| Author |
Message |
    
bruce
Frequent Contributor
Username: bruce
Post Number: 239
Registered: 04-2008
Rating: N/A
Votes: 0 (Vote!) | | Posted on Tuesday, 02 February, 2010 - 07:02 pm: | 
|
Hi,
I have a query about electrolytics that I dont imagine gets discussed in text-books. It's about the maximum working voltage.
I'm presently making a PSU of 100V. I need to charge some capacitors to this voltage and I wanted to have about 1000uF. I dont have anything like that to hand, so I thought of putting two 50V ones in series. Actually I dont have any 50V ones either, but I do have a handful of 1000uF/35V ones. Two in series would have a nominal 70V rating. But would they survive 100V? I know the capacity of these things is very wafty; plus or minus about 30% isnt it! How tight can they make the working voltage? Or is it a generous margin?
One more thing: in older SPMS the mains was rectified to give around 400V. They would have 2 electrolytics of 200V WV in series. I never knew if this was because they wanted a split supply or because they didnt have 400WV devices. These days its all a single rail.
Across each capacitor is a resistor. Why? It is usually high, 400-500K. This could be a bleed resistor, but I guessed it might be necessary to even out the rails. If the second is right, then it seems rather a high value. What would the design parameter be here ?
bruce |
twintub
Frequent Contributor
Username: twintub
Post Number: 61
Registered: 02-2007
Rating: N/A
Votes: 0 (Vote!) | | Posted on Tuesday, 02 February, 2010 - 10:14 pm: |
|
I'm not a 'proper' designer but I would think that if a manufacturer said a cap is 35V max then it would be foolish (indeed dangeorous) to exceed it! I don't believe that electrolytics would have much (if any) overvoltage tolerance - manufacturers surely control the manufacturing process to a fine degee!
From a safety aspect it is always a good idea to put bleed resistors across high voltage caps and every decent designer/manufacturer would do so, and there may be EN standards (law) that insist upon it.
Maybe reversed biased diodes are advisable across each cap - this will reduce the chance of an electrolytic becoming reverse charged if it's own charge became depleted before any other series series cap.
Best advice: Don't use caps in series - Go and buy caps with a voltage rating at least 20% over the maximum RIPPLE/NOISE voltage that will be expected: don't forget that an electrolytic will really have a higher voltage across it than a DC meter (i.e. average reading) would suggest.
(Message edited by twintub on 02 February, 2010) |
terrym
Frequent Contributor
Username: terrym
Post Number: 708
Registered: 05-2005
Rating: N/A
Votes: 0 (Vote!) | | Posted on Tuesday, 02 February, 2010 - 11:06 pm: |
|
Twintub is more or less right.
Caps voltage rating should be taken as printed, otherwise, any safety margin is lost.
If using caps in series to make up the voltage rating, high value resistors across the caps help to balance out the voltage across them. Also don't forget that that capacitance value reduces for caps in series.
The ealier smps were designed to work on 110v or 240v. The reason for 2 x 200v caps was that in 110v mode, they formed a voltage doubler to reach the 345v of the main rail. In 240v mode, they are connected in series across the rail.
TM |
ant
Frequent Contributor
Username: ant
Post Number: 570
Registered: 05-2006
Rating: N/A
Votes: 0 (Vote!) | | Posted on Tuesday, 02 February, 2010 - 11:15 pm: |
|
Hello Bruce,
I would err on the generous side with voltage rating. Don't forget that two 1000mfd will give only 500mfd. At 500k those are bleeder resistors, balancing resistors would be much lower - vague memories of past circuits suggest perhaps 100k maximum, to halve the voltage seen by the junction without being too affected by leakage currents.
Regards Ant |
zeitghost
Frequent Contributor
Username: zeitghost
Post Number: 1431
Registered: 01-2006
Rating: N/A
Votes: 0 (Vote!) | | Posted on Wednesday, 03 February, 2010 - 08:42 am: |
|
Over volting electrolytics often gives an intimate acquaintance with the internal structure.
Don't do it.
Two 63V 1000uF caps in series would be ok, with bleeder resistors across them. |
bruce
Frequent Contributor
Username: bruce
Post Number: 242
Registered: 04-2008
Rating: N/A
Votes: 0 (Vote!) | | Posted on Friday, 05 February, 2010 - 11:20 am: |
|
Hello guys,
Thanks for the replies, although I was a bit taken aback. I rather thought someone would say " theres a huge margin of error on the WV rating", but, instead, I get lots of dire warnings that my hair will fall out and that all my pigs will die of the flux.
It's a tricky one, because I dont yet know for sure whether 100V will do it, or, even if it does, whether I need 100uF or 1000uF, so I wouldnt know what to buy. And, as Maplin is the only place here that does components, and theyve now stopped selling electrolytics, I'm stuffed. Unless I take hideous risks like overvolting.
Recently I made a flasher from an old disposable camera. To speed up the flash-rate I substituted a smaller reservoir capacitor; 50uF instead of 200uF. But I only had a 250WV component. The design automatically fires when the voltage reaches 250V ( which is handy! ), but, as the auto-fire mechanism was a bit temperamental, the voltage frequently overshot to 350V. As the capacitor didnt seem to suffer any ill effects, I certainly started to believe that the manufacturers err on the side of caution.
The whole thing is fitted in the nose-cone of a rocket, so, if it does explode, it should look really spectacular. And expensive.
Oh, BTW, there is a site ( sorry I dont have the URL ) which describes an incident in which an American gentleman, who was no doubt after a Darwin Award, tried to increase the fun of sledding by fixing a rocket to his back. The rocket was a 'pyro' type and home-made. He, the story goes, used a motor-cycle exhaust pipe as the casing which he filled with black powder, gasoline and match-heads. A friend ( who might now be charged with attempted manslaughter ) kindly lit the fuse, but, rather than propel him, it simply detonated and put him in ER with severe facial burns. He may yet be charged with something, although God knows what. Creating a noise in a public space? You couldnt make it up.
bruce |
joe
:::: Super User ::::
Username: joe
Post Number: 737
Registered: 05-2005
Rating: N/A
Votes: 0 (Vote!) | | Posted on Friday, 05 February, 2010 - 12:15 pm: |
|
Bruce,
My personal opinion and experiance is that Elect. Caps are one of the weakest part of any circuit. They are nearly always the component that fails in PSU's and I've seen more than my fair share of their insides splattered all over the PCB.
They are also the only component I can think of that has a shelf-life besides batteries.
I would never run a cap at or near 100% rated voltage and I've always opted for a 25% - 30% margin when using them unless it's one of the very low voltage ones like these 5v super caps; I don't want to hurt anybody, including me.
You also have to factor in operating temp as this has an affect on the caps parameters.
Just my humble thoughts.
Joe
Read my ramblings - www.techbites.com/joe-farr
|
gajjer
Frequent Contributor
Username: gajjer
Post Number: 289
Registered: 05-2007
Rating: N/A
Votes: 0 (Vote!) | | Posted on Friday, 05 February, 2010 - 01:56 pm: |
|
Hi Bruce
can you tell us a little more about what you are trying to do. What is the source of power? What is the consumption of the circuit?
cheers
gaj |
bruce
Frequent Contributor
Username: bruce
Post Number: 244
Registered: 04-2008
Rating: N/A
Votes: 0 (Vote!) | | Posted on Friday, 05 February, 2010 - 02:43 pm: |
|
Thanks for that Joe. I will, as they say in all Bruce Willis movies, "take it under advisement." Quite when Americans are going to learn to speak the King's English, I dont know.
Hello Gaj. You probably remember I wrote in a previous thread about a blocking oscillator. It was the flasher I was having trouble with. As the thing was basically a hacked camera circuit, and ran off 1V5, I couldnt ( easily ) operate the part of the circuit which controlled its rate. This sub-circuit was a 393 comparator and it used a photo-transistor. I had noticed that the little neon lit up at 250V and I wanted to use the light to trigger the photo-transistor. I thought it was rather clever, although you can bet your a**e someone else already thought of it.
