Upgrading the power supply capacitors in the Crack

[jrebman]

The vishay mkp-1832 caps are very nice, but unfortunately, due to the voltages in the crack, there is really no place for them -- definitely not in the power supply, and they are most likely not even rated high enough for use as bypass caps ion the output caps.  We discussed their use as output coupling cap bypasses some time ago in the earliest thread on cap rolling for the crack.  They're also fairly small.

-- Jim

 

[deltaunit]

Thought I should update this thread with my findings on the o-scope seeing as I brought it up earlier.

Well the results are in and well, there was no measurable amount ripple on the last cap - the stock design works perfectly!

N.b. The spikes with the regular interval in this image appear whether the scope is connect to anything or not at maximum sensitivity and so can be discounted. This was taken at the maximum sensitivity and shows a straight a line as can be.

Here is the second cap at a much lower sensitivity.

That doesn't mean any upgrade on the last cap won't reduce any residual noise further though, so I'll still be upgrading it!

 

[HF9]

I initially purchased a Solen 220uF film cap to replace the last capacitor, however it was just too large to fit without extremely creative mounting. I ended up bypassing with a Auricap 2.2uF I happened to have in the parts bin. I felt that there could have been a slight difference in the smoothness of the presentation, but it could have easily been placebo in this case. People like John Broskie recommend film bypasses in the power supply, so it can't be a bad thing

 

[deltaunit]

I think Doc should offer an separate blank case kit for just this requirement! Just a solid metal top and wooden sides ready to filled with big poly and oil caps.

 

[Grainger49]

Outboard power supplies have been around for quite a long time.  I built my first two Bottlehead kits with the power supplies in outboard boxes.  Having all AC three feet away reduces any chance of hum. 

That leaves room in the audio chassis for all sorts of goodies where the power supply was designed to be.

 

[Doc B]

I think Doc should offer an separate blank case kit for just this requirement! Just a solid metal top and wooden sides ready to filled with big poly and oil caps.

That's quite available right now. We can sell you a wood base kit and you could order a custom plate from Front Panel Express.

 

[MxT]

Hi Guys stupid question but i have to ask it. Lets say I want to bypass the last cap in the power supply. What value would you suggest? Must it match the 250v output of the original cap. as in I would sourche a 0.22uf 250v cap to do this? Or can the voltage be higher as in 400v 0r 600v ie 0.22uf 600v. And is there a recomended value for the UF value. as in 0.1uf etc. That being said IF i used a 250v cap could I aplly that on the output caps as well. ie I have the same make and model  0.22uf 250v cap bypassing the 2 output caps and the last power cap- or am i missing the boat here??

IF i am correct in my understanding then I would have two options availbe to me
Option A replace the two stock output caps with 2 ERSE 100.0μF 250v PulseX - Metallized Polypropylene Film Capacitors
and
BYPASS the last Power cap with a 1.0μF 250v PulseX - Metallized Polypropylene Film Capacitor

Option B
BYPAS the two stock output caps and the stock power cap with a 1.0μF 250v PulseX - Metallized Polypropylene Film Capacitor on each?



for the sake of interest should the above be viable could I then bypass the first two power caps as well with the same  1.0μF 250v PulseX - Metallized Polypropylene Film Capacitor? or should I push up the power caps bypas to something in the 2.2uf range since I read somewhere that a good rule of thumb is using a 1/100 value for bypassing?

 

[Laudanum]

Ok, I'll take a stab but just to get you started.  Im not an expert at this by any stretch so take it as food for thought, not gospel.  

There is no advantage in using a bypass cap of a higher voltage than required ( than the cap you are bypassing). In fact, if Im not mistaken, higher voltage caps may even be a disadvantage over using a cap with a lower (but still ample) voltage rating (related to higher inductance of the larger/higher V caps compared to the smaller lover V caps).
There is the possibility of audible resonance when bypassing a larger value cap with a small value (because of interaction with the inductance of the larger cap) . Not sure if it's really, normally a problem because bypassing seems to be the "norm".  Ive read it argued that it would only occur at frequencies well above the audible range and then arguments that it can indeed be audible.
The ratio between bypass cap and cap that is being bypassed is of much debate.  1/100, 1/10, 1/1000, multiple bypass caps ... seems like there is an argument for any and all of them.

Remember, this is all general stuff from my laymens grasp.  Im sure Ive missed a lot and I dont understand all the tech. aspects of it.  Others will jump in and correct me or elaborate, or both.

But here is a thread on the forum that discusses some of this to get you started ...  http://www.bottlehead.com/smf/index.php/topic,1723.0.html  

 

[Grainger49]

To piggy back on Laudanum's excellent post . . .  higher voltage caps often have higher ESR, equivalent series resistance, and inductance than a similar series cap of a lower voltage.  For instance, if you measure two Mundorf M-Caps of the same value and of different voltage ratings you will probably find a higher ESR and inductance from the one with the higher voltage rating. 

When a circuit is called upon to release the energy stored in the capacitor the series resistance slows that release.  So that is a reason not to select a higher voltage.  A reason to select a higher voltage is it is in your parts box and costs nothing. 

Parallel capacitors can cause resonance because of the capacitance, inductance and resistance that they contain.  That is just one reason to try first using jumpers.

That said I have never had a problem with a bypass cap that I could hear.  I haven't measured the circuit with a sweep to check.