section of our website. Check back often!!!
Re: Improved DRSSTC Current Monitor Circuit
Steve Conner, Sat Jun 10 2006, 05:42PM
If you don't like expensive op-amps, my current monitor/limiter does all that Dan's does using a TL074 and half of a LM339.
The only drawback of my circuit is that it's not at all obvious how it works, but trust me it does read and limit the actual peak primary current accurately. Right down to zero amps, too. 
It doesn't have the "view setpoint" feature, but you can add that with one diode and one SPST push button. (again I leave this as an exercise for the reader :P)
If you're looking for other rail-to-rail op-amps to substitute in Dan's circuit, try the LMC6484 or LMC660.
Steve Conner, Sat Jun 10 2006, 05:42PM
If you don't like expensive op-amps, my current monitor/limiter does all that Dan's does using a TL074 and half of a LM339.
The only drawback of my circuit is that it's not at all obvious how it works, but trust me it does read and limit the actual peak primary current accurately. Right down to zero amps, too. 
It doesn't have the "view setpoint" feature, but you can add that with one diode and one SPST push button. (again I leave this as an exercise for the reader :P)
If you're looking for other rail-to-rail op-amps to substitute in Dan's circuit, try the LMC6484 or LMC660.
Re: Improved DRSSTC Current Monitor Circuit
Marko, Sat Jun 10 2006, 05:55PM
Huh?
Nice addon for a DRSSTC.
I wonder if anything more simple than that can be done, like LM3914 (for LED's) with rectifier and
burden (like OCD) with some passive calibration?
Probably circuits won't be hard to build but they end to have bigger number of conponents than entire coil (steve ward's at least )
Marko, Sat Jun 10 2006, 05:55PM
The only drawback of my circuit is that it's not at all obvious how it works
Huh?
Nice addon for a DRSSTC.
I wonder if anything more simple than that can be done, like LM3914 (for LED's) with rectifier and
burden (like OCD) with some passive calibration?
Probably circuits won't be hard to build but they end to have bigger number of conponents than entire coil (steve ward's at least
)Re: Improved DRSSTC Current Monitor Circuit
Reaching, Sat Jun 10 2006, 06:39PM
im still on breadboarding this circuit and the first part works fine, but the lm741, lm3914 story wont work, dont know why,too many resistors . now im looking for a way to simple connect the lm3914 to view the primary current.
okay, i got it, but handle with care, its still a breadboard design, but works fine. the only thing you cant do with it is to view the overcurrent setting on the bar graph, but you can connect a digital voltmeter on U Ref to view the setting
Reaching, Sat Jun 10 2006, 06:39PM
im still on breadboarding this circuit and the first part works fine, but the lm741, lm3914 story wont work, dont know why,too many resistors . now im looking for a way to simple connect the lm3914 to view the primary current.
okay, i got it, but handle with care, its still a breadboard design, but works fine. the only thing you cant do with it is to view the overcurrent setting on the bar graph, but you can connect a digital voltmeter on U Ref to view the setting
Re: Improved DRSSTC Current Monitor Circuit
Marko, Sat Jun 10 2006, 08:11PM
Exactly what I tought of.
You could maybe make R7 and R8 variable for calibrating/offseting the meter (just for led's this seems fair), and voltage on burden depends on current directly.
Maybe you can be evil and use 3914 led output itself as easily readable overcurrent protection, altough it won't be fine-tunable and practical as comparator.
I wonder what was the reason for more-complex design?
Marko, Sat Jun 10 2006, 08:11PM
okay, i got it, but handle with care, its still a breadboard design, but works fine
Exactly what I tought of.
You could maybe make R7 and R8 variable for calibrating/offseting the meter (just for led's this seems fair), and voltage on burden depends on current directly.
Maybe you can be evil and use 3914 led output itself as easily readable overcurrent protection, altough it won't be fine-tunable and practical as comparator.
I wonder what was the reason for more-complex design?
Re: Improved DRSSTC Current Monitor Circuit
HV Enthusiast, Sat Jun 10 2006, 09:10PM
Sure. If all you want is an LED current meter, a single diode, cap, and an LM3914 IC is all you need. However, for this particular design, two separate filters are required for the current sample. One is sized to reset every pulse burst, while the LED display is designed to hold for a second or two (orders of hundreds of bursts). Remember, if you use only one RC filter (for hold-up) and size it too large, the current limiting will be too slow to work properly.
You'll need to correct for the diode drop though, unless your full range current voltage is large enough that the Vfwd of the diode is neglectable.
Firkragg,
If you compare just the active current limiting portion circuits of Steve Wards vs. this circuit, Steve Ward actually uses more components. Neither circuit is better or worse, they both accomplish the same exact thing.
Sounds like you have an error in your breadboard circuit. I'm using the same circuit in my DRSSTC IV system now and it works perfectly (setpoint indicator)
This does work, but given Reaching's circuit as is, it will only provide a true reading during 100% duty cycle. The 100us time constant at the input to the comparator is much too slow provide proper hold-up between bursts, especially at lower PRFs. I use about a 5-10 sec time constant on my RC filter (10uF / 1Meg) for the LED display.
As stated before, you really need two separate filters for proper operation.
1. If you use a large time constant (enough to provide proper hold-up for the display to provide accurate readings over entire PRF range), then it will be TOO SLOW for current limiting. A few hundred bursts could go by with a 5-10 sec time constant filter before the current limiting kicks in.
2. If the time constant is set to very short, the active current limiting will react fast enough, but the hold-up time will be too short to maintain a proper level (again at entire PRF range) to be accurate on the LED display meter.
Hence the need for the buffer op-amp circuits in there.
They act more as buffers, than scaling / offset.
That is the reason for the more-complex design.
Of course, you could buffer with diodes as well (come up with a pair of diodes) into two separate circuits.
BTW, if you are afraid of a few extra components in your DRSSTC design, then you should probably stick to traditional spark gap coils!!!
HV Enthusiast, Sat Jun 10 2006, 09:10PM
wrote ...
I wonder if anything more simple than that can be done, like LM3914 (for LED's) with rectifier and
burden (like OCD) with some passive calibration?
I wonder if anything more simple than that can be done, like LM3914 (for LED's) with rectifier and
burden (like OCD) with some passive calibration?
Sure. If all you want is an LED current meter, a single diode, cap, and an LM3914 IC is all you need. However, for this particular design, two separate filters are required for the current sample. One is sized to reset every pulse burst, while the LED display is designed to hold for a second or two (orders of hundreds of bursts). Remember, if you use only one RC filter (for hold-up) and size it too large, the current limiting will be too slow to work properly.
wrote ...
It doesn't have the "view setpoint" feature, but you can add that with one diode and one SPST push button.
It doesn't have the "view setpoint" feature, but you can add that with one diode and one SPST push button.
You'll need to correct for the diode drop though, unless your full range current voltage is large enough that the Vfwd of the diode is neglectable.
wrote ...
Probably circuits won't be hard to build but they end to have bigger number of conponents than entire coil (steve ward's at least
Probably circuits won't be hard to build but they end to have bigger number of conponents than entire coil (steve ward's at least
Firkragg,
If you compare just the active current limiting portion circuits of Steve Wards vs. this circuit, Steve Ward actually uses more components. Neither circuit is better or worse, they both accomplish the same exact thing.
wrote ...
okay, i got it, but handle with care, its still a breadboard design, but works fine. the only thing you cant do with it is to view the overcurrent setting on the bar graph, but you can connect a digital voltmeter on U Ref to view the setting
okay, i got it, but handle with care, its still a breadboard design, but works fine. the only thing you cant do with it is to view the overcurrent setting on the bar graph, but you can connect a digital voltmeter on U Ref to view the setting
Sounds like you have an error in your breadboard circuit. I'm using the same circuit in my DRSSTC IV system now and it works perfectly (setpoint indicator)
wrote ...
Exactly what I tought of.
