the 5 ohm resistor died because it could not handle the peak amps (i have seen it glowing through the housing) - NOT because it could not handle the energy. it got partial damage and after a few uses it totally gave up it job. A same sized type with 50 Ohm or above probably will work fine together with high volts and big controllers, but in this case the connection should not be closed quick as it takes a few seconds until the precharge is fully done
Big spark when connecting battery to bike
- Thread starter andrenoites
- Start date
Alan B
100 GW
You are right in that increasing the resistor value reduces the current and instantaneous heating, but is that enough to save the resistor? Let's look at the process a little deeper.
There are different ways of looking at any problem, and this one is a bit tricky. At high current the I^2 R heating is high, but the current drops off more quickly so the heating time is reduced. So it is not obvious what the total energy absorption (and heating in the resistor) is. The interesting thing about charging a capacitor bank with a resistor is that the thermal energy dissipated in the resistor doesn't change with the resistor's value. Only the time over which this occurs changes. The fully charged energy in the capacitor bank is 1/2 * C * V^2, and this same amount of energy is dissipated in the resistor when charging the capacitor bank. The same thing happens for discharging the capacitor bank with a resistor, only in that case it is easier to see the total stored energy ends up in the resistor.
The resistor has a certain thermal mass and temperature rise per amount of heat energy. If the resistor's resistance is large enough the heating will be stretched out over a long period of time and allow cooling processes to occur. But are we patient enough for that to do much good? The difference in cooling between a few milliseconds and a couple of seconds is not very much in these poor thermal conductivity resistors. If we set the value of the resistor to take minutes to charge the cap bank then the dissipation would have enough time to cool the resistor, but are we that patient?
What we need is a resistor with enough thermal mass to absorb the capacitor's energy without heating the material to a damaging temperature, then we can charge the capacitor bank quickly and not wait for slow cooling processes to help the resistor. High temperature pulse rated resistors are the best choice, but other types can be adequate as well. A physically small resistor must have very high temperature capability because it is going to heat that small material a lot. Physically larger resistors, especially with wire elements that can handle high temperature are well suited. The inductance they may have is also helpful in slowing the current rise. Ceramic materials can handle high temperature, those are also well suited for this application.
Another approach is to use a bunch of resistors in series or parallell instead of one component. Each dissipates part of the energy, increasing the effective thermal mass absorbing the energy.
So increasing the resistor value helps a little but not a lot until the charging time gets to be long enough for cooling processes to take place, and we may not be patient enough for that to help out.
There are different ways of looking at any problem, and this one is a bit tricky. At high current the I^2 R heating is high, but the current drops off more quickly so the heating time is reduced. So it is not obvious what the total energy absorption (and heating in the resistor) is. The interesting thing about charging a capacitor bank with a resistor is that the thermal energy dissipated in the resistor doesn't change with the resistor's value. Only the time over which this occurs changes. The fully charged energy in the capacitor bank is 1/2 * C * V^2, and this same amount of energy is dissipated in the resistor when charging the capacitor bank. The same thing happens for discharging the capacitor bank with a resistor, only in that case it is easier to see the total stored energy ends up in the resistor.
The resistor has a certain thermal mass and temperature rise per amount of heat energy. If the resistor's resistance is large enough the heating will be stretched out over a long period of time and allow cooling processes to occur. But are we patient enough for that to do much good? The difference in cooling between a few milliseconds and a couple of seconds is not very much in these poor thermal conductivity resistors. If we set the value of the resistor to take minutes to charge the cap bank then the dissipation would have enough time to cool the resistor, but are we that patient?
What we need is a resistor with enough thermal mass to absorb the capacitor's energy without heating the material to a damaging temperature, then we can charge the capacitor bank quickly and not wait for slow cooling processes to help the resistor. High temperature pulse rated resistors are the best choice, but other types can be adequate as well. A physically small resistor must have very high temperature capability because it is going to heat that small material a lot. Physically larger resistors, especially with wire elements that can handle high temperature are well suited. The inductance they may have is also helpful in slowing the current rise. Ceramic materials can handle high temperature, those are also well suited for this application.
Another approach is to use a bunch of resistors in series or parallell instead of one component. Each dissipates part of the energy, increasing the effective thermal mass absorbing the energy.
So increasing the resistor value helps a little but not a lot until the charging time gets to be long enough for cooling processes to take place, and we may not be patient enough for that to help out.
megacycle
100 kW
Suppose it can't hurt to repeat the same information over and over, if no one is reading posts.
megacycle
100 kW
A 100 Ohm, 1W, carbon resistor in the end of a 5.5mm bullet.
Drilled 3.5mm hole in end of a 5.5mm bullet, down to about 2/3rd resistor length, held bullet together for drilling by cutting down a 5.5mm socket into a ring, to go around the bullet.
The carbon resistor doesn't appear to be fragile, after removing its coating, it's got a spiral pattern on it, I think that's the actual resistor and the body is a substrate.
Detached one resistor end cap and filed off outer coating left about 1/3 covered and exposed other end cap by filing.
Placed a tiny insert of TRS rubber against the resistor side, when pushed into bullet end, to hold it in place, it must be some kind of compressible material otherwise when the bullet is pulled out, the resistor will fall out the end and when connecting, if the insert is too hard, the bullet end won't compress for insertion into the socket.
The resistance value changed a bit, to 133 Ohms when measured.
Did the same with a 1000 Ohm, 5W, had to drill a 4mm hole this time and put a tiny bead of silicone close to the exposed tip, for both insulating and holding resistor in place and have left the ring around the bullet till the silicon dries.
Things to bear in mind, be careful when drilling to drill straight down, the bullet seemed to guide it naturally.
The bullet is now thinner walled, due to drilling, it doesn't appear to affect the inserting or the bullets stiffness as the resistor is inside it, but it should affect the current rating of the bullet a bit, because it's volume has decreased, but most of the current should flow through the gold plated layer and my setup is only 20A max.
How will affect the resistor power rating, not to sure, if anything the brass inner bullet should provide some heat sinking for the resistor, possibly increasing it's power rating.
Drilled 3.5mm hole in end of a 5.5mm bullet, down to about 2/3rd resistor length, held bullet together for drilling by cutting down a 5.5mm socket into a ring, to go around the bullet.
The carbon resistor doesn't appear to be fragile, after removing its coating, it's got a spiral pattern on it, I think that's the actual resistor and the body is a substrate.
Detached one resistor end cap and filed off outer coating left about 1/3 covered and exposed other end cap by filing.
Placed a tiny insert of TRS rubber against the resistor side, when pushed into bullet end, to hold it in place, it must be some kind of compressible material otherwise when the bullet is pulled out, the resistor will fall out the end and when connecting, if the insert is too hard, the bullet end won't compress for insertion into the socket.
The resistance value changed a bit, to 133 Ohms when measured.
Did the same with a 1000 Ohm, 5W, had to drill a 4mm hole this time and put a tiny bead of silicone close to the exposed tip, for both insulating and holding resistor in place and have left the ring around the bullet till the silicon dries.
Things to bear in mind, be careful when drilling to drill straight down, the bullet seemed to guide it naturally.
The bullet is now thinner walled, due to drilling, it doesn't appear to affect the inserting or the bullets stiffness as the resistor is inside it, but it should affect the current rating of the bullet a bit, because it's volume has decreased, but most of the current should flow through the gold plated layer and my setup is only 20A max.
How will affect the resistor power rating, not to sure, if anything the brass inner bullet should provide some heat sinking for the resistor, possibly increasing it's power rating.
Alan B said:
I'm right there with you. Thats the point.
as you mentioned resistor with wire elements: i'm quit sure the type in AS 150 and also XT 90 is such wire-wound type and the part i've seen glowing like a small lamp through the housing was exact this wire so it seems its only the wire itself that must take the whole thermal impact at the beginning rather than the entire resisitor. this is how it looks:
wire resistor or metal should be normally the better choice for higher operating temperatures as you mentioned, but i have not found anything about how well they can handle short bursts.
it would be interesting if a carbon type would be better choice or not - at least the resistive material would have more mass:
like this:
as the resistor in AS 150 only is plugged in and not soldered or anything else, it would not be hard to replace it with one that holds. maybe we can find one
alsmith
100 kW
Does anyone know what type of resistor is it that's used on the Hobbyking type of connector? And how are they constructed? They are neat so modifying might be an option.
wesnewell
100 GW
The ones I bought were worthless for my application with 24s lipo once I found out the resistor was only 6 ohm. They shouldn't be used with voltages over 24V imo, though some have used them with up to 12s lipo. It all depends on the controller. They might work ok with a low voltage controller with small caps inside. I used them behind my own precharge circuit.
megacycle
100 kW
wesnewell said:
Yeh if you go from 6S to 18S Lipo energy dumped due to voltage is 9X and if the controllers caps are eg 4X bigger, giving 28X difference, resistor makes for a poor fuse, if its only a 1W :?
I'd go with the 5W for higher voltage bike controllers, that way the occasional 10-15W the resistor would experience in a short time constant should be okay because it's only occasional.
If you want to be on the safe side then if CR is much bigger than 1s, waiting a few seconds might allow the resistor to dissipate the heat into into its metal connector.
I'm going to go for a 220 Ohm 5W, I just measured my charger caps and they are 6000uF and my bike controller I found was around 5000uF, so if I use the same resistor for both, around 67% of energy is dissipated in the resistor in 1RC, 7.2W & 6W will be dissipated in 1.3s & 1.1s resp'.
markz
100 TW
Whats a proper fuse for 72V 65A setup?
I assume you are talking about the "Ignition" wire below?
Probably why they have an "Ignition" wire right!
wesnewell
100 GW
One rated for at least 90VDC and at least 65A, like a CNL fuse and holder. That said, I'd just use a cheap maxi fuse holder with a 70A maxi fuse for under $10 total with spare fuses.markz said:
markz
100 TW
Whats the issue with twisting the ignition wire straight to the red power cable?
Just the simple fact if batteries are connected, the Controllers Caps are charged?
maxi fuse holder like this
I got one of those, just use automotive fuses. Quick, easy and cheap...if not free.
Just the simple fact if batteries are connected, the Controllers Caps are charged?
maxi fuse holder like this
I got one of those, just use automotive fuses. Quick, easy and cheap...if not free.
Alan B
100 GW
Those fuse holders have been known to melt.
Most automotive fuses are not rated for the voltage in use.
Solar AC/DC circuit breakers are a good option here, and give an on/off switch that is handy.
Most automotive fuses are not rated for the voltage in use.
Solar AC/DC circuit breakers are a good option here, and give an on/off switch that is handy.
megacycle
100 kW
That's because the automotive fuses are rated for automotive use up to around 30-40V max.Alan B said:
When the fuse blows, say due to a controller short circuit fault or battery lead short, a higher voltage than rated will sustain the current through an arc, just like a welder does, these cheap flat air break type, fusible link type fuses are unsuitable for higher voltage ebikes and if using a fuse, it should be the barrel type, with arc quenching material inside.
Yes they double as a battery main switch and put a precharge resistor across the breaker, so when it's switched on there's no inrush. Standard voltage ranges are between 125-250V.Alan B said:
I use a 40A and it is good for handling up to around 100A surges for up to 1 minute and over 60A for a few minutes.
I've got plenty of them if anyone wants any PM.
Attachments
markz
100 TW
How do you rate your "Solar AC/DC circuit breakers" to your controller?
The highest I could find from local source is 63A 150VDC, and right now I got 65A controller, but later will mod the traces for higher amperage.
The highest I could find from local source is 63A 150VDC, and right now I got 65A controller, but later will mod the traces for higher amperage.
markz
100 TW
How do you rate your "Solar AC/DC circuit breakers" to your controller?
The highest I could find from local source is 63A 150VDC, and right now I got 65A controller, but later will mod the traces for higher amperage.
The highest I could find from local source is 63A 150VDC, and right now I got 65A controller, but later will mod the traces for higher amperage.
Alan B
100 GW
I was directed to 100A and higher rated units on Alibaba.
The 63A unit would probably work for you. I used one on my 80A system and it only popped on a major climb. They are rated to hold overcurrent for minutes, depending on ambient temperature. But get a 100A unit, the link is in my Borg thread a few pages back, or search Alibaba, they were under ten bucks shipped.
The 63A unit would probably work for you. I used one on my 80A system and it only popped on a major climb. They are rated to hold overcurrent for minutes, depending on ambient temperature. But get a 100A unit, the link is in my Borg thread a few pages back, or search Alibaba, they were under ten bucks shipped.
megacycle
100 kW
They are rated on the curve of the breaker, most are B, B/C or C, Like I've got.markz said:
They need to be rated higher than the rating of-
Continuous, Long Surge and Short duration surge values.
So you can see that a breaker rating can be a lot less than the surge rating, eg a 40A could do 100A short surge and on a 70V ebike this equates to around 45A continuous 3100W, and uo to 7000W surges
byebyepetrol
100 W
On my wee 14.5AHr 36v Dolphin pack I used to get a spark if the charger was off at the mains before I plugged it in
but if the charger is already on, it plugs in with no spark.
but if the charger is already on, it plugs in with no spark.
megacycle
100 kW
On, Charger output caps charged, so small voltage differential between charger and battery.byebyepetrol said:
Off, output caps discharged, large differential, so large current battery to output caps.
Rodney64
100 kW
I've been using a 12v DC 200amp breaker, but now it's time to change to on of these solar DC breakers.
KINNINVIEKID
1 kW
Simply buy a pair of XT90 Anti spark connectors to replace your standard ones.andrenoites said:
[urlhttp://www.hobbyking.co.uk/hobbyking/store/__61690__XT90_S_Anti_Spark_Connector_2pairs_bag_.html][/url]
markz
100 TW
I believe we went through this already, those are good for 48V, but when you are rocking 60V+ you just need to buy a 150ohm 5 or 10% resistor and connect it up first, then connect the main leads.
icecube57
10 MW
Yeah they don't work at higher voltages. I make precharge dongles.
Matt Gruber said:
Is there any merit to this?
I want to put a resistor on a small 12 volt power supply that will be connected permanently as a precharge resistor. I will use an On/Off switch on the main power wire to turn the power supply on and off.
Does it make sense to use like a 10,000K resistor or higher for safety as the precharge resistor? This way if I have the unit powered off, and there is a short it would take a very long time to drain the battery through the precharge resistor?
llile
1 kW
- Joined
- Dec 18, 2010
- Messages
- 457
Here's how I solved this whole issue a couple of ways.
First, some battery makers include the precharge resistor, and they cheat the anderson connectors so that the precharge connection makes first. I have one battery made this way and it works pretty well.
On another bike, after I got tired of the big spark, and also some parasitic drain that was killing batteries between charges, I put a switch in the battery supply cable. It is a three-position switch - ON-OFF-ON. I was able to find a heavy duty 30 amp switch at an auto parts store - although it is rated 30 amps at 12 V it seems to work OK at 36 V. At least it hasn't failed yet!
One of the ON positions goes straight through, the other goes through a 5K resistor. I think I used a 2W resistor. Wrapped it all up in enough tape and hot glue that the resistor lead won't bend and break off.
When I park the bike, I turn it to OFF, then when I need the bike again, I turn it to CHARGE for a few seconds then ON. So far it is working pretty well. ( If I forget and turn it directly to ON, I can hear the big spark destroying the insides of the switch, POW! So far it hasn't failed.)
First, some battery makers include the precharge resistor, and they cheat the anderson connectors so that the precharge connection makes first. I have one battery made this way and it works pretty well.
On another bike, after I got tired of the big spark, and also some parasitic drain that was killing batteries between charges, I put a switch in the battery supply cable. It is a three-position switch - ON-OFF-ON. I was able to find a heavy duty 30 amp switch at an auto parts store - although it is rated 30 amps at 12 V it seems to work OK at 36 V. At least it hasn't failed yet!
One of the ON positions goes straight through, the other goes through a 5K resistor. I think I used a 2W resistor. Wrapped it all up in enough tape and hot glue that the resistor lead won't bend and break off.
When I park the bike, I turn it to OFF, then when I need the bike again, I turn it to CHARGE for a few seconds then ON. So far it is working pretty well. ( If I forget and turn it directly to ON, I can hear the big spark destroying the insides of the switch, POW! So far it hasn't failed.)
