I think I might have burn up the controler

[wiredwrong]

  I've hit another snag on this project, the motor and transmission did not provide the speed I was aiming for, it had plenty of low end but was very slow, and had no range to speak of.

 I had gotten a couple of power chairs from the local scrap collector and decided to try the motor from one of them, so I cut out the old one and welded in the new motor mounts, assuming that the output shaft was 5/8"(I didn't measurer, just guessed), I headed to the mower shop to get a sprocket, who would have ever guessed that a 17mm shaft was so uncommon. Could not find one, I decided to make one, that went without further setbacks.

 I hooked everything up to a 24v 40A controller and went for a test drive, top end was perfect but was a little sluggish on take off. I decided I needed a bigger controller, and I knew that I had one off of another power chair that I had scraped out years ago.

It turned out to be a Curtis 1208-24xx, 55A peak 70A w/reverse and all kinds of neat little features, but I had not kept any of the controls for it, and had no idea what wire went where.

 After many hours of searching I found a manual online in .pdf format. So after reading the whole thing and making a parts list, I went to radioshaft, and got me a 5k-ohm potentiometer, and a couple of various resistors, came home and hooked up the controller to a set of battery's and a motor, IT WORKS!!!!

I was going to head to the shop and put it all together on the bike, when the doorbell rang, I pushed everything aside on the table, only to find out it was one of the kids in the neighborhood, who had already been told repeatedly, my little one was away at her grandma's for the weekend. When I got back to the table, I began to unhook everything when a loose hot wire hit the board somewhere, for just about 1/2 a second or less, I decided to test it again, its fried. It still goes forward/reverse and will stop, but it goes full throttle in both directions and when it stops the boards safety feature shorts the motor like a brake, there is also a resistor on the board that gets very hot very quick now.

 Could it be the potentiometer?? I tried to find a TV or radio repair shop to talk to me about how to test a certain parts of the board like resistors and diodes and such, but they don't exist here anymore. I'm reading all I can and learning, but have a long way to go. I'll snap more pictures of the bike, motor, and controller in the morning. If you have any thoughts on what it might be let me know.

[wiredwrong]

Here's a link to th first entry about this project.

http://www.fieldlines.com/story/2007/12/31/0187/1809

[wiredwrong]

Sorry, forgot to post picture of controller

[Bruce S]

RandyOKC;

I've been following your post of the bike/car with great interest:-)

   There's a couple thing you can test. The 5k pot. You should be able to put a meter on the two outer leads, most pots I have is 3 leave the one in the middle alone for know. Put your meter in Ohms , my cheapo one has a 20k point, this will be just fine. Turn the pot one way then the other,go slow, see if the meter changes. It should go from (0) to reading 5k. The middle one is usually 1/2 the full range. Make sure and test it going slow, if for some reason it goes to (0) anywhere other than the (0) beginning point , it toast.

The HOT resistor: The picture you have up isn't close enough to check, you can send me a pic to my private email of any size , I can look closer then.

I would hazard a guess that if the resisitor is getting hot, more than likely one of the voltage regulators is "toast" too.

Curtis doesn't use any strange parts. Mostly off the shelf items, which is good as you can back track and replace the affected ones.

You in a last ditch effort contact Curtis and ask them about a rebuild (:-]

Send pic of the front and back and I'll have a good close look.

Good luck

Bruce S

 

[wiredwrong]

  Thanks for your interest and reply, Ive got a meter that only has a 200k range here and I'm getting readings from the pot that would suggest to me that its bad, but I have a fluke meter at the shop and I'll test it over there next time I get that way, the readings I'm getting here are all over the place. Ive never trusted this meter here at the house.  

here's the back of the board..

   And the front...

And a shot at the hot resistor...

I think possibly I may use this dc motor as the genny to charge it ...

spinning by hand it easily lights a 24v bulb to full brightness, and I hope that it will make a good hawt.

[wiredwrong]

Got the wrong tag for the front...

[Bruce S]

RandyOKC;

   Took as best a look as I could. The Resistor should a warm, and may get hot to the touch. I have a feeling that the three legged bugger just below it is shorted closed internally. The resistor is probably good for a full watt it's a 46 or 47  Ohm,can't tell if that's blue or violet, so it's possibly for regulation purposes.

There is also a couple interesting digs aourna couple places on the back along with a few black/ burn marks.

If your good at soldering, you could pull the transistor out and test it.

How fast does the resistor heat up? if the transistor to shorted internally it'll get hot quick.

Hope this helps

Bruce S

[oztules]

RandyOKC,

 I agree with Bruce on this. The MPSA06 (NPN general purpose transistor) will be internally shorted. This needs replacement.(or use an equivalent eg (2n4401 etc.)

This transistor shapes the inverted drive for the fets that act as freewheel diodes. (the ones closest to the relays). When it shorts, it places the 470r directly between b+ and b-..... gets hot quick, and will cook if left running.

Once that is replaced, and you have checked your pot, things should get better.

So far from your tests, the forward / rev relays are working, your switch-on relay is working, but we have yet to find the pulse width control working. If the Fets were shot, the relays would not have pulled on, and the 1W resistor next to the three transistors at the top left of the board would get very hot as well.... luckily this has not happened yet. When testing it is useful to have a current controlled power supply, or a resistor in line with the b+ to limit the current to a low level..... stop from blowing fets for no reason when fiddling.

Be warned, this is not as easy as it looks, the thin lands on the board makes for difficult removal of the components... without pulling off the tracks.

Best of luck with it, been a long time since I fixed this style of board. As Bruce says, Curtis use off the shelf components, so blind replacement of all ic's and diodes is a possible answer to fix the thing for only a few dollars in parts ... probably $3.00 plus FETS, (except the PMC-98500 PWM chip... it is a curtis special, but I have never seen one blown).

The problem is getting the parts off the board sucessfully without ripping the spiderweb tracks.... and the dip they moisture-proof the board with is a pest as well.

I find it odd that the thing runs full bore, usually with the freewheel fets not turning on, it runs slowly.... Hope you have a scope handy.... almost necessary for this enterprise.

..........oztules

[oztules]

One more thing,

The apparent burn marks on the bottom are where the 100r resistors are in series with the outer pot lines.one is grounded... if 24v was flashed onto the other end of this resistor, it may go high resistance so check the 100r for integrity.

Also, some of this voltage may have gotten into the lm339 (bottom left) and given it some problems with one of it's  input op amps.

This may help explain why you have direction control, but no speed control.

This may be difficult to determine, and will require tracing the op amp's output with changes to the pot position.

If it doesn't function properly, it will cost about 30cents....but be a dog to replace.

........oztules

[wiredwrong]

Oztules, you are amazing, how did you know the numbers on all these things from the pics I posted? Do you have the schematic? if so can I have a copy?

  I'm still trying to find a replacement for the MPSA06, I hope to source one from the mountains of stuff laying around. That way I can practice the soldering skills a bit before I attempt to fix the Curtis. Your warning about pulling off the tracks, has me a little bit worried about attempting this. After my neck surgery 3 years ago I have almost no dexterity left in my hands and I don't think that I'll be steady enough to do  it, but that won't stop me from trying, and if I give it a go early on a good day I may be able to.

The apparent burn marks on the bottom, I think may just be corrosion, the board had a lot on it when I first opened the case, but most came off with a soft brush.

[oztules]

Sorry RandyOKC,

I don't have schematic for these things, but fixed a fair few of them 10 years ago before we built our own version.

So I was familiar with parts of the board where they usually fail. The MPSA06 failed regularly, as did the totem pole transistors beside it(and the fets, zeners not so much, and the odd lm339 (the bottom right one) I have not come across the symptoms you describe.... mostly they just fail completely and so are pretty basic to repair.

Usually the fets have a few failures among them. (the first one to blow short, protects the others) Just test the   continuity between the drain and source of them. But usually they preclude the start relay and direction relays from operating.

It is possible that if the drive fets are shorted, it would give full speed both directions. The instant motor stop you are experiencing is due to the relay dropping out fast and shorting the plus and minus leads of the motor like shorting a windmill. Usually it waits till it sees the motor stop before releasing the relay.

It makes more sense that the marks on the back are not burn marks, as it is a reasonably high impedance part of the circuit so seemed odd that you could flash the 100r resistors. ( but it is in the drive line for speed control and direction)

You can test for continuity between the collector and emitter of the MPSA06 before you pull it out, but it must be toast to punish the 470r they way you suggest....I can't recall another way to cause it to act like this (apart from the drive fets and flywheel fets both shorted, but that is much more spectacular...)

Good luck bushranging with it, however a scope is almost a must... unless you just keep replacing the semi's till it works. (this I expect is near impossible without going spare trying to get the ic's out without board damage.... but deperate times.... desperate measures I guess.

The lm339 on the bottom right is involved in the current regulation of the motor (and drive pulse width). Upon thinking about it, it is possible that if that chip is damaged, it could also give odd results.

From memory, the paths are convoluted and interdependent between the motor current reg, sync pulses and deadtime and pulse width. Never really unravelled it, but could give misleading symptom /cure problems. It may use current sense (across the drive fet rds on, synced with pulse drive) to control the relay fw/rev characteristics. (are we stopped yet?) The interrelation between functions is still not clear to me as I never followed it up. Maybe someone else here knows more about this.

