Author Topic: Trying to determine what voltages should be in a circuit  (Read 2149 times)

0 Members and 1 Guest are viewing this topic.

Offline gmcTopic starter

  • Regular Contributor
  • *
  • Posts: 136
  • Country: gb
Trying to determine what voltages should be in a circuit
« on: September 26, 2019, 04:35:32 pm »
I'm busy trying to build a arduino based Hakko T12 soldering iron.

I have been using the design from: https://www.hackster.io/sfrwmaker/soldering-iron-controller-for-hakko-t12-tips-on-arduino-f7a888

Wired everything up and there is some issue with the heating circuit....it's not working.

I've measured the voltages as shown below:



The problem I am getting is the 100R resistor in the middle of the pic is getting very hot. Voltage across it is 8.2V, so 0.6W. The circuit design uses a 1/4W resistor so I'm guessing its not supposed to get that hot and I have messed up something. Emitter voltage of 21V is incorrect, or gate voltage of 12.8 is wrong...or both.

Double checked everything and still not getting anywhere. Looking at the circuit is it possible to work out that the voltages should be.
I've tied the input low for testing, from the PWM (Not connected to the arduino)

I'm using a IRF540, with general purpose transistors. BC547 for NPN, and BC327 PNP.
Also the iron is not connected.

The purpose of the pnp/npn connected to each other also has me confused.

« Last Edit: September 26, 2019, 05:47:12 pm by gmc »
 

Offline bob91343

  • Super Contributor
  • ***
  • Posts: 2675
  • Country: us
Re: Trying to determine what voltages should be in a circuit
« Reply #1 on: September 26, 2019, 05:18:46 pm »
You can't have 12.7V on the base and 21 V on the emitter.  You say both transistors are KSP10 but one is NPN and the other PNP.  Looks like a miswire somewhere.
 

Offline AVGresponding

  • Super Contributor
  • ***
  • Posts: 4931
  • Country: england
  • Exploring Rabbit Holes Since The 1970s
Re: Trying to determine what voltages should be in a circuit
« Reply #2 on: September 26, 2019, 05:36:42 pm »
The pnp is an S9012H, not KSP10.

Are you measuring the voltages on this circuit with no PWM input? If so, that's part of the problem, though there's another issue as the gate of the fet shouldn't be drawing enough current to cause an 8v drop on the gate resistor. Gate current should be effectively 0.

Check the drain and source are the right way around also
nuqDaq yuch Dapol?
Addiction count: Agilent-AVO-BlackStar-Brymen-Chauvin Arnoux-Fluke-GenRad-Hameg-HP-Keithley-IsoTech-Mastech-Megger-Metrix-Micronta-Racal-RFL-Siglent-Solartron-Tektronix-Thurlby-Time Electronics-TTi-UniT
 

Offline gmcTopic starter

  • Regular Contributor
  • *
  • Posts: 136
  • Country: gb
Re: Trying to determine what voltages should be in a circuit
« Reply #3 on: September 26, 2019, 05:47:58 pm »
Made a typo with the voltages. Should be 21.7V on the base.

I've re-checked the drain and source and they are correct.
Circuit is with no PWM connected.
« Last Edit: September 26, 2019, 05:52:20 pm by gmc »
 

Online fourfathom

  • Super Contributor
  • ***
  • Posts: 2005
  • Country: us
Re: Trying to determine what voltages should be in a circuit
« Reply #4 on: September 26, 2019, 06:19:05 pm »
there's another issue as the gate of the fet shouldn't be drawing enough current to cause an 8v drop on the gate resistor. Gate current should be effectively 0.

That current through the resistor is flowing through the 18V zener, which I assume is there to protect the FET from excessive gate voltage (I haven't checked the FET specs).  There's no good reason that gate resistor needs to be 100 Ohm.  It could probably be 10K, and then it wouldn't be burning up.
We'll search out every place a sick, twisted, solitary misfit might run to! -- I'll start with Radio Shack.
 

Offline Audioguru again

  • Regular Contributor
  • *
  • Posts: 115
  • Country: ca
Re: Trying to determine what voltages should be in a circuit
« Reply #5 on: September 26, 2019, 06:29:15 pm »
The original IRF9540 is a P-channel Mosfet but your replacement is an IRF540 which is an N-channel Mosfet. Then the circuit works wrongly like you have.
The two transistors are complementary emitter-followers to drive the high gate capacitance of the Mosfet with enough current to quickly charge and discharge the capacitance quickly.
 

Offline gmcTopic starter

  • Regular Contributor
  • *
  • Posts: 136
  • Country: gb
Re: Trying to determine what voltages should be in a circuit
« Reply #6 on: September 26, 2019, 06:33:20 pm »
The original IRF9540 is a P-channel Mosfet but your replacement is an IRF540 which is an N-channel Mosfet. Then the circuit works wrongly like you have.

Aggh...yes you are right! That's it. I wrongly assumed they could be interchanged.  |O
 
The following users thanked this post: Ian.M

Online fourfathom

  • Super Contributor
  • ***
  • Posts: 2005
  • Country: us
Re: Trying to determine what voltages should be in a circuit
« Reply #7 on: September 26, 2019, 08:16:41 pm »
The original IRF9540 is a P-channel Mosfet but your replacement is an IRF540 which is an N-channel Mosfet. Then the circuit works wrongly like you have.
The two transistors are complementary emitter-followers to drive the high gate capacitance of the Mosfet with enough current to quickly charge and discharge the capacitance quickly.

Yes, that N-channel would be a big problem!  But you still want to increase the value of that 100 Ohm resistor, since with the zener current the 100 Ohm resistor will be dissipating as much as 360mW, and the zener as much as 1W (depending on duty-cycle).  How fast is the PWM signal?  You want fast switching at the FET, but I would think that you could go to 1K for the resistor which would reduce the power dissipation by a factor of 10.  With the 1500pF gate capacitance, that 1K Ohm would give you an RC time constant of 1.5 uS which should be comfortable.  You do want the zener, since the max gate-source voltage on that FET is only +/- 20V.
We'll search out every place a sick, twisted, solitary misfit might run to! -- I'll start with Radio Shack.
 

Offline Ian.M

  • Super Contributor
  • ***
  • Posts: 13217
Re: Trying to determine what voltages should be in a circuit
« Reply #8 on: September 26, 2019, 09:38:05 pm »
Component values in the drive circuit appear to be chosen for 3.3V logic.  Assuming a 5V Arduino driving it, decrease the 5K1 resistor to 3K3 to reduce the gate drive to about 14V.  Then the Zener wont conduct (but's still there to clip any gate voltage spikes) and the 100R gate resistor should stay cool.
 

Offline edavid

  • Super Contributor
  • ***
  • Posts: 3464
  • Country: us
Re: Trying to determine what voltages should be in a circuit
« Reply #9 on: September 26, 2019, 11:48:17 pm »
The voltages shown are not possible.  There must be a wiring problem.  Maybe the zener cathode is wired to the wrong place.
 

Online fourfathom

  • Super Contributor
  • ***
  • Posts: 2005
  • Country: us
Re: Trying to determine what voltages should be in a circuit
« Reply #10 on: September 27, 2019, 12:14:39 am »
The voltages shown are not possible.  There must be a wiring problem.  Maybe the zener cathode is wired to the wrong place.

The OP had substituted an N-channel fet for the P-channel.  That's the reason for the strange voltages.  "Audioguru again" spotted it.
We'll search out every place a sick, twisted, solitary misfit might run to! -- I'll start with Radio Shack.
 

Offline edavid

  • Super Contributor
  • ***
  • Posts: 3464
  • Country: us
Re: Trying to determine what voltages should be in a circuit
« Reply #11 on: September 27, 2019, 12:50:56 am »
The voltages shown are not possible.  There must be a wiring problem.  Maybe the zener cathode is wired to the wrong place.

The OP had substituted an N-channel fet for the P-channel.  That's the reason for the strange voltages.  "Audioguru again" spotted it.

The MOSFET type is irrelevant (well, not to the circuit actually working, but to the gate resistor heating up).

The only conduction path from the gate resistor to ground would be through the MOSFET gate and the FR107 (both would have to be blown or miswired).

However, that's not possible, because the MOSFET drain is at 23.7V.

Therefore we know that either the schematic or the voltages are wrong (or maybe both).
« Last Edit: September 27, 2019, 12:59:43 am by edavid »
 

Online fourfathom

  • Super Contributor
  • ***
  • Posts: 2005
  • Country: us
Re: Trying to determine what voltages should be in a circuit
« Reply #12 on: September 27, 2019, 01:43:40 am »
The MOSFET type is irrelevant (well, not to the circuit actually working, but to the gate resistor heating up).

Oops, I was looking at the drain-source diode but you're right, that's not going to conduct to the gate.  What *does* happen if you exceed the gate-source breakdown voltage?  Does the gate start conducting to the channel?  Is this permanent, or potentially recoverable?

So I finally looked at the voltages noted on the schematic, and yes, there is no current path that would bring the resistor/gate connection down to 12.8V.  Given that, we can't trust that the schematic, and/or the voltages are correct.  The 100 Ohm resistor is getting hot so we know there is indeed current flowing through it, but where is it going?  We can't tell from what we have in front of us.
We'll search out every place a sick, twisted, solitary misfit might run to! -- I'll start with Radio Shack.
 

Offline gmcTopic starter

  • Regular Contributor
  • *
  • Posts: 136
  • Country: gb
Re: Trying to determine what voltages should be in a circuit
« Reply #13 on: September 27, 2019, 05:52:40 am »
Thanks for all the feedback. I've ordered the correct mosfet and will replace and re-measure all the voltages and re-check all the wiring.

I see a few people have build this circuit without issues so I'm hoping the circuit is correct. Msybe my dodgy wiring. I'll triple check everything again.

 

Offline gmcTopic starter

  • Regular Contributor
  • *
  • Posts: 136
  • Country: gb
Re: Trying to determine what voltages should be in a circuit
« Reply #14 on: September 27, 2019, 07:33:16 am »
Problem solved.  :-[

Traced everything again. Lesson learned. Trace where things should be going, also where they  should not be going. I built the circuit on the stripboard and missed in cutting one of the tracks so there was a connection on the fet going where it wasn't supposed to. 

Thanks again for the pointers. When I get the proper fet I'm sure it will all work.  I'll retest all the components as well, hopefully it didn't blow anything.
 
The following users thanked this post: AVGresponding

Offline AVGresponding

  • Super Contributor
  • ***
  • Posts: 4931
  • Country: england
  • Exploring Rabbit Holes Since The 1970s
Re: Trying to determine what voltages should be in a circuit
« Reply #15 on: September 27, 2019, 12:52:22 pm »
As Dave is fond of saying "you don't learn anything if it works first time".

Some weird and twisted individuals have more fun when things stop working. I'm one of them  ;D

Glad you figured it out!
nuqDaq yuch Dapol?
Addiction count: Agilent-AVO-BlackStar-Brymen-Chauvin Arnoux-Fluke-GenRad-Hameg-HP-Keithley-IsoTech-Mastech-Megger-Metrix-Micronta-Racal-RFL-Siglent-Solartron-Tektronix-Thurlby-Time Electronics-TTi-UniT
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20363
  • Country: gb
  • 0999
Re: Trying to determine what voltages should be in a circuit
« Reply #16 on: September 27, 2019, 12:59:58 pm »
I'm glad you got it working. I hope you realise there isn't much point in regulating the current to the element. Simply turning it on and off will do. The inductor and freewheeling diode, used to make it into a buck converter are unnecessary.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf