Increasing Voltage to controller - how much can it take?

gardener

10 mW
Joined
May 23, 2007
Messages
22
Location
Hampshire, UK
Hello all

I've been a lurker on this forum for a while now and have found lots of useful advice and ideas. Many thanks to all for that.

I have a 24V brushless geared hub motor kit with a 15Amp controller and I am thinking of running the kit at 36V. My question is what components of the controller are affected by the increase in voltage.
From my (limited) understanding the main components would be the MOSFETS. I looked up the MOSFETs and they seem to be rated at 60V and 42A so I imagine they will handle 36V.
Are there any other things I should check before putting 36V through it and possibly destroying my controller? I'd rather confirm in advance whether the controller can take it before having a go.
 
You want to check the voltage rating on the main capacitor(s) and try to determine what kind of voltage regulator it has.

The voltage regulator is typically a transistor or a 78xx type linear regulator. The 78xx (usually a 7812 and a 7805) regulators have a maximum input voltage of 35v, and won't be too happy with a fully charged 36v battery.

The transistor based regulators are dependent on the transistor rating. These are often mounted on the heatsink and resemble a FET (TO-220 case).

If you post a good picture of the guts, we might be able to identify it.

If the regulator is not sufficiently rated, you can often power it from the 24v tap on the battery pack. This usually keeps the low voltage cutout feature intact as well. The regulator is usually powered through the on/off switch, so rewiring it is not difficult.

Here's a diagram of the basic idea:
 
Thanks for the quick reply. I have a couple of pictures, admittedly not great ones but they may help with identification. I will take some more later on tomorrow.

Sorry about the gigantic photo, I got the other one ok but don't know what happened with the capacitors pic :oops:

There are 6 MOSFETS and I believe based on the writing on them that they are STP60NF06 N-CHANNEL 60V - 0.014W - 60A TO-220/TO-220FP
STripFETâ„¢ POWER MOSFET.
 

Attachments

  • overview.JPG
    overview.JPG
    421.3 KB · Views: 2,412
  • capacitors.JPG
    capacitors.JPG
    333.8 KB · Views: 2,497
Wow ftecher, how do you know all that stuff?

My suggestion, blast the thing with as many volts as you want, then if it breaks, buy a new one that's made to run 36v for $35:

http://www.tncscooters.com

My calculations:

My time is worth $80/hr. Messing around with a controller takes 60 hours (if you're like me). Total cost = $4800. By my calculation, you'll be saving 99.3% by buying a new one.
 
gardener said:
Thanks for the quick reply. I have a couple of pictures, admittedly not great ones but they may help with identification. I will take some more later on tomorrow.

Sorry about the gigantic photo, I got the other one ok but don't know what happened with the capacitors pic :oops:

There are 6 MOSFETS and I believe based on the writing on them that they are STP60NF06 N-CHANNEL 60V - 0.014W - 60A TO-220/TO-220FP
STripFETâ„¢ POWER MOSFET.

The caps are 50 volts, so on the safe side you would probably blow them with a 48 volt setup (since fully charged they will be more, probably around 56 volts)

A 42 volt setup might work (36V + 6V), the controller might get really hot, so that's another thing to check on to prevent overheating.
 
Capacitors usually have a fair amount of overvoltage tolerance. Running a few volts over usually won't blow them. Increased ripple current blows them. Using higher voltage rated capacitors would be better, but might not be worth changing unless you have easy access to the right materials.

Beagle's point about similar controllers being dirt cheap is a good one. Not much risk in trying.

From the picture it appears that controller uses a switching regulator. I've not seen that in many, but it's a good feature. I'd check the transistors near the inductor to see what their voltage rating is. The inductor is that red cylindrical thing in the lower left of the pic.

If you use the 24v battery tap, you don't need to worry about the regulator.
 
You want to check the voltage rating on the main capacitor(s)

The two big caps are 50V, the other 3 smaller ones next to them are 50V, 50V and ? (I guess it'll be 50V as well). There are 3 caps on the right hand side of the board next to the 2 chips. Those ones are 16V, 25V and 25V only.

Beagle's point about similar controllers being dirt cheap is a good one. Not much risk in trying.
I'm not sure if the controllers mentioned will work with my motor. I have the following connections on the controller:
(near the MOSFETS) yellow, blue and green. These wires go to the motor.
(next to the big caps) red and black to the battery
(bottom of the board on left) 5 wires to motor (hall effect wires I am led to believe) red, black, yellow, green and blue.
3 wires go to the throttle (yellow (-ve), red(+ve) and green(marked sp on the board)
2 other blue wires go to a pot which seems to adjust the speed the motor spins at (one of these goes to where the big caps are and the other goes near the hall sensor wires to a place marked se on the board).

If someone can suggest a controller they think will work with my connections then as suggested I can get a backup controller in case I blow mine up.

I don't know much (anything really) about electronics but am I right in thinking that the part of the board with the chips on would be powered at a low voltage (say 5V) and the only high voltage bit is where the MOSFETS and big caps are?

When the board is in the case the MOSFETS are pressing against the outside of the aluminum case and there is also a fin from the top cover pressing against the front of them when the cover is on, so I think the heat sinking is good.

As the big caps are 50V then could it be that I could run the kit at 36V then?
 
Oh yeah, you're right, your controller is for a brushless motor, so the super cheap controllers won't work. The 20 amp Crystalyte controller looks very similar.

36v should be no problem for the caps. If you use the 24v tap method, then everything else should be fine. I'm just not sure if you can pump 36v into the existing connections, but it appears that should be OK too.

The battery voltage gets stepped down to around 12v and 5v for everything except the output stage.

Take a look at the two black parts pointed out below. See if you can read the numbers off them.
 

Attachments

  • t_overview_2.jpg
    t_overview_2.jpg
    40.6 KB · Views: 3,893
Sorry for the delay in replying, I've been too busy to use the bike for a week or so.

Only the black thing on the left of the photo has any writing on it. As far as I can make out it says A1275 YK535. Hope this helps with identification.

Regards

Paul
 
Thanks to all the good advice given I decided to try out the kit at 36V. I used a 12V SLA from a server UPS and wired it in series with my 24V Nimh. Wow! The speed and power increased considerably. Unassisted speed is now 20+ mph on the level (15.5 mph on 24V) and with pedalling it does 22-25mph. On a downhill slope on full throttle I could feel the motor at 24mph.
These speeds are on 26" * 1.5 tires and with the speedo set to 1965mm wheel size.

I did 4 miles and the controller is still in one piece! It did get hot however. I estimate the aluminium case was 35-40 C. Is this too hot? The controller was in a pannier so it wasn't getting any airflow to cool it. Would it help substantially if I strapped it to the frame with good airflow over it?

Interestingly the three lights (min, mid and max) on the throttle finally lit up at 36V. At 24V I have only ever seen the min light. This makes me think that the throttle is from a 36V kit. I wonder also if the controller is really a 36V one also? Perhaps my kit was sold as 24V to keep the cost (of batteries) down? Certainly the motor felt so much better running at 36V, it makes me realise how underpowered it was at 24V. The motor was not perceptibly warm after the run either so I don't think there's any problems there.
 
There is never a good reason for covering up the controller; it's always best to expose it to air. Since you're a moving vehicle, you have all this free rapidly-moving air that's great for cooling, so take advantage of it. To keep the controller waterproof, all you need to do is use tape or a soft glue to seal the entrances around the wires. Even that shouldn't be necessary if you don't ride in a downpour.
 
40C is OK. Around 60 is the danger point. More air is better.
Cooling fins, like an old CPU heatsink, glued onto the side with silicone will help quite a bit too.
 
fechter said:
40C is OK. Around 60 is the danger point. More air is better.
Cooling fins, like an old CPU heatsink, glued onto the side with silicone will help quite a bit too.
Thanks for the tip. Do you mean glue the fins to the outside of the controller case?
 
Yep. I suppose it's nicer to bolt the fins on with some heatsink goop inbetween, but silicon RTV glue is nearly as good as a heat conductor as long as the glue layer is very thin. If needed, just pry them off with a screwdriver for service.
 
Thanks for the advice everyone. I have strapped the controller on the chainstays just behind the seat tube. I've done a couple of 5 mile runs flat out and the controller case is not perceptibly warm at all. I may not need the extra fins after all, however they'll do no harm so I will get one. Just got to work on waterproofing it now.
 
CGameProgrammer said:
There is never a good reason for covering up the controller; it's always best to expose it to air. Since you're a moving vehicle, you have all this free rapidly-moving air that's great for cooling, so take advantage of it. To keep the controller waterproof, all you need to do is use tape or a soft glue to seal the entrances around the wires. Even that shouldn't be necessary if you don't ride in a downpour.
Hi

There is a good reason to cover the controller, even if you seal it with silicon goo water will still get in to the controller, its better to keep them in the airflow if you can but not 100% necessary, my KMX controller is in a saddle bag on the side of the bike and is fine at 42V 35A, it is warm but nowhere near the cut-off for the controller, it totally depends on the ambient temperature though.

My controller on my main bike sits on a beam rack, I made up a clip on cover out of one of those camping dry bags, its a tough vinyl material, it works great as its not enclosed, the air goes through and out of the back but no rain gets in. Water is very searching it will find its way in and its not a good mix with those controllers I can tell you! ha ha, the on/off buttons on them are very poor and exposure to rain and dirt will kill them in no time, its better to cover these controllers than expose them.

Cheers

Knoxie
 
I removed the on/off button on mine, since the controller is located in an inaccessable spot on my scooter anyway. I plugged the hole by gluing a small piece of metal over the inside. The blue wire in the throttle cable goes to the same place as the switch, so I switch the power from the blue wire to the battery (+) through my main relay. You could also just run a pair of wires from the existing switch to a new one (water resistant) mounted in a convenient spot.
 
I was just thinking of asking about airflow vs waterproofing the controller.

It is not that uncommon for me to get caught out in the rain so it is always a toss up air flow or raincover.

Since hanging around this forum I have been pushing my bikes speed and have heated up the controller more than ever while at the same time shifting its position to a seat bag. I may need to rethink that location

Hey Knoxie, any chance of posting a picture of your rain cover?

Cheers Greg
 
Back
Top