Connect with us

Triac Problem

Discussion in 'Power Electronics' started by andyx, May 4, 2018.

  1. andyx

    andyx

    11
    0
    Nov 12, 2015
    I have a MAC97A6 Triac that's blown, it's from a timer module that's part of an extractor fan. This has happened on two identical fans so I suspect the triac isn't tough enough for the motor. I suspect the manufacturer is just penny pinching with the components.

    The triac is triggered from a logic (NOR) gate via a resistor.

    Both fans lasted approx 18 months to 2 years, but at £200+ each I'd rather fix the timer board if I can.

    Can anyone suggest a better triac that would work? Doesn't even have to be the same TO92 package.

    I appreciate any advice you can offer.
     
  2. Robert_fay

    Robert_fay

    124
    34
    Jun 15, 2017
    Q401E3 from Littelfuse might be an option assuming that it is a problem of not having a tough enough triac. If there is something else failing that is causing this part stress beefing up the triac may just cause the other part to fail.
     
  3. andyx

    andyx

    11
    0
    Nov 12, 2015
    Thanks Robert, I'll try that. If something else blows up as a consequence I'll maybe have to just replace both fans.

    Well known ventilation manufacturer who were no help at all, only interested in selling me new extractor fans.

    Thanks for your help.
     
  4. globecollector

    globecollector

    27
    9
    Jun 27, 2011
    Can you determine if the TRIAC has failed due to over voltage (i.e, back E.M.F. or other inductive effects from the motor), over current or simply overheating. If the original was in a TO-92 package, it can't be much of a TRIAC, or a motor for that matter either.

    I once had to fix a fan speed controller for a fan that ventilated a chicken shed....2000 chooks died because of an underrated (T0-220 Packaged) TRIAC with inadequate snubbing...I'd go for a sensitive gate TO-220 packaged TRIAC and ensure it is well snubbed...i.e. 100nF 250VAC X2 capacitor in series with 100Ω 1W resistor all in parallel with a 275v MOV. Make sure the TRIAC current is about 3A minimum. If the fan idles or motor hums because of the 100nF capacitor, reduce it to 47nF and increase the resistortp 150Ω.

    The Q401E3 above is a smaller member of a much larger family of TRIACS, I'd be looking for down the list for one in a TO-220 package with similar gate sensitivity to the original and a current rating of 3-6A.

    Q4006N4 looks like a nice robust one, 400v, 6A, 25mA gate in all quadrants except IV which is 50mA. You need only quadrants II and IV.

    Q6004V3 is twice as sensitive on the gate, (10mA QII, 25mA QIV same as Q401E3), 4A rated, higher blocking voltage, (600v) so better able to resist inductive crap from the motor but is a TO-251 package with the flag cut off...so less able to heatsink alone. A Q4004V3 would probably be O.K. too, it is the same as Q6004V3, but a little less open circuit voltage headroom (400v) and cheaper.
     
    Last edited: May 5, 2018
  5. Petkan

    Petkan

    19
    2
    Feb 9, 2011
    Petkan:
    The triac may need a RC snubber across its output to limit the rate of voltage changes (like 100R in series with 0.1uV/400V)
     
  6. andyx

    andyx

    11
    0
    Nov 12, 2015
    Checked circuit, absolutely nothing between the triac and motor. Manufacturer saving a penny I suspect.
    I don't think it's an overheating problem, just simply too small current carrying capacity of the triac itself.

    I can't find Q4006N4 or Q6004V3 on fleebay here in the UK, what about any of the BTA06 alternatives. Would be safer with isolated tab, it's fairly close to the adjustment pot for the timer.
     
  7. andyx

    andyx

    11
    0
    Nov 12, 2015
    What about a BTA08-600TW or BTA08-600SW?
     
  8. globecollector

    globecollector

    27
    9
    Jun 27, 2011
    If there is nothing between the TRIAC and motor, then there is no snubber. The snubber needs to be across MT1 and MT2 of the TRIAC. Use the values I or others have given you, but to be really safe, place a 275v MOV across MT1 and MT2 also.

    Some comments for "young players" too...if you choose a TRIAC that is in a different package from the original TO-92...or even if you DO choose an equivalent TO-92 packaged TRIAC with much the same specs as the original ..MAKE SURE THE PIN OUTS ARE CORRECT...often different numbered (but electrically equivalent) devices will have different pin-outs and devices in different packages will have different pin-outs...so be aware of this. Gate firing pulses are always applied between MT1, or A1 and the Gate, so MT1/A1 will be connected to the driving logic gates chip's Vdd/GND line, (which in a non-isolated circuit such as this might well be the the Arctiv/Line/Phase of the mains...so it is only a local "GND" with respect to the electronics inside the unit, not real TRUE ground or Earth.) MT2/A2 will go to the load, (Motor).

    To select a replacement TRIAC,, you need to first try to find the specs of the original. The important parameters are:-

    1) Hold-off voltage, (i.e. the voltage that appears between MT1 and MT2 before the TRIAC is fired, this must be greater than the peak value of the mains....for the U.K. (same as Australia or South Africa) 230v R.M.S. at 50Hz. So the prak is 1.414 x 230 = 326 volts so choose a 400v TRIAC, (if you want a greater safety margin, choose a 600v TRIAC, but it will be more expensive)

    2) Current, i.e, the current that flows through the TRIAC once it has been triggered, this will be the (instantaneous) motor current, so connect the motor directly to the mains and put a 1Ω resistor in series with it, measure the voltage drop with a multimeter on the AC volts range across the 1Ω resistor and this will be the current. Select a TRIAC with a current rating somewhat greater than this to accommodate instantaenous values during the AC cycle...but not too much greater or it may unlatch...(i.e. the holding current may be too small, look up the Wikipage on TRIACs to find out more.) So if the motor draws 500mA, any TRIAC up to 3A or so will do...try bigger ratings if you like and see if they will hold or look in the datasheets of the BTA family.

    3) gate trigger current this varies dependant on the "firing quadrant"...look up the Wikipage, it explains this clearly.
    https://en.wikipedia.org/wiki/TRIAC I would suspect your original TRIAC had a 10mA QII and 25mA QIV rating, so find a BTA device with the same ratings. You did say it was triggered from a logic gate, correct? Make sure this logic gate was not damaged when the original TRIAC failed...if so, or even just as a precaution, change the gate chip anyhow. Try to get a gate chip of the same brand, particularly if it is a Schmitt type, (like a 4093) as different brands can have different trip voltages.

    If you don't know, or cant find the specs of the original device, make an EDUCATED GUESS, I'd say that from the application, 400v, 1A, 10-20mA Gate.

    Is there a resistor between the output of the logic gate and the TRIAC's gate, if not, a good practice is to put 47-100Ω there. For extra chip protection place two diodes, one with its anode to the chip's Vss/GND rail and its cathode to the output of the chip driving the TRIAC, put a second one with its anode to the output of the chip and its anode to the chip's Vss/RAIL pin...crappy 1N4004's will suffice...if the mains then comes out the gate of the failing TRIAC, the resistor will go up in smoke and the chip will be saved by the diodes!

    Some points of note...if the voltage rating of a TRIAC, (or any other power semiconductor for that matter) is exceeded, even for an instant..the device will be destroyed.
    If the current rating of a TRIAC, or most other power semiconductors, is exceeded by say 50%, the device will usually handle it...it depends in how much greater the current spike is over the device's rating and for how long it lasts...the greater it is, the shorter duration and repetition frequency the device will handle, but you have some leeway with current..but NO LEEWAY AT ALL with voltage.

    Other members have provided you with URL links to component suppliers and datasheets....use tham to choose the device that will solve your issue RELIABLY, so you will never need to pull those fans apart ever again!

    The snubber has been deliberately left out because the accountant was leaning over the designing engineer's shoulder and whispering "sweet nothings" into his ear! Believe me, it was a very deliberate act that goes against all the rules of good, reliable design.
     
    bushtech likes this.
  9. Petkan

    Petkan

    19
    2
    Feb 9, 2011
    Petkan:
    It is possible the triac needs an RC snubber across (100R in series with 0.1uF/400V). The motor is inductive load and can produce voltage spikes. Fans I have seen use TO-92 packaged triac - small indeed...but the motor is small too. Over-Current, Over-voltage or over-heating? Steady state current is limited by the motor, its inductance limits the rate of current rise (we eliminate over-current as suspect). You can measure the triac temperature after max time operation to confirm or reject over temp hypothesis. My guess - over-voltage spike due to lack of snubber (if confirmed the is none).
     
  10. andyx

    andyx

    11
    0
    Nov 12, 2015
    Sorry guys but my electronics knowledge and experience is basic to say the least, you're all far better and more knowledgeable than me. I can replace the triac and even work out replacing it with one with a different pin configuration but that's about it. I just hoped someone would just say "Replace the triac with Blah", and that would be that.

    Here's a diagram of the output to the motor:
    triac.jpg
    The motor is fairly beefy, it's an industrial extractor fan from on of the market leaders in ventilation. I know it looks a cheap rubbish design, 'cause it is.
    I thought I could just replace the MAC97A6 with something bigger/better.

    My apologies, I didn't mean to mislead you into thinking I knew more than I do. You're all brain surgeons, I'm just the guy that wheels the injured into A&E.

    Thanks for taking the time to try and help, I really do appreciate it.
     
  11. Bluejets

    Bluejets

    3,542
    719
    Oct 5, 2014
    If it lasts 18 months - 2 years, I'd just replace what is already there. Couple of dollars for the part and you're away.
    Chances are a line spike or similar took it out and I doubt any snubber will help there. Commercial installations will be chockers woth all sorts of line spikes and interference.
    Quite simply not worth the effort as the manufacturer obviously concluded.
    Perhaps look at it as "guaranteed job income" every so often.
     
  12. andyx

    andyx

    11
    0
    Nov 12, 2015
    Thanks to everyone who's responded. I've replaced the triac with a 2N6073AG which works a treat, just needed to change the pin connections.
    I'd like to add-in the protection mentioned but unsure about the MOV, never used one before. Is the circuit diagram correct?
     

    Attached Files:

Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day

-