geekhack

geekhack Community => Keyboards => Topic started by: dante on Sat, 02 November 2013, 20:52:58

Title: IBM Model M2 capacitors replacement
Post by: dante on Sat, 02 November 2013, 20:52:58
From this archived thread: http://geekhack.org/index.php?topic=5065.0

The OP specs these replacement caps:

2.2uF 63v electrolytic
47uF 16v tantal

The thread is a few years old - and before I take a soldering iron to my M2 I just want to make sure nobody has come up with better/alternative capacitors for this.
Title: Re: IBM Model M2 caps replacement
Post by: terran5992 on Sat, 02 November 2013, 20:56:47
Unicomp sells a few types of BS caps
Title: Re: IBM Model M2 caps replacement
Post by: dante on Sat, 02 November 2013, 20:58:51
Unicomp sells a few types of BS caps

Sorry - I meant capacitors.  The M2 is notorious for a couple capacitors dying.
Title: Re: IBM Model M2 capacitors replacement
Post by: tp4tissue on Sat, 02 November 2013, 21:42:51
ooo... I'd recommend Panasonic FR.....  (http://www.cute-factor.com/images/smilies/onion/4e5dd44a.gif)
Title: Re: IBM Model M2 capacitors replacement
Post by: kolonelkadat on Sat, 02 November 2013, 21:50:23
I wouldnt use tantalums. That they typically fail as short circuits kinda bothers me.

if it were me id use some electrolytics from nippon chemi con.

just make sure the voltage  rating is equal to or higher than the ones youre replacing.
Title: Re: IBM Model M2 capacitors replacement
Post by: mich on Sun, 03 November 2013, 14:59:53
SMD ceramic caps work great and are less prone to failures than electrolytics.

http://geekhack.org/index.php?topic=5065.msg376771#msg376771
Title: Re: IBM Model M2 capacitors replacement
Post by: dante on Sun, 03 November 2013, 15:07:26
SMD ceramic caps work great and are less prone to failures than electrolytics.

http://geekhack.org/index.php?topic=5065.msg376771#msg376771

Does anyone else want to get into the cage with this guy?  Otherwise ceramic capacitors it is.

As I understand it dis/reassembly is a pain in the ass thus I prefer this be a one shot deal and don't mind spending an extra .50 or whatever to get the best quality capacitors possible.
Title: Re: IBM Model M2 capacitors replacement
Post by: 0100010 on Sun, 03 November 2013, 21:08:53
+1 for ceramic vs electrolytic, for tolerance and longivity.
Title: Re: IBM Model M2 capacitors replacement
Post by: tp4tissue on Sun, 03 November 2013, 23:21:41
+1 for ceramic vs electrolytic, for tolerance and longivity.

Even if it's true that ceramics are better... There's a certain  charm and mysticism to electrolytics' that makes electronics containing them soul-ful...  (http://www.cute-factor.com/images/smilies/onion/023.gif)


same thing with Modern cars.. they're absolutely superior in in terms of technological improvements... but when you see an old Lincoln town car... there's a special feeling....
Title: Re: IBM Model M2 capacitors replacement
Post by: larrymoencurly on Sun, 03 November 2013, 23:42:37
Is the 2.2uF capacitor just for bypass, or is it a timing capacitor for a power-on reset signal?  If the latter, there should be a resistor in series with it, probably going to the +5V power for the keyboard, and I wouldn't change its type or value, but it shouldn't hurt to replace a 63V part with one rated anywhere from 16V - 50V.   

I don't see how the 47uF can be critical, unless the circuit isn't very stable and can oscillate, in which case it may help to add a 1uF tantalum or 0.1uF ceramic across it.  However sometimes adding a ceramic capacitor can make oscillation worse.  Also tantalums can be really sensitive about excessive voltage, so you may want to stick with those rated for at least 25V - 35V.
Title: Re: IBM Model M2 capacitors replacement
Post by: tp4tissue on Mon, 04 November 2013, 05:43:46
Is the 2.2uF capacitor just for bypass, or is it a timing capacitor for a power-on reset signal?  If the latter, there should be a resistor in series with it, probably going to the +5V power for the keyboard, and I wouldn't change its type or value, but it shouldn't hurt to replace a 63V part with one rated anywhere from 16V - 50V.   

