0
\$\begingroup\$

I've got a circuit which is powered using a 24VAC transformer fed into a KBU606G full bridge rectifier to generate 24VDC. Originally there was a 470uF smoothing capacitor, however I was unaware that it would make the output rise to 33/34V, and I don't think that having smoothed DC is necessary for the fan and relay so the cap was removed. This 24V is then dropped to 5V using an L7805 regulator and powers a PIC18f4520 and two HE3621 relays.

The PIC then powers two BC337, switching the HE3621 relays which are then used to switch 3 larger 24V relays (HF92F-024D) to ground. There is also a 24V 0.8A DC fan powered from the bridge.

The issue I'm having is that the L7805 dies whenever the three HF92F relays are powered - it's not too consistent, sometimes it dies when two are powered, sometimes it doesn't blow at all for a while. The fan also runs incredibly slowly when the PIC isn't connected, and at full speed when it is plugged in. The regulator and PIC also blow even when the fan isn't plugged in.

I know that the issue is obviously power/current related, but I can't see why the regulator is dying so much, I would've thought that the diodes would prevent any back-emf issues that could be blowing it.

I don't have much experience with relays so it's probably something obvious but any ideas on the best fix for the problem? I've also tried using an A8498 switching regulator with no luck. Should I be using a current limiting resistor on the input to the regulator?

I've attached a schematic of the circuit and the datasheets to the relays/regulator. Hopefully everything makes sense, let me know if anything needs clarifying.

Thanks.

https://docs.rs-online.com/8941/0900766b814f1e8d.pdf
https://m.littelfuse.com/~/media/electronics/datasheets/reed_relays/littelfuse_reed_relays_he3600_datasheet.pdf.pdf
https://www.st.com/resource/en/datasheet/l78.pdf
https://www.wontop.com/uploadfiles/56/sort_excel/pdf/kbu600g.pdf

enter image description here

\$\endgroup\$
4
  • 3
    \$\begingroup\$ I would calculate how much power the L7805 dissipates and also check its maximum voltage rating. I see the maximum voltage is specified as 35V and your input goes up to 34V, assuming the mains voltage is exactly as specified... if it's 5% higher then so is the regulator input and now the regulator is getting 35.5V peak \$\endgroup\$
    – user20574
    Commented Aug 16, 2021 at 10:14
  • 2
    \$\begingroup\$ You need to re-instate a capacitor across the bridge to a) absorb any brief higher-voltage fluctuations coming from the mains and b) for regulator stability. However, this will make your regulator heating worse. Your regulator is right on the theoretical edge of its voltage input - bad - that's too close. Can you drop some of the voltage before the regulator, with a 10 V zener, or a 7812 or 7815 regulator? \$\endgroup\$
    – Neil_UK
    Commented Aug 16, 2021 at 10:22
  • \$\begingroup\$ I think I have some zener diodes in, I'll try add one into the circuit and see how it goes. Thanks. \$\endgroup\$
    – M_Stenton
    Commented Aug 16, 2021 at 10:48
  • 2
    \$\begingroup\$ A filter capacitor is required from 4-diode bridge to GND for another reason: Relay clamping diodes D1, D2, D3 have nowhere to dump current from those three relay coils...normally this energy goes into the filter capacitor, charging it ever-so-slightly. With no filter capacitor, transient energy raises the L7805 input voltage FAR above 35V momentarily. In that moment....poof! \$\endgroup\$
    – glen_geek
    Commented Aug 16, 2021 at 15:51

2 Answers 2

Reset to default
1
\$\begingroup\$

I was unaware that it would make the output rise to 33/34V, and I don't think that having smoothed DC is necessary for the fan and relay so the cap was removed.

Yes it will rise to 34 volts minus a couple of diode drops so maybe 32.7 volts but, if the transformer isn't running into a full load situation, the secondary voltage might be up to 10% higher so, that makes the peak voltage more like 36 volts and the 7805 absolute maximum rating is 35 volts.

Add to this are generator AC variations of maybe another 5% and it's quite easy to see that the 7805 isn't going to survive very long. I haven't begun to add on surge voltages from other equipment on the AC line or what happens when there are surges due to load drops or indirect lightning.

A 7805 also needs an input capacitor and an output capacitor and will not work well with no bridge rectification at all.

You should get a transformer that outputs the right AC voltage on the secondary to produce 24 volts DC, use a reservoir capacitor and use resistive droppers to limit power dissipated by the 7805.

\$\endgroup\$
2
  • \$\begingroup\$ That makes sense, I've been tunnelling on the relays so much I've completely ignored the basic parts. I'll add in the 7805's in/out capacitors when I next test. The reason the 24V transformer is used is because it is required for the rest of the circuit, but seems like its unavoidable so I'll have to figure out how to make it work with an 18V. Thanks. \$\endgroup\$
    – M_Stenton
    Commented Aug 16, 2021 at 10:34
  • \$\begingroup\$ @glen_geek that's not correct; the diodes will circulate current through their respective relay coil holding it on for a further several milliseconds until current falls to zero. There is no splash onto the power rails. \$\endgroup\$
    – Andy aka
    Commented Aug 17, 2021 at 8:17
0
\$\begingroup\$

Originally there was a 470uF smoothing capacitor, however, I was unaware that it would make the output rise to 33/34V

The capacitor charges to the peak of the rectified sine wave, which is 24*1.4.42 = 33.94V. Not surprising.

I don't think that having smoothed DC is necessary for the fan and relay so the cap was removed. This 24V is then dropped to 5V using an L7805 regulator

I don't understand the thinking behind this at all. Sure, maybe the fan can work with a rectified sine wave, but an LM7805 requires that the input voltage always be 3V or so higher than the output voltage, which isn't the case since the rectified sine wave goes to zero 100 times a second. This causes the regulator output to go to 0V periodically, shutting down all the circuits powered by it.

...using an L7805 regulator...

By L7805, do you mean LM7805 or 78L05? They have different power ratings. If you are using the latter, ensure that your circuit draws less than 0.1A from it. Also, try to measure the temperature of the regulator (even feeling it will do). If it's too hot to touch, it's too hot to run. That tells you that the regulator is overloaded.

The fan also runs incredibly slowly when the PIC isn't connected, and at full speed when it is plugged in.

Im not quite sure why this happens, because the fan is not controlled by the PIC or a relay. It should run at a constant speed independent of whether the uC is there or not.

If 24V DC is what you want, consider using a 230V-18V step-down transformer. You can calculate its peak on your own.

\$\endgroup\$
1
  • \$\begingroup\$ Hi, the rectifier is an L7085CV, and you're right it is unsurprising - it just slipped my mind. I have considered an 18V transformer so that the caps can be re-added, but the rest of the circuit (not shown) requires the 24V output, and originally the PIC wasn't involved in the circuit so there was no regulator so it didnt matter as much what I was getting on the output of the bridge. I'll have to see if I can rework the circuit somewhat. The fan issue is confusing me though, I'll have to get a hold of another fan and see if it's just an issue with that specific one. Thanks. \$\endgroup\$
    – M_Stenton
    Commented Aug 16, 2021 at 10:41

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.