EEVblog Electronics Community Forum
Electronics => Repair => Topic started by: murta79 on September 15, 2022, 09:15:01 pm
-
Hello,
I got this Fluke 8845A that I am troubleshooting, from the scrap bin :)
The symptoms thus far is as follow. When plugged in the first time the screen was showing "OVERLOAD" and the transformer got very hot, to the point that streams of smoke came out and I quickly shut off the power. My first thought was that the voltage selector was in the wrong position, but it is in 240V which is correct here in Sweden.
I have done some basic troubleshooting:
- I disconnected the power-plug that goes to the FPGA-side of the board and measured the voltages on the "communication/VFD-display-side", these voltages looks OK (...and also no smoke when I only have this side connected).
- I then tried and connect only 1 of the secondary voltages on the FPGA-side at a time. I found that if I leave +/-20 volt disconnected (violet cables) the instrument stops showing "OVERLOAD" (The transformer still gets hot but at least no smoke this time either).
- Voltages on the secondary side of the transformer:
Violet_1 = 16,4VAC
Red_1 = 13,9VAC
Violet/White = 0V
Violet_2 = 12,3VAC
Red_2 = 12,6VAC
Measuring resistance (on PCB) between +/-20V->0V gives Mohm readings and between +/-15V->0V gives 50-200Kohm readings (fluctuating).
Grey_1 = 3,7VAC
Grey/White = 0V
Grey_2 = 3,7VAC
This is as far as I have come. I was hoping someone here has some good ideas on how to proceed in my troubleshooting. Regarding the transformer, I´m pretty sure some of the windings on the +/-20 volt has shorted out, probably because of a faulty component on that rail, maybe also at the same time shorted windings on the +/-15V ?
The question is, is it necessary for me to buy a new transformer to continue the troubleshooting or can I use the one I got until I figure out what other components are faulty on the board and then buy a new transformer?
-
..Forgot to mention, I did look trough a bunch of other posts on this forum regarding this multimeter but I could not find any with similar problem symptoms as mine. If anyone knows of any posts that can help me (on this forum or other), please notify me. :D
-
Check the power supply caps and diodes first. Something is shorting.
-
add some load (resistor ~1k/incandescent bulb ~50W) in series with multimeter cord to save transformer from destruction
observe voltage drop
-
First, the 'OVERLOAD' warning is very common when the unit is turned on with nothing connected. It is just the result of noise and autoranging, a quirk to be sure but one that should be easily resolved by shorting the inputs or selecting DCI (current) immediately upon turning it on. Not having power to the guarded power supply also might cause the 'OVERLOAD' to go away, but for now I would just not worry about it.
If the transformer is smoking, can you measure the amount of current the unit is drawing? And if that is excessive, what happened to the fuse? Does the unit appear to have been worked on or is there any physical damage?
Your transformer has not likely been improved by smoking, but if the current draw is normal (low) when the secondaries are disconnected, then it may be OK enough for now. Possibly the unit was not set for 240V, connected to 240 volts long enough to overheat or overvoltage and damage a diode, regulator or other component in the guarded power supply and then finally set to 240 volts, but too late and without replacing the fuse--which is 1/4A slo-blo for 120V but only 1/8A slo-blo for 240V. Check which you have...if you see a 1/4A fuse in there, this is the likely scenario.
If it sort of works but draws too much current, you'll have to map out the guarded power supply because AFAIK there are no schematics available. I've worked on these a bit, but don't remember the exact setup for the guarded PSU. Perhaps post some good photos?
-
Is it possible to verify that your regulated voltages are correct and stable? 5V,-5V,15V,3.3V,1.5V,55V,20V-20V,7Vref
If not, check the ohms between ground and regulated voltages.
-
add some load (resistor ~1k/incandescent bulb ~50W) in series with multimeter cord to save transformer from destruction
observe voltage drop
This is an absolute MUST. NEVER take unknown equipment and plug it unprotected into the mains, it's quite likely you will make matters worse. Always do a resistance measurement of the mains side before plugging in and ascertain you have the right resistance for expected power consumption - that will dismiss absolute shorts in the winding, plus check there is high resistance to earth.
