Need Advice for High Current, High Voltage, High Speed, Switching

[townsend]

Hello everyone!

I'm new here - just freshly registered. I have a PhD in Neuroscience, with other degrees prior in Mathematics and Computer Science, and prior to that I earned diplomas from college in the areas of Electronics and Electrical. I don't profess to be a formally recognized "Engineer" and do not have the PEng designation, but my formal study of the area at the college level was carried through as I worked on my Undergrad, Master's and Phd, and I do have two patents registered to my name in the area of EEG acquisition electronics.

Put simply, I might not be familiar with all the jargon that PEng's use, but I have a good grasp of the principles and am able to design my own stuff and debug/troubleshoot other peoples' designs. So, please excuse me in advance if I don't seem to follow what you are saying - its just a matter of us speaking different working languages.

So, here's my question:
I am in the middle of trying to implement an idea, but I'll spare you all the details and only include what is necessary to help describe my problem. I have a system operating at about 680V and about 40A, and I need to be able to switch a component in and out about 120 times per second. I am under the gun to get a prototype completed fast, and so instead of working on my own switching mechanism, I chose to use a solid state relay (SSR). I chose one with random on/off (i.e. it doesn't wait for a zero crossing, but switches when you tell it to). I also need optical isolation to simplify the connection to the microcontroller that will be handling the timing.

To guarantee perfect operation, I really need sub-millisecond timing accuracy. I also don't want to break the bank with the project, so I wanted to keep the cost of the switch under $50. Unfortunately, the only solution that I found indicated in the datasheet that the device would switch on (and off) within 1 millisecond of being told to do so.

I was disappointed but checked my calculations and found that although not ideal, this would produce "acceptable" results. Without providing all the details, the potential for the delay will result in the output voltage of my project being generated to be a little higher or lower than desired, but still within acceptable limits.

1/ Can anyone point me to a SSR that can respond within 100usec, is within budget, and meets the voltage and current requirements?
2/ Can anyone point me to a resource to help me design my own switching system so I don't need to use a ready-made SSR?

Regarding question 2 above, I specialize in low current low voltage analog and digital applications, so "power electronics" is not a familiar area for me. That's why I'm here seeking some advice. Thanks in advance!
-gt-

 

[townsend]

Respectfully, your "clever " circuit has been done a million times before, it is easy to construct a circuit that allows charging and then puts the C in series with the source - this is commonly done in industry at a variety of power levels,

as suspected - you have overlooked the currents that may flow in the cap at charge up ( due to source inductance and other reasons ) - and potentially high peaks of current in the load - also if this equipment is operated continuously it needs to meet the requirement for no DC in the mains - this is for earth stake corrosion reasons and others, so there had better be a bridge rectifier in the source side circuit ....

Yes of course, such as the standard voltage double, which everyone is familiar with. The "clever" part is what I do (that I haven't shown) to minimize the number of components required. In cases where you want the capacitor to charge fully during half the cycle and discharge fully during the other half, it is trivial. Only diodes are required to "steer" the current flow to achieve that. If you require a smaller amount of voltage and you don't know the "clever trick" you will need more than two switches to move the capacitor to the places it needs to be to charge and discharge to do the job. Try it yourself and your see. It is not obvious how to achieve the required behavior with only two switches. In other words, the "clever" part is not charging a capacitor to a particular voltage and then discharging it in in series through the load. As you say, that's an old idea. The clever part is getting away with only two switches to do it.

Yes, there is a bridge rectifier supplying that "hump" in my diagram.

--- Updated May 21, 2023 ---

Congratulations for posting the first images after 9 longwinded text messages.

In terms of SSR specs this is an AC/DC type (AC = bipolar output voltage, DC = on/off capability). They are typically implemented by anti-serial MOSFET or IGBT pairs. If I'm misinterpreting the description and the voltage across the switch is only unipolar, a DC SSR type would suffice.

I don't see off-the-shelve 500V/50A (or more) AC/DC SSR on the market, at first sight the effort exceeds your initial $50 mark by a large factor.

