OK, here goes if you can follow this picture....
The ST has 4 poles.
Each pole makes 60v AC
If connected as follows, and I tap points A,C for 120v, and C,E for 120V, and A,E for 240, and install the bridge rectifiers as shown, I'll also have rippled 60v DC.
I can then filter this DC into a charge controller-grid tie inverter setup.
I wouldn't need switches, breakers, etc.
I'd have every combination at all times!!!!
Does anyone see flaws in my thinking?
Yes. :-* you don't need to jumper the connections to b,c & D. If you do so you're back to ensuring that all the coils are in correct polarity. Sending an individual coil to an individual rectifier eliminates that need.
If I don't jumper B,C and D, how do I get normal 120/240?
Short answer: from the inverter, Long answer: I was thinking to wire my st5 with a 4 pole marine transfer switch to allow a simple change of operation. That was until I saw the price of the switch. How important is it to be able to do both at the same time.
I got a mess of squiggly lines on paper here and muddling myself so not much luck at trying to explain it even if I do have a bit of a grip on it. I think I see a conflictthough. All 4 coils must be in series with the phase correct to give 240 V scenario; i think that is a given, but for the 60V scenario you want to take all of each coils current but cannot have the voltages additive or opposing. Since all 4 coils produce simultaneously they need to be in same polarity and parallel connected.
When I try to draw to suit 120 or 240, it seems to have the bridges trying to conduct ~ current in different directions at the same time. If I draw it to satisfy DC configuration it screws up the other.
Lead thou me on, the night is dark and I am far from home.
On a normal setup ST, if you have 240, you simply center tap for two legs of 120.
this would simply center tap each set of 120 for two legs of 60 (total of 4)
How is this any different?
I'm not arguing, I just can't wrap my head around why this won't work.
Looks like it will work fine. I basically wanted to do the same thing til I realized I have a nice transformer that was given to me.
If you take the two 120 legs of the center tapped 240 you cannot combine them into a single load as that creates a dead short. I get screwed up when I try to visualize FW rectification and combining.
Brett,
I don't want to pour water on your parade but I think you've diagrammed a big spark and fried rectifiers. It would take me a few minutes to wrap my mind around it, but first of all take your negative output: Isn't the negative rectifier output of A-B going to be 60 V different than the neg output of B-C? But I'm not an electrician or electrical engineer....
Marcus
The DC bus that ties the bridge rectifier units output together also would connect internal AC bridge components to a common reference voltage, yet the series connections of the windings would attempt to have them at different voltages.
I think you could have four separate floating DC outputs and still maintain the stated 120 / 240 connectability, but to put full output into a common DC bus the series connections of the windings would have to go and be separately connected to each bridge.
QuoteThe DC bus that ties the bridge rectifier units output together also would connect internal AC bridge components to a common reference voltage, yet the series connections of the windings would attempt to have them at different voltages.
Yes.
You have diagrammed your outputs in parallel, not in series. The parallel circuit ties your reference voltage together, but the windings of the generator are tied in series (at B, C, and D). That's what I was trying to say last night after midnight.
Marcus
My brain simply refuses to see it, but I will bow down and accept that I can get 60v DC from each separated pole, or 120/240, but not both.
Oh well, it was worth a try.
Thanks for keeping me from doing something 'smokey'
Ye Gads! I would like to see someone come up with a more definitive description of what the exact problem is. Dont give up on my say so as I admit to being weak on the internals of solid state devices.
I am sure though, that you can combine all the rectified ~60V outputs of the bridge rectifiers into one of 60 x 1.44 VDC output. You also might be able to get away with putting the coils all properly phased and parrallel connected through the rectifiers for similar output, but that risks one or more of the diodes running away with a larger portion and overloading if they are not perfectly matched. I just dont think you can have the coils interconnected for either/or 120/240 at the same time.
I like what you did with bringing the ends of all coils out so they could be connected such to get rid of the single leg load groan. It is something my gen. could benefit from too, but that is very elective surgery.
OK, lets break it into pieces. Bear with me.
Take out the red & black leads from the diagram for a minute. You're left with a 4 poles wired in series, with a bridge rectifier across each pole, correct?
Across A-B there's 60 VAC. Same for B-C, C-D, D-E.
Across A-C there's 120 VAC. Same for B-D and C-E.
Across A-D there's 180 VAC. Same for B-E.
Across A-E there's 240 VAC.
Everything just like in the ST manual so far.
Now, take a multitester and read the output voltage of each of your rectifiers (with the black and red leads removed).
Across each rectifier from "pos" to "neg" there should be approx 60 x 1.44 volts DC.
Now measure from "neg" output from bridge 1 to "neg" output on bridge 4. There's going to be a 180 volt difference. Same thing from "pos" on bridge 1 to "pos" on bridge 4. Each rectifier will have a different "ground" or "negative" voltage reference because the gen poles are connected in series. It doesn't matter that it is called negative, rectifier 1 has a different "ground" voltage with respect to each of the others. Get what I'm saying? Now if you put connections (red and black wires in your diagram) across all the "pos" and all the "neg", there's a short.
You've got the generator poles in series but the rectifiers in parallel. If you want to keep the poles in series, you can't parallel the outputs.
On the other hand, if you separate out the poles (i.e. don't connect them at B, C, or D), and hook up each separately to the rectifiers, then you can connect the rectifiers in series.
Output of "pos" on rectifier 1 to "neg" on rectifier 2, and so on. Then you can select 60-120-180-240 volts.
Of course there is the caveat that rectified DC = 1.44xAC. You'll have to slow down your engine or put the appropriate resistor in the ST excitation circuit to decrease the output V AC. Also, if you're drawing off the 240 V DC circuit, if one of the rectifiers burns up the whole 240 V leg is dead.
Hope I got that right....
Marcus
Yes, Jens is saying exactly the same thing but got his thoughts on "paper" before I did.
QuoteSuffice it to say, your concept will work if the generator windings are situated in parallel instead of the current series arrangement.
Exactly, see my post above.
QuoteOf course this would also mean no 120/240 at the same time.
For AC, that is correct, but not for DC. You could series connect the rectifiers and get different combo's. Analogy: think stacked batteries with leads coming out between them. What would you do with 180 V DC, though? :)
QuoteNote that I simplified things so please don't tell me that the voltages are all wrong by a factor of 1.44.
If you use a large choke on the output, it will smooth out the pulses to near RMS, IIRC. Forgot to mention that in the above post.
Marcus
I wish I had the parts here to test it like that. I think he would still get what he originally thought. The reason I say that is because i believe the way he has it setup forces 60vac input per rectifier. I have been wrong many times before tho. Lol.
it will force 60vac to each individual rectifier, however where you get into trouble is paralleling the output
of each of the rectifiers, there ends up a problem with feedback loops iirc
you can do it as drawn only if you use each of the rectified outputs for separate applications, with no interconnects
basically all outputs with floating grounds, or you end up with the ever increasing voltages becoming additive as referenced
to the ground.
the only way to get the job done otherwise so that you can retain AC output 120/240 is to use isolation transformers
on each of the coils, the secondaries of each transformer could then be rectified and then connected in parallel without feedback loops
referenced to ground.
thats as good as i can see it, but i may well be wrong.
got a bad ass cold and i feel like crap, on cold meds so my thinking is even slower than usual.
with a little time and about 3 bucks in diodes from radio shack, (the 50pack of in4001's) one could make up 4 bridge rectifiers
and test the output of each coil set as drawn, that should work no problem
then connect all the bridges in parallel and see what the result is, certainly it won't hurt the engine or the genhead, but i am not
very optimistic that any of the diodes will survive?
sure would put an end to the discussion in either event?
bob g
At the very least, I see some pretty bright minds trying to think through the same problem, and still not 100% sure of the result. It isn't completely cut and dry. I feel better. Thanks guys.
The experimental (sacrificial) diode experiment may be in my near future.
First on the list is to get the 4 Evergreen 210w panels mounted on the workshop roof.
Even if the generator isn't doing what I want at the time, at least I know the sun will be shining (this is southern New Mexico after all).
