It’s the mid 1990’s and I’m a design engineer for a corporation that designs and manufactures customized measurement techniques and excessive energy electronics. Our clients vary from experimental fusion reactor operators to electrical utilities.
In the future my boss tells me to affix him within the convention room to satisfy with some individuals from a mass transit automotive producer. He says they’ve points with one in all their suppliers’ product and request our help.
We meet the subway automotive producer mission supervisor and an engineer from one in all their suppliers. The product is a brand new crimson tail mild put in on subway vehicles. Opposite to its title, the tail mild is put in at each ends of the vehicles as a result of most subway vehicles run in each instructions alternatively. The issue is that the sunshine works OK a couple of instances after which fails.
The sunshine meeting incorporates each the LEDs and the digital provide to energy it. They present us the schematic, elements of it are right here:
A part of the unique schematic, drawn from reminiscence.
The enter provide is 74 volts DC and comes from a battery. The unit is a non-isolated DC to DC step-down flyback converter that powers the built-in LEDs. It makes use of a controller IC, U1, with a separate energy MOSFET, the MOSFET isn’t proven in my schematic. The MOSFET switches the first of transformer-inductor, T1. There are two secondary windings on the transformer. The primary winding goes to a rectifier, D2, and filter, not proven, after which to the LEDs, not proven. The auxiliary winding is rectified by D2 and filtered by C1. It then powers the controller IC. The primary load is a series-parallel association of crimson LEDs consuming 15 watts or so.
The controller can’t be powered straight from the battery provide as a result of the voltage is above the IC most provide worth. Additionally, because it sends gate sign to the MOSFET, it can’t be powered at greater than 20 V or it would destroy the MOSFET.
The tail mild designers figured that, in the event that they join U1’s Vcc pin to the provision by way of a collection resistor-shunt Zener circuit, the ability misplaced to feed the IC will probably be greater than 2 watts and it’ll decrease the overall effectivity by about 15%.
In order that they determined to make use of a start-up circuit that will energy the IC momentarily and that when the converter is operating, the IC will probably be powered from the converter’s output. In addition they must shut-off that start-up circuit. They used a collection resistor, R2, linked to the collector of a small sign excessive voltage BJT, Q1, with a enough voltage and present score. Transistor Q1’s emitter is linked to the IC provide pin internet, the identical that the second secondary winding powers. The bottom of Q1 is linked to a Zener, D1, biased by a excessive worth resistor, R1, from the enter energy. As soon as the converter has began and is operating, the auxiliary winding voltage is rectified by D2 and filtered by C1. The Vcc will increase above D1 voltage and biases Q1’s base-emitter junction off. This cuts off the present and brings the start-up circuit losses to virtually zero. General, half a dozen small sign elements value a lot lower than one or two excessive energy elements, so the answer they selected for the start-up circuit is sensible.
The issue is that Q1 fails repeatedly. The converter begins up quickly and the designer assures me that not one of the elements overheat.
I instantly see the supply of the issue, however I preserve my mouth shut. If I inform them the answer now, they may pack up and never pay us for our experience, arguing that the time we spent isn’t well worth the paper work to pay us. My boss says we’ll work on it urgently. They depart us a working unit and the schematics.
As soon as they’ve left the constructing, I speak to my boss. Designing the start-up circuit is a collection of compromises: choose a Zener worth that’s above minimal working voltage for the IC, compute the variety of activates the transformer to provide a voltage above the Zener however not too excessive or it would harm the MOSFET. Including to the issue is that the converter is regulating the principle output present and never the auxiliary output voltage and has to function over the entire vary from the minimal to the utmost battery voltage.
If the operating voltage from the auxiliary winding is way greater than the Zener voltage, Q1’s reverse-biased base-emitter junction can breakdown and destroy the transistor. Most Vebo is usually 6,0 V for the MPSA42, a transistor just like Q1, and is typical of many BJTs [1]. If the emitter-base junction reverse present stage is low, the junction behaves as a Zener diode. With a better present and an extended period, the beta drops and the noise will increase when regular operation is resumed. If the reverse present is extreme, then the transistor fails [2] [3], and that’s the reason Q1 fails on this product.
I inform my boss that the answer is easy, add a collection diode, D3, between Q1’s emitter and the controller Vcc pin, see the modified schematic:
A part of the modified schematic, drawn from reminiscence.
He agrees with me. I modify the unit. I take a look at it many instances with on and off cycles, over the entire battery voltage vary and the unit runs completely.
Late the following day my boss telephones our shopper to say that now we have recognized the issue and have a easy and low value answer for him. After every week our shopper informs us that the producer has confirmed our answer to the issue and all they’re each very happy with our work.
The teachings discovered are:
1- watch out for reverse biasing the base-emitter junction of a BJT, 2- do implement rigorous design opinions and use a listing of things to test, the record right here is an efficient begin [4], 3- suppose earlier than giving solutions to clients, there’s at all times cash concerned, your boss pays your wage with buyer’s cash, don’t give it away, and most significantly 4- by no means let the client appear to be a idiot.
[1] ONSEMI, datasheet MPSA42: NPN Bipolar Small Sign Transistor, February, 2013 − Rev. 8. https://www.onsemi.com/pdf/datasheet/mpsa42-d.pdf [Page retrieved 2022-06-21].
[2] Pease, Robert A. Troubleshooting Analog Circuits. Butterworth-Heinemann, 1991. Web page 77.
[3] Motchenbacher C. D. and J. A. Connelly. Low-Noise Digital Programs Design. John Wiley & Sons, Inc. 1993. web page 133.
[4] Wallace, Hank. Electronics Design Guidelines, http://www.jldsystems.com/pdf/Electronicspercent20Designpercent20Checklist.pdf [Page retrieved 2022-06-19].
Daniel Dufresne is a retired engineer and has labored in telecommunication, mass transit, client merchandise, and excessive energy digital design. He additionally was a professor at a Cegep de Saint-Laurent. He earned his bachelor’s diploma from Ecole Polytechnique de Montreal. He lives in Montreal, Canada and nonetheless works on digital tasks and repairs digital take a look at tools.
Associated Content material
