Author Topic: Transistors not delivering full power issue  (Read 1586 times)

0 Members and 1 Guest are viewing this topic.

Offline mitrynicolaeTopic starter

  • Contributor
  • Posts: 45
  • Country: ro
Transistors not delivering full power issue
« on: March 31, 2019, 08:43:45 am »
Hello everyone,

I have an issue regarding the output stage of a power supply. Regardless the output transistor used or the configuration of the regulator the transistors refuse to let the full power through them. I have attached a schematic which include both of the connections variants(using BTJ or N channel mosfet). Please note that the transistors used are MJ11028G or IXTK90N25L2 (which are more than adequate for this job) but the design application does not have them in the library. I have abstracted the regulator since the issue is not in that part (or at least that's what I suppose).
Few things regarding the setup:
- the regulator has a loop option in which it tries match the input voltage with the set one (yes using op amps). No mater how high I set the input the output won't pass certain limits described in the schematic.
- the regulator has also a buffer option (open loop) in which the input is ignored and a proportional voltage to the set one is delivered to the output transistors. Again setting the full throttle won't do the job.
- you will now say "yeah, but its normal, your external power supply cannot deliver the required amps and the voltage drops". This is not the case. The external power supply is a Korad 3005p which can deliver 30V at 5A which is more than enough for the job. Using the BTJ (when the measured voltage is 12.1V) the external PSU report a 1.21 amps used. Also changing the 10 ohms resistor with a car headlight bulb rises the current drained to over 4.5amps but the voltage is still low which again indicates that the output transistors are not the bottleneck.
- there is no measurement issue. I have measured the voltage with both an external multimeter and also using the internal circuitry of the regulator. Both reported the same values (+ - some milivolts which can be ignored)

1. At 20V I am expecting 2 amps consumption but still the voltage should not drop. Why it drops?
2. Why changing output transistors should have any effect on the output voltage?
3. Is there any issue in how I am measuring the voltage?

P.S. Again I believe there is no issue with the regulator since it can deliver on the output @15V at 100mA which in my view should be more than enough to drive the transistors (at least the mosfet).
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: Transistors not delivering full power issue
« Reply #1 on: March 31, 2019, 09:01:38 am »
It appears the regulator is going open loop. What's the maximum voltage it can output?

You have an emitter (BJT) source (MOSFET) follower. The voltage on R1/R2 will always be lower, than the base/gate voltage. The transistor starts to turn on, when a certain threshold between the base and emitter/gate and source is exceeded. The voltage on R1/R2 rises, thus reducing the potential difference between the base-emitter/gate-source and turning the transistor off. The circuit settles with the voltage on the emitter/source a bit lower than the voltage on the base/gate.

BJTs have a turn-on threshold of about 0.7V (double that to 1.4V for a Darlington pair such as the MJ11028G) and MOSFETs a couple of volts, hence why the voltage across R2 is lower, than that across R1.
« Last Edit: March 31, 2019, 09:03:19 am by Zero999 »
 

Offline mitrynicolaeTopic starter

  • Contributor
  • Posts: 45
  • Country: ro
Re: Transistors not delivering full power issue
« Reply #2 on: March 31, 2019, 09:11:09 am »
Nice catch @Zerro999. It can be. I will test it in open loop configuration with common emitter (load connected to collector) and I will let you the result.
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: Transistors not delivering full power issue
« Reply #3 on: March 31, 2019, 09:24:56 am »
Have you measured the gate/base voltages?

I estimate they're about 13.5V.
 

Offline mitrynicolaeTopic starter

  • Contributor
  • Posts: 45
  • Country: ro
Re: Transistors not delivering full power issue
« Reply #4 on: March 31, 2019, 09:33:58 am »
I have good news and bad news.
The good news is that you spotted the bug correctly. Moving the load to the collector fixed the bug and now I the amps are at the correct value. The bad news is that now I cannot measure the input voltage because the circuit will enter in short circuit. Please check the attached schematic. Maybe you have a proposal. Differential amplifier is not an option since the circuit is already on PCB.
 

Offline mitrynicolaeTopic starter

  • Contributor
  • Posts: 45
  • Country: ro
Re: Transistors not delivering full power issue
« Reply #5 on: March 31, 2019, 09:38:51 am »
@Zero999 Actually the voltage is 14 volts. Anyway you have found the problem. Now maybe you can help me solve the sensing issue that is now raised.
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: Transistors not delivering full power issue
« Reply #6 on: March 31, 2019, 10:19:18 am »
What are you trying to do?

It's not clear.

Are you trying to boost the current or voltage output of your regulator? An emitter follower can only ever boost the output current. It will reduce the maximum output voltage by a few volts, as you've just discovered.
 

Offline mitrynicolaeTopic starter

  • Contributor
  • Posts: 45
  • Country: ro
Re: Transistors not delivering full power issue
« Reply #7 on: March 31, 2019, 10:38:42 am »
Actually both should be boost up. The regulator, as its name suggest should only be concerned of keeping the output voltage or the output current at a set level. The final design should be a PSU. That is why I have abstracted the regulator on the schematic. His job is to measure the voltage across the load and adjust it through the base of the transistor in order to make sure that it is according to the set value so how it is working internally it is irrelevant for the new bug. If maybe you are doing OOP (object oriented programming) you can see this as an interface, where the regulator has a common terminal (to ground) a positive output and a positive input. My concern now is how to link this interface to the output interface (which should also have a common terminal, an input and a feedback terminal). If you have any suggestion of how to link the output transistor or how should I connect the input terminal don't hesitate to tell. Even if it is wrong, it could lead to other ideas that maybe can solve the issue.
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: Transistors not delivering full power issue
« Reply #8 on: March 31, 2019, 06:46:01 pm »
Actually both should be boost up. The regulator, as its name suggest should only be concerned of keeping the output voltage or the output current at a set level. The final design should be a PSU. That is why I have abstracted the regulator on the schematic. His job is to measure the voltage across the load and adjust it through the base of the transistor in order to make sure that it is according to the set value so how it is working internally it is irrelevant for the new bug. If maybe you are doing OOP (object oriented programming) you can see this as an interface, where the regulator has a common terminal (to ground) a positive output and a positive input. My concern now is how to link this interface to the output interface (which should also have a common terminal, an input and a feedback terminal). If you have any suggestion of how to link the output transistor or how should I connect the input terminal don't hesitate to tell. Even if it is wrong, it could lead to other ideas that maybe can solve the issue.
No, how the regulator works is important. If its maximum output voltage is only 14V, then the maximum output voltage will be a couple of volts less with an emitter follower and much less with a source follower. An emitter/source follower cannot boost the output voltage, only the current!

To get a higher output voltage, a voltage gain stage is required. The easiest way would be to connect the regulator's feedback pin directly to the output and as another op-amp with the appropriate gain, before the Darlington pair.

https://electronics.stackexchange.com/questions/107045/whats-wrong-with-an-opamp-as-a-voltage-regulator?rq=1

The trouble now is there will be no over-current protection. The emitter follower will always provide as much output current as the power supply or Hfe will allow.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf