They don't have less inrush; there is nothing inherent to shared windings that reduces inrush.
Take a 96VA 120/24V transformer connected as an isolation transformer. Measure inrush, say that comes out to 8A when the 120V winding is energized. Now, connect it as a step-up auto transformer. It will still exhibit 8A of inrush.
It is possible it appears inrush to be smaller, since one could consider the auto transformer connection to be rated for 576VA instead of 96VA, therefore instead of inrush being 10x rated, it is now 1.67x rated. In other words, the auto transformer is designed to only require enough core and winding to accomodate the effective VA transformation difference, not the entire primary or secondary VA.
At the end of the day, inrush still comes down to air core inductance, DC resistance, the degree to which the core is operated near saturation, and remanence. The fact there is a shared winding is irrelevant. The amount of copper is really no different with an auto rated for transformation difference; copper is based on desired winding losses.
Autos do have special construction to accomodate unique short circuit forces and BIL requirements, so that will cause leakage inductance (and therefore air core inductance) to be different from a standard two winding xfmr. The degree to which inrush is changed based on this becomes fairly trivial.
OK then here is a better possibility . When I get my auto transformers built I always say be generous with the core as the weight is dramatically less than an isolating type ( 230 % greater power possibility is typical if 2: 1 ratio and in the KVA ratings ) . I think that is why . Also the simple design is reasonably cheap for labour so if absolute size is not a problem cost is reasonable . A 4 KVA auto costs less than a 300 B transformer . The latter has 1.5 hours of work to build .
A 4kVA auto former could be built from a 400VA 230:10+10Vac .
The 10Vac could be split into 4 windings of 5Vac each.
This, gives an autotransformer with nominal input tappings of 220, 225, 230, 235, & 240Vac and the 4kVA output could be taken from any of the input tappings to suit the downstream equipment.
What does that do wrong?
Post101 says that autoformers are made different. What is the difference and how do these affect performance?
The 10Vac could be split into 4 windings of 5Vac each.
This, gives an autotransformer with nominal input tappings of 220, 225, 230, 235, & 240Vac and the 4kVA output could be taken from any of the input tappings to suit the downstream equipment.
What does that do wrong?
Post101 says that autoformers are made different. What is the difference and how do these affect performance?
Post 101 is talking about autotransformers sized in MVA and voltages in kV. Look up BIL and surge testing with 8/20 wavefronts and you will start to appreciate some of the challenges with autotransformers, and why they are treated with special consideration; it's really out of scope of this thread.
I doubt there is anything unique about little popcorn autotransformers measured in a few kVA. And the real point is that when hooking up an isolation xfmr as an auto, inrush doesn't change.
I doubt there is anything unique about little popcorn autotransformers measured in a few kVA. And the real point is that when hooking up an isolation xfmr as an auto, inrush doesn't change.
If anyone is still following this link a question ( Andrew T )
I have 2.3 KVA transformer which I will start on a 16 A 10 R inrush thermistor made by Amathermic . Simply putting a relay on the output side ( to short the thermistor ) seems to be to be a possible solution against the usual 555 timer . The reaction time of the relay is 7 mS and has an AC operated coil ( 115 V ) . I would imagine 7 mS already helpful ? The transformer I have in mind will only infrequently be switched off . It would allow a 10AT fuse . The only issue is the fuse . I have even considered fitting a switch with LED reminder of status .
I have 2.3 KVA transformer which I will start on a 16 A 10 R inrush thermistor made by Amathermic . Simply putting a relay on the output side ( to short the thermistor ) seems to be to be a possible solution against the usual 555 timer . The reaction time of the relay is 7 mS and has an AC operated coil ( 115 V ) . I would imagine 7 mS already helpful ? The transformer I have in mind will only infrequently be switched off . It would allow a 10AT fuse . The only issue is the fuse . I have even considered fitting a switch with LED reminder of status .
Hi Audio San . It might just be OK as it is a 16 A relay . I did think the same so value any suggestions . I will try it as it is cheap to find out . >40 mS would be better .
I was researching the arcing of relays . I was using the same 16 A relays in a similar application which has proved to be reliable . I took the top off of the relays and noticed reasonable levels of arcing . Like the old contacts of a motorcycle points system ( HT coil LT side ) . On a motorcycle these were usually damped with a 0.1 uF paper capacitor if memory is correct ? Reading up suggested the reason for arcing is air being ionized . Solution was to use oil to exclude air . I only had cooking oil at hand , sparking seemed identical .