Fixing one when it is in front of me is fairly straight forward..... doing it from half a world away and no scope etc is difficult, but I will try and be as helpful as I can.... which may not be much.

I may be more an animal than an expert on these things.... but I have never failed to fix them. ( but the first few took a fair while.)

.........oztules

[ghurd]

This won't help get the replacement part out, but about removing the faulty TO-92 without damaged tracks.

Squish it with pliars.  Remove 1 leg at a time.

Use a solder sucker to get the holes open.

Hope that helps.

G-

[wiredwrong]

   Thanks, what you say helps me a lot, I really believe I can fix this hopefully without a scope, I totally believe that MPSA06 and the pot are the only 2 things that are damaged, I was almost sure the hot wire had flashed the 470r and that was the next item in line, I forgot to bring home the fluke when I went to the shop yesterday. But Ill go get it right now.

  I'm going to test for continuity between the collector and emitter of the MPSA06 before I pull it out, but I don't know whether it should show continuity or not, my guess is no?

  I know your probably going "hey you should know these these things" but the only electronics knowledge I have is from working in Cable Tv, Alarms, and Telephone. This is way beyond anything I ever had to do at any of those jobs. I have my CompTia A+ cert in computers but this wasn't covered there either.

   I understand most all of what you are talking about and what I don't understand I will research and learn it. I would love to say I was able to fix this and with the help of the people here I know I can, as soon as this project gets finished I'm going to start on the genny to charge it and I'm sure I'll have some questions about that too.

   Ill post progress as it happens, Thanks again to everyone who's chimed in..

[Bruce S]

RandyOKC;

   When you check these they're pretty much just like testing diodes. One way will read a very high resistance while switching the leads will show very low, but not (0). A dead short between the two means toast.

 You could set your meter on the diode setting and test it this way, it would be a better indicator.

Caution:

The fairchild ones are good, but I've seen datasheets show at one time the collector and emmitters where produced backwards from one year to the next , then back again.

While in circuit other items could give a false reading, due to gound or signaling loops.

Now worries about asking all these questions, you're doing just like we did when we began learning too:-)

BTW: I have a long time friend that is out in McAlister,OK. Big time Linux guru, my mentor in fact.

Do you ever get out that way?

He's the face behind mccooler.net :-)

Cheers

Bruce S

[wiredwrong]

Update:

I thought the MPSA06 was bad, as indicated by my Fluke

I kept trying to get my hands steady enough to remove and replace by removing components from other junk, but realized I would not be able to fix it so I went looking for a TV repair guy I once knew. He still has the shop, but just is not listed in the phone book anymore.

When I found him he agreed to help. We went to the workbench and he tested it and said it was good, handed back to me and indicated h didn't want to troubleshoot it...

So I started calling all the wheel chair repair shops asking if they repaired controllers. Most of them kept asking "what kinda Chair it was for?" and when I told them it was just the board, that it was not in a chair they had little more to offer other than "can't help ya" I eventually got referred to a guy named "Joe", when I called him he seemed very knowledgeable and egger to help, and did not ever ask what chair it was for. I told him I had a Curtis 1208c that wasn't functioning correctly.

He said "I have pretty good sized bone yard come on down, we'll match it up." I asked how much he wanted for on he quoted me $175, I talked to him a few more minutes and he then said that I could have a good used one for $75, and my old board. This sounded like a pretty good deal so I printed out some pics of the bike, got in the car and drove out there.

When I got there I found Joe in the back workshop, he was sitting around a small table with 2 other guys playing cards and drinking coffee, he looked at the controller in my hand, then picked up a box off the floor, and pulled out a Curtis 1208a, said that should work just fine, wrote the word bad on a piece of paper and taped it to the controller I had, tossed it into another box, then he asked what I was building.

I pulled out the pictures I had printed and passed them around, we talked for a long time about the project and after about an hour I reached in my pocket and said let me pay you before I forget, he told me to bring it home, make sure it will work and if it dose come back sometime in the future, and pay him then.

We also talked about battery's and he offered to save me some that still have life in them just bring him my dead ones for for the scrap yard, no need to send some with a little life, if there are completely dead ones lying around.

[DamonHD]

Nice to run across the 'good people' in life.

I'm fortunate to know a few, including one (in the Internet biz) who's also one of the smartest and best-connected guys, though you'd never know it to see him.

Rgds

Damon

PS. Probably he reckoned your hour's chat was pretty much worth the $75!  Most people don't make an effort to do interesting things...