So I used a little B.O. to generate 5V @ 5mA. In the event, I had terrible trouble getting the whole thing to work reliably ( it had to be isolated ) I tried opto devices and a 1:1 trigger transformer. God knows why, but it kept trashing the transistor.
Then I had a 'eureka' moment ( I remembered I'd left the bath running ). Anyway I got the wretched thing working and it is, as we speak, sitting atop the rocket, just like Apollo 13. I'm hoping to post the circuit sometime. Just a bit of fun.
The next mission ( should I choose to accept it ) is a 'module' which contains a small parachute. That's the easy bit. Getting it to deploy is the tricky bit. Somehow you have to open the door. I tried using a small solenoid, but they're awfully weak. The solenoid pulls the 'pin' which holds the door closed against thye tension of a piece of knicker elastic, and I wont divulge where I got that from. Anyway, I had a tiny 12V solenoid ( actually I made it from a relay ). Even at 40 - 50V it couldnt easily pull the pin. Next step 100V ( I'm using a boost SMPS; very basic, low power, just to fuel the capacitor ). Failing that, I shall go for 200V, but get one of the kids to do the testing in case it all blows up. ( I never said I was a good father ).
Despite all the dire warnings of frogs and snails, I'm going to use two 5000uF 50V caps in series. Fingers crossed, eyes shut and hands over genitals.
Bruce |
gajjer
Frequent Contributor
Username: gajjer
Post Number: 290
Registered: 05-2007
Rating: N/A
Votes: 0 (Vote!) | | Posted on Friday, 05 February, 2010 - 05:10 pm: |
|
Hi Bruce
well it seems it is rocket science after all!
The problem with solenoids is that the force isn't up to much until the armature gets really close to the electro magnet. Limiting the travel to that last mm might help.
As regards the size of the capacitor, thats a bit difficult to estimate. I would guess that the solenoid would act in say 0.2 sec. You know the resistance of the relay, so I would pick the C to make the CR time constant about 0.2sec.
You are going to give it a big hit initially. As the armature starts to move it will be getting more force from the coil.
Personally I would rethink the mechanism. Perhaps a geared down motor driving a thread with a nut on it.
Let me know when you are going to fire this thing and I'll set up the seismogragh! It could be one of those "We'll laugh about this after we've cleared the mess up" moments.
cheers
gaj |
magnum4
Frequent Contributor
Username: magnum4
Post Number: 828
Registered: 04-2005
Rating: N/A
Votes: 0 (Vote!) | | Posted on Friday, 05 February, 2010 - 08:13 pm: |
|
Hi Bruce.
Maybe your going the wrong way about it? Now dont point that missile at me yet 
What about going the other way, have your solonoid powered to hold door closed, and release power to open door?
The advantage of course is that the max power of the solonoid is when it is an its closed position, and if it were holding a compressed spring, ( which would be pre stressed by hand), you would have more power for activating the latch?!!
Just a thought
Regards,
Jim
|
alec_t
Frequent Contributor
Username: alec_t
Post Number: 176
Registered: 03-2009
Rating: N/A
Votes: 0 (Vote!) | | Posted on Friday, 05 February, 2010 - 10:29 pm: |
|
Even simpler than the suggested nut-on-a-thread would be a twisted ....er...thread. Motor twists thread more, thread gets shorter and pulls pin.
Regards, Alec |
bruce
Frequent Contributor
Username: bruce
Post Number: 245
Registered: 04-2008
Rating: N/A
Votes: 0 (Vote!) | | Posted on Saturday, 06 February, 2010 - 12:05 pm: |
|
Joe, you have to live dangerously. Man is only truly alive when he faces death. Of course, he can also be truly dead.
OK guys, thanks for the suggestions.
