1. OCAU Merchandise is available! Check out our 20th Anniversary Mugs, Classic Logo Shirts and much more! Discussion in this thread.
    Dismiss Notice

Automatic Potentiometer?

Discussion in 'Electronics & Electrics' started by RyoSaeba, Feb 15, 2021.

  1. OP
    OP
    RyoSaeba

    RyoSaeba Member

    Joined:
    Sep 11, 2001
    Messages:
    13,392
    Location:
    Perth
    Ooo thx again mate! :D My first lot I'll be buying off jaycar so I can get some testing going. Then I'll buy in bulk on ebay or aliexpress once I know what I'm buying.

    I'll probably be using a printed resin box. I can make slots cut out for air flow if required.
     
    Last edited: Feb 16, 2021
  2. mtma

    mtma Member

    Joined:
    Aug 12, 2009
    Messages:
    6,110
    TBH if you plan on making a number of these longer term, I'd spend some time looking at directly modding supply modules or ccfl drivers for a ramp instead of relying on this simple series pass method. The method will work but it's definitely low tech.

    ED- if you're aiming to put it in resin get one of these too for good measure.
    https://www.jaycar.com.au/to-220-heatsink/p/HH8514

    In the simple solution the main transistor without cooling help will be about +70 degrees above ambient at the transistor long term. It will work for some time but not really a good mix application wise to house it in resin as the situation will worsen.
     
    Last edited: Feb 16, 2021
  3. OP
    OP
    RyoSaeba

    RyoSaeba Member

    Joined:
    Sep 11, 2001
    Messages:
    13,392
    Location:
    Perth
    Low tech is fine with me. :D I buy the Inverter off ebays pre made. It'll be difficult trying to mod them, not that I know what's involved either lol. This method also allow me to customize it to people who wants the strike up mode. Some people are happy just to have it on/off. I can also sell it to existing people who has bought them previously.
     
  4. mtma

    mtma Member

    Joined:
    Aug 12, 2009
    Messages:
    6,110
  5. OP
    OP
    RyoSaeba

    RyoSaeba Member

    Joined:
    Sep 11, 2001
    Messages:
    13,392
    Location:
    Perth
    Ah good idea. I'll try a few things out I guess. I'll test the temperature first to see how hot it gets. I'll probably go get an IR thermometer to check it. Been wanting to get one for awhile hahaha
     
  6. OP
    OP
    RyoSaeba

    RyoSaeba Member

    Joined:
    Sep 11, 2001
    Messages:
    13,392
    Location:
    Perth
    So my jaycar shopping list is so far:

    Transistors
    https://www.jaycar.com.au/tip31-npn-transistor/p/ZT2285
    https://www.jaycar.com.au/tip41c-npn-transistor/p/ZT2291
    https://www.jaycar.com.au/bd140-pnp-transistor/p/ZT2190
    https://www.jaycar.com.au/bd136-pnp-transistor/p/ZT2188

    Resistors
    https://www.jaycar.com.au/47-ohm-10-watt-wire-wound-resistor/p/RR3360
    https://www.jaycar.com.au/1k-ohm-5-watt-wire-wound-resistor/p/RR3298

    Capacitor
    https://www.jaycar.com.au/10uf-35vdc-tantalum-capacitor/p/RZ6655

    Is that right? I chose larger ones cause I wasn't sure which ones I need. So go larger just in case. I also decided to grab both types of transistors just to see how they would differ.
     
    Last edited: Feb 16, 2021
  7. mtma

    mtma Member

    Joined:
    Aug 12, 2009
    Messages:
    6,110
  8. OP
    OP
    RyoSaeba

    RyoSaeba Member

    Joined:
    Sep 11, 2001
    Messages:
    13,392
    Location:
    Perth
    Ah ok. Thanks again mate! :D I'll post back results hopefully not flaming ones hahhhaha
     
  9. mtma

    mtma Member

    Joined:
    Aug 12, 2009
    Messages:
    6,110
    Sorry I need to facepalm for brushing over this between work.

