Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.
Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.
Hi pg,
This looks like your favourite book again.
The comment you make against the 'power gate trigger' is OK. What is important with the SCR is that the gate circuit is tied to ground through a low impedance. The component should be a resistor and possibly a diode if there were high negative voltages around. generally, there is no need for a diode, but there needs to be a resistor.
With regard to the second question in blue, this isolation is important (necessary) if the trigger circuit is driving two SCR's, for example in a bridge circuit arrangement. For example, I have a DC motor controlled by a full wave bridge. On one half of the bridge are two diodes. On the other half of the bridge are two SCR's. Each SCR gate is driven from a separate winding on the coupling transformer just like in your diagram with the uni-junction transistor pulse generator. In my drive circuit, the SCR gates are grounded effectively by the DC resistance of the transformer windings and no diode OR resistor is used. This drive runs on 480 volt single phase.
Question 3 I cant answer cos the switch ON time of an SCR is uncontrollable. It is due to the design of the device but is VERY short.
Question 4. This is again a bit confusing. The gate needs voltage and current to force the SCR into conduction. The driver circuit needs to supply enough power to do the job. Simply put, a small power source may not do the job but a big enough one will do the job. I suggest you go back to the BT 150 series datasheet and have a look at the gate drive requirements.
Hope this helps.
Q1: The diode looks to me like a protection diode to prevent negative pulses from getting to the device. That seems to agree with the book says too.
Q2: I think you don't want to have interaction between the gates. There are capacitive effects and other parasitics in these circuits. Isolation is a general method of reducing ground problems, transient issues and coupling interactions. This kind of stuff is not going to be apparent just by looking at simplified schematics. You encounter the problems and appreciate the solutions in real work where you try to build the design and find the SCRs triggering randomly or not at all. Then you trouble shoot and discover these problems and implement solutions.
Q3: It looks like the output voltage is driving a transistor base-emitter directly which gives an exponential nonlinear response on the output of the transistor. This will naturally speed up the pulse and decrease the rise time.
Q4: I'm not sure enough to answer this. I find the wording confusing relative the the circuitry shown. But, this can be tricky stuff that requires hands on experience. Hopefully someone else here can make sense of this section.
I love your book cos its yours and with the others on this site, maybe we can one day write an 'addendum/errata' volume. BUT, the book makes us think.
Your question raises the ugly side of SCR's. "very difficult to turn off".
In control circuitry using SCR's where they are driven from DC supplies, the switching OFF of these monsters requires a thing called 'FORCED COMMUTATION'.
The ONLY way to turn them OFF is to force a reverse current through the thyristor to reduce the anode current to zero, AND, at the same time make very sure that the gate circuit is well and truly shutdown, locked off, shortcircuited, expunged, whatever; because when the SCR actually goes into its forward blocking state, the anode voltage will rise, at a very great rate, and this rise in forward voltage will allow parasitic currents to flow in the SCR which, MAY trigger the gate circuit into conduction.
Having a look at the data sheet for the BT151, it says, under 'characteristics';
"rate of rise of OFF state voltage that will not trigger any device" Rgk = open circuit; dV/dT<50 v/microsec.
AND for Rgk = 100 ohm; dV/dT <200 V/microsec.
These two values given for gate to cathode resistance plays into your previous question about the reverse diode. An SCR needs a LOT MORE than a diode to shut it off in forced commutation duty.
Now to your question.
When the pulse is applied to the transformer, the current does flow in the direction of the red arrow. The pulse transformer has its own secondary circuit comprising the winding upward through the SCR and through the capacitor (which is discharged because the SCR is conducting) and then back to the pulse transformer. When the anode current falls to less than the holding current, the SCR blocks and the anode voltage rises to Edc. The pulse current also stops flowing because the SCR is now open circuit.
You can see that the capacitor is required to return the current to the pulse transformer.
As an aside, in the diagram the load is shown as a resistor. BUT, frequently the load is an inductor or a motor, and iF there is a commutation failure, (the SCR doesnt switch off)then the inductive load looks like a short circuit and then the fuses blow and if its a big motor, even the street light go dim.
This explains the beauty of IGBT's. (SCR's that can be turned OFF)
Hope this helps.
So, it means the SCR in that circuit does let the current flow in reverse direction for some time so that the current pushed by transformer has a complete path to flow, transformer->SCR->capacitor->back to transformer. I had thought that reverse current (from cathode to anode) cannot flow through a SCR.
So, it means the SCR in that circuit does let the current flow in reverse direction for some time so that the current pushed by transformer has a complete path to flow, transformer->SCR->capacitor->back to transformer. I had thought that reverse current (from cathode to anode) cannot flow through a SCR.
It will for a while, lookup 'reverse recovery time' in your text. After that you have a 'forward blocking recovery time' to wait before the device is completely 'OFF' and ready for the next conduction cycle. It's limited by the recombination speed of charge (the electrons and holes ) at the layers. A 'deep' recombination impurity like gold can speed things up. https://www.onsemi.com/pub_link/Collateral/HBD855-D.PDF
About page 13 on 'Switching Characteristics"
also Section 5 about page 60.
Before you updated your post, I was looking up the term 'forward recovery time'. I think it's the same as 'gate recovery time'.
"Turn−off time is a property associated only with SCRs and other unidirectional devices. (In triacs of bidirectional devices a reverse voltage cannot be used to provide circuit−commutated turn−off voltage because a reverse voltage applied to one half of the structure would be a forward−bias voltage to the other half.) For turn−off times in SCRs, the recovery period consists of two stages, a reverse recovery time and a gate or forward blocking recovery time, as shown in Figure 2.7."
I think you have it quite well now.
As nsaspook says when he quotes an ONsemi publication, its a 'process' that goes through several stages, and the semi conductor manufacturers are the ones who know quite a lot. I try to quote data sheets for you. The BT 151,2,3,4 series are a mixed bag of fast and slow devices. There's heaps for you to read in just those 4 datasheets.
Terminology is a real problem. eg I use the term thyristor interchangeably with SCR. You ask about 'gate recovery, forward blocking recovery, reverse recovery". These terms are all part of the general need to describe dynamic states or transitions from one state to another. Often different manufacturers use different terms for the same thing and sometimes there is a special need to describe a specific parameter which falls outside the usually expressed or defined terms. So one needs to be on guard.
Q4: As the frequency increases the time available for device recovery (the time needed to recover increases with the power level of the device due to recombination effects from stored charge as current flows in the semiconductor junctions) becomes smaller than the required recovery time for proper device operations even with zero current flow. The effect is similar to 'slew rate' limitations in OP amps.
As usual there is a trade off between speed and power in device design and the type of material processes needed to make a device that's fast and powerful. That's usually more expensive than just a couple of highly doped (slow) silicon layers on a large die needed for 60hz lamp dimming.
With Q1, it seems to me to be garbled. 'Source failure during Inversion' sounds to me like 'commutation failure'. A commutation failure occurs when the device is intended to be switched off but it fails to do so. In your previous examples, the method of forced commutation was by a pulse. But in some situations say, with 3 phase motor variable speed inverter drives, it is common for the DC supply to be short circuited through an inductor for a short time interval, and during this time all the SCR's in the 3 phase switching bridge, switch off. If some of the conducting SCR's don't switch off, then when the supply reappears after the short interval, then SCR's are conducting which should be off. Then it is possible to get short circuits. See the last bit of my post #5 about the street lights going dim. I think this is what the author means by 'source failure during inversion'.
Question 2 looks to me like the last three lines are in their wrong columns.
I would say 'Turn on and turn off durations are very small', for a transistor.'
I would say 'Power consumption is very low' for a transistor.
I would say 'Level of output voltage can be controlled' for a transistor.
and similarly, what is in the transistor column should be under the SCR column.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
