I have a 512mb stick of BH5 (corsair branded) ram. Two of the SMD capacitors on the top of the stick broke off so I need to find out what size/type they are. I am able to re-solder some new caps easly enough. The capacitors were small SMD ones, with the middle bit brown coloured, with no codes on it. There are plenty more on the pcb if there is a way to test them for size.. I can get pics of the ram PCB/capacitors, if needed.
Hmmmm... I don't think I've ever even used a dmm with a cap meter. None of mine have them - heck none of the ones at work (a bunch of flukes, and a couple other random brands) have them either... Also I heard that cap meters are for some odd reason normally fairly inaccurate, especially when you try to test a component already in a circuit. But yeah - measuring it is probabaly the best way - but unless the crack is getting to me - I think he needs to desolder one lead from one of the caps he's measuring (well - with the size of those things it's probabaly easier to desolder both) - and find a cap meter.
Well I dont know of any devices that measure it, although when I was at college they had a plotter rigged up to show chrage/voltage etc.. I dont remeber any of the calcs to work it out They will be fairly hard to measure. The ram works without them, but gets errors over 150mhz on the two chips missing the caps - the others chips seem to be fine. My other two sticks can do 220mhz+ so its quite a large difference.. so Id really like to get them fixed. I might send a mail to corsair..
You could to a comparative messurement, you would need to desolder one more cap thou. What you do is you set up an R/C oscillator with the original cap and messure frequency, then you try values until you hit the right one. You could allso use a multimeter with capacitance messurment and screw the value it gives, just look for a cap that gives the same value on the display.
As the ram does still work - 150mhz 2-2-2-7 instead of 230mhz 2-2-2-7 I dont really want to desolder any more caps. Im not sure if I am even capable of re-soldering them however I have some old stuff to practice on. Also these capacitors are cheap, but min order is usually 100.. so I really need a guess of a few values so I can buy some, to compare against...
surely you can find a better distributor? on this side of the big pond - you can order them in much smaller quantities from digi-key, mouser, or newark. Surely there's a distributor over there that can handle small quantities.
wow you guys must still be in the dark ages. even radioshack's (the shitty end of american electronics) basic meter http://www.radioshack.com/product.a...y_name=CTLG_011_008_002_000&product_id=22-811 has a cap meter
I assumed www.rapidelelectronics.co.uk would sell the right type but it seems they do not I bought some smd resistors from them at 40p/100 so I assumed the price would be the same.. anyway, this pic has a few similar looking capacitors, can you tell me what type they are: The one marked c11x9 (near the top, directly above the U which is overlayed onto the pic) looks the most similar.
no- capacitors arent marked like resistors. unless you have the original schematic, the only way to tell is by measuring them. like i said before, you can get a DMM with a capacatince function, or if you have an oscope you can see how many mS it takes to discharge thrugh a resistor of a known value.
Are you looking in the right place? http://www.rapidelectronics.co.uk/rkmain.asp?PAGEID=80008&CTL_CAT_CODE=30151 Check the width & length, there are around 3 common sizes IIRC.
okay, if you say so, both of my craftsman (automotive) meters have it, my radioshack meter has it, and even the no name "PRO-50" meters that my school buys by the crate full for 10$ each have it. i dont know where you buy your meters, but next time you need one, youll probably end up with it.
Capacitance measurement is fairly common on cheap crap instruments, but since the values you get are so far off that you are better off rolling a dice... That's why expensive meters don't use that feature, because the cheap way of measuring caps don't work, you need an expensive meter to get good values. Some semi-expensive instruments have it too, but at least the one i have goes from 4nF to 40µF, and arn't those small ones down in pF? Anyway, i have 2 shitty instruments at $6 a piecce one descent instrument at about $35, one good one at $85 and one even better at $110 or so. The $85 one has cap measurement, but as usual, you need to stick the cap into 2 slots on the instrument, you can't use the probes. But for a comparative measurement any capacitance meter should be fine as long as you desolder the cap before measuring.
That would explain why none of my or my work's meters have it - cheap dmms are not worth all the extra hastle that bad readings can cause, so I always make sure to get higher end meters.
I was betting it would end up being 100n, the most common cap size in the world... Do your own search, but for a Samsung PC3200 module,
You think the lack of a decoupling cap could cause that much of a hit in performance? I'm not much for the high speed digital stuff - but that seems pretty surprising to me.
I foudn a pic, which is not the same PCB as my ram, but the same chips. It has the capacitors so I hope you can guess what they are for The capacitors missing are C1 and C55 which are in the top corner on both sides of the stick, at the same end. If you look in the pic ablve, the very top row of capactitors.. its the end ones which are repeated above each chip. Heres some more info from my sticks: Pcb: TOPSEARCH TS-M-8V03C Corsair 50-00107 Rev A1 Ram chips: Winbond 2382D W942508BH-5
Well there must be some reason why the capacitors are put on the pcb. I'm quite surprised that the circuit would work even at those speeds without decoupling. There must still be enough decoupling for the IC's even if the nearest capacitors are missing. For high speed circuits it's very crucial to have decoupling caps as close as possible to the IC power pins. When you think about it there must be quite huge currents needed to be able to drive the PCB traces and other circuitry to the correct logic levels at those speeds. The current surges are short though but there are many of them. Without the decoupling caps close to the IC power pins there is too much inductance in the PCB power traces (planes) to get the currents from the current source. There might have been mistakes in the above and I'm sure someone with better knowledge can explain it better. EDIT: And back to the original question. For decoupling something like 0.1uF or 0.047uF (100nF and 47nF) or something in that range could work. 0.1uF can be found from many circuits. Then you just need to figure out the capacitor physical size like 0603, 0402, etc.
