AC rectifier: Using Power diodes, thyristors, or SCR's

[SnickersFS]

In 3 phase and poly phase fullwave rectification as seen here:

http://www.allaboutcircuits.com/vol_3/chpt_3/4.html#03267.png" class="bbc_link" target="_blank" rel="noopener noreferrer">http://http://www.allaboutcircuits.com/vol_3/chpt_3/4.html#03267.png

Suppose 1 phase is 31 volts and 12 amps pre rectification.

Can I use either Power Diode, a Thyristor, or an SCR for rectification?

How do I choose each for the requirements of the rectification; like 600volts 25 amps?

According to allaboutcircuits.com textbook, SCR's are not the best choice for fullwave AC control as it alters the waveform. But unsure.

Any help greatly appreciated,

SnickersFS

[scottsAI]

I assume you read about diodes: http://www.fieldlines.com/story/2005/3/10/202925/513

Other devices, like SCR, MosFET, IGBT etc can do the same job, typically with two issues:

First, requires control circuit to actively trigger conduction called active rectification.

Secondly, cost more than a diode.

Diode current ratings must be 2x worst case. Life tends to surprise you. 2X helps. If you really think you know what your doing then 1x, as a designer the cost difference must be large for me not to use at lease 1.5x.

Diode voltage rating, this can be surprising. 12V wind turbine with 5mph cut in. Again diodes must be rated for worst case voltage. Wind turbine is outputting 12v at 5mph wind, at 10 24v, 20 =48v etc. Furling hopefully will limit the top RPM at some point in the 20's MPH. Lets say 25, voltage is 60v, Use NO less then 120v diode. Price of 200v or 400v is not much. Exceeding the voltage rating burns up the diodes in most cases. With wind turbine connected to the battery, the diodes will never see the larger voltages. Wires break, stuff happens. With a clear mind could not used diode below 200v in a 12v system. 400 in a 24v,  600v in 48v. Diodes rated above 600v start costing more.

With line losses and wind turbine matching issues, worrying about the voltage drop in the diode is wasting you time/money. To understand why will require understanding the whole system. Complex.

Have fun,

Scott.

[SnickersFS]

From what I read,

the SCR's Gate is used to turn off the conducting of current.

Applied voltage leaves the SCR "always on" unless there is a voltage drop and polarity change (as in AC).

In a DC circuit, the Gate would be required to shut off the SCR to prevent current from flowing.

SCR's use about 20W for the 2n6508G.

Thyristors seem to be a heavy duty version of a power diode, but unsure.

I haven't come across much info on them.

I'm using 500rpm's as upper limit for the calculations. 31 Volts and 12 Amps per phase.

On the other side would be 62 Volts and 36 amps.

Am I sizing Diodes for pre-rectification or post-rectification?

[scottsAI]

SnickersFS, check again: http://www.allaboutcircuits.com/vol_3/chpt_7/5.html

SCR are a latching device, requires triggering to turn on, Once ON the normal way to turn off is to stop current flow. Above link suggest is can be done “sometimes” by the gate.

Thyristor = two back to back SCR, used in lamp dimmers, each cycle it's triggered to conduct. By controlling when triggered the effective power can be controlled. Works fine on Resistive loads and some motors.

PM alternators, output voltage is directly proportional to RPM. 5Mph cut in, then at 10mph the output voltage is twice. Wind turbines, power is determined by the wind. The RPM will NOT double in this case because the battery holds the voltage and RPM down, the blades will spin a little faster at 10mph yet not at double their speed.

Diode voltage rating should be based on worst case. Freewheeling wind turbine blades.

AC voltage is measured in RMS by most meters (or it pretends to) actual peak voltage is 1.4x.

This peak voltage is how a wind turbine can start charging at lower RPM, and is the voltage the diode must be rated to handle.

Diodes current is based on line current or the current it sees, not the batteries.

Have fun,

Scott.

[SnickersFS]

http://www.allaboutcircuits.com/vol_3/chpt_7/5.html#03222.png

I guess the picture does not show a trigger circuit that is required.

Quote-->SCR----"In the normal "off" state, the device restricts current to the leakage current. When the gate-to-cathode voltage exceeds a certain threshold, the device turns "on" and conducts current. The device will remain in the "on" state even after gate current is removed so long as current through the device remains above the holding current. Once current falls below the holding current for an appropriate period of time, the device will switch "off". If the gate is pulsed and the current through the device is below the holding current, the device will remain in the "off" state."

http://en.wikipedia.org/wiki/Silicon-controlled_rectifier

Quote-->Thyristors----"In a conventional thyristor, once it has been switched on by the gate terminal, the device remains latched in the on-state (i.e. does not need a continuous supply of gate current to conduct), providing the anode current has exceeded the latching current (IL)."

http://en.wikipedia.org/wiki/Thyristor

Apparently I had misread/misunderstood something.

I guess that leaves just Diodes.

So for the Diodes I am looking for 200 volt, 400 volt or 600 volt with an amp rating of at least 25 Amps?

Should I look for something with a higher amp rating, 35A, 40A, 50A?

