-
Hello,
I recreated circuit from this video () in LTSpice and it doesn't work. And C1 would be destroyed due to overvoltage. But on the video it appears to be working...
Can someone help me fix the schematic and/or explain how this circuit is supposed to work? I know you can buy those, but that's not the point - I'd like to learn.
I'd like to use it to SS a 1000W vacuum.
Thanks in advance
MK
-
Hello,
I recreated circuit from this video () in LTSpice and it doesn't work. And C1 would be destroyed due to overvoltage. But on the video it appears to be working...
Can someone help me fix the schematic and/or explain how this circuit is supposed to work? I know you can buy those, but that's not the point - I'd like to learn.
I'd like to use it to SS a 1000W vacuum.
Thanks in advance
MK
Hi,
Where does that ground connect to, was that in the original drawing?
Circuits like these have connections to either the line or the neutral, with ground just for protection not for the circuit itself.
-
Hi,
Where does that ground connect to, was that in the original drawing?
Circuits like these have connections to either the line or the neutral, with ground just for protection not for the circuit itself.
Nowhere - it is there for the simulation to work at all. Without it you either get an error or garbage. Can be connected anywhere and it basically says '0V is here' so all other can reference it. -
For the voltages from the simulation to make sense, they need to be referenced to Neutral, not somewhere in the middle of the firing delay circuit. Also you have MT1 and MT2 reversed and are using far too high gate and timing resistors for the TRIAC model you are using, which is based on worst case (max.) gate trigger current, not the typical value. Furthermore the TRIAC library contains the note:Quote
For a correct triac behavior, the "Maximum step size" must be below or equal 20µs.
If you fix all of the above, it sort of works. However its startup is asymmetric, resulting in DC current through the load, so it wouldn't be advisable to use this circuit for anything other than incandescent lighting. It would cause major problems with inductive loads like AC motors and transformers, probably worse than not having a soft start circuit at all.
Sim of 'sort of working' circuit attached. Requires same libraries as O.P's sim.
-
I agree, stupid place to connect 'the reference' and a very piss poor circuit for anything EXCEPT an incandescent lamp or possibly a universal motor with a commutator that can run on A.C. or D.C.! I sadly see many non-working or un-feasable circuits drawn up and presented on YouTube videos by folks who claim to be 'in the know' about circuit design!!! I think YouTube should auto-delete any video that starts with 'Free-Energy' like the idiots with an induction motor driving an alternator and being not only self sustaining but then also sourcing power to external loads!!! The very description of a bullshit post!!! Don't you just love the advertisers of 9900MAH 18650 cells!!! They should be banned from Ebay!!
-
Taken from my trusty General Electric’s SCR manual, 5th edition.
You will require an UJT and a trigger transformer. Basically you have to substitute the circuit arrangement shown in red on the first image and substitute it with the circuit in the second image, with the caveat that you should still connect the trigger transformer on the UJT’s B1 terminal instead of the resistor.
-
Hi,
Where does that ground connect to, was that in the original drawing?
Circuits like these have connections to either the line or the neutral, with ground just for protection not for the circuit itself.
Nowhere - it is there for the simulation to work at all. Without it you either get an error or garbage. Can be connected anywhere and it basically says '0V is here' so all other can reference it.
Of course that is what I suspected, but I like to be sure.
As others have mentioned, this circuit is highly questionable. If it ever works at all, it looks like it charges a cap which turns the triac on then the cap discharges so on the very next half cycle the triac turns back on again. As the cap probably does not discharge all the way, eventually it stays high enough to keep the triac on. That looks like the way it is supposed to work, if it does in fact work at all.
One of the problems is that we don't know which half cycle the triac will turn off and which half cycle it will turn back on again. This could lead to a non symmetrical AC waveform which means there will be a net DC for a time until the triac turns on fully. For example, if it turned on with the first half cycle then turned off after that half cycle was over, it would end up applying a net DC to the load. If this behavior kept repeating, it would look almost like a half wave rectified sine getting to the load, at least until the triac turned on and stayed on.
There is a chance though that you could mitigate this by making sure the time constant while it stays on and when it stays off is long enough such that it allows several cycles to get through to the load. That way the net DC should be small. The problem then is that a slow start circuit would not be as slow start as we would like it to be because it would have to let maybe 10 full cycles get to the load as soon as it turned on. This means it could be hard to create a slow start circuit without accurate timing such as from a microcontroller. That's of course unless you plan to run this with resistive loads only or other loads that can tolerate that unusual behavior.
The way this is usually done is a small value resistor is placed in series with the load, then after a short time the resistor is shorted out so the load get the full power of the line. A relay is probably best to short out the resistor here. If you decide you need multiple steps, you could use one larger value resistor and one smaller resistor, and short out the larger value resistor first then short out the smaller value resistor.
Of course this has to be tested also on the load you intend to use with it, and on any other load if that load is different.
If you use a microcontroller you can time it very precisely by syncing to the line voltage. It's always a good idea to keep the positive and negative half cycles equal in duration though to reduce the DC offset that could emerge. That simply means that if you trigger the positive half cycle turn on at say 1ms after the first zero crossing, then after the start of the negative half cycle turn it back on after 1ms also.
-
Thank you gentlemen for your input and adjusting the schematic. I definitely felt the "it's not that simple" vibe, so for now I'll go and buy a ready made module to unblock myself (and possibly not damage my vacuum cleaner) and go get that GE SCR manual and read it - definitely looks interesting.
I already have the resistor/relay solution: 10x220Ω/5W over ~1 sec that gives roughly 50% power, but I wanted to replace the resistor with something more elegant, smaller and not heating up so badly (but probably I'll try reducing the time). -
Vacuum cleaners usually use brushed series-wound DC motor, so called universal motor, which runs on DC or AC. Phase-angle triac control, randomly firing PWM, rectification to DC then PWM, are all valid strategies for universal motors.
Most optimal would be rectification to DC plus filtaration with a large capacitor, then PWM with current sense and pulse-by-pulse current limiting, to limit torque, resulting in start-up ramp without having to explicitly tune the ramp speed.
Series resistor with bypass relay circuit is lossy. In your electricity bill it probably doesn't matter as it's activated only for some seconds, but be careful with thermal design. You don't want the resistors to cause fire, especially in corner cases when the relay control fails, or with repeated on/off cycles. Use thermal fuse, thermally coupled to the resistors, at very least. -
Thank you gentlemen for your input and adjusting the schematic. I definitely felt the "it's not that simple" vibe, so for now I'll go and buy a ready made module to unblock myself (and possibly not damage my vacuum cleaner) and go get that GE SCR manual and read it - definitely looks interesting.
I already have the resistor/relay solution: 10x220Ω/5W over ~1 sec that gives roughly 50% power, but I wanted to replace the resistor with something more elegant, smaller and not heating up so badly (but probably I'll try reducing the time).
Oh that's interesting. If you can take it apart you can examine the motor to determine what type it is.