Next, measure the resistance across the secondary side windings and check its proportional to expect voltage per winding if you can. If it is low, could be bad winding or bad diodes, cap, or regulator. if it is open, could be bad winding...
Also check the resistance across the rails as delivered from the power supply with it disconnected from what it's supplying.. if these are short, you need to get rid of the short before you proceed. Also, this is a good time to check the resistances of what it is supplying in isolation, as a short could be there as well.
Basically powering it up usually comes last before taking a whole load of measurements to reduce the likelihood of what you experienced. This also assists in your understanding of what the power supply is doing in a safer way. Plus you can use this as an opportunity to do ESR measurement of caps to see if any have gone bad. I usually spend a good 30 minutes checking power supplies in isolation before any power gets applied, it's time very well spent.
Now you might be lucky and have got away with this, but it certainly reduced the life of the transformer if that was smoking.
-
Now you might be lucky and have got away with this, but it certainly reduced the life of the transformer if that was smoking.
A replacement transformer will be needed IMO at some point because inguard isolation depends on its integrity and being smoked could affect that. If it is still producing the correct voltages, I'd wait on replacing it until the other issues are dealt with. Testing to make sure that all the inguard, outguard and primary windings are at least nominally isolated is probably enough to proceed for now.
-
I would expect the primary-side fuse to be sized to protect the transformer. Was the fuse correct for the mains voltage?
-
I would expect the primary-side fuse to be sized to protect the transformer. Was the fuse correct for the mains voltage?
nails are not marked with mains voltage ::)
-
The main purpose of the primay fuse is to prevent a fire after the transformer is blown or when used with the wrong mains voltage setting.
The normal cheap fuses are not very accurate: they may blow just above the nomunal current, but could also last long at twice nominal. The tolerace is too large to really protect the transformer without-over sizing the transformer and use a relatively small fuse. It may help with the wrong mains setting, but subtile over load, like just 1 winding seeing a near short (e.g. Ta capacitor causing a short) can over load it. With frequent hot - cold cycled the transformer should also be the primary source of failure, like getting some windings shorted out.
-
fast /slow /delayed
and cheap made out of ill-proper materials like aluminum
-
Answering and closing my old thread...
I was cleaning out some storage at home and found this old broken instrument. I decided to have one last look at it and see if I could fix it before scrapping it.
This time I decided to take out the board from the casing since I was going to throw it away anyway. On the underside I discovered a burnt component, a rectifier for the +/- 20 volt supply rail. After a closer inspection it turned out not to be the rectifier itself that was burnt but the de-coupling capacitor before the rectifier. The capacitor was not only burnt, it was completely missing, as in, it had basically exploded and had left metal residues on the board and pads. I \$\Omega\$-measured across where the capacitor had been and it showed ~500 \$\Omega\$. I cleaned the board with acetone as much as I could and ohm-measured again, this time it read >5k \$\Omega\$. I figured that some of the capacitor residues must have left a low- \$\Omega\$path between the pads and this was why the transformer had started smoking, drawing too much current on the +/-20V rail (I also de-soldered the rectifier and measured it, it was OK). After that I re-soldered the rectifier back in place and powered on the unit, ready to pull the plug. To my surprise this time the smoke did not come from the transformer but from the underside of the board (I had to flip it right side up when powering it on). I immediately turned the power off and after that ohm-measured across where the capacitor had been situated. This time it showed almost a dead short ~6 \$\Omega\$. The PCB between the capacitor pads had charred quite a bit so I started to carve off the sot. When I have carved off and cleaned I ohm-measured again and to my surprise the measurement had climbed to ~100 \$\Omega\$. I repeated the procedure, carving out more and more of the charred PCB between the pad and every time the resistance kept climbing. I was basically scratching out charcoal until I hit the ground plane ~1mm down in the PCB, after cleaning with acetone this time the measurement showed infinite resistance. I again powered on the board and this time... no smoke! And the instrument seems to work just fine! :) :) :)
Now the question is... Why had the ceramic capacitor exploded? Was it just component failure?
