EEVblog Electronics Community Forum
Electronics => Repair => Topic started by: hammil on June 02, 2012, 03:54:44 pm
-
Hey everyone
I bought a PM 3315 125MHz DSO from eBay for £30 - on account of it having a 'power supply fault'.
So I powered the unit up, and no surprises it doesn't work. The LEDs were flashing intermittently and the character displays flashed 'E' (for error, presumably). However, when I turned it off, I did see a flash on the CRT, so something's working at least.
Opening up the unit, I saw that there were two power boards, one feeding into the other. The secondary one has four transformers in a line, presumably for the high voltage CRT supply. I believe the fault lies with the primary board. See attached pics.
I was certain the DC filter caps would have blown, but they seem fine to me. The 'dints' you can see there appear to be regular, and thus by design. The rectifier tests OK.
Now, the big transformer to the right was emitting a high pitched intermittent noise, fairly loud, although I could probably deal with it. Apparently this can show degradation, but it looks fine to me. Maybe a few hairline cracks in the laminating material, but this thing is 20+ years old.
Moving to the far left, the mains cap there is definitely busted. Huge cracks all over it, however it is just a noise filter cap, it shouldn't affect the operation of the scope too much. And it is of course self healing, there's no short across it.
That DIP to the left is the switch-mode controller - a TDA1060. I can find these on ebay, so if it is busted, I can replace it without too much trouble. And that's certainly a possibility - on the underside, around a couple of passive pins connected to the chip (the transistor output and ground, as it turns out) the solder mask is cracked and swollen. Thoughts?
Between the two transformers there's a load of passives. The underside of the board is blackened around there, however I have tested the components to the best of my abilities and found them to be working.
One problem I think, is that on the main output header, all rails of which appear to come from the transformer output, a lot of the pins are shorted to one another. Is it possible for a transformer to catastrophically short? Wouldn't that blow a fuse somewhere?
On the output header, left to right on the topside, the pins are shorted in groups of 4, 8, 3, 1.
The solder shorts on the underside suggest they should be grouped 2, 2, 1, 1, 1, 1, 1, 1, 2, 3, 1.
On the connector to the secondary board, the matching pins are grouped 2, 2, 1, 1, 1, 1, 1, 1, 6.
Of course, this could be a feature of it being tapped, in which case I need not worry.
If it turns out the transformer is busted, I have no idea where to get a replacement. Maybe I'll just have to jury-rig several with the same ratios. Ideally I'd replace it anyway, as that whine, while bearable, is somewhat distracting. Maybe I can find some (fireproof, non-conductive) noise insulation foam somewhere.
It has some markings on it: 4022 369 68571, then DY 565. The first three groups look like a Philips part number, judging from the service manual of the 3305 (by the way, the service manual for the 3315 would be a great help, but I can't seem to find it anywhere). DY 565 seems to be a part number, but I can't find any reference to it.
So that's where I'm at. I've taken some pics from the other boards as well, for your viewing pleasure - it really is a work of art inside.
My current equipment consists of a £5 pocket meter - AC, DC, resistance/continuity, capacitance, frequency, diode. Soon, however, I will get myself a Fluke 87.
-
no way with old electronics devices : check the capacitors
better use an esr meter ( lots of schematics and links to buy here : http://kripton2035.free.fr/esr-repository.html (http://kripton2035.free.fr/esr-repository.html)
caps that don't bulk are not always good...
also caps that bulk doen't always have bad esr ...
check also semiconductors with an ohmmeter
but you certainly have a cap in the secondary part that is bad
-
As a first step: thou shall check voltages. Check the various voltage rails while loaded, both voltage and ripple (it's fairly common to need a scope to repair a scope). Dried out caps will show up as excessive ripple, shorted rectifiers as a pure AC voltage and shorted tantalum caps/semiconductors across the rail as a way too low voltage. Is the 3305 service manual similar enough so you can identify the various power rails?
