Need help diagnosing H-Bridge

[eKretz]

Hi all, I have a computer FFB steering wheel that was recently damaged by an overvoltage condition. The components are rated for 40V and I accidentally fired it up with 52V. Now, the motor drive for one direction isn't working. I thought I had blown one of the driver MOSFETs (Fairchild FDD4141 and FDD8447L) so I grabbed some from Mouser and replaced the ones I thought were bad, to no avail. I then replaced all of the MOSFETs and still one direction doesn't work. The circuit is using 540 Schottkys but when I checked them they all seemed to give continuity in one direction and blocked the other, so I think they're OK. I don't know where to start, can anyone suggest some steps to take in diagnosing this? I can see no visible damage anywhere on the board or components after careful inspection under a loupe.

 

[(*steve*)]

The mosfets were possibly a good guess, but it sounds like a wrong one :-(

My next guess would be the drivers. You need to have a look at the voltages on the gates of the mosfets. The ones that are not working probably have zero volts with respect to the source even when they're supposed to be turned on.

Trace these signals back to a chip or a transistor, etc, and see if the input of that is OK. Keep going back until you find a good input and no output.

It may help to post images of both sides of the board and tell us which mosfets are not switching.

 

[KrisBlueNZ]

What's a "computer FFB steering wheel"?

What's a "540" Schottky? Do you mean an MBRS540 or something?

What's the model number?

Do you have a link to user documentation, service manual, or schematics?

 

[eKretz]

A computer FFB steering wheel is a steering wheel used in computer simulation racing that has Force Feedback. The 540 Schottkys I believe are barrier diodes, I'm not sure what the proper name is, you can see them in the photos of the bridge below. There is no service manual or schematic available, and the user manual is a joke. I am somewhat new to the electronics world but I am hoping you guys can help me out with this, I know the basics and I learn fast. I can do hot air rework etc. I have already replaced all the MOSFETs on this board.

Here are the voltages I found by checking so far, it looks like everything is switching correctly, right? But voltage is not right. So would the controller be the next item to check? I'm having a hard time following the traces. The first pic below is the wheel trying to turn clockwise but nothing happens. The second pic it's trying to turn counter-clockwise (I have unplugged the motors so I could check the voltages; when plugged in they function correctly in the CCW direction). (Pictures moved to below).

 

[KrisBlueNZ]

Thanks for the clarifications. I'm afraid I don't think I'll be able to help. I know nothing about force feedback steering wheels. I recommend Steve's advice. Good luck!

 

[(*steve*)]

It's a complex board, possibly more than 2 layer, and with a solder mask that makes it really hard to see the tracks. All of these things are going to make things difficult.

Can you post a larger image as the images of the top and bottom are too small to make out the details. Preferably we need to be able to read the markings on all the components, but at the very least, on the semiconductors. Concentrating on the RHS of the board would be OK for now.

Also, where do the motors plug in? Or is there just the one motor?

It would be handy to know where each pin on the motor connector is connected to on those 4 mosfets.


Is the only fault that it won't turn the motor in one direction?

 

[eKretz]

Hmm tried to attach a larger version, they are 8mp photos originally but photobucket shrunk them.

Yes the only fault is that the motors won't turn in one direction. Everything else works fine, I even re-flashed the drive chip firmware and it worked fine. The motors plug in to the socket on the lower right that says MOTOR. I will check the pins and note where they go shortly.

 

[eKretz]

See below for better pics

 

[eKretz]

Here's a crop of the bridge showing the continuity of the motor pins to the MOSFETs and diodes:

On the larger photos: let's see if this is any better:

Attempting Clockwise rotation (dead motors):

Attempting Counter-clockwise rotation (motors work):

Back Side:

 

[(*steve*)]

Can you tell if a high current flows when you are attempting to rotate the motor in the direction it doesn't go?

Looking at the 2 diodes D1000 and D1003, when the motor is turning, one diode reads 0V and the other 24V. This indicates the appropriate mosfets are turned on.

When attempting to turn the motor in the other direction, the voltages should be 24 and 0, but you get 10.6 and 0.

If there is a heavy current flow, this would be indicative of something shorted.

However I think it is more likely that one of the mosfets isn't being turned on properly.