Let's assume you want to assemble the circuit from discrete components, then switch selection and driver design are the major tasks, additionally mechanical integration/heatsinking and overvoltage/overcurrent protection as far as necessary.

You already mentioned switching speed requirements, isolation voltage is another important parameter. In case of doubt, separate isolated DC/DC and gate driver are the way to go. There are however gate driver IC with built-in power supply isolation.

--- Updated May 21, 2023 ---

The device I purchased for my prototype was actually a little lower rated: 40A at 480V. It was an Altran Magnetics, LLC product from Digikey and was $50.29. I'd like to "beef up" the design to 500V/50A ... Revisiting Digikey, I see you are right - that will cost a LOT!!!

Thanks for the input.
-gt-

Hi,

with a capacitor in series you can´t avoid those "current peaks" by referring the timing to the AC input voltage.
The correct timing needs to be done via the voltage across the switch.

And still then you need to take care that the capacitor voltage does not drift away. If there is a mismatch in positive and negative halfwave average current, then the voltage of the capacitor may drift (integrate) in one direction. Making the capacitor voltage bigger and bigger ... and the switching currents may become bigger and bigger.

Klaus

You said that the capacitor voltage could get larger and larger, and I have a question about that: Even if the capacitor was not fully discharged when it was placed in the "charging" configuration, how can it exceed the voltage of the charging supply? If the charging voltage is lower than the current voltage on the capacitor, won't the capacitor "discharge" until it reaches the value of the charging voltage, and then follow the charging voltage up until it is disconnected? I'm having trouble understanding how it could ever rise above the source supply during the charging cycle - I might be missing something obvious and perhaps you can help me see it. It behaves as I imagined in the circuit simulator - but that's just a simulator.
-gt-

 

[FvM]

Altran Magnetics SSR is an AC type and can't switch of charging as intended according to post #17 waveform.

 

[townsend]

Altran Magnetics SSR is an AC type and can't switch of charging as intended according to post #17 waveform.

I thought the difference between an AC and DC SSR was just that the AC version allowed current to flow either way through it, while the DC version only supported current flow in one direction. Under that interpretation, you can use an AC relay for either purpose, but a DC relay can't be used on AC. Was I mistaken and if so, what about the internal circuitry would prevent an AC relay from being used in a DC application? Remember, the one I chose was NOT a zero crossing variety, but was the "random on/off" so it isn't the case that it would be waiting for a null in the AC cycle to switch. Rather, it can be turned on or off and any arbitrary time in the AC cycle.

If I had an AC circuit and I only turned the SRS on during the positive alterations of the AC cycle, how would it tell the difference between that and the situation where there are only positive humps? It is turned off where the negative humps would be in that case, so it can't know anything about the polarity of the voltage at those times.

Fedex will be delivering these tomorrow morning, so I can test, initially at a much lower voltage and current, to see what happens, but if I'm wrong about what I expect, I guess I'll need some help understanding what is happening inside of the device.
-gt-

 

[KlausST]

how can it exceed the voltage of the charging supply?

I did no say this.

Example: Simple RC circuit:
Put a series RC on a DC voltage. The voltage at the C will rise and rise.
But it will never exceed the DC voltage.

How your circuit reacts on this ... only you can answer, because we don´t have the according application details.

Klaus

 

[townsend]

I did no say this.

Example: Simple RC circuit:
Put a series RC on a DC voltage. The voltage at the C will rise and rise.
But it will never exceed the DC voltage.

How your circuit reacts on this ... only you can answer, because we don´t have the according application details.

Klaus

Okay, thanks - I understand now. In the application there is only one capacitor, and when in charge mode, it only charges to the specified voltage (first diagram) and in the discharge mode, it adds that specified voltage in series with the source to increase the source voltage output by that amount (initially at least - as it discharges, the added amount diminishes). There's no evidence in the circuit simulator that the voltage creeps higher over time. But again, its just a simulation.