I don't see how the 47uF can be critical, unless the circuit isn't very stable and can oscillate, in which case it may help to add a 1uF tantalum or 0.1uF ceramic across it.  However sometimes adding a ceramic capacitor can make oscillation worse.  Also tantalums can be really sensitive about excessive voltage, so you may want to stick with those rated for at least 25V - 35V.

2.2uf 63v is still a pretty big cap bro...
Title: Re: IBM Model M2 capacitors replacement
Post by: tipo33 on Mon, 04 November 2013, 12:08:33
This is a keyboard controler not a vintage amp, just solder in a capacitor that meets the requirements.
Title: Re: IBM Model M2 capacitors replacement
Post by: tp4tissue on Mon, 04 November 2013, 13:47:31
This is a keyboard controler not a vintage amp, just solder in a capacitor that meets the requirements.

of course I know that..... still...... electrolytics' are just m0re sexy as far as electronics go....
Title: Re: IBM Model M2 capacitors replacement
Post by: larrymoencurly on Tue, 05 November 2013, 04:10:09
Is the 2.2uF capacitor just for bypass, or is it a timing capacitor for a power-on reset signal?  If the latter, there should be a resistor in series with it, probably going to the +5V power for the keyboard, and I wouldn't change its type or value, but it shouldn't hurt to replace a 63V part with one rated anywhere from 16V - 50V.
2.2uf 63v is still a pretty big cap bro...
5mm diameter X 8mm length?
Title: Re: IBM Model M2 capacitors replacement
Post by: tp4tissue on Tue, 05 November 2013, 05:59:07
Is the 2.2uF capacitor just for bypass, or is it a timing capacitor for a power-on reset signal?  If the latter, there should be a resistor in series with it, probably going to the +5V power for the keyboard, and I wouldn't change its type or value, but it shouldn't hurt to replace a 63V part with one rated anywhere from 16V - 50V.
2.2uf 63v is still a pretty big cap bro...
5mm diameter X 8mm length?

yea.. that's big as far as "missing" goes...  if you miss a few ceramics, 0ohm resistors,, that's fine.. they don't often put the bigger ones in that could go without.. they represent significant cost in mass produced products.
Title: Re: IBM Model M2 capacitors replacement
Post by: dante on Fri, 08 November 2013, 19:37:30
Can any of you give me a hand ordering these?  Places are also asking for: package/case, tolerance, termination style(?wtf) I don't know about any of those things.  I just want some damn ceramic capacitors to solder in :/

Maybe if you have free time post some links?  I just want to buy a few of each.
Title: Re: IBM Model M2 capacitors replacement
Post by: tp4tissue on Fri, 08 November 2013, 20:22:02
Can any of you give me a hand ordering these?  Places are also asking for: package/case, tolerance, termination style(?wtf) I don't know about any of those things.  I just want some damn ceramic capacitors to solder in :/

Maybe if you have free time post some links?  I just want to buy a few of each.

through Hole.. that's all you need to know..

tolerance doesn't matter... this isn't the space shuttle.. but lower the number the better.. HOWEVER realize it's not a linear relationship

so 5% tolerance an a 2000v part is different compared to 5% tolerance on a 200v part.
Title: Re: IBM Model M2 capacitors replacement
Post by: rootwyrm on Fri, 08 November 2013, 21:19:05
Sigh.

I have many, many boards of much older vintage, electrolytic equipped, with no failures. Low quality capacitors fail, abused capacitors fail, defective formula capacitors fail. Keyboard controllers are not sensitive enough to generally care unless it's a catastrophic failure. And if it's catastrophic, well, yeah.

Upgraded Aluminum for 2.2uF 50V (http://www.digikey.com/product-detail/en/UWT1H2R2MCL2GB/493-2217-1-ND/590192)
Upgraded Aluminum for 47uF 16V (http://www.digikey.com/product-detail/en/UCL1C470MCL6GS/493-3916-1-ND/2300343)
I recommend double checking because I don't remember the exact diameters, but I'm fairly sure those are the correct diameters. These are direct swap. If you want to go overkill, same specs in Cornell-Dublier Alchip family.