I can't tell you all how much I appreciate being able to bounce ideas around.
I only hope I am somehow contributing to help others as much as they help me.
-Brett
there really is no reason to risk 16 diodes when 8 will prove the concept as viable or not
just do it across 120 vac and another across the other 120vac output and then parallel the outputs
that wouldn't take long to find out what the results might be
bob g
The transformer I have will allow me to center tap 120v. If I get time tomorrow I will test it.
QuoteThe transformer I have will allow me to center tap 120v. If I get time tomorrow I will test it.
Excited to hear the results.
Great thread. I could see the problem but not articulate it. Thanks guys!
Ok guys I did the test and have some interesting results. This was done using a dual primary dual secondary transformer with 120vac input from the grid (I took the parts to my office with me). All tests were done with no load.
First test:
Input 121vac
Secondary A output 60.5vac
Secondary B output 60.5vac
Rectifier ouputs connected in parallel = 56.5vdc Frequency = 59.9Hz
Rectifier outputs in series = 113vdc Frequency = 59.9Hz
I switched the hot/neutral on one rectifier only and had the same results
Frequency on dc side = 59.9Hz
Second test:
Input 121vac
Secondary A output 60.5vac
Secondary B output 60.5vac
Secondaries series connected 121vac
Rectfiers connected in parallel = 113vdc Frequency = 120Hz
Rectifiers connected in series tripped the breaker (with only 1+ and 1- connected and the others not connected to anything)
I switched the hot/neutral on one rectifier only and had the same results
EDIT: I still have the parts at my office, If anyone would like me to try any other setups let me know.
Without diagrams, I'm a little fuzzy on the results of your test, so bear with me.
Do these tests show that it will work the way I originally drew it?
No. If u had each 60v coil separate it would.
OK, than that proves that this is a dead horse. I have to choose between normal 120/240 AC, or 4-60v AC legs separated and rectified/filtered for the grid tie setup.
Oh well. This proves I can't have it all :(
Thanks for the time you put in to proving my theories junk ::)
Thats not a problem cause I am gonna use this setup for my charging. So I needed to test it anyway. I do have a few questions tho. When I first hooked up the rectifier (not this test, this was done a few weeks ago) I had 106vac input and around 103vdc out with no load. But when I hooked it up to the mx60 the vdc went up to 148vdc. The only thing that was different was the rectifier was mounted to the cabinet when hooked to the mx60. Does the voltage change when the casing is mounted/grounded on a rectifier? This also shows in this test that with no load and the rectifiers not mounted that the vdc was a little lower than the input. ???
d34,
The decreased voltage across the bridge rectifier is easily explaned:
Each diode has an almost 1.x V drop across it, and a bridge rect has 4 diodes. So 2 diodes to rectify each pos or neg wave gives the drop from 60.5 to 56.6 V.
For the other question:
The transformer output is electrically isolated from the input, so couldn't it be a ground problem? Is the neutral of the secondary grounded? I'm not the best one to answer that question though because that was a similar problem I had when routing 80 V AC to DC. The neg lead was then 40 V higher than ground. ??? Watch your Voc into the MX60 because it will only tolerate up to 140 V DC, I think.
Marcus
Quotethe only way to get the job done otherwise so that you can retain AC output 120/240 is to use isolation transformers
on each of the coils, the secondaries of each transformer could then be rectified and then connected in parallel without feedback loops
referenced to ground.
So, this just hit me. Thank you Bob. I realize this thread was dead, but I'd like to go back and stress that you can't take the neg output of a bridge rectifier and connect it to neutral or ground of the generator without a isolation transformer! Last night I hooked it up like that and got 200+ V DC when I thought it should be 80 V DC. Couldn't figure out what was wrong, but like mobile_bob said, "you end up with the ever increasing voltages becoming additive as referenced to the ground." Now it makes sense after reading and re-reading.
I'm at work again, but ideas always come at the wrong time: In a spare moment today, I'll make a diagram and see what you all think of something.
Marcus
You have to watch the actual internal connections in transformers, some share coils (autotransformer) some have 2 coils but no isolation, some have 2 coils and are truly isolated to 1KV.