In any case, I DO have the problem you eluded to earlier during the charging cycle. I naively thought that since the charging voltage starts at zero and not some high voltage, that the charging current would gently increase as the voltage rose from zero. This turns out NOT to be the case (at least in the circuit simulator) and even though the voltage is very low, the current is still unacceptably high. It very quickly jumps up to 60 amps and stays parked there during most of the charging cycle. So just as you suspected, this is problematic. During the discharge cycle, the current is only what the load draws and is well within acceptable limits. So the discharge cycle is not a problem. If the device is rated at 40A and I draw 60A for less than 50% of the time, is that going to be okay? In other words the average current is about 30A.

 

[FvM]

Charging current is C*dI/dt, thus highest near voltage zero crossing when starting with discharged capacitor. Capacitance must be < 400 uF to keep 50A Imax.

 

[townsend]

Charging current is C*dI/dt, thus highest near voltage zero crossing when starting with discharged capacitor. Capacitance must be < 400 uF to keep 50A Imax.

Thanks for the formula! A formula is worth a thousand words. So my dilemma now is being able to store enough energy to provide the higher series voltage (big capacitor) to feed the load while not exceeding the desired current during charging (smaller capacitor). It seems I may have to use a larger switch and accept a larger current. However what about the duty cycle? The charging current is only present under 50% of the time, so MAYBE I can get away with 60A on a 50A device? I guess it depends on the detailed ratings of the device. Anyway thanks for that.
-gt-

 

[dick_freebird]

A waveform (or pair of) would be worth more than the 1000+
words thus far.

Is 120Hz only coincidence, or part and parcel of the scheme?
Is there perhaps a half-cycle (or close to it) that could just
use a time-delayed TRIAC or SCR to charge? Of course then
your timing is slave to the line voltage cycle and you don't
get your fine timing interest. But how real is that in fact?

 

[townsend]

I'm trying to find a cheap but effective way to change 220VAC into something that can power a 380Vrms three-phase power supply. The supply uses a three-phase bridge rectifier where the output gets filtered and passed on to a traditional high frequency switching power supply. Clearly the equipment doesn't care about the 120 degree phase shift between phases, and does not REALLY require three-phase power, which is not available in my area of town. I was quoted (from multiple sources) upwards of $4000 for a solution to change 220 single phase into 380 three phase. All so that the three phase bridge can turn it back into DC inside the computer!!!!! Well scrap that! I have 10 of these units and I don't want to spend $40K to power them all. If they were three phase motors that actually require three phases, that would be different, but in the end, I just need to provide something that looks like a 380V rms bridge rectified output, which I will feed into the target equipment bypassing the six diode bridge. I'm aware that more filtering will be required than with the original three phase bridge, but that's a separate issue to be dealt with later. So this is my motivation and target. Attached are the waveforms of what I am starting with (from 220 bridge) and what I WISH I could produce overlapped with the "good enough" version that I can produce. The equipment is happy with 342 to 418 RMS. My target is 380 RMS, but my circuit produces 342 which is fine.
(NOTE: Diagram was replaced after editing)

 

[Easy peasy]

Respectfully, a classic boost PFC ckt running off the 220V line will easily provide the 380VDC you say you require - this will then go thru 2 diodes in your downstream psu.

A benefit is that the current from the 220Vac line will be sinusoidal and hence the minimum rms current draw

The classic, ubiquitous, PFC booster has plenty of output capacitance too.

--- Updated May 21, 2023 ---

Also - an auto-transformer, or any transformer with an 160V output can be energised and the output put in series with the 220Vac to provide 380Vac - this is a very quick solution

 

[townsend]

Well, my first thought was to just use a 220 to 380 step up, but in short order I realized I could instead step down to 160 and put that in series with the mains to produce the required 380. However I abandoned that solution the moment I priced out transformers. The rating of the transformer in the case of the 220:380 needs to be upwards of 6.6KVA ... those are not cheap (into 4 digit prices). The 220:160 is not as bad but still not cheap. It would still need to pass the same current, but the voltage is much lower, so it would need to be almost 3KVA ... still be around the $600 mark.

