[eat]

Hi I was hoping some one would help me figure out this power supply problem

I cant find my documentation that I know I once had and most likley still do some where on this computer

I would like to install some fuses and some leds so i know what is going on

I seem to remember the leds connecting to the fuses on one end somehow but it was some time ago

I have uploaded some pictures here
http://www2.whidbey.com/ror/power%20supply.html

please have a look at the parts that blew and what I am thinking of doing about it

any comments about if there is a better way or what have you feel free to let me know

[eat]

Woke up in middle of the night 3 am
The only way that makes sense to me is
On the positive side after the fuse the resistor connects
then the positive side of the led then a wire to the negative side before the connector
Does this sound right ?
Well back to bed

[Al_The_Man]

It look like there is more than one unit, I would reverse engineer one and use it to repair the blown one, you need to draw out the schematic.
It is almost impossible to tell anything from the photo's except general layout.
Al.

[keebler303]

Its pretty simple. You overloaded the board and the trace (not sure which) from the bridge rectifier to the screw terminal blew out. If there was a sustained short it may have caused the trace to heat up and pop. In any case, fuses are necessary as you now know. Chances are that a fuse would have blown before the trace on the board. There was nothing else so the trace acted like the fuse. I would check to make sure that the load on the power supply is not larger than it is rated. If you are stalling your motor with a G320 you might be trying to pull 20A out of a supply that looks about like 10A or so. Short term (like acceleration) the components can handle the overload. When the overload continues, the components start to fail.

Matt

[eat]

Thanks the stalling was a mechanical problem with the gear reduction that has been fixed

I replaced all the rectifiers with 35 amp units
I have seen a guy run a full size milling machine on one of these power supply’s with the same motors

I can use the fuse module he sells but would have to build
a way to see inside the case and now it sits under the machine with just the plug in side visible

Truthfully I have no where else to put it
Currently I have to pull it out from the other side and open it if there is a problem

I have contacted the manufacturer to see if he would sell me some of his leds and resistors he has not replied yet
or if he would tell me where or what to get
I am reading the resistors as 23 Ohms
and am having trouble finding them they look like 1 watt ?

I am on page 8 of a huge electronics book but went back to read page 2 again

So it seems to me my best bet is to cannibalize a couple of his fuse modules of his and I have a couple the wires will be real short but I should be able to desolder the resistors
And leds and connect them with some wire

Should have taken electronics in collage
Instead of diesel heavy equipment mechanics

Well with the parts I have I should be able to put something together to indicate what’s going on from the side I can see

And I will start studying electronics in my spare time thank you for the replies

[bharbour]

Ordinary Radio Shack LED's are fine for this. From your pictures, the resistors look like 10K 1/2 watt. The resistor values are somewhat dependent on the power supply voltage. LED's will usually light up with 1 or 2 milliAmps of current and tolerate up to 20 milliAmps, so there is a 10:1 range to work over. The formula for the resistor is:

R = (PowerSupplVoltage - LEDVoltage) / LEDCurrent

PowerSupplyVoltage = what your power supply puts out
LEDVoltage = voltage across LED, 1.7V is usually a good value.
LEDCurrent = the amount of current through the LED. 10 milliAmp or 0.010 is a good number.

The wattage rating for the resistor is:
P = (LEDCurrent^2) x R

LED Current is the value you used in the formula above
R is the value you calculated

The power rating of the resistor must be larger than or equal to the value P, I usually get resistors 2X the calculated power.

For example, with a 48 Volt power supply and 10mA of LED current:

R = (48 - 1.7) / 0.010 = 4630 ohms, 4.7K is the next standard size.

P = (0.010 ^ 2) x 4700 = .470 Watts. 1/2 watt is usable.

If you want the LED to turn on when the fuse blows, connect the cathode of the LED to the end of the fuse that goes to the load, one end of the resistor to the annode of the LED and the other end of the resistor to the end of the fuse that the power supply is connected to.

If you want the LED to be on if the fuse is good, connect the cathode of the LED to the minus side of the supply. Connect the annode of the LED to one end of the resistor. Connect the other end of the resistor to the load side of the fuse.

Both of these explanations assume that the fuse is on the + power wire, which is generally where you want it.

To tell the annode from the cathode of an LED, hold it with the wires facing you and look at the plastic at the base of the LED. There should be a small flat side on the plastic edge on the cathode side. It is sort of D shaped, but the flat is not that pronounced.

Good Luck,
Bob

[eat]

thanks I think you saved me alot of trouble I will get this
figured out when I get home from work tonight

one question lets say I try experimenting with the
setup some

with this being on the power supply whats the best way to
discharge the stored power or the best way to be safe
before connecting the stuff if I have all ready ran it with another setup

normally I just wait a few hours but some times this is not convenient

how do you guys go about this thanks

[Al_The_Man]

Usually the quick way is to use a large wattage incandescent lamp, it possesses a very low resistance when cold that will rapidly dissipate stored DC and you can also see the decay visually.
Al.

