Turning on a transformer at zero crossing is the worst you can do!
Can you explain why? I thought that's the best thing you can do. Voltage is at minimum. Some manufacturer even sell parts with built-in zero crossing detection.
The problem switching inductive relates to the fact that most of these SCR circuits use pulse triggering of the TRIAC gate.
Primary current abruptly stops when falling below the holding current. This is the source of audible noise.
The best you can do is pump a continous DC current of 50mA into the TRIAC gate.
Having this problem in the past I took 3 Teccor TRIACs of 25 amps that where driven by 30kHz generated by NE555 through isolating gate drive transformers. These TRIACs switched three 2kVA toroidal transformers powering a 5kW MOSFET amp. This was in the early eighties...
Primary current abruptly stops when falling below the holding current. This is the source of audible noise.
The best you can do is pump a continous DC current of 50mA into the TRIAC gate.
Having this problem in the past I took 3 Teccor TRIACs of 25 amps that where driven by 30kHz generated by NE555 through isolating gate drive transformers. These TRIACs switched three 2kVA toroidal transformers powering a 5kW MOSFET amp. This was in the early eighties...
The problem switching inductive relates to the fact that most of these SCR circuits use pulse triggering of the TRIAC gate.
Primary current abruptly stops when falling below the holding current. This is the source of audible noise.
The best you can do is pump a continous DC current of 50mA into the TRIAC gate.
Having this problem in the past I took 3 Teccor TRIACs of 25 amps that where driven by 30kHz generated by NE555 through isolating gate drive transformers. These TRIACs switched three 2kVA toroidal transformers powering a 5kW MOSFET amp. And there was no noise at all. This was in the early eighties...
Primary current abruptly stops when falling below the holding current. This is the source of audible noise.
The best you can do is pump a continous DC current of 50mA into the TRIAC gate.
Having this problem in the past I took 3 Teccor TRIACs of 25 amps that where driven by 30kHz generated by NE555 through isolating gate drive transformers. These TRIACs switched three 2kVA toroidal transformers powering a 5kW MOSFET amp. And there was no noise at all. This was in the early eighties...
Hi,
I have been using the zero crossing to power ON/OFF in all my amplifiers. This will prevent the inrush current on power ON/OFF. I used a micro to ramp the incoming AC on power ON/OFF at zero crossing with no problems. The problem it is that you must turn ON/OFF the amplifier at zero crossing. The inrush current occurred when the polarity of the AC it is the same as when you power ON/OFF. Since I ramped the AC on power UP/down at zero crossing it will keep the magnetization of the transformer core at minimum. I used this method for 2 years, with excellent results. Reading the incoming current of the AC it is showed a low current on power ON/OFF.
I have been using the zero crossing to power ON/OFF in all my amplifiers. This will prevent the inrush current on power ON/OFF. I used a micro to ramp the incoming AC on power ON/OFF at zero crossing with no problems. The problem it is that you must turn ON/OFF the amplifier at zero crossing. The inrush current occurred when the polarity of the AC it is the same as when you power ON/OFF. Since I ramped the AC on power UP/down at zero crossing it will keep the magnetization of the transformer core at minimum. I used this method for 2 years, with excellent results. Reading the incoming current of the AC it is showed a low current on power ON/OFF.
Voltwide... I'm confused. Why are you pulse triggering the gate if you're just turning on a power transformer?
Pulse triggering the gate is usually done if you're trying to vary the AC voltage applied to a load, as in speed control of a motor, or dimmer lamp applications, or temperature controlled heaters.
We're just turning on a power transformer, and do it hopefully near the zero crossing. Am I missing something?
Hi,
BY ramp the incoming AC you suppress the inrush current and will allow to slowly charges the capacitors. Also prevent the "thump" in the speakers.
BY ramp the incoming AC you suppress the inrush current and will allow to slowly charges the capacitors. Also prevent the "thump" in the speakers.
Yes, its a bit counter intuitive but you should switch inductive loads on at the mains peak rather than zero point. Something easier said than done, at least simply.
What I didn't mention on my design is a further refinement of having the triac perform the soft start feature (with either a series resistor or thermistor) and then having the SS switch bypass both. It works beautifully and the proof is being able to use a small fuse for the primary
I tried HF pulse drive but found the transformer was always more noisy when the triac alone was in circuit.
Whether you are using a solid state relay or a TRIAC with Opto-DIAC (MOC3010 or similar) - in each case this is pulse triggering of the TRIAC.
The gate current is derived from the anode - and after the TRIAC has turned on, this voltage drops to 1V. Consequently gate currents stops immediately after turn-on of the TRIAC.
This is what makes it a pulse driven TRIAC.
Mooly, I cannot say why you encountered problems with RF-drive. You might check whether your TRIACs operate in all 4 quadrants.
Mooly;4074086 What I didn't mention on my design is a further refinement of having the triac perform the soft start feature (with either a series resistor or thermistor) and then having the SS switch bypass both. It works beautifully and the proof is being able to use a small fuse for the primary.[/QUOTE said:
You might have a point there. I used a snubberless triac BTA16-600SW. Looks like its not specified for operation in Q4
Some explanations relating to synchrone switching mains transformers:
What we have to avoid is drive a transformer into saturation, because in that case primary current explodes and might blow some fuses.
So we have to take care of magnetic flux in the transformer core.
Magnetic flux is proportional the applied voltage times the duration of that voltage, the "voltage-time-area", measured in Vsec, mVsec, uVsec etc.
The bigger the area, the more flux is built up in the iron core.
During normal operation, the magnetizing current is zero at voltage peak and maximum at voltage zero crossings.
