First off i want to say thanks to those who have taken the time to read this and help me out.

Okay so onto the problem, i had a precor 9.23 tread mill break on me and a friend gave me his so i decided to take mine apart.
It would randomly not work for weeks at a time then just start one day working and stop.
I took it apart and with a basic meter probed around, i got the motor to work with a 24V psu at greatly reduced speed but the motor runs just fine.
So i decided to short out the 2 mosfet's (d-s) on the driver just to see what would happen. and of course the motor ran full speed, then i unshorted them and the driver was able to work on its own, vary the speed with out any problems.
So i went to bed and woke up and again the driver would not work and it was not plugged in over night. So again i short out the mosfet's and turn it on and again the motor goes full speed. I then turned it off and unshorted them and then the controller was running just fine for the whole day.
Im not to sure what part would be bad. maybe im discharging a faulty cap or the mosfets are bad and somehow go back to working after being shorted for a bit.
Seeing as most the parts on the motor driver circuit are ranging around 5$ each i wanted to try to narrow it down besides replacing all mosfets, caps, rectifier and diodes.
From my testing it seems the mosfets are fine as they do work and stay working for long periods of time.
The caps are nice and flat
Im thinking maybe its the rectifier
Some details on the motor its a 90V 25A 3hp
Its a big motor the largest i have ever seen on a treadmill.
My goal is to get the original controller working and mod it to run on my mill or lathe. It also has a actuator that would be kinda nice to use to raise and lower the head on my round column mill but thats just me getting lazy.
Below are images of the board and 2 mosfet's and rectifier

The first time i shorted it out it made a nice good spark.

 

While circuit is powerd ON, shortening the FETs terminals is not a good practice. I think you need to heat the whole board with a heat gun. Defiantly you’ll do it carefully and would not burn the components at all. (I have taken advantage of this many times in my practice). It’ll ensure you that your board has no more dry joints and accumulated moisture. Because of extremely high input impedance, a little amount of charge can hold an FET for hours. Secondly, you should pay special attention to gate resistors used for biasing and other gate drive circuitry. Consider the potentiometer too. Have you checked the FETs using a good digital multimeter? Flat and nice looking capacitors are not always “good”.

 

Well i have a nice hot air station solder station combo, i also have a nice toaster i use in my garage to harden my powder coated parts might that be better?
I can understand why shorting the fets is a bad idea as it getting unregulated power into my motor.
MY meter is a bit dated the old days of the brown and white plastic units, but i will try to do what is described here.
http://forum.allaboutcircuits.com/showpost.php?p=30000&postcount=5
To check both caps.
Im going to draw up the circuit for the drive half then double check every thing again.
Im just not used to things coming back to life in such a repeatable manner.
The fact is every the on the board works 100% for a whole day after the fets have been shorted, even the fets work and vary the speed. So it threw me for a bit of a loop as to what would be holding it up.

 

I agree, that behavior is very odd and Im sure it tells a story. We just can't understand what it means.

Instead of bypassing - shorting - the MOSFET, I wonder if a high ohms resistor, say 10kΩ, might accomplish the same thing. It would cause a small waste of current but that might be a small price to pay if it allowed the board to work 100% of the time. Not exactly an elegant solution but cheap and easy to try. I suppose you could even put a low wattage light bulb across the MOSFET instead of your short.

I think I'd be reluctant to cook the whole board unless you can find bad solder joints with careful inspection. Rifle instead of shotgun.

 

I followed the link in your post. It describes the normal checkings. Did you notice that, in your case, you are fighting against something leaky. You might find every thing ok when checking with a simple multimeter. We can’t say that “Every thing is 100% ok but not working”. As from picture of your board, it is double sided print and surely has many vias. Mostly you can’t inspect a via even with a magnifying glass. I’m not asking to burn and roast your board (and “Cook” as said the above poster) by OVER HEATING it with a hot gun. Do it with open eyes and control. I have worked in a TV manufacturing plant for some years and honestly, trouble shooted many PCBs by carefully cooking and even re-soldering them through soldering machine, also 2 or 3 computer mother boards, but again with a lot of care.
Anyway, it was just an opinion and you might not dare to follow it. Just try it as a last remedy when you decide to throw it up.
Don’t ignore the moisture while working with FETs. Is there any feed back that is fed back to the driver circuitry?
Wish you Good Luck