You could maybe make R7 and R8 variable for calibrating/offseting the meter (just for led's this seems fair), and voltage on burden depends on current directly.
Exactly what I tought of.
You could maybe make R7 and R8 variable for calibrating/offseting the meter (just for led's this seems fair), and voltage on burden depends on current directly.
This does work, but given Reaching's circuit as is, it will only provide a true reading during 100% duty cycle. The 100us time constant at the input to the comparator is much too slow provide proper hold-up between bursts, especially at lower PRFs. I use about a 5-10 sec time constant on my RC filter (10uF / 1Meg) for the LED display.
As stated before, you really need two separate filters for proper operation.
1. If you use a large time constant (enough to provide proper hold-up for the display to provide accurate readings over entire PRF range), then it will be TOO SLOW for current limiting. A few hundred bursts could go by with a 5-10 sec time constant filter before the current limiting kicks in.
2. If the time constant is set to very short, the active current limiting will react fast enough, but the hold-up time will be too short to maintain a proper level (again at entire PRF range) to be accurate on the LED display meter.
Hence the need for the buffer op-amp circuits in there.
They act more as buffers, than scaling / offset.
That is the reason for the more-complex design.
Of course, you could buffer with diodes as well (come up with a pair of diodes) into two separate circuits.
BTW, if you are afraid of a few extra components in your DRSSTC design, then you should probably stick to traditional spark gap coils!!!
Re: Improved DRSSTC Current Monitor Circuit
Marko, Sat Jun 10 2006, 09:29PM
I actually more tought for steve conner's current sense circuit, and I never want complicating things that aren't complicated. Your one isn't that bad really.
Marko, Sat Jun 10 2006, 09:29PM
BTW, if you are afraid of a few extra components in your DRSSTC design, then you should probably stick to traditional spark gap coils!!!
I actually more tought for steve conner's current sense circuit, and I never want complicating things that aren't complicated. Your one isn't that bad really.
Re: Improved DRSSTC Current Monitor Circuit
HV Enthusiast, Sat Jun 10 2006, 09:38PM
Firkragg,
The current limit circuit i use is about as simple as they get. Its no more complicated than anyone elses current limit circuit including both the Steve's. Just a current transformer, rectifier, and comparator. Heck, its actually less complicated than Steve Conners as he is using (2) active components.
The complex part of the circuit is just for display purposes and has no important function other than lighting up a few LEDs. And as i stated before, you could eliminate those op-amps completely and use just a couple of diodes or'd together. I'll still finishing up this document, so i'll probably add a few "simpler" alternatives as well.
HV Enthusiast, Sat Jun 10 2006, 09:38PM
Firkragg,
The current limit circuit i use is about as simple as they get. Its no more complicated than anyone elses current limit circuit including both the Steve's. Just a current transformer, rectifier, and comparator. Heck, its actually less complicated than Steve Conners as he is using (2) active components.
The complex part of the circuit is just for display purposes and has no important function other than lighting up a few LEDs. And as i stated before, you could eliminate those op-amps completely and use just a couple of diodes or'd together. I'll still finishing up this document, so i'll probably add a few "simpler" alternatives as well.
Re: Improved DRSSTC Current Monitor Circuit
Marko, Sat Jun 10 2006, 09:47PM
This IS the part of circuit i'm talking about. 'thing about lighting few leds', current snse circuit used to provide constant voltage for LM3914 input.
As I figured out OCD is the same for all of us, I didn't say nothing for it.
Marko, Sat Jun 10 2006, 09:47PM
The complex part of the circuit is just for display purposes and has no important function other than lighting up a few LEDs. And as i stated before, you could eliminate those op-amps completely and use just a couple of diodes or'd together. I'll still finishing up this document, so i'll probably add a few "simpler" alternatives as well.
This IS the part of circuit i'm talking about. 'thing about lighting few leds', current snse circuit used to provide constant voltage for LM3914 input.
As I figured out OCD is the same for all of us, I didn't say nothing for it.