Stay away from the high voltage rails unless you have the proper equipment (high voltage probe).
-
I checked them unloaded - we've got various voltages, .5 - 30V, all AC of course. DC conversion must be done in the secondary board.
I haven't checked them loaded yet, the display is a little too unpredictable for my liking. But believe it is a loading problem - it looks like it works for a bit, then flashes the 'E's. There's even some screen flashes.
Either way, the busted mains cap is getting pretty toasty, especially when loaded, so there's definitely a lot of ESR and/or leakage. I'm going to desolder it for now, might replace it with one of the spares I've got to hand.
EDIT: Just checked them loaded. The 30V rail is now 34V, but all the others have dropped to 3V or practically 0. Problem there I think.
I think I'll invest in an ESR meter: http://www.ebay.co.uk/itm/ESR-Meter-Electrolytic-Capacitor-Meter-DC2-1x5-5-Blue-LCD-gray-cover-/140750932752 (http://www.ebay.co.uk/itm/ESR-Meter-Electrolytic-Capacitor-Meter-DC2-1x5-5-Blue-LCD-gray-cover-/140750932752)
It's a fairly common design. This one is really the only option for me unless I want to spend many moneys or wait a month for one to be sent from China - unless anyone knows of a better solution?
-
IMO measuring ripple and replacing suspect caps (it's not like they're that expensive) can often replace an ESR meter. If the power supply is stable and with low ripple, than the ESR of the filter caps will probably be acceptable. If the power rail looks suspect, putting a known good cap in parallel is a good way to check. If the situation improves, than it might be a bad cap. If it does not, something else may be going on.
It sounds to me like one of the rails is pulled down by something (shorted tant/semi?). If the voltages are fine with the load disconnected, then the problem is most likely not on the power supply board or the transformer. The 30 V rail going up is likely because the power supply increases its duty cycle in an attempt to maintain regulation on the shorted rail. The voltage of some of these rails may be derived from a single rail. Do any of the rails that measure very low measure low resistance with the power supply disconnected? Are you able to measure current (either by convenient series resistors or putting a DMM in series) on the various rails? Does the PM3305 service manual give current ratings for the rails?
-
You could be right. It looks like the 30V rail is shorted to mains earth - but that could just be the high-voltage transformers of course. Maybe that's why the two 30V rails appear shorted by the secondary board - they go to the same transformer chain.
No tants here, fortunately!
Measuring current would be tricky as the boards are connected together with a header-pin-cable setup. Plus my current DMM doesn't measure it. I might try when I get the 87, if I can't fix it by then.
Oh, it turns out that the other rails are 11V or 5V AC under load. *shrug* Me being stupid, probably.
Also - the main caps are outputting 170V with a 5V ripple - which is interesting considering there's 250VAC in. Maybe something to do with those power resistors there.
Looks like I'll have to check the secondary board.
-
if you are planning to buy a 87v ... The fluke 87v doesn't give you absolute best, the U1272A gives more bang-for-buck ..
-
Unfortunately I can't find the U1272A second-hand anywhere, and new it's over £200. I can, however, pick up a used 87-III for £100 :)
-
Unfortunately I can't find the U1272A second-hand anywhere, and new it's over £200. I can, however, pick up a used 87-III for £100 :)
You should know a 87-III have major flaws fluke didn't acknowledge and that capacitance range only goes up to 5uF
-
What flaws might these be? I can't find anything on google.
Regardless, I have found an 87V for £150 - however I have found a few other meters just marked '87', no series number. Presumably these are series 1 or 2? Are they any good?
-
What flaws might these be? I can't find anything on google.
Regardless, I have found an 87V for £150 - however I have found a few other meters just marked '87', no series number. Presumably these are series 1 or 2? Are they any good?
... there's the 87 thread that was revived not long ago. Flaws that concern safety
Once again, Fluke is not perfect, nobody is.