Thus the magnetizing vsec-area is a quarter of a full cycle sinewave. And this is the area the transformer is designed for.
If you turn on at zero crossing, the first voltage magnetizing voltage area ends at the next zero crossing i.e. after one half cycle. This results in twice the vsec-area compared to normal operation, and so there is potential danger to saturate the core at the end of the first half cycle.
Turning on and off on zero crossing relies on the assumption, that the remanent field in the core is 100% present at turn-on.
Personally I would not go for that.
And practicallly these things vastly depend on the specific transformers and the headroom to saturation they provide.
What we have to avoid is drive a transformer into saturation, because in that case primary current explodes and might blow some fuses.
So we have to take care of magnetic flux in the transformer core.
Magnetic flux is proportional the applied voltage times the duration of that voltage, the "voltage-time-area", measured in Vsec, mVsec, uVsec etc.
The bigger the area, the more flux is built up in the iron core.
During normal operation, the magnetizing current is zero at voltage peak and maximum at voltage zero crossings.
Thus the magnetizing vsec-area is a quarter of a full cycle sinewave. And this is the area the transformer is designed for.
If you turn on at zero crossing, the first voltage magnetizing voltage area ends at the next zero crossing i.e. after one half cycle. This results in twice the vsec-area compared to normal operation, and so there is potential danger to saturate the core at the end of the first half cycle.
Turning on and off on zero crossing relies on the assumption, that the remanent field in the core is 100% present at turn-on.
Personally I would not go for that.
And practicallly these things vastly depend on the specific transformers and the headroom to saturation they provide.
Yep 😛
(Have you seen Eva's circuit for a slow start with a triac. Diagram on post #43 Its not an SS relay but its one interesting circuit non the less.
http://www.diyaudio.com/forums/powe...rt-circuit-no-relays-no-aux-transformers.html)
(Have you seen Eva's circuit for a slow start with a triac. Diagram on post #43 Its not an SS relay but its one interesting circuit non the less.
http://www.diyaudio.com/forums/powe...rt-circuit-no-relays-no-aux-transformers.html)
Yes, interesting. Although the link to the schematics seemed to be broken, eva just did it with these snubberless TRIACs based on the same concept I applied in the early eighties.
Btw, it is a real pity that Eva is gone, she was a real nice contributor in this forum.
Theoretically, I/III-quadrant operation should be ok with pure resistive load.
But we are talking of transformers here...
Btw, it is a real pity that Eva is gone, she was a real nice contributor in this forum.
Theoretically, I/III-quadrant operation should be ok with pure resistive load.
But we are talking of transformers here...
Last edited:
The diagram was also here,
http://www.diyaudio.com/forums/powe...relays-no-aux-transformers-3.html#post1150055
Very talented for sure.
http://www.diyaudio.com/forums/powe...relays-no-aux-transformers-3.html#post1150055
Very talented for sure.
Hi,
I made very good experiences with Sharp´s S102/S202 Series.
The different affixes stand for different approvals (UL, TÜV, CSA).
Apart from that the suffixes stand for zero-crossing or non-zero-crossing Versions.
And last, there are two versions differing in the minimum LED current.
I typically choose the zero-crossing, low current (8mA) version and run them between 10mA and 20mA LED Bias.
The Sharps switched reliably (and even wo. cooling) from the 230V power line of 3kW switchmode amps and associated SMPS supplies to very small loads, like a automatic Electrostatic speaker supply.
jauu
Calvin
I made very good experiences with Sharp´s S102/S202 Series.
The different affixes stand for different approvals (UL, TÜV, CSA).
Apart from that the suffixes stand for zero-crossing or non-zero-crossing Versions.
And last, there are two versions differing in the minimum LED current.
I typically choose the zero-crossing, low current (8mA) version and run them between 10mA and 20mA LED Bias.
The Sharps switched reliably (and even wo. cooling) from the 230V power line of 3kW switchmode amps and associated SMPS supplies to very small loads, like a automatic Electrostatic speaker supply.
jauu
Calvin
Calvin
From my recollection the sharps are opto-coupled DIACs with improved dV/dt immunity. In that case gate drive is fed from anode circuit and stops in the moment of conduction.
i.e. gate is pulse driven.
I understand that you made good experience switching large smps with that circuit. But doing the same with a transformer is another cup of tea.
From my recollection the sharps are opto-coupled DIACs with improved dV/dt immunity. In that case gate drive is fed from anode circuit and stops in the moment of conduction.
i.e. gate is pulse driven.
I understand that you made good experience switching large smps with that circuit. But doing the same with a transformer is another cup of tea.
Hi,
yes, but the thread started with a general Q about the use of SSRs as power switches 😉
SMPS will gain more and more importance and become more and more common, as european and other ´green´ regulations will increasingly restrict the use of massive iron.
When I tested the Sharp 8A SSRs (there are 16A Versions too) I was surrised that these compact devices handled even high power SMPS/class-D so effortlessly, since even small SMPSs may draw incredily high inrush current peaks.
For example is the cold start inrush current(max) of the 25W Meanwell PS-25 specced to be 40A(@230V) !!
jauu
Calvin
yes, but the thread started with a general Q about the use of SSRs as power switches 😉
SMPS will gain more and more importance and become more and more common, as european and other ´green´ regulations will increasingly restrict the use of massive iron.
When I tested the Sharp 8A SSRs (there are 16A Versions too) I was surrised that these compact devices handled even high power SMPS/class-D so effortlessly, since even small SMPSs may draw incredily high inrush current peaks.
For example is the cold start inrush current(max) of the 25W Meanwell PS-25 specced to be 40A(@230V) !!
jauu
Calvin
- Status
- Not open for further replies.