[eat]

wow thanks so much for the information
I have already used the formula
once to figure out the resistors for the leds I have
As i live on a island and the radio shack here only has 2
of each led they carry and said it would take some time if I order from them

then I called them and they did not have the resistors
so since I have to order the resistors I am ordering some panel mount leds with wires attached

here is how new I am at this
the place I am ordering from had about 15 types of resistors and it took me a wile
to figure out I wanted to be on the carbon film resistor
ordering page

well soon as I use the formula again with the specs from the new leds I will know what resistors to get

thanks again for making it easy enough for me to understand
I will try to learn more about this stuff it seems a little more fun now
and thanks for the light suggestion sounds like a great idea
I will go make me a new tool with a light socket and some clips to add to my collection

[eat]

the new parts seem to work great the led lit up and was working for a wile after the new parts blew out
I believe the resistor and led parts and fuses will be fine
you can see in the pictures below when I tested the first one after I got it soldered
to make sure it worked before soldering the others

but the bad news seems to be that the lack of fusses was not the problem
soon as I flipped the breaker the 2 rectifiers in the same place as the ones before blew out
I have them circled in blue
I think they where overloaded
like matt said but it must be from something on the
toroid side of the rectifiers

all i did when i got these was connedct the wires per instructions to run as 240 that will be one of the first places I check then I will check whatever elce I can get a look at
to see if I can see or smell out any problems

they both blew out right by where the green wires enter the rectifiers

i did run this for about 3 hours in the last few months
with no problems untill the sudden blow out the first time

if I cant find anything I may test them seprately
I have included some pictures if anyone would like some others of certian areas let me know

it looks a little funny as I had to put it part way under for the cord to reach

[eat]

I have not been able to find ant thing wrong so far
the power coming in checks out

the onley thing the 2 rectifiers have in common is the ground and both of them have the green and blue wire reversed
from the other 2 that is how they have always been

any one have a idea of what I should be looking for if I test the blue and green wire on one of the parts that blew

AN15452 1500 watt
this is the part number on the 2 units I will try to track down some more info I guess I will try to run some tests now

[eat]

well I disconnected one of the blown rectifiers and connected a voltage tester

thinking that would give me a good reading of what it was getting when it blew

turned on the power and it blew the other 2 rectifiers at the blue wires

so i figured that was about all the work on it I could
afford for now

and decided to call it a night

[nnpepe]

Looks to me like the resistors are 10K ohm 1/2 W. Could be wrong going just from the picture (Not sure what the voltage is) LED's drop about 1.4V @ about 14-17ma in order to be seen. The way it works is that the fuse "shorts" the resistor when its good (current takes the path of least resistance) and you get no LED light (fuse good, no current flow through resistor), when the fuse blows the path of current is through the resistor to the LED (remember that LED's are polarized). LED lights up to indicate that fuse is blown. Looks like maybe you had catastrophic failure (no snubber action to bring current up slowly) but maybe I missed the bottom of the board to see if you blew out the tracks, everything has limitations.

[eat]

yes it seems to keep blowing out tracks
I am wondering if there is just to mutch of a rush of current
for the boards to handle I believe there is somthing you can do about this but I cant remember where I read it

the manufacturer said to send them back and he had new higher capacity rectifiers he could fit

I hate to send them clear across the united states again
but I may

when I put in the sub panel I believe I used #2 wire
and I used the biggest cord to connect it to the 220 plug
anyone know if it could just be burning out the boards
from the inrush of power to fill the capacitors

I had nothing plugged in for the power to go out to the last time I turned it on and that most likely did not help
but it seems to me you should be able to plug in your power supply with nothing else without it blowing something out if it is working right

[eat]

I was reading this thread and sounds like the same problem
http://www.cnczone.com/forums/showthread.php?t=29290

I believe I will remove and check the toroids for shorts and possibly remount them
but it also sounds like a good idea to use something like this

http://www.specsensors.com/scl.asp

digikey has some of these surge current limiters

to bad I did not understand a lot of the thread

but for a problem to just start up like that
it must be the same thing a short under one of the toroids
I don't think a inrush of current just starts after a few months
I still cant understand how they both seem to blow parts at the same time always
unless it sends a spike back into the line then across the lug I have them tied into for power

[bharbour]

It is possible that it is an inrush problem, but I kind of doubt it.

Can you disconnect the transformer from all of the boards and measure the AC voltage going into the rectifier/filter boards? Also, measure from the transformer outputs to the chasis ground. To figure out what the AC voltage should be, take your expected DC voltage and divide it by 1.4. For example, a 50VDC output supply should show:

50/1.4 = 35VAC or so.

If it is within a few volts of this, it should be OK. If it is 2X the calculated value, check the transformer input wiring.

Measuring between the transformer leads and chasis ground should show 0 VAC with all the boards disconnected from the transformer. If it shows a voltage, it means that the transformer is probably shorted internally or you still have something connected to the transformer.

If the transformer is not set up right for the 240 V input, you can get twice the voltage that you are expecting on the output which can cause the filter caps to short out. This will pop rectifiers.

In general, fuses are much too slow to protect electronics. Unless the overcurrent is much much larger than the fuse rating, it can take seconds to pop a fuse. It only takes milliseconds to pop rectifiers and transisitors. Fuses are still useful to protect things like wiring and board traces and prevent fires.

The heavy wire that you put in on the AC power side could make an inrush current problem worse, but it would have to be pretty close to a problem without the heavy wire.

If your power supply filter caps are rated for the unloaded power supply voltage, (and they should be), there should be no problem with powering up the power supply with no load. I do it all the time.

If both the power supply boards blow diodes at the same time, it makes me thing that their outputs may be connected together incorrectly.

Good Luck,
BobH

[eat]

you can see in the picture what happened I have the unit mounted on a shelf under the machine and I believe the vibration caused it to wear threw the rubber

the other link I read gave some good information on how to fix this I will follow those instructions and possibly find a different place to put this

Looking back is easy but I should have checked everything I could with the power off first and ran what tests I could without power to it first would have saved me over 100 bucks well hopefully I learned something again
thanks for everyones help