I was researching the arcing of relays . I was using the same 16 A relays in a similar application which has proved to be reliable . I took the top off of the relays and noticed reasonable levels of arcing . Like the old contacts of a motorcycle points system ( HT coil LT side ) . On a motorcycle these were usually damped with a 0.1 uF paper capacitor if memory is correct ? Reading up suggested the reason for arcing is air being ionized . Solution was to use oil to exclude air . I only had cooking oil at hand , sparking seemed identical .
At 50Hz a full cycle is 20mS, so 7mS is not even a half cycle.I'd go with at least 10 cycles; 200mS.
See figure 2-G.
http://relays.te.com/appnotes/app_pdfs/13c3206.pdf
See figure 2-G.
http://relays.te.com/appnotes/app_pdfs/13c3206.pdf
Last edited:
Thanks for the link . As it points out it could hit at excatly the worst time .
A NE 555 timer is fine for the job and cheap . About 1 uF and > 200 K should do fine for the timing ( 1 M ? ) . Use a polyester 1 uF as they are cheap and reliable . Nice thing about 555 is the internal crowbar resets it very quickly if the mains has a brown or blackout . Although most will know , fit a 1N4001 or whatever reverse biased across the relay coil to protect the 555 from back EMF of the relay coil .
A NE 555 timer is fine for the job and cheap . About 1 uF and > 200 K should do fine for the timing ( 1 M ? ) . Use a polyester 1 uF as they are cheap and reliable . Nice thing about 555 is the internal crowbar resets it very quickly if the mains has a brown or blackout . Although most will know , fit a 1N4001 or whatever reverse biased across the relay coil to protect the 555 from back EMF of the relay coil .
Very Off Topic reply
The coil is an inductor passing current when the points (switch) are closed.
The points open on the cam and break the inductor current. That starts a back emf and that emf is multiplied by the other part of the coil into kV of emf that jumps the gap across the spark plug.
If that was all there was to it, that final spark is so short that it does not reliably ignite the mixture.
So they add a cap in parallel to the coil primary. That creates a resonant circuit and instead of damping, it multiplies the output but much more importantly it oscillates and creates a series of sparks . It's the relatively long duration of the series of sparks that ignites the mixture reliably.
not how the switched coil works.
The coil is an inductor passing current when the points (switch) are closed.
The points open on the cam and break the inductor current. That starts a back emf and that emf is multiplied by the other part of the coil into kV of emf that jumps the gap across the spark plug.
If that was all there was to it, that final spark is so short that it does not reliably ignite the mixture.
So they add a cap in parallel to the coil primary. That creates a resonant circuit and instead of damping, it multiplies the output but much more importantly it oscillates and creates a series of sparks . It's the relatively long duration of the series of sparks that ignites the mixture reliably.
Good point if the unintended pun can be allowed Andrew . I suspect I even knew that once upon a time as I used to read endlessly about HT coils . My Honda 900 was said to work better if given 2 extra volts . There were expensive coils that worked on 13.8V as replacements . It was said Honda were being crafty and using it to restrict performance ( the 107 BHP originals were trouble I read ) . I can believe it as with new plugs mine would approach 130 MPH ( in Germany , honest ) . When the plugs were a bit used 110MPH was more typical . Bike magazine raced one and used the extra battery cell . They switched to standard for road use .
The difference of note is the use of a snubber on a switch . The points is just a capacitor ( ? ) and as you say a resonant circuit . Anyone have views on ideal snubber circuits ? 0.1 uF and 100 R is common .
My observation was the spark levels give a clue as to how healthy the relay is in use . The devices are very similar . Relays say > 50 000 operations . A typical motorcar will do that in a few minutes . I suspect relays are good for millions of operations if used with care .
The difference of note is the use of a snubber on a switch . The points is just a capacitor ( ? ) and as you say a resonant circuit . Anyone have views on ideal snubber circuits ? 0.1 uF and 100 R is common .
My observation was the spark levels give a clue as to how healthy the relay is in use . The devices are very similar . Relays say > 50 000 operations . A typical motorcar will do that in a few minutes . I suspect relays are good for millions of operations if used with care .
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.