-
Did you take any pictures of the damage & repair? I'm not familiar with this Fluke DMM, but the HP equivalent is know for surface mount tants burning up, multilayer ceramic capacitors aren't much better, I've had them burn up in older things at work, they tend to crack & short internally.
David
-
Capacitor probably leaked and the board soaked up the dielectrics and got conductive. Had that happen a few times. Also, burnt board can be conductive, too.
-
@Murta79
Even if the meter still work, the x-former has heated and fumed, it is a future failure .... i would order one
because layers of varnish / insulation may have heated up and efficiency has reduced
Acting like this is kinda dumb no offence
-
@drhex
It was a small ceramic capacitor so should not contain any dielectric. My bet is also on burnt board becoming conductive.
-
@coromonadalix
Yes, while this is true.. This transformer is not something you just order from distributor. I might be lucky to find a cheap broken instrument on ebay which can act as a donor, I´ll keep looking.
Another possible solution would be to order a standard transformer just for the +/-20V rail.
-
MLCC can fail short from cracks caused be mechnical stress. Tantalum capacitor may also fail with a short, but as an additional danger a power full supply (e.g. > 2 A) could heat the capacitor so much that it ignites internally and produces some really intense heat / flame to cause more damage and smoke.
That extra fire is specific to tantalum capacitors, especially the old style drop form factor.
-
@Murta dont forget the same one Tektronix branded dmm 4050 if i recall ....
ask Tek if they have this transformer available, you never know ....
a thread about it
https://www.ittsb.eu/forum/index.php?topic=1532.0 (https://www.ittsb.eu/forum/index.php?topic=1532.0)
-
coromonadalix
Thanks! I will check this.
-
Here are some photos of the "crater" I carved out.
Btw, even though the transformer is not smoking any more it becomes really hot after 10-15 minutes. Not sure how hot it should get since I don´t have a non-faulty instrument to measure on, but I guess it should not be hot to touch :-\
-
Do you have anything to measure the temperature with? and/or a plug in mains power meter? both could be used to determine if the transformer has shorted turns.
David
-
Don´t have my power meter here at the moment, will get it tomorrow. But I just measured the temperature with a thermocouple, and it reads ~92degrees C after 5 minutes. And the even though the rate of the rise gets slower it was still going up at 5 minutes.
-
That seems high to me.
David
-
Did some more measurements...
Disconnected the 20V output from the transformer.. still goes above 95 degrees C after 5 minutes.
Disconnected the VFD-display / external interface power supply connector.. still goes above 95 degrees C after 5 minutes.
Disconnected the 12V output from the transformer.. still goes above 95 degrees C after 5 minutes.
Disconnected all outputs from the transformer.. still goes above 95 degrees C after 5 minutes
What could be causing this, short on the secondary side in the windings ?
-
Probably shorted turns on the winding that got overloaded by the shorted MLCC. I've got several vintage HP TE with transformers that have failed in similar ways, one had the primary winding burnt open (from a trapped wire).
Some AC voltage measurements of the secondary windings might reveal if the output is lower than expected. No load voltages should be higher anyway, if another forum member with this DMM could help with normal reading that would be very helpful. ;) I'm assuming it's too new to have a detailed service manual? :-//
Having said that, you definitely need a replacement transformer.
David
-
I would suggest that you change that bridge rectifier, and while it's out check the PCB area under it.
-
@Paceguy
I did de-solder the rectifier to get underneath and clean, at the same time I also checked that it was ok.
Since the transformer gets to hot even though I have no load connected to it I fear that some windings have shorted and that I would need a new transformer to be able to use the instrument :(
-
Begun the transformer "repair" has...