Talk:Active rectification

Bridge power dissipation graph
The graph in this article is misleading. The power lost in the non ideal FETs used in the graph will scale with I2R, as shown. However, the assumption is made that the diodes in the graph are ideal (unlike the FETs) - in reality the resistive element in a nonideal diode will have the effect of increasing the diode forward voltage, causing extra power loss. In the end, the diode's internal resistance will always be in the same ballpark as that of a good MOSFET, so coupled with the forward voltage, lines showing the power loss in a well designed diode-bridge rectifier below those of a well designed MOSFET active rectifier would never be the reality.

For the same cost, a given diode rectifier may well waste less power, as a high RDS(on) FET would be required to compete with the diodes on cost. However you would not look at using an active rectifier based on component cost in the first place (although the extra efficiency might save overall system design cost due to the lower heatsinking requirement, less forced air cooling, and lower space requirements that a more efficient design brings) Therealscaryjeff (talk)  —Preceding undated comment added 21:40, 11 May 2010 (UTC).

BJT
Could anyone verify use of BJTs in this function? It seems unlikely that they should offer any advantage over a diode since they obviously contain an internal junction and cause a voltage drop much as diodes do. —Preceding unsigned comment added by 77.252.115.250 (talk) 09:30, 24 March 2011 (UTC)


 * A BJT's saturation voltage can be significantly below the typical 0.7V forward voltage drop of a diode. That said, in saturation a BJT's base current can also be significant (and you can't just draw the base current from the collector node Darlington-style because the collector voltage is below the base voltage in the saturated condition). Bernd Jendrissek (talk) 02:16, 23 October 2011 (UTC)


 * Exactly so. It is wrong to think that there's a junction between emitter and collector, since there are two, of opposite polarities that cancel, to first order (the collector junction runs reverse biased).  Dicklyon (talk) 02:41, 23 October 2011 (UTC)

History
I'm not sure if this is the right page to explain historic mechanical rectifiers, that were active and normally synchronous. What do you think? Tabby (talk) 05:20, 5 December 2011 (UTC)
 * Sure, why not?  Oh, we do have Mechanical rectifier; could use a brief mention here and more there.  Dicklyon (talk) 06:10, 5 December 2011 (UTC)

Suitability of MOSFETS at higher voltages
In my opinion it must be stated that the practical use of a MOSFET is only useful when low voltages or low currents are used. MOSFET resistance increases significantly faster than their voltage rating, so using them at high voltages (say 200V or more) only delivers dissipation reductions at a fairly low current. 82.139.114.136 (talk) 15:22, 27 February 2012 (UTC)
 * So put more in parallel if you want to save more power at high currents. Dicklyon (talk) 16:10, 27 February 2012 (UTC)
 * SiC MOSFETs could be used in high voltage situations, but they usually have high RDS_on, often about 25mOhm, for a 90A, 1200V rated MOSFET. 10-13mOhm for modern new designs from around 2018 and 2019. But in high voltage applications I see it is often beneficial to stuck just to the SiC Schottky diode, instead of doing active rectification, as the diode will usually drop about 1.7V (1.4-1.8V) at 25 deg C, which in 1000V application isn't much. Other option is to put Multiple mosfets in series to handle high voltage, but obviously that is extremely tricky, and very prone to failure (if only one mosfet turns on, the second mosfet will see full voltage on DS!), plus the high side mosfet is hard to drive, because drive voltage itself needs to be high (from DC). I have seen 8kV systems with 10 MOSFETs in series for example in real life, but it requires a lot of design attention, filtering and balancing, and often is only used for pulse applications. In some other applications, it is possible to use them for continues loads. It is similar to Cascode. Usage of MOSFET + diode in parallel + extra diode in series, could help a little bit, and in some middle range could be beneficial. 2A02:168:F609:0:6EB4:8A5F:9FC4:E8FB (talk) 14:25, 22 July 2019 (UTC)