I have considered many of these alternatives. I've also collected a sack full of tiny mechanisms from discarded DVD ROMs, FDDs, whatever, and have spent many a happy hour staring at them trying to decide the best way to proceed. Part of the difficulty is that there are too many options, and each one will take some time to 'work up'. I made an executive decision ( wow ) to go with the relay and just see if I can get a result. It may not be the best option, but you have to start somewhere. If it works OK, then at least I've got something to work with. After that I can investigate other possibles in my own good time. Ah! Time. Where does it all go?
Dont forget that as WW2 loomed and we needed a functioning early warning system asap, Watson-Watt cobbled together a quick and dirty fix. Our radar stuff wasnt a patch on the stuff the Germans had, but considered better than nothing. He said at the time " second best tomorrow".
You heard it here first!
Gaj, you're welcome to come and see them gracefully soar ( and then crash ) if you're anywhere near Leicester. It wont happen until Easter anyway and first I have to find a local farmer who'll let me use a field. You darent fire them in a park in case some kid gets hit on the head. Although it could possibly knock some sense into them, the parents dont see the funny side.
bruce |
magnum4
Frequent Contributor
Username: magnum4
Post Number: 830
Registered: 04-2005
Rating: N/A
Votes: 0 (Vote!) | | Posted on Saturday, 06 February, 2010 - 06:02 pm: |
|
I love executive decisions 
Be interested in how you get on.
A camera for the maiden flight might be a good idea 
Cheers
Regards,
Jim
|
alec_t
Frequent Contributor
Username: alec_t
Post Number: 177
Registered: 03-2009
Rating: N/A
Votes: 0 (Vote!) | | Posted on Saturday, 06 February, 2010 - 06:36 pm: |
|
....if you know of a camera able to withstand several 'g' without disintegrating!
Regards, Alec |
ant
Frequent Contributor
Username: ant
Post Number: 574
Registered: 05-2006
Rating: N/A
Votes: 0 (Vote!) | | Posted on Saturday, 06 February, 2010 - 10:11 pm: |
|
Hello Bruce et al,
Lateral thinking - what about a short link of thin wire holding a catch, which could be blown out like a piece of fuse wire?
Regards Ant |
741
Frequent Contributor
Username: 741
Post Number: 132
Registered: 08-2005
Rating: N/A
Votes: 0 (Vote!) | | Posted on Monday, 08 February, 2010 - 12:29 pm: |
|
Hi Bruce,
Could you use a supercap in your application? I think they hold charge reasonably and so you could charge the cap externally if need be prior to launch.
(I can already see the LOX smoke and the gantries..) |
bruce
Frequent Contributor
Username: bruce
Post Number: 246
Registered: 04-2008
Rating: N/A
Votes: 0 (Vote!) | | Posted on Wednesday, 10 February, 2010 - 10:03 am: |
|
Hello guys.
The fusible link is a nice idea. One wrocketeer used a solenoid in reverse. He employed a camera flash cct to charge a capacitor to 400V and this ( somehow ) ejected a slug of metal away from a solenoid at high speed. Rather like those explosive bolts NASA used. I'm tempted to use black powder and just blow the thing to bits mid-air. Well, plenty more bottles where they came from ( hic! ).
Supercaps are only 5V aren't they? ( mind you, I did come across a 10V one ) Are you thinking of them in series? It's not the energy in the capacitor as much as the voltage that achieves the effect.
Lots of wrocketeers have cameras these days. You can see footage of this on You Tube. In the USA they can get their hands on nifty little video cameras about the size of a match-box. You can even get an aerial video of a rocket as it crashes through the roof of your house.
Why do we do it? Because we can.
Bruce |
741
Frequent Contributor
Username: 741
Post Number: 134
Registered: 08-2005
Rating: N/A
Votes: 0 (Vote!) | | Posted on Wednesday, 10 February, 2010 - 02:11 pm: |
|
Following from gajjer's comments, I'm guessing you go for a high voltage to create a high initial current and thus magnetic field. I wonder if there is some way of profiling the shape of the current vs time to give some minimum area.
Could it be that the admittedly much lower supercap current (5V/relay impedance) would still work as it is on longer?
I suppose not.
