The SCR is probably as good as it gets - there may be a case for very low RDSon MOSFETs.

Your trip voltage is around the same ball park as a diac - but most can only handle about 2A peak pulse.

An SCR will boost the current handling, just hook the diac from anode to gate - I'd suggest at least 10R in series with the diac to limit the current until the SCR takes up the load.

diacs are getting harder to come by - many CFLs have one in the start up circuit.

Yes, I wonder how many amps a 40,000 microfarad cap would produce on the initial surge into the 24v battery bank. Is a 10R resistor making this a "wasteful" type of trigger?

 

Yes, I wonder how many amps a 40,000 microfarad cap would produce on the initial surge into the 24v battery bank. Is a 10R resistor making this a "wasteful" type of trigger?

It's difficult to say without knowing the ESR of the cap, the internal resistance of the batteries and the resistance of the wires. But I would say less than 300 amps and it doesn't last long. Probably in the order of 2 ms.

 

It's difficult to say without knowing the ESR of the cap, the internal resistance of the batteries and the resistance of the wires. But I would say less than 300 amps and it doesn't last long. Probably in the order of 2 ms.

Is there an SCR that can handle this discharge? Or somewhere else a suggestion was for an IGBT.

 

Tell us some more about the source of the voltage to charge the cap. It's still not clear to me why you need to store it first unless you just want to pulse charge the batteries.

This project is not mine. I'm enquiring for a contact I have who lives, literally, in the middle of the Pacific Ocean and has limited internet access - so I'm doing the research for him! He definitely wants to pulse charge the batteries from his big cap.

 

Is there an SCR that can handle this discharge? Or somewhere else a suggestion was for an IGBT.

The circuit with the inductor I posted will limit the peak current through the SCR.
The larger the inductor, the smaller the peak current.
For example, 1mH limits the peak current to about 47A.
The bonus is higher efficiency also.

 

The circuit with the inductor I posted will limit the peak current through the SCR.
The larger the inductor, the smaller the peak current.
For example, 1mH limits the peak current to about 47A.
The bonus is higher efficiency also.

Please bear with me - I really don't have much knowledge. Someone here as suggested this cap (40,000 microfarads) could delivery 300 amps when discharging into the battery bank. If your circuit limits current, does it just slow it down through the inductor, but it "gets there in the end"? I don't want to loose any current that should be going into the battery.

 

If your circuit limits current, does it just slow it down through the inductor, but it "gets there in the end"?

Yes, the inductor rather acts like an inertia to the current to lower its peak value.
All the charge still gets transferred, and with better efficiency as a bonus.

A small power transformer low voltage secondary can be used to provide the inductance, if you have one lying around.

 

Yes, I wonder how many amps a 40,000 microfarad cap would produce on the initial surge into the 24v battery bank. Is a 10R resistor making this a "wasteful" type of trigger?

The 10R resistor only protects the diac for the short time it takes the SCR to start fully conducting. The initial current pulse probably wouldn't do the SCR gate much good either.

 

Yes, the inductor rather acts like an inertia to the current to lower its peak value.
All the charge still gets transferred, and with better efficiency as a bonus.

A small power transformer low voltage secondary can be used to provide the inductance, if you have one lying around.

I always thought power transformers weren't gaped and as such were subject to saturation.
It would seem that by the time you account for the cap ESR, the wiring and the internal resistance of the battery the winding resistance would need to be in the order of 50 milli-ohms or less?

 

I always thought power transformers weren't gaped and as such were subject to saturation.
It would seem that by the time you account for the cap ESR, the wiring and the internal resistance of the battery the winding resistance would need to be in the order of 50 milli-ohms or less?

Toroidal mains transformers aren't gapped and are easily prone to saturation if mis used. Laminations usually have scale from heat treating, I think that gives a small amount of gapping on a lamination by lamination basis.

Toroidal transformers usually have a wound core - a long strip of soft iron coiled up to make the donut shaped core.

 

I always thought power transformers weren't gaped and as such were subject to saturation.

Good point.
You would likely need a gapped or air core inductor to handle the high peak current without saturating.

It would seem that by the time you account for the cap ESR, the wiring and the internal resistance of the battery the winding resistance would need to be in the order of 50 milli-ohms or less?

The winding resistance should not be a big factor unless you are really concerned about efficiency.

 

. Laminations usually have scale from heat treating,

That 'scale' is actually done on purpose. It is an oxide coating that has nothing to do with any heat treating, though it is done in a furnace. It's done on purpose to help prevent eddy currents.

 

That 'scale' is actually done on purpose. It is an oxide coating that has nothing to do with any heat treating, though it is done in a furnace. It's done on purpose to help prevent eddy currents.

The books told me its from annealing because work hardened iron would have too much remanence..

The scale comes in handy though...............

 

The books told me its from annealing because work hardened iron would have too much remanence..

The scale comes in handy though...............

A few years ago I bought a core set from Temple, they had a real good tutorial on the process. Just had a look and they no longer have the tutorial on their site. It explained how and why each step was done.

 

Below is the LTspice sim of my take on a circuit with an SCR and an inductor.
V1 and R1 simulate the charge source for the capacitor.
The 7.5V Zener diode causes the SCR to trigger when the capacitor voltage is about 8.8V above the battery voltage.
Diode D2 prevents reverse gate current which showed up in the simulation.
Due to the inductive resonant charging effect from the 1mH inductor, C1 discharges to <18V, about 6V below the battery voltage.
This improves the energy transfer efficiency and also insures that the SCR will be reverse biased and turn off after each charge pulse.

View attachment 128778

Here's a short article I wrote on inductive charging of a capacitor, if interested.

Hey crutschow - Just been revisiting your response here and read your short article (again, I think). It all sound very good. At the end of the article you say, "The tradeoffs are charge efficiency vs charge time and physical inductor size and cost." Roughly how much longer does the inductive process take? I mean, if the process is nearly twice as efficient but nearly twice as slow then the battery charge rate (time) would be nearly the same as without the inductor. I still can't quite get my head around what you say are 50% losses in DC switching. Are you sure about that - how can I verify this - is there are formula? Please pardon my ignorance.

 

I still can't quite get my head around what you say are 50% losses in DC switching. Are you sure about that - how can I verify this - is there are formula?

A capacitor stores a charge of Q = V*C.
The energy from the voltage supply to charge the capacitor is E = Q * V.
Substituting the two equations gives E = CV².
But we know that the energy stored in a capacitor is 1/2 CV².
So where did half the energy go?
It was dissipated in the resistance of the charging source and the wire between the source and the capacitor.
This loss is minimized by using an inductor in series with the capacitor as it stores the extra energy during the charge and then supplies it to the capacitor at the end, minimizing the dissipation in the resistance.

 

But we know that the energy stored in a capacitor is 1/2 CV².

So is 1/2 CV² a fixed "law" that is nothing to do with actually switching the DC? That is, does this apply to a constant charging DC, and to AC?

 

So is 1/2 CV² a fixed "law" that is nothing to do with actually switching the DC?

Yes, it's the energy stored in a capacitor, independent of how it's charged.
The equation is derived by integrating the voltage times the current over time as it discharges.

 

Yes, it's the energy stored in a capacitor, independent of how it's charged.
The equation is derived by integrating the voltage times the current over time as it discharges.

Ah, OK. So your inductor and SCR/diode arrangement reduces the losses in charging. 90+ percent efficient at charging as opposed to the regular 50% efficient - that's obviously the way to go.