Re: Improved DRSSTC Current Monitor Circuit
Reaching, Sat Jun 10 2006, 10:00PM
mhh, why do you think that my circuit will not work properly? i looked at your shematic and copied the important parts, such as the comparator etc which is nearly the same as in your circuit except the shorter pulse on the ocd out on the ne555 to clear the sync flipflop on the driver board. R7 and R8 is a voltage divider by ~8 to provide a exact current reading on the bargraph. for example, if you get 1 volt on the comparator input, you should have 100A primary current, this gives 0,12volts on the input pin 5 of the lm3914 so the first led lights on, cause the voltage range of the lm3914 goes from 0-1,2volts to light all 10 leds.nothing too complex here.
btw dan, what should i change to make it work properly in drsstc duty, on the breadboard it works just fine.
Reaching, Sat Jun 10 2006, 10:00PM
mhh, why do you think that my circuit will not work properly? i looked at your shematic and copied the important parts, such as the comparator etc which is nearly the same as in your circuit except the shorter pulse on the ocd out on the ne555 to clear the sync flipflop on the driver board. R7 and R8 is a voltage divider by ~8 to provide a exact current reading on the bargraph. for example, if you get 1 volt on the comparator input, you should have 100A primary current, this gives 0,12volts on the input pin 5 of the lm3914 so the first led lights on, cause the voltage range of the lm3914 goes from 0-1,2volts to light all 10 leds.nothing too complex here.
btw dan, what should i change to make it work properly in drsstc duty, on the breadboard it works just fine.
Re: Improved DRSSTC Current Monitor Circuit
Marko, Sat Jun 10 2006, 10:10PM
The circuit will work, but leds will undergo 'PWM' effect dependent on interrupter duty cycle.
Maybe a filter cap on input or something should help a bit, but I don't know how good is speed going to be.ž
You can always try-and-see.
Dan's circuit sems to use active filter, current setpoint linked to OCD etc. for a bit higher conponent count.
Marko, Sat Jun 10 2006, 10:10PM
The circuit will work, but leds will undergo 'PWM' effect dependent on interrupter duty cycle.
Maybe a filter cap on input or something should help a bit, but I don't know how good is speed going to be.ž
You can always try-and-see.
Dan's circuit sems to use active filter, current setpoint linked to OCD etc. for a bit higher conponent count.
Re: Improved DRSSTC Current Monitor Circuit
Steve Conner, Sat Jun 10 2006, 10:37PM
This is true. I got round that in a sneaky way though. I do the current limiting with a window comparator working on the unrectified CT signal, so it's as fast as can be since there's no filter whatsoever. So if the current goes over, the drive will be stopped on the very next cycle.
For the meter, I used a precision rectifier with a very fast attack time constant (0.1uF/10 ohm) and a very slow decay (0.1uF/10 Meg ohm). So it grabs the peak current within about 2 cycles and holds it for long enough to give a steady reading above about 10bps. It was designed for an analog meter but would also drive a bargraph fine.
Steve Conner, Sat Jun 10 2006, 10:37PM
wrote ...
1. If you use a large time constant (enough to provide proper hold-up for the display to provide accurate readings over entire PRF range), then it will be TOO SLOW for current limiting. A few hundred bursts could go by with a 5-10 sec time constant filter before the current limiting kicks in.
1. If you use a large time constant (enough to provide proper hold-up for the display to provide accurate readings over entire PRF range), then it will be TOO SLOW for current limiting. A few hundred bursts could go by with a 5-10 sec time constant filter before the current limiting kicks in.
This is true. I got round that in a sneaky way though. I do the current limiting with a window comparator working on the unrectified CT signal, so it's as fast as can be since there's no filter whatsoever. So if the current goes over, the drive will be stopped on the very next cycle.
For the meter, I used a precision rectifier with a very fast attack time constant (0.1uF/10 ohm) and a very slow decay (0.1uF/10 Meg ohm). So it grabs the peak current within about 2 cycles and holds it for long enough to give a steady reading above about 10bps. It was designed for an analog meter but would also drive a bargraph fine.
