voltage drop with scr

[ejonesss]

i dont know if this is the right place to ask so you mods are welcome to move it to the right place.

i was wondering do scr's have a voltage drop like transistors,fets and diodes have?

because remember http://www.endless-sphere.com/forums/viewtopic.php?f=31&t=16085&start=0&hilit=vsx40MD23

i was having the problem that if the power goes out the battery would discharge and eventually fry the converters.

and i found a relay would work to cut off the load to prevent the modules from shorting the cells.

i think an scr would do the same thing .

 

[heathyoung]

Yes, of course. Use your SCR as a latch to drive the relay.

 

[ejonesss]

i dont know if you looked at the post correctly i was asking if i a power scr has any voltage resistance like other switching semiconductors like transistors and fets.

because i found that scr's are cheaper than relays.

the voltage drop is what makes diodes NOT sutable for charging because then the cells are not being charged as well.

Yes, of course. Use your SCR as a latch to drive the relay.

 

[Jeremy Harris]

Yes, SCRs do have a small forward voltage drop. This isn't your problem though, as I think you want a means of turning off the supply and SCRs are very difficult to turn off. If you remove the gate drive an SCR just stays on if there's current flowing though it. The only way to turn them off is to reduce the current to near-zero, when the SCR will reset to it's non-conducting state.

Jeremy

 

[ejonesss]

i think you are thinking of a a triac it is made of 2 scr's with their a and k pins crossed and will latch with dc.

the problem i discovered is that the outputs do not turn off on them converters and i thought i could use a scr instead of a relay because they are cheaper.

Yes, SCRs do have a small forward voltage drop. This isn't your problem though, as I think you want a means of turning off the supply and SCRs are very difficult to turn off. If you remove the gate drive an SCR just stays on if there's current flowing though it. The only way to turn them off is to reduce the current to near-zero, when the SCR will reset to it's non-conducting state.

Jeremy

 

[heathyoung]

So you want to use the SCR as a switch rather than a latch. They do have a forward voltage drop, like any semiconductor.

When the power is off to the +ve terminal of the SCR, they will isolate, but you would need to increase the charging voltage to compensate, (and then compensate for temperature etc etc).

So, short answer is NO. SCR is great for a latch - put a current on the gate, and remove, it will stay in the conducting state until you disconnect the +ve - same as a triac.

 

[Jeremy Harris]

i think you are thinking of a a triac it is made of 2 scr's with their a and k pins crossed and will latch with dc.

the problem i discovered is that the outputs do not turn off on them converters and i thought i could use a scr instead of a relay because they are cheaper.

Nope. I'm not thinking of a triac at all. An SCR will remain in the conducting state for as long as a certain minimum current (the holding current) is flowing between the anode and cathode, once it has been triggered on by a gate pulse. Just take a look at the Wikipedia entry if you don't believe me: http://en.wikipedia.org/wiki/Silicon-controlled_rectifier

If you want to turn an SCR off when it's conducting DC, then you need to find a way to reduce the current flowing through it to a value below the latching threshold. One way to do this is to use a commutating circuit; essentially a capacitor and second SCR that can be triggered to momentarily reverse bias the main SCR.

Jeremy

 

[fechter]

The other way to turn off a SCR is to pull down on the gate with a little more current than the load. This would only work for relatively light loads since the maximum gate current will be a fraction of what the main current is.

When an SCR is on, the forward voltage drop is about the same as a regular diode. You might as well use a regular diode in that case, which would automatically prevent the batteries from draining back into the converter. You can use a Shottkey diode that has less of a voltage drop and will throw off less heat.

Don't those converters have a 'standby mode'?

The other approach would be to use a good FET. These have a very low forward voltage drop and can be turned on and off with very little current. The tricky part is driving all the gates since they would all be at different voltage levels and 3.7v is not enough to really make sure they are on. By the time you get all that working, relays would probably be cheaper.

 

[dnmun]

thanks alot jeremy. also the link in wikipedia to dimmers has the diagram of an AC dimmer showing the circuit diagram and explaining how they work. they also talked about how new dimmer are using IGBT to do pwm of the sine wave to adjust the current that way, and talk about how it is effectively like a switch mode power supply.

i can see how a charger built using this principle to charge batteries in the 90V range such as the current 72V lifepo4 packs using only the IGBT switching the house 120V AC and with fat capacitors, maybe even go use the 240V with full wave bridges to create the same front end as a switch mode. or charging multiple packs in series with integrated microprocessor communication between pack controllers to allow efficient charger use, in this case multiplexing by going in series for the charging.

if that could be developed, it would be a singular step forward in solving the power supply AC/DC conversion for universal charging spots, without the EFI problems.

 

[heathyoung]

You can do 'chopping' in this fashion, but its an ugly ungainly way of doing it. Not to mention an extremely (electrically) noisy way either.

When I was just starting out in SMPS electronics, I made one like this - wiped out FM reception in a 30m (100ft) radius. You need an inductor to do this properly (buck topology) - and if you are thinking of doing it unisolated - life insurance.

Fetcher is closer to the mark with his suggestion - use a logic-level FET, use the output of the converter (through a diode) to switch it on. Logic level FET's are pretty cheap,

 

[dnmun]

i was just following up on the stuff in the wikipedia article on dimmers.

i assumed that the IGBT is somehow switched fast enuff, or maybe it could be slower than SMPS, and that is how they get the sine wave potential reduced. i did not really know how it would work, would be neat if there was someone familiar with what they were talking about in that article.

if they are switching the input to the IGBT, then the bipolar could be modulated but would not have to be switched on and off as they do with the power mosfet in the front end of the SMPS. and then if they did not need the inductor to push current into the back end, the efficiency could be higher, parts count lower.

30 meters is pretty good, i have to keep my chargers away from my digital tv antenna or it loses signal too.

but if it is possible to use IGBT in some simple form to manage the charging such that the costs and the complexity of the charger are minimized, i think that would really be significant. jmho