Thank you for your response, it is helping my understanding very much,

SnickersFS

[betwixt]

EEK - Thyristor is NOT back to back SCRs.   Thyristor is an alternative name for an SCR.

The component that IS like a back to back SCR is called a Triac.

Brian.

[scottsAI]

Yes. Changed what I was going to say, forgot to change it to Triac!

[scottsAI]

SnickersFS, what's your power source?

"Suppose 1 phase is 31 volts and 12 amps pre rectification."

Using that as an example 25a diodes would be a good choice. 200 to 400v PIV rating.

Diodes can be used in parallel best if the same, sum current rating of each for total rating.

At these lower values 25 to 35a diodes are fairly cheap.

Diodes dissipate heat, 25a = 25w, Heat sink is REQUIRED.

Size of heat sink will depend on airflow, actual current etc.

Have fun,

Scott.

[dnix71]

I don't trust triacs. I used to pick up microwave ovens and repair them as a hobby and give them away to thrift shops or a local preschool.

One of the bigger ovens marketed by Sears had a triac feeding the transformer instead of a mechanical microswitch. As soon as I shut the door, magic smoke began to rise from a control board just behind the touch panel. The smoked resistors were easy enough to replace, but there was nothing burnt looking about the triac. It just didn't work.

The gate insulation had failed and it was allowing the line side to backfeed the gate supply, which was only supposed to be a few volts.

The triac didn't cost much to replace, but I didn't like the design because it wasn't failsafe.

[SnickersFS]

Hmmm...25 600v 1A diodes in parallel.

Hows this look:

http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=VS-25F40-ND

Has a stud which can be mounted to a heatsink for cooling.

Another style:

http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=VS-HFA25PB60PBF-ND

Fun is my middle name!:D

[scottsAI]

Yes, they will work fine. Other diode configurations:

Consider bridge rectifiers: contain 4 diodes for single phase or 6 for 3 phase.

Single phase:

http://cgi.ebay.com/new-DIODE-BRIDGE-RECTIFIER-model-KBPC2506-25A-600V_W0QQitemZ270456306666QQcmdZVi
ewItemQQptZLH_DefaultDomain_0?hash=item3ef873bbea&_trksid=p3286.c0.m14

OR: http://www.forcefieldmagnets.com/catalog/product_info.php?cPath=22_30&products_id=98

3 phase requires two single phase bridges rectifiers.

3 phase bridge:

http://cgi.ebay.com/35-AMP-3-THREE-PHASE-BRIDGE-RECTIFIER-WIND_W0QQitemZ250508004446QQcmdZViewItemQQ
ptZLH_DefaultDomain_0?hash=item3a5370cc5e&_trksid=p3286.c0.m14

Best yet!

Have fun,

Scott.

[SnickersFS]

Thank you for answering. I have been unsure about sizing diodes.

My goal is to attempt more than 3 phases, is why I want to build my own AC rectifier.

This has been most enlightening.

A couple of other forums I have a post in got no answers unless I made reference to free energy project, I was baiting really; which got it moved out of the project section and still no answers to my original question. They are nutty in the opposite direction of the free energy crowd.

SnickersFS

[scottsAI]

Your welcome.

The more you tell us the better we can help. Without writing a Book!!

All to often a new guy will not tell us what he is doing, thinking it's some great new thing. Later find it was done a long time ago, so so results.

What you see here is the very well though out solutions.

Once you truly understand "why", then going off with your ideas may work. (at least a better chance).

Have fun,

Scott.

[Ungrounded Lightning Rod]

Except that it's not exactly two back-to-back SCRs.  It's more like (but not precisely) a parallel combo of an NPNP and a PNPN SCR, so the trigger electrode is referenced to the same power electrode for both current directions.

[Ungrounded Lightning Rod]

Triacs don't have "gate insulation"  Perhaps it was some other sort of switch, such as an insulated-gate bipolar transistor (which is essentially a bipolar transistor with a FET to drive the base), its triac equivalent, or some other sort of composite device.

[Ungrounded Lightning Rod]

Paralleling diodes, because of their wildly nonlinear voltage/current response and slight process variations, requires some care in design and a little balancing resistance to keep the current from concentrating on some of them and exceeding their rating.

Search the board for other posts where this issue is described.

Or get diodes that have a high enough current rating to stand alone, so you don't have to spend time engineering this issue.

[Ungrounded Lightning Rod]

Diode voltage rating, this can be surprising.

Don't forget that diodes see the SUM of the peak generated voltage with the load voltage as the voltage they have to block.  So take what the wind genny would produce in a storm and double it.  This lets it handle the case where you've overcharged the battery into running dry or polarization and it's acting like a capacitor and riding the voltage peaks.

There's no point, once you've destroyed the batteries and maybe the load with overvoltage, of THEN also destroying the diodes, having them fail shorted, and having the high current resulting from the diodes shorting the mill in a high wind burn out the genny too (or of starting a fire in the battery shed once the hydrogen bubbles rise off the plates and the battery starts driving hundreds of amps into the diodes).