Re: Improved DRSSTC Current Monitor Circuit
HV Enthusiast, Sat Jun 10 2006, 11:45PM
That method alone gives you at least twice a fast response alone as the conventional rectified comparator circuits.
I was playing with the notion of using a precision rectifier, but decided i didn't need it since signals below 1V (50A to 100A in the coils i've done) really don't mean much as far as DRSSTCs go.
HV Enthusiast, Sat Jun 10 2006, 11:45PM
That method alone gives you at least twice a fast response alone as the conventional rectified comparator circuits.
I was playing with the notion of using a precision rectifier, but decided i didn't need it since signals below 1V (50A to 100A in the coils i've done) really don't mean much as far as DRSSTCs go.
Re: Improved DRSSTC Current Monitor Circuit
Hazmatt_(The Underdog), Sat Jun 10 2006, 11:49PM
Thank you Dan for posting the current limiter.
I have not yet attempted building a SSTC of any variety and am clueless at such high currents. These types of posts and schematics really help me see what these systems should be doing and how to approach problems like HF current.
I hope to have a small SSTC halfbridge going soon. Only trouble is that its not in my budget, nothing is.
So, Ill just take care of my Linear Systems class, sit tight, and try to cease some thunder some day.
Hazmatt_(The Underdog), Sat Jun 10 2006, 11:49PM
Thank you Dan for posting the current limiter.
I have not yet attempted building a SSTC of any variety and am clueless at such high currents. These types of posts and schematics really help me see what these systems should be doing and how to approach problems like HF current.
I hope to have a small SSTC halfbridge going soon. Only trouble is that its not in my budget, nothing is.
So, Ill just take care of my Linear Systems class, sit tight, and try to cease some thunder some day.
Re: Improved DRSSTC Current Monitor Circuit
Steve Conner, Sat Jun 10 2006, 11:57PM
If you'd wanted to build Tesla coils, you should have taken the "Completely Useless But Kind Of Fun Systems" class instead.
Steve Conner, Sat Jun 10 2006, 11:57PM
If you'd wanted to build Tesla coils, you should have taken the "Completely Useless But Kind Of Fun Systems" class instead.
Re: Improved DRSSTC Current Monitor Circuit
HV Enthusiast, Sun Jun 11 2006, 12:04AM
Actually, most of my classes I took at university would fit under the category:
"Completely Useless and DEFINITELY Not Fun"
Heck, the only useful class i took in high school was typing.
HV Enthusiast, Sun Jun 11 2006, 12:04AM
Actually, most of my classes I took at university would fit under the category:
"Completely Useless and DEFINITELY Not Fun"
Heck, the only useful class i took in high school was typing.
Re: Improved DRSSTC Current Monitor Circuit
Hazmatt_(The Underdog), Sun Jun 11 2006, 02:46AM
Steve, I can't just build a TC for a class. In fact, the last class 400D, a design class, wouldn't have made it possible.
You think "Sure you could its design"
Sorry bud. This is a TEAM effort. I wanted to do something as important as power supplies, well, nothing doing. I got thrown into a group to solve their power problem.
Coming up with a viable electric 4 wheel off-road vehicle was good, and I got an A, but I would have rather done a power supply.
The point is, even if you have a design class, its all by group and I was out voted 4:1 and wasn't allowed to do a project by myself.
Bonus: since it was all calculations, it didn't cost me anything.
Hazmatt_(The Underdog), Sun Jun 11 2006, 02:46AM
Steve, I can't just build a TC for a class. In fact, the last class 400D, a design class, wouldn't have made it possible.
You think "Sure you could its design"
Sorry bud. This is a TEAM effort. I wanted to do something as important as power supplies, well, nothing doing. I got thrown into a group to solve their power problem.
Coming up with a viable electric 4 wheel off-road vehicle was good, and I got an A, but I would have rather done a power supply.
The point is, even if you have a design class, its all by group and I was out voted 4:1 and wasn't allowed to do a project by myself.
Bonus: since it was all calculations, it didn't cost me anything.
Print this page