I have just designed a PCI12F508 based cold start circuit.
Just waiting for pcb's.
Just waiting for pcb's.
An externally hosted image should be here but it was not working when we last tested it.
It is a soft start circuit.
On power up the phase angle to a triac is slowly ramped up over a few seconds to limit the start up current.
The input is the mains and the output is phase controlled mains to a amplifier transformer.
Once the soft start is over the triac is turned on permanently.
The mains is monitored and if it disappears the triac is turned off until mains reappears and then the soft start sequence is restarted.
On power up the phase angle to a triac is slowly ramped up over a few seconds to limit the start up current.
The input is the mains and the output is phase controlled mains to a amplifier transformer.
Once the soft start is over the triac is turned on permanently.
The mains is monitored and if it disappears the triac is turned off until mains reappears and then the soft start sequence is restarted.
How do you ensure 100% balance of the positive and negative phases during start up and transient conditions such as mains glitches ?
Any slight imbalance will cause the equivalent of 'DC' to appear in the primary and result in either a substantial rise in current or even lead to core saturation and essentially fault current conditions with massive currents being drawn. Ensuring the triac doesn't misfire (correct snubbing) also seemed impossible to me given varying load and toroid conditions.
I've not been successful with triac controlled soft starts on large toroids and went down a different route entirely.
Any slight imbalance will cause the equivalent of 'DC' to appear in the primary and result in either a substantial rise in current or even lead to core saturation and essentially fault current conditions with massive currents being drawn. Ensuring the triac doesn't misfire (correct snubbing) also seemed impossible to me given varying load and toroid conditions.
I've not been successful with triac controlled soft starts on large toroids and went down a different route entirely.
The PIC is very accurate between the positive and negative half cycles.
The triac has a snubber although according to the spec it doesn't need one for inductive loads.
A triac isn't the cheapest option by far as it needs a good sized heat sink.
But it does have the advantage it has no contacts to burn out over time.
To be honest I haven't got as far as testing the circuit yet as I am waiting for pcb's.
The triac has a snubber although according to the spec it doesn't need one for inductive loads.
A triac isn't the cheapest option by far as it needs a good sized heat sink.
But it does have the advantage it has no contacts to burn out over time.
To be honest I haven't got as far as testing the circuit yet as I am waiting for pcb's.
Hmmm... I've been down this road with triacs (not the PIC though) and wasn't successful I'm afraid. I was using a 650va toroid. When it wasn't happy with any imbalance it growled fearsomely and absolutely exploded the mains fuse. The triac always survived though.
I guess you'll just have to try it. Hope it works 🙂 Keep us posted.
I guess you'll just have to try it. Hope it works 🙂 Keep us posted.
I have seen problems before with triacs with motors.
A customer asked me to look into controlling the speed of a shaded pole motor using skipping half mains cycles. The motor was very unhappy and got hot very quickly.
I know toroidal transformers are fussy about balanced mains. Any DC and they rattle. They are more fussy than E I cored transformers.
A customer asked me to look into controlling the speed of a shaded pole motor using skipping half mains cycles. The motor was very unhappy and got hot very quickly.
I know toroidal transformers are fussy about balanced mains. Any DC and they rattle. They are more fussy than E I cored transformers.
Why do you want to use a triac, with a voltage drop of about 1-2V thus causing power loss. What is wrong with a oversized relay?
Oh I think so. VTM is typically 1,5 V.
So with a normal load of an amp or two , say 1,5-3 W loss continously.
So with a normal load of an amp or two , say 1,5-3 W loss continously.
Last edited:
My "thinking" is bad.
Looks like they do lose some volts.
That surprises me.
How do they switch LV supplies?
Like the 160kW motor drive in an F1 car?
or a 10kW motor driven from a capacitor storing only 5V?
Looks like they do lose some volts.
That surprises me.
How do they switch LV supplies?
Like the 160kW motor drive in an F1 car?
or a 10kW motor driven from a capacitor storing only 5V?
I would guess antiparallell thyristors, FETs when using DC , IGBTs when applicable. I guess there is evwn triacs with lower voltage drop but AFAIK notwith less than 1 V.
A lot depends on both the triac and the quadrants it is fired in together with how it is fired (such as pulse or DC).
To calculate the power dissipated in the triac requires knowing the triac 'knee voltage' and the 'slope resistance' and also the average and RMS current of the load.
To calculate the power dissipated in the triac requires knowing the triac 'knee voltage' and the 'slope resistance' and also the average and RMS current of the load.
The triac I am using has 1.55v drop across it.
With 13 amps mains that is 20 watts.
20 watts in 3000 watts is minimal.
A triac has the advantage that it has no contacts or mechanical parts to wear out.
But on the other hand you have to dissipate 20W, which is a considerable amount considering that most triacs mounting tab is at live potential, thus demanding a thoroughly isolated heatsink or isolating wafers etc. Or using a isolated tab triac, with far worse Rtjc.
I wonder which is the most reliable solution in the end?
(insulated or isolated, english is not my first hand language...)
I wonder which is the most reliable solution in the end?
(insulated or isolated, english is not my first hand language...)
I would go for longevity of a triac over a mechanical part.
A triac isn't the cheapest solution because of the heat sink but it will be the most reliable.
A triac isn't the cheapest solution because of the heat sink but it will be the most reliable.
I have repaired my fair cut of dimmers etc to not trust a triac. Either it overheats or it will break due to overvoltage conditions from a voltage transient. It probably survives any overcurrent situation though unless to long duration.
Your 2X0VAC can have considerable transients even in densly populated areas. In parts of Sweden where long lines in air is common designing for full 4000Vrms transient is wise. Sometimes not even that is sufficient.
Your 2X0VAC can have considerable transients even in densly populated areas. In parts of Sweden where long lines in air is common designing for full 4000Vrms transient is wise. Sometimes not even that is sufficient.
The triac does need a decent heat sink especially if not in a vented enclosure.
There is a snubber that will help with short transients.
There is a snubber that will help with short transients.
There is very little chance you will succed removing 20W of heat in a not vented enclosure. I would recommend either forced air circulation or an external heatsink, thus increasing the safety demand.
I would also recommend you to put a thermal switch or similar temperature protecton at the heat sink if you are planning to waisting 20W or so.
Snubber or varistor is good.
A relay sized properly in a soft start will switch quite a low load if the turn on is delayed sufficiently. The normal load should of course not be the problem. I expect life expectancy to be very long. Actually I think that the power resistors are the weak point, subsequent power on might overheat these. I have repaired a few hefty power amps where actually the soft start resistors have blown and protected the mains fuse...
(it is not only semiconductors that are used to protect fuses, any component in series with a slow blow fused can be used to protect that expensive little part if properly undersized... 🙂 )
I would also recommend you to put a thermal switch or similar temperature protecton at the heat sink if you are planning to waisting 20W or so.
Snubber or varistor is good.
A relay sized properly in a soft start will switch quite a low load if the turn on is delayed sufficiently. The normal load should of course not be the problem. I expect life expectancy to be very long. Actually I think that the power resistors are the weak point, subsequent power on might overheat these. I have repaired a few hefty power amps where actually the soft start resistors have blown and protected the mains fuse...
(it is not only semiconductors that are used to protect fuses, any component in series with a slow blow fused can be used to protect that expensive little part if properly undersized... 🙂 )
Last edited:
- Status
- Not open for further replies.