You mentioned the boost power factor correction circuit ... is it adjustable to exactly 380? It can't exceed the required voltage otherwise I would have used a simple voltage doubler and fed it with 480V rms "humps".

I did a quick google search, but I need to spend more time looking at this proposed solution. Thanks for the tip.
-gt-

--- Updated May 22, 2023 ---

I guess I should mention that I can produce a sinusoidal output (which I have done in my circuit simulator) by doubling the parts and using two separate capacitors charging one on every odd numbered "bump" and discharging on the even "bumps" and doing the opposite with the other. This of course requires four switches and opto's and two caps. The cost is double, and since I don't think it is important for the output to be sinusoidal (its just being filtered by the host device) I thought I was happy enough with the waveform I got for half the cost. However, admittedly, the double-parts version does produce 380 rms ... actually a bit more since the start and end of the "sine" (not quite a sine) have much steeper slopes and the area under the curve is a bit larger even though it reaches the same peak value.
-gt-

--- Updated May 22, 2023 ---

Diagram for post #31 attached. This is the version with four switches and optoisolators and two capacitors.

--- Updated May 22, 2023 ---

Hello again, everyone. I have to say I really appreciate how helpful you folks are being!

I wanted to ask about transformers. Member "Easy peasy" suggested just using a transformer, but as I responded, that was my very first attempt at a solution but I abandoned it because the high KVA requirement makes them prohibitively expensive. I had also thought (as the member suggested) about putting the output of a 160V transformer in series the the mains to get 380, but found that expensive too. This is why I went the route of the topic of my posting instead - to avoid a costly transformer. However, the transformer solution is so simple that I hate to give up on it prematurely. So I'm reaching out to see if anyone knows where I can get a 3kVA transformer that steps down from 220 to 160V at a reasonable price? Does anyone have a favorite website they know of where I could look? I've been hunting, but must not be looking in the right places. Even the 3KVA units are looking like 4 digit prices, and I can't even find a 220:160 transformer even if I was willing to pay that much. Surely, there must be some cost effective units out there someplace with the right turns ratio. I only spent $200 total on my prototype. I would be willing to spend more on a nice clean transformer solution, but not up past the $600 mark for sure.
-gt-

 

[Easy peasy]

p.s. an auto transformer is lower sized for the same power, as the primary is tapped above the 220V and not isolated, essentially 220Vac + 160Vac, the current in the 160Vac winding will be 17.34A ( for 6.6kW @ 380Vac ) and the current in the 220Vac wdg will be 12.63 amp, so the total VA of the auto transformer is 5.55VA, for 6.6kVA in total, as some of the energy comes direct from the 220Vac source.

 

[townsend]

p.s. an auto transformer is lower sized for the same power, as the primary is tapped above the 220V and not isolated, essentially 220Vac + 160Vac, the current in the 160Vac winding will be 17.34A ( for 6.6kW @ 380Vac ) and the current in the 220Vac wdg will be 12.63 amp, so the total VA of the auto transformer is 5.55VA, for 6.6kVA in total, as some of the energy comes direct from the 220Vac source.

Ah, yes - I didn't realize that. Thanks a lot for bringing that to my attention. However I'm still struggling to find an affordable transformer. I assume an autotransformer might be cheaper since it sounds potentially easier to construct since it is one big coil with a tap. There's a place I was talking to in Canada close to where I live that quoted me a custom transformer. It was expensive but was a three phase transformer intended on being driven by a 220v-to-380 single-to-three phase converter with a PWM output. The purpose of the transformer was to help smooth out the PWM. I guess I'll reach out to them and get a price on exactly what I want - now that you've enlightened me, I guess I will now ask for an "auto transformer with a 160V winding at 6.6kva". I'll see what the cost would be. I hope someone will point me to a good site that has an already made stock unit since I'm sure it won't be cheap to have one made.