As to the quality and reliability of Nichicon - that's who you find in million dollar servers alongside Kemet for tantalum. In fact, those exact parts I linked above. Spending more does not necessarily make something better - usually it just means 'it's more expensive for no good reason.'
Title: Re: IBM Model M2 capacitors replacement
Post by: tp4tissue on Fri, 08 November 2013, 21:31:53
Sigh.

I have many, many boards of much older vintage, electrolytic equipped, with no failures. Low quality capacitors fail, abused capacitors fail, defective formula capacitors fail. Keyboard controllers are not sensitive enough to generally care unless it's a catastrophic failure. And if it's catastrophic, well, yeah.

Upgraded Aluminum for 2.2uF 50V (http://www.digikey.com/product-detail/en/UWT1H2R2MCL2GB/493-2217-1-ND/590192)
Upgraded Aluminum for 47uF 16V (http://www.digikey.com/product-detail/en/UCL1C470MCL6GS/493-3916-1-ND/2300343)
I recommend double checking because I don't remember the exact diameters, but I'm fairly sure those are the correct diameters. These are direct swap. If you want to go overkill, same specs in Cornell-Dublier Alchip family.

As to the quality and reliability of Nichicon - that's who you find in million dollar servers alongside Kemet for tantalum. In fact, those exact parts I linked above. Spending more does not necessarily make something better - usually it just means 'it's more expensive for no good reason.'

you linked to smd.. I generally always get through hole, because "just in case" the size is diff, I could sorta wiggle it in some where  else..

I agree on not spending HUGE bux on caps..  I always just go for panasonic fr, they're not expensive and are supposed to last forever...
Title: Re: IBM Model M2 capacitors replacement
Post by: kolonelkadat on Fri, 08 November 2013, 23:22:26
as much as I hate to agree with tissue, I do agree that buying through hole "radial" electrolytics takes a lot of the guesswork out of ensuring you get a cap that will fit.

go here, select your capacitance and voltage, click apply filters, then pick whatever one you want from the list. Ive included panasonics and United chemi con.
http://www.mouser.com/Passive-Components/Capacitors/Aluminum-Electrolytic-Capacitors/Aluminum-Electrolytic-Capacitors-Leaded/_/N-75hqw?P=1z0z7l5Z1yzvjj0Z1z0zl7s&Ns=Pricing|0

or choose to go the smd route, and do some measuring and paying attention to package sizes and stuff. your call.
Title: Re: IBM Model M2 capacitors replacement
Post by: larrymoencurly on Sun, 10 November 2013, 21:22:43
Just buy general purpose radial, through-hole electrolytic capacitors of the same diameter, length, and pin spacing.  +- 20% tolerance is fine   DigiKey.com can be cheap for small orders because they charge so little for shipping under 8 oz.
Title: Re: IBM Model M2 capacitors replacement
Post by: tp4tissue on Sun, 10 November 2013, 21:32:50
as much as I hate to agree with tissue, I do agree that buying through hole "radial" electrolytics takes a lot of the guesswork out of ensuring you get a cap that will fit.

go here, select your capacitance and voltage, click apply filters, then pick whatever one you want from the list. Ive included panasonics and United chemi con.
http://www.mouser.com/Passive-Components/Capacitors/Aluminum-Electrolytic-Capacitors/Aluminum-Electrolytic-Capacitors-Leaded/_/N-75hqw?P=1z0z7l5Z1yzvjj0Z1z0zl7s&Ns=Pricing|0

or choose to go the smd route, and do some measuring and paying attention to package sizes and stuff. your call.

99% of the time.. people agree with me...

99% of the time.. I will not agree with anyone.. it is my philosophy to challenge in-the-box thinking.. even if what you protest would 'mostly" be true.

for example... I could just leave the fact that you agreed with me be.. OR I could disagree with the fact that you agree...

and say something along the lines of.. well.. even if you purchase smds, you can still solder leads to them to fulfill the flexibility of radial-through-holes..

While getting radial-through-holes from the get-go would eliminate all chances of getting an exact replacement part, fitting in the exact intended way..

I hope this is clear...(http://www.cute-factor.com/images/smilies/onion/8f337f1c.gif)