Still, there must be an optimun capacitance and initial voltage in the sense that the cap and any charging circuit is as physically small and light as possible.
(Message edited by 741 on 10 February, 2010) |
magnum4
Frequent Contributor
Username: magnum4
Post Number: 838
Registered: 04-2005
Rating: N/A
Votes: 0 (Vote!) | | Posted on Tuesday, 16 February, 2010 - 10:17 pm: |
|
Alec. one on the ground would be fine
That way Bruce would not have any worries about G`s 
Regards,
Jim
|
bruce
Frequent Contributor
Username: bruce
Post Number: 257
Registered: 04-2008
Rating: N/A
Votes: 0 (Vote!) | | Posted on Wednesday, 17 February, 2010 - 07:33 am: |
|
You can buy these video cameras which are about the size of a pack of Wrigleys gum. Lots of wrocketeers in the USA use them. I've seen many ads on Ebay for cameras of all sorts and, for what they are, they seem jolly good value; about £20. Most come from Hong Kong, so, if anything goes wrong with the transaction, you could be in a fix!
I dont know how well they would survive the descent. ( Spike Milligan once said that he wasnt afraid of flying; it was the crashing he objected to. ) The Holy Grail is to deploy a sexy little parachute, although the impression I get is that they dont always deploy on request and, as with rock-climbing, it doesnt matter how many times you get it right, you only ever get to make one mistake.
Bruce |
gajjer
Frequent Contributor
Username: gajjer
Post Number: 291
Registered: 05-2007
Rating: N/A
Votes: 0 (Vote!) | | Posted on Wednesday, 17 February, 2010 - 05:13 pm: |
|
Hi Bruce
a friend of mine put a camera on his model plane and got some great footage. Worth a go.
Just another thought about the release mechanism. How about gravity operated? A catch that actuates by a weight while it is in descent.
I agree with you about the supercap. Its the initial current that will do the work. The solenoid has to work against a big gap initially. Once you have overcome that its easy.
I hope you post any footage you get.
cheers
gaj |
gajjer
Frequent Contributor
Username: gajjer
Post Number: 292
Registered: 05-2007
Rating: N/A
Votes: 0 (Vote!) | | Posted on Wednesday, 17 February, 2010 - 05:39 pm: |
|
Hi again
more thoughts on a gravity catch.
How about a tube say 1cm diameter in the middle of the capsule. The tube is vertical and holds a round weight in it. Radiating from the tube would be sprung loaded rods. When the weight is in place it pushes the rods out against their springs and the outer ends engage with the latch. When the rocket is inverted, the weight falls out, the springs retract the rods and the latch is free. The larger the diameter of the tube and weight, the more travel you get from the rods.
The only thing you have to worry about is if the weight can overcome the friction on the rods. ie the weight doesn't drop when inverted and is held in place by friction of the rods due to the springs.
cheers
gaj |
bruce
Frequent Contributor
Username: bruce
Post Number: 258
Registered: 04-2008
Rating: N/A
Votes: 0 (Vote!) | | Posted on Wednesday, 17 February, 2010 - 05:56 pm: |
|
Hello Gaj,
Nice to hear from you again.
Footage from wrockets is a bit dissappointing, to be honest: You get a picture of the ground which rapidly gets futher and further away. Then, as the wrocket rolls over, you get an uninterrupted view of the sky. You can see plenty of these things on You-Tube. I imagine that putting these cameras in R/C aeroplanes would give terrific pics.
My first priority is to get a reasonably reliable parachute deployment. Not an easy task, it seems. Still, what do we always say? If it was easy, everybody would be doing it.
I've just knocked up a Mickey-Mouse cct. to generate that 100V from a 4-pack of NiCads. The output was dissappointing, and, when I checked, I found the inductor was saturating. Now, saturation comes as no big surprise if you use dodgy ferrites, which I do all the time, but this was a proper ferrite pot with air-gap
( FX2236, I think ) Small, but perfectly formed and only required to go to half an amp. I was gobsmacked to find it saturated so easily. Honestly, you cant get the staff.