    The 47 ohm, should be a higher power resistor. Although it will probably work as is, the impulse load is well over what that 1/2W jaycar resistor really should be handling. If you repeatedly played with it for a block of time it would be liable to fail.

    https://www.jaycar.com.au/47-ohm-5-watt-wire-wound-resistor/p/RR3266 would be the closest fit from jaycar.

    If you've already bought some what you can do is actually grid them with two parallel and each pair in series.
     
  10. OP
    OP
    RyoSaeba

    RyoSaeba Member

    Joined:
    Sep 11, 2001
    Messages:
    13,392
    Location:
    Perth
    Nope haven't gone shopping yet. Probably later this week. Once again thank you for giving clear instructions and help. :D
     
  11. OP
    OP
    RyoSaeba

    RyoSaeba Member

    Joined:
    Sep 11, 2001
    Messages:
    13,392
    Location:
    Perth

    Ok just tested it with an empty load and measuring the output with my multimeter. I think it worked for a bit then the capacitor popped. I checked the temp of everything else they seem ok. Now the power just passes through 12V directly. The Capacitor I got is 25V 10uF TS14. Do I need something bigger?
     
  12. mtma

    mtma Member

    Joined:
    Aug 12, 2009
    Messages:
    6,110
  13. OP
    OP
    RyoSaeba

    RyoSaeba Member

    Joined:
    Sep 11, 2001
    Messages:
    13,392
    Location:
    Perth
  14. OP
    OP
    RyoSaeba

    RyoSaeba Member

    Joined:
    Sep 11, 2001
    Messages:
    13,392
    Location:
    Perth
    Ok I think the circuit is right. But don't see any ramping. Unless it happened too quickly I didn't noticed it. Nothing blew this time. I think I followed the diagram correctly, but I am probably wrong. This is what I have. Not the prettiest thing but I need to understand it myself first before I refine it on a PCB. Sorry about this spaghetti, hopefully it makes sense. Also Altronics apparently didn't have 10uF 25V caps so they sold me a 10u 50V cap instead.
    upload_2021-2-22_16-3-52.png
     
  15. mtma

    mtma Member

    Joined:
    Aug 12, 2009
    Messages:
    6,110
    BD136 is the wrong way around. Swap C and E. See the circuit from before.

    I apologise, I may have jumped quite a few steps in terms of assuming your electronics knowledge. It would be that you might want to go through some electronics tutorials but unfortunately I'm not sure which ones are good enough to recommend as the best electronics learning is done by starting out doing simplistic things like these.

    The PNP transistor is like the polarity reverse of the NPN one, meaning that the emitter of a PNP transistor should be towards the positive voltage for correct operation. It actually does kind of work the other way around too but it's very limited (sorry, explanation is involved and not too relevant right now)

    Other potential issues which you' may want to experiment with:

    As you'll see in the simulation, the ramp I set is about a couple of seconds to full voltage but the voltage rises abruptly at first then flattens out. This is because the filter formed by the 1000 ohm and 10uF capacitor has what's called a time constant, which is calculated tau = R * C and the rise time represents the time taken to get to 63% of the applied voltage and the voltage ramp is exponential in curvature. Assuming you bought the 8 pack of 1000 ohms, you can make the ramp time longer by just chaining more 1000 ohms in series up to 8 times the time constant for trial adjustment purposes.

    The CCFL driver probably doesn't respond completely the linear way, in terms of the light saber ramp effect. You might find that most of the ramping only happens from 0-3V but the rest is pretty much full fill. To test this you can also use your adjustable power supply, and plot out what voltages correspond to what level of fill, and chart this. This way you will know how mismatched the theoretical ramp curve of the basic R-C ramp this circuit generates compared with the effect you're after. If the deviation is severe, we may need to try a different kind of solution.

    You can also test this circuit with a lower input voltage to manipulate this, once it is actually ramping. The RC formula mentioned before you will note contains no parameter relating to voltage - and indeed the relation is that the voltage rises to 63% within RC time. Therefore applying a lower voltage you will ramp proportionally to the lower voltage in the same overall time.
     