When the motor is working, the top 2 mosfets are on. It appears in both cases that Vgs is 6V

In the case when the motor isn't working, the bottom 2 mosfets should be on.

We're seeing only 1.8V and 0.4V Vgs. I suspect that there's a driver in common for the bottom pair of mosfets and it is damaged.

See if you can trace where the gate of these mosfets (T1002 and T1009) are connected to (they are isolated from each other, but it would not be unusual for them to go to the same driver chip). There may be a low value series resistor too.

Actually I note a diode and a resistor near each mosfet. These are possibly connected in parallel and connected between the gate and the driver.

I also note that there is not a pair of similar devices close by the mosfets. The driver may be the 8 pin or the 14 pin device.

The 8 pin device appears to be an LM358 -- I'd not suspect that at the moment.

The 14 pin chip is a quad NAND gate -- probably not that either.

That really only leaves the 4 transistors just below it. These *may* be used to switch the mosfets, and *may* be worth checking. -- I'd have a quick look at the voltages on their base, emitter, and collectors in the two cases and see if there's any clues there.

Maybe you can determine if they connect to the gates of the mosfets.

Also, I note that some of the capacitors are leaning over. Check the bottoms of them to make sure they're not swollen or leaking. It is unusual that they're not flush with the board.

 

[eKretz]

I'm fairly sure there's no current flow, nothing is getting warm anywhere. I will check out the items you mentioned and let you know what I find. As for the caps they looked that way from new, but I will take a closer look at the bottoms.

Thank you very much for the help Steve!

 

[(*steve*)]

Great, I was just trying to spot as much as I could. The solder mask makes following the traces pretty difficult.

But great photos though. Makes checking things so much easier.

It was only half way through my post that I discovered the size of the images -- that answered a lot of my questions as I could read everything.

I hope it allowed me to be reasonably specific.

 

[eKretz]

You have been a great help. The photos are made easier by a nice camera and a mom who's a photographer, so it was easy to learn good practice and ask for advice in the past when it was needed! =)

Upon closer examination, the caps look fine underneath and I can see no signs of any leakage or swelling. I did discover that my MM has a cap mode and some of the uF readings were a little high, some caps say 330uF but I was getting like 530uF, is that a problem? the 100uF caps were right on. Now for some bad news. As I was about to start checking BCE voltages I shorted a leg of one of the J3Ys and fried it. They are SO tiny! I wish I had a better set of smaller leads. So I guess I'm going to need to order up a few of those and the 2TYs and replace the borked one before I can get any useful measurements.

Again, thanks very much, your help is much appreciated!

To be continued!

On edit: I have only found a UK source to order the transistors, Mouser and DigiKey don't seem to have them. Do you know of any better U.S. sources?

 

[(*steve*)]

Your mum is giving you good advice.

Caps reading over their value is not a problem.

Eeek! for the SMD transistors. I'll look up 2YT for you...

It could be an S8550

 

[eKretz]

Looks like s8050 and s8550. Any opinions on picking them up from eBay? They are both available there for a decent price.

 

[(*steve*)]

Yep, that would be my guess. They're a complimentary pair.

You never quite know what you're going to end up with on eBay. Especially from a Chinese seller. But these are probably fairly non-critical, so if you're willing to take the risk, it should be fine.

Where are you (country, state)? There may be good mail order or other local options.

 

[eKretz]

I'm in Indiana, USA. It looks like breakdown on price is $.10 eBay and $.45 DigiKey. Are the DigiKey transistors better quality you think?

 

[(*steve*)]

In theory, Digikey will supply known genuine parts. The same can't be said for eBay.

There are other options, look at Mouser, for example.

 

[eKretz]

Ahh, I was searching for the 3-digit codes, I didn't realize I needed to look up the 5-digit code. Such a noob!

 

[eKretz]

Steve, is there a way to tell whether I need s8050L/s8550L or s8050H/s8550H? Mouser has different p/ns for them. Here's a link to the datasheets:

http://www.mouser.com/ds/2/258/MMSS8050(SOT-23)-31837.pdf

http://www.mouser.com/ds/2/258/MMSS8550(SOT-23)-12703.pdf