Dell Optiplex SX260 Computer Won't Boot Up

Feb 14, 2013 24 Replies

That would require a lot of filtering to keep hum out of the audio and harmonics out of the RF. :(

Where is the fun in that? it's more fun to start with a blank page. ;-)

I want to run at least 100 KHz, but would prefer a higher frequency. The estimated average current is 10 mA from the datasheets for the tubes. A torritdial transformer will raditae less noise, be more efficient and the entorecirccuit should fit a deep drawn box the size ioof an Altoiods box, or one of the hundred or so NOS bathtub caps with bad seals I have in the shop. They are easy to open, with a propane or aceteleyne torch. Melt the solder around the edge, and find the soldered over vent hole. Shove a dental pic in, and pop it open. The lazy way is to solder up a box out of scrap PC boards. :)

Not sure exactly what that meter is telling you, but if you have high ohms (more specifically, high ESR) in parallel caps then they are ALL high. If those numbers are anything like the ESR then they are definitely U/S.

MORE UPDATE:

I failed to mention that there is no sound (beeps, etc.) when I turn on the power. Also, I removed the CPU to do some of the tests.

Using an external variable power supply (with current meter) connected across this bank of 2200UF 6.3V caps in parallel (with proper polarity), at 2 volts, less than 250ma. It looks as though this problem has nothing to do with a "shorted" cap/s.

Again, Thanks in advance, John

Nichicon goofed up the production of models HM(M) and HN(M) from about

2002-2004. The date codes are in the form LYYMM, where L is a letter, YY = last 2 digits of the year, and MM = week of the year. So H0324 means the year 2003, week #24. 4-5 of those capacitors are 1500uF, 6.3V, 10mm diameter and located around the DIMM sockets and AGP socket. Behind the CPU are about ten 1800uF, 6.3V, 8mm caps and another 3-4 1800uF, 16V, 10mm caps on the other side of the MOSFET heatsink. Those may be Rubycon brand (big "K" stamped in the tops) model MBZ or MCZ instead of Nichicon (big "+" stamped into tops). The Rubycons were never defective, but TheCapKing.com says the Pentium4 was so power hungry that even those caps can go bad in 5 years. Most caps in motherboards can't be tested with an ESR meter except by unsoldering them because so many are in parallel.

caps, which are more durable in heat. Here's some information about doing that:

quality caps.

pretty obvious fakes (Teapo brans substituted from Sanyo or Nichicon), and apparently only one Ebay dealer sells only genuine caps:

Hi Larry,

Thank you very much for this very useful information.

Fortunately I have other computers so I am in no rush to fix it.

John

Hi,

UPDATE:

Sure enough, I have found several electrolytic capacitors (tested out of circuit) which include every one of the eight 2200UF 6.3V (next to each other and connected in parallel). As for the brand on these bad (2200/6.3) caps, I can see what looks like "2DIVDK" and there is a "M" in a circle on another part of the cap.

There were no bulges in any cap, but a loss of capacitance such as measured via a digital cap tester showing around 1500uf for these 2200/6.3 caps. Often an increase in ESR can cause a lower cap reading on this tester. My ESR meter is not useful on large capacity caps.

I will have to order replacements, but since I have other good working computers, I am in no hurry.

Thanks again to everyone for your replies, John

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required