[larcenasb]

mcandmar, could you please elaborate -- I'm trying to be thorough before I set my mind on a judgment about the power supply caps. I've been reading older threads and getting a lot of conflicting info. When you say, "if anything it may be a disadvantage as you are putting more of a strain on the power supply having to charge up such a large reservoir, especially on startup," what exactly might happen to the power supply? Please explain.
 
Food for my jostling thought on the subject:
 
1.) Paul Joppa, of Bottlehead, said, 
 
"The output audio current flows through the power supply, mostly the last capacitor, as well as through the output coupling cap and the output triode. So the final power supply cap should, in theory, be as important as the output coupling cap with regard to the sound of the amp. This is a characteristic of cathode follower circuits, such a the Crack and the Foreplay." 
 
Later, he went on to say, after a member said he liked the difference using a higher uF cap,
 
"I post this with some frequency: 'I never met a power supply improvement that was not also a sonic improvement.'"
[from the Bottlehead Forum thread, 'Upgrading the power supply capacitors in the Crack,' page 1, replies #4 & #9]
 

2.) Grainger49 said, in response to a member's usage of 1x 1000uF and 2x 470uF PS caps,

"Wow, I like a lot of big capacitors! There are a number of engineers here and they would also suggest more power supply capacitance and a choke in the power supply filter."

[from the Bottlehead forum thread, 'A Crack Across the Ocean,' page 1, reply #7]

Concluding questions:

- Is it then in fact safe to use a 1000uF cap in the power supply of the crack?

- Apart from smoothing ripple and filtering hum, is it possible for a larger electrolytic in the PS to affect dynamics and power delivery?

 

Thanks and take care!

 

[ScottFW]

  I bought Cree Schottky diodes for the power supply ...

You have a board to mount those on? I'm pretty sure those only come in a TO-220 package, so would require some bending of the legs to mount them to the stock terminal strips without an adapter board (I used the one from... Parts Connexion I think it was?). Also, the mounting tabs on those are electrically live so take some care that they don't touch anything else under there.

 

[larcenasb]

Hi again ScottFW! Yes, I ordered from Parts Connexion too. I saw how others installed it on that board. I'll post pictures once mine is all set up!
 
P.S. About the heat concern, I was thinking of some way to cage the board and diodes, still thinking about a solution.

 

[Tom-s]

Mine are just mounted in free space on the stiff short wiring. I've never noted any heating issues.

 

[JamieMcC]

Rather than go for a big expensive 220uf film for the last power cap you can replace the last resistor with a $15 choke and use a less expensive 50uf -100uf film cap instead the smaller cap will be easier to fit in and The C-7X choke is probably the best value upgrade for the Crack the difference it makes is clearly audible. This would be my own prefered option and gives more bang for your buck its also a straight forward mod thats well documented on the BHF.
 
http://eu.mouser.com/ProductDetail/Triad-Magnetics/C-7X/?qs=%2fPiZ59IM4y3T0O8Zu9aP3g%3d%3d

 

[larcenasb]

Thanks JamieMcC, I'm trying to weigh the pros and cons before I decide on which route to take. How do you think the choke ($20) and a 100uF film cap ($30 for a Dayton) would fare against a 1000uF Nichicon KX electrolytic ($8)  bypassed with a 10uF Mundorf Evo aluminum + oil ($20).
Can you explain what the advantages and differences would be of each route? Thanks!

 

[larcenasb]

And thanks Tom-s, I'll not worry about the diodes!

 

[JamieMcC]

 

From my understanding and experiences I have found film is preferable sonically to electrolytic for anything the signal passes through.  Also I am pretty sure I have seen mentioned on the Bottlehead forum by one of the team that there is the potential when fitting an overly large value electrolytic a chance of damaging the diodes and over taxing the transformer.
 
As an exaggeration think of the transformer as a tap filling a bucket (capacitor) up when there is no load the tap has time to fill all the buckets fully as the last bucket (the third cap) is put under load and emptied its contents is topped up from the two dc smoothing reservoir capacitors before it and the transformer tops up the first.  With equal values the 220uf can do this easily but if you swap the last cap with higher values such as the 10000uf it’s going to put a strain on the components before it including the transformer supplying the.
 