    Last edited: Feb 22, 2021
  16. OP
    OP
    RyoSaeba

    RyoSaeba Member

    Joined:
    Sep 11, 2001
    Messages:
    13,392
    Location:
    Perth
    Not at all mate. I'm very grateful for you helping. :D I remember back in highschool being taught this stuff, but that was about 30 years ago lol. I've since returned most of what I left highschool with back to the school. :lol: Once I get this sorted I'll buy you a carton of your favourite beverage and have it delivered. :D Ideally I would love to sit through tutorial and build stuff and learn. But time doesn't really permit. So it's kinda just learning on the job like this lol. Hands on with experimenting which I don't mind at all. Start simple. :D

    I did notice the transistors have reversed pin out and the diagram too. I thought I got it right too. Oh well. It's all part of learning lol.

    I've just flipped C and E on BD136 and still no ramping at all. Even at lower voltage I just see full brightness as if it's just a direct passthrough.

    Just tried it with all 8 resistors. It's still doing the same thing. If one of the transistor is burnt, would it still work? I'm wondering if I burnt it with the reversed cap. Although none of them were even warm when I checked them when the cap popped. Only the cap was warm. I tested it with lower voltages too but all resulted in the same thing. The CCFL starts to dimly light up at one end at 3.5V. Then it achieve full light at about 7V and at 8V it's fully lit and it just gets bright as it hits 12V. But even with all 8 resistors, it's just straight to full brightness. Tested it on a multimeter and it's basically telling me the same thing.
     
  17. mtma

    mtma Member

    Joined:
    Aug 12, 2009
    Messages:
    6,110
    Typically when transistors burn, they will let you know because they get really hot and usually let out smoke. The PNP transistor may have also burned by being connected as it was.

    If you have a multimeter handy, check that the voltage at the base of the BD136 is starting from (relative to 0V of course) some low value close to 0V (there may be some residual). If it's not then that's the initial problem. The circuit relies on the load to drain down the voltage of this cap after it is switched off. You can manually drain it with a 1000 ohms when the circuit is not connected to the supply and restart it and check if there is a ramp there. If no ramp exists here under that condition, there's a problem.\

    The next step is to check at the emitter of the BD136 - it should be copying that ramp with a 0.6V or so difference. The TIP31 then copies that value again with a 0.6V or so difference.

    ED- if your multi-meter is a slow updating digital one you might want to try with all the 1000k's in place, as it will give you a few seconds of ramp time to look at.
     
    Last edited: Feb 22, 2021
  18. OP
    OP
    RyoSaeba

    RyoSaeba Member

    Joined:
    Sep 11, 2001
    Messages:
    13,392
    Location:
    Perth
    Yup slow updating digital multimeter checked. :lol: I am trying with all the 1000k in place right now and don't see anything. I think the BD136 is cooked. I'll go grab another tomorrow. I thought I had got a BD140 too but it turns out it's a TIP32.
     
  19. mtma

    mtma Member

    Joined:
    Aug 12, 2009
    Messages:
    6,110
    See the attached for some notes to check around the circuit.

    Importantly there should also be a voltage drop across B-E and C-E of the TIP31 as well, regardless of whether or not the ramping is working. So give those points a measure too. If there's no difference that trasnsistor is suspect too as it should have at least one P-N junction worth of drop.

    ED- the TIP32 should work in place of the BD136 as well, but beware that the BD136 and TIP32 have different pin orders with respect to their tab or part markings.
     

    Attached Files:

    Last edited: Feb 22, 2021
    RyoSaeba likes this.
  20. mtma

    mtma Member

    Joined:
    Aug 12, 2009
    Messages:
    6,110
    Oh they're 1000 right? Brown black black brown brown?

    Or are they actually 100k, Brown black black orange brown or 1000k, Brown black black yellow brown (doesn't look like this one).

    (last brown actually represents tolerance in % but just added it for simplicity)

    I hate these blue body resistors for this reason.
     
    Last edited: Feb 22, 2021

Share This Page

Advertisement: