No Brakesmart, No Maxbrake... Now what?
Registered User
Joined: Nov 2011
Posts: 4,479
Likes: 211
From: Isanti, MN
I am using an fqp47p06 Mosfet. Originally I started with 10kHz PWM, but had to redesign due to the heating problem, and also because of too much temperature instability with my circuit. After that, I decided to use Adruino for control. I tried 490Hz and 245Hz or whatever the slower output is. I am using a base biased 2N3904 bjt to pull the mosfet gate low and a resistor to drive it back high when the bjt is off. I don't remember the value of the resistor, but I think I chose something to flow about 100mA. I also used part of a 50A bridge rectifier as a freewheel diode, which does help with the heat. The heat is basically just split between the rectifier and mosfet. I feel like the whole thing still disapates too much heat. The cheapo off the shelf controllers don't seem to even get warm. I opened one of cheapos up to see if I could get a clue how they do it, but it looks like just the mosfets and freewheel diode. I will say they are using two 50A mosfets and they are oddly enough N-type. I was thinking of doubling mine up and seeing what happens.
Without putting a scope on it, I'd bet you're fighting energy dissipation from the field collapse in the brake magnets. Again, with nothing but imagination for data, I would suspect the others use lower current, higher voltage, faster mosfets and let the back emf ring free, or maybe snub it at 100v or so with a zener.
On the other hand, doing some quick calcs, you should have 3.5W or less heat to handle. Sumthin's fishy. Heck, I drive the third axle (second trailer) with a darlington emitter follower on about a 10 inch heat sink, and it don't get very hot at all. That's full linear brute force.
Registered User
Joined: Sep 2009
Posts: 5
Likes: 0
That's a fairly heavy switcher, but slow. I imaging is almost smoked at 10Khz. It'll do the job but needs a fairly decent heat sink to dissipate the heat. It's rated 160W if kept cool, probably about 2 watts or less in free air.
Without putting a scope on it, I'd bet you're fighting energy dissipation from the field collapse in the brake magnets. Again, with nothing but imagination for data, I would suspect the others use lower current, higher voltage, faster mosfets and let the back emf ring free, or maybe snub it at 100v or so with a zener.
On the other hand, doing some quick calcs, you should have 3.5W or less heat to handle. Sumthin's fishy. Heck, I drive the third axle (second trailer) with a darlington emitter follower on about a 10 inch heat sink, and it don't get very hot at all. That's full linear brute force.
Without putting a scope on it, I'd bet you're fighting energy dissipation from the field collapse in the brake magnets. Again, with nothing but imagination for data, I would suspect the others use lower current, higher voltage, faster mosfets and let the back emf ring free, or maybe snub it at 100v or so with a zener.
On the other hand, doing some quick calcs, you should have 3.5W or less heat to handle. Sumthin's fishy. Heck, I drive the third axle (second trailer) with a darlington emitter follower on about a 10 inch heat sink, and it don't get very hot at all. That's full linear brute force.
But you're right, it's all imagination for data without seeing it on a scope.
Is your darlington emitter follower being used to drive a second trailer from the signal of the first? (as in pulling a double or 'road-train') Just trying to picture the configuration in my head...
Registered User
Joined: Nov 2011
Posts: 4,479
Likes: 211
From: Isanti, MN
Nah, all the Tekonsha's, Draw-Tite's, and MaxBrakes I've taken apart run high current MOSFETs and a high current freewheeling diode to keep the ringing from the magnets under control... I think they used high current MOSFETs to protect the warranty department, as they'd probably take enough abuse to pop the fuse before the transistor. I've been reverse engineering these things to try and build and possibly commercialize my own brake control since all the 'good' TBC manufacturers seemed to have gone the way of the dodo without much information about their demise (lawsuits or patent infringement suits?). Maybe I should release mine under OpenSource Hardware and sell the boards or somethin...
But you're right, it's all imagination for data without seeing it on a scope.
Is your darlington emitter follower being used to drive a second trailer from the signal of the first? (as in pulling a double or 'road-train') Just trying to picture the configuration in my head...
But you're right, it's all imagination for data without seeing it on a scope.
Is your darlington emitter follower being used to drive a second trailer from the signal of the first? (as in pulling a double or 'road-train') Just trying to picture the configuration in my head...
Here's the diagram so you don't accidentally strain yer brain.
Registered User
Joined: Sep 2015
Posts: 5
Likes: 0
I did scope the output to the magnets, and while it's been awhile, I remember the the rising edge of the signal was nice and sharp. I think the falling edge was a bit slow, but I wasn't sure if that was the mosfet slowly turning off or if it was some sort of feedback/ringing from the coils. I did look for the recommended LT1154 and couldn't really find anywhere to buy it. The LT1910 is more to my liking, but it appears to be only be available in surface mount. I realize my power dissipation will be 5 times higher with my current mosfets, but is there any way to use the components I already have? I was planning to double up the mosfets, which should half my mOhms, but then I'll have twice the gate capacitance to deal with. I feel like I need some sort of push/pull transistor configuration on the mosfet gate. I was also thinking, since my load is inductive, if one of the edges needs to be slow, maybe I would be better to slowly turn the coils on.
Registered User
Joined: Sep 2015
Posts: 5
Likes: 0
The mosfets the cheap controller uses are fqp50n06 which are 21mOhm 24nC and it doesn't get hot. The ones I'm using are fqp47p06 26mOhm 84nC. So I guess the biggest difference is just the gate capacitance amd the fact that they use 2 mosfets.
Registered User
Joined: Sep 2009
Posts: 5
Likes: 0
I guess I'm not opposed to doing SMD work, you could get a 'Surf-board' to convert the driver chip to a through-hole package (like DigiKey 9163CA-ND, or 9082CA-ND). Or a prototyping board like this one: Universal Prototyping board
Typo on the LT1154, try LTC1154, Digi-Key stocks them (also only available in an SMD package).
All the brake controllers I've opened up are using parallel N-channel output devices (don't know any of the part numbers off the top of my head), which I'd always assumed was for redundancy, but maybe they actually need it for power dissipation.
Typo on the LT1154, try LTC1154, Digi-Key stocks them (also only available in an SMD package).
All the brake controllers I've opened up are using parallel N-channel output devices (don't know any of the part numbers off the top of my head), which I'd always assumed was for redundancy, but maybe they actually need it for power dissipation.
Registered User
Joined: Apr 2008
Posts: 764
Likes: 0
From: Flagstaff, AZ
I too regret not buying a Maxbrake or Brakesmart previously, I didn't think they would go out of business. I have been using a prodigy p2 since 2007, I never had problems with it (not much ice in AZ and I only tow a few times per year). Now there is something wrong with the electronics in the p2 and it doesn't shut off, it is draining my battery if the truck sits for a couple weeks.
Thread
Thread Starter
Forum
Replies
Last Post
klx650a2
Towing and Hauling / RV
13
Sep 23, 2010 07:59 PM
wallawalla
3rd Generation Ram - Non Drivetrain - All Years
16
Mar 1, 2008 10:28 AM