Another option would be to go with the choke but keep the 220uf electrolytic and bypass it with a Russian Teflon or small film bypass, examples of both on the BHF.
 
Having done it a couple of times now fitting a choke and a standard mkp 50-100uf film cap is going to make a very noticeable improvement. It would be a must do mod for me if building a Crack again.

 

[Jeb Listens]

I agree with Jamie - I found replacing the last electrolytic with a film cap a very worthwhile exercise - perhaps slightly less impactful than doing the output caps but still very good, especially when it comes to producing a smoother more natural treble presentation. 
 
I used the choke and a 100uf/250v film cap.   Previously I tried bypassing the last electrolytic with a selection of different small film caps and found that they all changed the sound but not in any way that I liked.  For my preferences and gear, that was a backwards step. 
 
After I had replaced it altogether with a film, I then tried bypassing it and found that to be a positive step forward.  By that point I had accumulated some quite nice small caps so used a jupiter cap, which is probably overkill - but I really like the results. 
 
In my experience, bypassing caps really requires a bit of experimentation - you can do this carefully with some suitable clip leads before finding what works for you and soldering into place. 
 
I found the cree diodes to be a really nice upgrade too -especially for the money (~£10).  
 
After going through the modding process, one thing I would recommend is that if your plan involves drilling the top-plate multiple times,   gather all your components, decide on the layout then do all the drilling in one go.  It gets pretty tight in there. 

 

[Tom-s]

  And thanks Tom-s, I'll not worry about the diodes!

As you see in my picture. The only thing i did was tape the top plate with for insulation right under the diodes. Just in case they'd ever touch the top plate.
 
I'm currently reviewing my film cap options as wel. For now the 220uf Ansar looks like a great option costing about 60€ shipped 82mm long and 60mm diameter. Any other "small" and cheap options? (diameter is the limiting factor in my case)
Please remember that the last film cap is in series with the output capacitor. So unlike grausch suggested it actually does influence the capacitance seen by the headphone.
 
Edit: http://suppression-devices.com/ansar/index.html

 

[larcenasb]

Thanks for all the replies. And that's a great analogy JamieMcC, that helps a lot!
 
In the Bottlehead forum thread, 'Choking the Crack," Paul Joppa said,
"I would never use less power supply capacitance (last cap) than in the output cap, and I'd prefer at least twice that - if it were my amp."
 
But how come with the choke the capacitance can be lowered? And how low is okay? I saw Mikey on the Bottlehead forum use a 22uF Solen cap and Doc's green crack looks like it uses a 100uF Audyn cap. I'm learning a lot, thanks

 

[Tom-s]

http://www.duncanamps.com/psud2/

 

[JamieMcC]

Basically a choke is hundreds of times more efficient at reducing the ripple but it costs a thousand times more than a resistor so if a choke is used then last capacitor doesn't need to be so big to get the same end value results.
From memory as a example I think the stock crack power supply has a ripple of 1.5mv or 1500uv when using a choke the final capacitor only needs to be around 22uf to achieve the same result but here using a say a 100uf capacitor will reduce the ripple to 150uv so a massive reduction in ripple is noise floor which is perceived as a blacker background and increased resolution and improved dynamics such as bass impact. I am sure there is much more to it but essential that is my limited under standing. using my phone to reply has been exceedingly frustrating grrr

 

[larcenasb]

Basically a choke is hundreds of times more efficient at reducing the ripple but it costs a thousand times more than a resistor so if a choke is used then last capacitor doesn't need to be so big to get the same end value results.
From memory as a example I think the stock crack power supply has a ripple of 1.5mv or 1500uv when using a choke the final capacitor only needs to be around 22uf to achieve the same result but here using a say a 100uf capacitor will reduce the ripple to 150uv so a massive reduction in ripple is noise floor which is perceived as a blacker background and increased resolution and improved dynamics such as bass impact. I am sure there is much more to it but essential that is my limited under standing. using my phone to reply has been exceedingly frustrating grrr

Haha, well thanks so much for taking the time to help me out I think I will get the choke and three Dayton 100uF 250V caps then bypass the PS cap w/ a nicer film cap. I'll post pictures when I'm done! And thanks, Tom-s, I'll play with that software a bit to understand further!

 

[ClintonL]

Anyone know where to find tubes which will ship internationally to Australia?