Crystalyte Controllers - Repair and Modification information
- Thread starter fechter
- Start date
25k is a wild ass guess. Be careful messing around with that one, since you could easily crank the limit up to self-destruct point.
Roy
100 mW
I've got the old FET's out of my 20a controller board and
have done my best to soak up the excess solder. But... the
holes are just too tight to allow the replacement 4110's to fit in
properly. Right now I'm considering: Filing the FET leads to make
them skinnier (sounds insane), tiny drill to open up the holes (likely to tear the trace?), clamp a paper clip to the tip of a solder gun and force it thru the
holes melting solder along the way. etc. Only 2-3 out of 18 holes are adequate to allow the new FET's to install easily.
Suggestions?
Thanks,
Roy
BTW, how does one get the privilege of starting a new topic?
have done my best to soak up the excess solder. But... the
holes are just too tight to allow the replacement 4110's to fit in
properly. Right now I'm considering: Filing the FET leads to make
them skinnier (sounds insane), tiny drill to open up the holes (likely to tear the trace?), clamp a paper clip to the tip of a solder gun and force it thru the
holes melting solder along the way. etc. Only 2-3 out of 18 holes are adequate to allow the new FET's to install easily.
Suggestions?
Thanks,
Roy
BTW, how does one get the privilege of starting a new topic?
the holes should be big enough for any TO-220 fet. have you used solder wick to get the last of the solder out?
the next step is to enlarge the holes with a sharp probe or drill. they can be soldered on both sides of the board so you do not need to rely on the through hole plating to carry the gate signal through, and the power leads must be soldered to the side where the bus is anyway.
be sure to use plenty of solder on the drain and source connections, as it is really the solder that carries the bulk of the current
the next step is to enlarge the holes with a sharp probe or drill. they can be soldered on both sides of the board so you do not need to rely on the through hole plating to carry the gate signal through, and the power leads must be soldered to the side where the bus is anyway.
be sure to use plenty of solder on the drain and source connections, as it is really the solder that carries the bulk of the current
Solderwick is good for clearing the holes.
Another method is "heat and beat". Heat up the solder, then quickly whack the board against a table or something to fling the solder out of the hole. Just be careful not to damage any of the protruding components.
This is about the only way to clear solder from a solder cup type pin on a connector.
Good luck.
Another method is "heat and beat". Heat up the solder, then quickly whack the board against a table or something to fling the solder out of the hole. Just be careful not to damage any of the protruding components.
This is about the only way to clear solder from a solder cup type pin on a connector.
Good luck.
Roy
100 mW
bobmcree said:
I don't think I can wick much more solder off w/o more trace damage. I will make sure S/D get plenty of solder.
Thanks,
Roy
Roy
100 mW
Yikes! Don't think I've the spunk to try "heat and beat". But what do you mean by "solder cup type pin"? To me these are just holes in the PC board. I've got a probe made from a bobby-pin that is flat like a component lead. Made by filing it down to a tapered end. That's my next try.fechter said:
Roy
Solder cup pins are like the kind installed in many connectors. Not usually found on boards. Since the "cup" is closed at the bottom, there's no way to vacuum out the solder.
Anyway, the bobby pin should work. I sometimes use a tiny screwdriver like a drill to ream out the leftover solder. Just try not to destroy the copper or lift the traces. Like Bob pointed out, if you rip out the plating that goes through the hole, you can fix that by thoroughly soldering the lead on both sides of the board.
Anyway, the bobby pin should work. I sometimes use a tiny screwdriver like a drill to ream out the leftover solder. Just try not to destroy the copper or lift the traces. Like Bob pointed out, if you rip out the plating that goes through the hole, you can fix that by thoroughly soldering the lead on both sides of the board.
TylerDurden
100 GW
Sometimes you can use a can of dust-off (compressed-air) to blow out the hot solder.
:?
:?
I agree with the `heat and beat' thing, when nothing else seems to work. It's not like you have to whack it real hard. Just get it hot and let the edge of the board fall lightly against the table using your wrist as a pivot. It's less likely to lift traces than other methods.fechter said:
If your controller board can't stand up to that it won't make it very far down a rough road.
(Note. I haven't tried this on a controller. Just on all kinds of other boards.)
Richard
rkosiorek
100 kW
TylerDurden said:
i usually use a small HSS drill bit but not as a drill i heat up the solder then i push thru the drill don't ask what size. i just use the small one from the number drills that fit in the hole. the black oxide coated ones are okay because the solder dosent stick to them.
rick
calinb
100 W
- Joined
- Oct 28, 2007
- Messages
- 102
Thanks for this excellent sticky post. I have the "newer" rev. "B" X-CT4840B 48V 40A Phoenix Speed Controller sold by ElectricRider.com. It's powered by a 16 cell Yesa 20Ah LiFeP04 pack and runs a 5304 on a 26" wheel. I'm considering upgrading my FETs to lower Ron components and also doing the shunt rework described here, while I've got the box open.
I see references to "20A", "35A", and "40A" controllers. Which one is mine? Is it really a 40A controller or is it actually only 20A (continuous) or something else?
Also, will this controller work for battery packs substantially over 48V? I may want to try a 72V or even 96V pack someday.
Thanks again,
-Cal
I see references to "20A", "35A", and "40A" controllers. Which one is mine? Is it really a 40A controller or is it actually only 20A (continuous) or something else?
Also, will this controller work for battery packs substantially over 48V? I may want to try a 72V or even 96V pack someday.
Thanks again,
-Cal
your 4840b is a 40A controller which most likely has 55v fets. you are probably right up at the limit with your pack fresh off the charger.
you can read the part numbers on the fets without taking the board off the metal base to verify what they are using in your unit. you might not reduce the on-resistance a lot by switching to the higher voltage IRFB4110s but you will gain the ability to go to 100v.
the 20a controllers are much smaller, the 30,35,40, and now 50A units are all the same as the one you have with different fets and different shunt resistor values.
you can read the part numbers on the fets without taking the board off the metal base to verify what they are using in your unit. you might not reduce the on-resistance a lot by switching to the higher voltage IRFB4110s but you will gain the ability to go to 100v.
the 20a controllers are much smaller, the 30,35,40, and now 50A units are all the same as the one you have with different fets and different shunt resistor values.
calinb
100 W
- Joined
- Oct 28, 2007
- Messages
- 102
Thanks! You answered all my questions, Bob. I'll be opening up my controller to install a 6-pin connector for a Cycle Analyst soon. I'll check my parts against the others in the FET Selection Guide.xls spreadsheet.bobmcree said:
Hey--I just noticed we're neighbors. I live in Portland.
-Cal
RevDon
100 W
Please see this post, and let me know what you think 
http://endless-sphere.com/forums/viewtopic.php?t=2868
Don
http://endless-sphere.com/forums/viewtopic.php?t=2868
Don
Hello, could someone help mod my 36/48v 20amp controller ---> 72v20amp?
(I'd like mosfets changed to 110v, so 4110 using.)
Do these mosfets cost $2-$3 when bought? (so labor will takes an additional 20 minutes or so? or is this charged by project basis?)
If Bob specializes in these controller mods, could you provide standard quote?
Also, do you happen to know RMS value on Crystalyte 72v20amp controller?(110v is single phase peak rated) Could 30s2p (or 99v) dewalts work alright on this controller? 99v is used most on trailer hauling or heavy carrying on bike.
I want to use 6 dewalt packs in these configurations:
15s2p x 2
10s2p x 3
20s2p + 10s2p
If it's best in using each pack frequently or one after the other. Then I'm leaning towards 15s2p x 2. This way, I could use one 15s2p pack one day, then switch to another 15s2p. So that both packs gets used regularly, with equal cycles between them.
I'm also needing just 36v-50v on these "neighborhood" rides. So using one pack is perfect. Then on my longer journeys (23-27km or more), I could series them together in getting 99v. Though I use the 99v more in range than speed, most of the time. With feather boosts when needed.
(I'd like mosfets changed to 110v, so 4110 using.)
Do these mosfets cost $2-$3 when bought? (so labor will takes an additional 20 minutes or so? or is this charged by project basis?)
If Bob specializes in these controller mods, could you provide standard quote?
Also, do you happen to know RMS value on Crystalyte 72v20amp controller?(110v is single phase peak rated) Could 30s2p (or 99v) dewalts work alright on this controller? 99v is used most on trailer hauling or heavy carrying on bike.
I want to use 6 dewalt packs in these configurations:
15s2p x 2
10s2p x 3
20s2p + 10s2p
If it's best in using each pack frequently or one after the other. Then I'm leaning towards 15s2p x 2. This way, I could use one 15s2p pack one day, then switch to another 15s2p. So that both packs gets used regularly, with equal cycles between them.
I'm also needing just 36v-50v on these "neighborhood" rides. So using one pack is perfect. Then on my longer journeys (23-27km or more), I could series them together in getting 99v. Though I use the 99v more in range than speed, most of the time. With feather boosts when needed.
3 DeWalt packs in series might be pushing it, but I bet it would work. I think you'd be better off with 2 series. If you mod the controller with 4110's, you can increase the current limit to get more power.
3 in series is somewhat pushing limits. (true)
which is 99v on a 110v controller.
(though I've already tried 84v on 72v 20amp controller no problems, so I guess 99v is an interesting next step)
My 36/48v 20amp controller is just my second controller I'm wanting modded.
If you use 99v with partial throttle, does (clyte) 407 motor/controller somehow operate at less than half efficiency? What is efficiency measured with? (some formula/equation?), how does this affect battery life?
Though I have done varying throttle amounts on 84v, with dual packs lasting several hours. Think mostly I'd be concerned with battery life.
which is 99v on a 110v controller.
(though I've already tried 84v on 72v 20amp controller no problems, so I guess 99v is an interesting next step)
My 36/48v 20amp controller is just my second controller I'm wanting modded.
If you use 99v with partial throttle, does (clyte) 407 motor/controller somehow operate at less than half efficiency? What is efficiency measured with? (some formula/equation?), how does this affect battery life?
Though I have done varying throttle amounts on 84v, with dual packs lasting several hours. Think mostly I'd be concerned with battery life.
irfb4110 are to my knowledge good for 100V, not 110V. I would not use the controller for anything more than about 80V anyway. You get peaks from the motor winding that exceeds the battery voltage.
The crystalite controllers are not 100% protected and not perfect. Many failed. Don't try your luck, unless you are prepare to replace your controller soon.
The crystalite controllers are not 100% protected and not perfect. Many failed. Don't try your luck, unless you are prepare to replace your controller soon.
Lowell's 4110 controller has survived 108v, so I guess there's a bit of headroom in the ratings. As soon as you put some load on the batteries, the voltage will sag a bit anyway.
At partial throttle, the efficiency does drop, but not to 50%. We don't have any actual test data on this. Efficiency is much more affected by load. If you are pulling full throttle on a really steep hill and the motor bogs down to half speed, then the efficiency will suck. At light load, partial throttle, I think it will still be pretty good.
At partial throttle, the efficiency does drop, but not to 50%. We don't have any actual test data on this. Efficiency is much more affected by load. If you are pulling full throttle on a really steep hill and the motor bogs down to half speed, then the efficiency will suck. At light load, partial throttle, I think it will still be pretty good.
when you write Lowell's 4110 controller, you mean Crystalyte 72v20a controller? (both uses same mosfet?)
so if I get on bike (160-170lbs), how much will voltage sag?
so if 30s (fresh off charger) will be 110-111v. is there a way in getting this down --> 99v nominal voltage? Do you need to let batteries rest a few hours or minutes? Or ride just one 15s pack till you reach 50v?
Or if 30s is too risky, maybe ... 15s2p + 12s2p? (27s) ... or is safest route: 12s2p x 2, then a 6s2p booster pack? (so either use dual 12s2p ... or 12s2p + 6s2p)
What about using 13s2p dual packs in series? (unless all cells should be in 6's?)or 15s2p + 12s2p or 13s2p? ...
Maybe dual 12s2p + 6s2p? (then I could choose either 24s or 18s... 12s+12s or 12s+6s)
Is someone also familiar with NG1 charging times on these dewalt packs? Could you explain how charging time works? Is there a way to modify charger in getting fastest charge time? Either that or build packs in such a way to get fastest charging?
If 99v is too risky with it's off the charger 110v, could I compromise and just get around 90v? (which is 27 or 28 cells in dual packs)
so if I get on bike (160-170lbs), how much will voltage sag?
so if 30s (fresh off charger) will be 110-111v. is there a way in getting this down --> 99v nominal voltage? Do you need to let batteries rest a few hours or minutes? Or ride just one 15s pack till you reach 50v?
Or if 30s is too risky, maybe ... 15s2p + 12s2p? (27s) ... or is safest route: 12s2p x 2, then a 6s2p booster pack? (so either use dual 12s2p ... or 12s2p + 6s2p)
What about using 13s2p dual packs in series? (unless all cells should be in 6's?)or 15s2p + 12s2p or 13s2p? ...
Maybe dual 12s2p + 6s2p? (then I could choose either 24s or 18s... 12s+12s or 12s+6s)
Is someone also familiar with NG1 charging times on these dewalt packs? Could you explain how charging time works? Is there a way to modify charger in getting fastest charge time? Either that or build packs in such a way to get fastest charging?
If 99v is too risky with it's off the charger 110v, could I compromise and just get around 90v? (which is 27 or 28 cells in dual packs)
shinyballs
1 kW
Is there a problem if the black wires going to the battery of the 35A 72v Controller is shorter by 20 inches? Red wire is longer cause I added a shunt and a fuse. thanks...
shinyballs said:
No problem, it won't make any difference. You just don't want lots of excess wire that you don't need.
Similar threads