BTW, you hit the nail on the head when you talk about triggering. It comes as a huge surprise to newbies at this game that what works on the ground rarely works in flight. For example, mechanisms which rely on gravity turn out to be useless; as the wrocket descends, everything goes into freefall. A favourite is the mercury switch: I even fell for this one myself. One really professional guy in SA has a wrocket so big he can do fancy experiments in it, rather like NASA and the shuttle! He fitted 3 mercury switches inside ( one was upside down ) plus a camera to record what they did, PLUS an altimeter so he could plot everything out. It really was thorough.
Anyway, the mercury switches behave nothing like they do on the ground. The wrocket motion isnt smooth like you expect; in fact it jerks and shudders all over the place and the mercury switches were going on and off like the National Grid in winter. If it was easy........
Cheers
Bruce |
bruce
Frequent Contributor
Username: bruce
Post Number: 259
Registered: 04-2008
Rating: N/A
Votes: 0 (Vote!) | | Posted on Thursday, 18 February, 2010 - 04:47 am: |
|
Hello again gaj,
Something blooming queer is going on with this website: the last 2 messages on this topic seem to have arrived out of sequence.
If I understand your latch idea, then, as per my last post, I think you'll find that the ball wont roll down the tube as expected. It will do so when it's on the ground, but not when it's in the wrocket; in free-fall, the wrocket and the ball fall at the same rate, and the ball stays put. Yes; isn't it!
Last night I finished the 100V inverter and it was a thumping disappointment. I had a good look at it again and I believe I know the reason. Someone has previously mentioned that the solenoid only really generates decent pull in the last few mm of travel. The solenoid I'm using has about 1cm of travel ( it sort of needs the pin to be that long ) and, as the armature is only 2cm long, then, even at 100V, the initial pull is so weak it cant get the armature moving against the drag from the elastic band. If it was easy.......
I actually have a Plan B, and it isnt, as it usually is, to take up a different hobby. I'll have a little 'play' and get back to you.
Onwards and upwards.
Bruce |
gajjer
Frequent Contributor
Username: gajjer
Post Number: 293
Registered: 05-2007
Rating: N/A
Votes: 0 (Vote!) | | Posted on Thursday, 18 February, 2010 - 05:19 pm: |
|
Hi Bruce
'free falling together' should have thought of that!
I suppose a light spring assisting the weight is out of the question.
Re the solenoid. 1cm is way too big a gap. You need to be thinking more like 1mm. Can you extend the armature so that the magnetic gap is smaller but the end of the armature moves a longer distance. Alternatively, how about modifying a servo? Cut out the electronics and go straight to the motor. You might have to limit the current for when it stalls on the end stop though.
cheers
gaj |
boris
Frequent Contributor
Username: boris
Post Number: 494
Registered: 05-2005
Rating: N/A
Votes: 0 (Vote!) | | Posted on Thursday, 18 February, 2010 - 07:46 pm: |
|
Perhaps worth considering...
Lower value electrolytics are usually available with higher voltage ratings. (And lower ESR). It might be worth considering using several small electrolytics in parallel, with appropriate voltage ratings. You add the capacitances and the ESRs are treated like parallel resistors, reducing significantly the total ESR.
EG: 4 * 50uF (300v) caps will give 200uF (300v). They will share the load, dissipate less energy in the ESR, produce more spontaneous current and last longer than single (or series) caps.
Thank you for not breeding.
|
bruce
Frequent Contributor
Username: bruce
Post Number: 270
Registered: 04-2008
Rating: N/A
Votes: 0 (Vote!) | | Posted on Monday, 22 February, 2010 - 12:10 pm: |
|
Hello guys,
I told you that the 100V PSU was disappointing. The armature travel was a problem, but maybe Boris has a point here: I used two old 50V caps in series. They certainly seemed to disharge much more slowly than I'd have guessed. Is that ESR?
I'm working, even as we speak, on a different method. I hacked a CD ROM and I'm using the head-read mechanism. Watch this space, but dont hold your breath
bruce |
|
Closed: New threads not accepted on this page