 

[Easy peasy]

Just a caveat about playing with the mains, while the theory may seem straight forward - the practicalities are not, switching any capacitor into another at differing voltages invites fairly large surge currents, turning off currents invites fairly large overvolt spikes for the un-initiated ....

--- Updated May 22, 2023 ---

re single phase auto transformer, be a good idea to quote the currents above - as even Tx wdg places do not calc this correctly ....

 

[Easy peasy]

there is very little new under the sun.

Q1,2,3, have integral internal diodes - or external as required.

 

[townsend]

View attachment 182896

there is very little new under the sun.

Wow, that's for sure:
One capacitor,
Two switches,
... and a "cleverly" placed "bypass" diode.
Yep, that's my "invention" alright! Just replace the IGBTs with black-box SSR's and that's my circuit.
(Actually, in the simulator I used photoisolators feeding into Darlington pair transistors for my SRS's)
Like you said - there is very little new under the sun.

Well, depending on how ridiculously priced a custom wound auto-transformer is, I may still need to pursue this. I'll call the shop tomorrow and see how much they want for it. Thanks for taking the time to dig this diagram up for me.
-gt-
p.s. I might need to use an isolation transformer because my shop has 220V - 20 amp outlets throughout. If I use an autotranformer, I need to install 30A breakers and wiring, since the solution will draw 24 amps. If I use an isolation transformer, I can run each unit from a pair of 20 amp circuits. One will feed 10 amps to the transformers primary and the other circuit will feed 14 amps to the series secondary and out to the circuit. I guess if I get the phase backwards on one of them compared to the other, it will put out 220-160 and no damage will be done. I can probably install a fail-safe relay that only lets the result through to the target unit when 380 is detected. Ironically, the earlier lower powered single phase version of the equipment I need to power comes with TWO power cords attached. So it isn't unheard of.

 

[Easy peasy]

you may have noticed that when Q2,3 turn on there will be significant currents into the equipment, if the following equipment consists of a bridge rectifier followed by storage caps ...

--- Updated May 22, 2023 ---

6600 / 220Vac = 30 amps, so 15 amp from each socket in toto, for perfect power factor of load,

the rms current will be somewhat higher as the waveshape of the current going into the psu's will be quite peaky and far from sinusoidal.

 

[townsend]

you may have noticed that when Q2,3 turn on there will be significant currents into the equipment, if the following equipment consists of a bridge rectifier followed by storage caps ...

Yes, I mentioned that in one of my posts in this thread. It draws 60 Amps during most of the charging cycle in the simulator, and I suspect the only thing limiting the current is the "on" resistance of my Darlington pair. I suppose I could "waste" power by putting in a resistor, or purposely using a switch with a higher "on" resistance to reduce the current. Is there no way to deal with this problem that doesn't create heat and waste power?
-gt-

--- Updated May 22, 2023 ---

you may have noticed that when Q2,3 turn on there will be significant currents into the equipment, if the following equipment consists of a bridge rectifier followed by storage caps ...

--- Updated May 22, 2023 ---

6600 / 220Vac = 30 amps, so 15 amp from each socket in toto, for perfect power factor of load,

the rms current will be somewhat higher as the waveshape of the current going into the psu's will be quite peaky and far from sinusoidal.

Oh, its actually 5300W ... 'sorry about that ... I used 6600W, but the actual load is less.

 

[KlausST]

Hi,

If it's the darlington that determines the current limit, then the darlington also will dissipate a lot of heat.

Klaus

 

[BradtheRad]

Voltage semi-doubler using capacitors. Just to show what may be possible without a transformer or autoformer. I present it since I don't expect it to be my million dollar idea. Anyway it's untested in hardware and needs time and effort before it can be made into a solution.

Capacitors charge in series during one half of the cycle...

Then discharge in parallel during the second half while adding voltage of the mains AC sine wave.

Transistors turn on at the proper time automatically, steer current through the load.

There is also the classic bridge voltage doubler, a simpler circuit using 2 caps and 2 diodes. It can be used in the same fashion by installing transistors which produce AC at the load. It puts greater stress on components, and draws greater current from house wiring.