Luxembourg Crack rebuilt

grausch

New member
Dear All,

I had detailed my previous Crack build here: http://bottlehead.com/smf/index.php?topic=6597.0, but since it was my first build I always felt that there were certain parts I could have done better. In order to redo those parts I was not happy with, I knew I needed to remove quite a lot of the components. Once I had gathered all the parts I needed, I stripped down the Crack until it was almost bare.

Here are some photos of the disassembled Crack after I redid some of the wiring.

QAlUKGH.jpg


ykB8lrV.jpg


bLEoH2t.jpg


33cI0Er.jpg


Almost all of the parts in the Crack have been upgraded over time. Full list of upgraded parts is as follows starting from the RCA jacks and working downwards:

[list type=decimal]
[*]REAN / Neutrik RCA Jacks - The original jacks were fine, but I love the second nut that really locks this jack in place. Bottlehead is now including similar RCA jacks, but if they did not I would definitely recommend this upgrade.


[*]Tube Depot terminal strips - I did not want to desolder all the terminal strips, so ordered some new ones from Tube Depot with some other parts. These ones look and feel better, and the upper terminal holes are larger, but the 6-terminal strip does not quite fit on the transformer mounting. This leads to a little bit of stress on the terminal strip itself, but I viewed is as acceptable. Not really an upgrade, but having extra space on 14U was nice. Downside to these is that with all the space on the top terminals you really need to ensure that you have a proper mechanical connection, otherwise you will have loose connections.


[*]Dale resistors - Except for the 10W resistors that will be removed for the Speedball, I upgraded all of the resistors to Dale resistors. Most of them have higher wattage, except for the 270ohm 5W resistors in the PS. The resistors all look nicer, but the higher wattage means that they don't fit as nicely into the terminal strips as can be evidenced from the bends in the resistor leads. The 270 ohm resistor across the transformer is also too short to fit, thus I needed to bridge it with some regular wire.


[*]18 awg cloth covered wire - First time I saw cloth covered wire, was in this thread http://bottlehead.com/smf/index.php?topic=6855.0. Really liked Big Al's work and wanted to do something similar. I decided on 18 awg as I considered the thicker wire an upgrade. I quickly discovered that the thicker wire is more difficult to bend in tight quarters and takes up much more space. If I were to build another Crack, I would use the thinnest wire I can get away with.

Edit: I forgot to mention that the twisted wire can not be twisted using the drill method. Spinning the drill slowly or manually turning the two lengths of wire does not work. The cloth will separate and it does not create uniform twists. I did each twist individually by hand until I had the necessary length for the build.


[*]Belton tube sockets - I wanted tube sockets where the nine-pin socket was flush with the top plate. Belton fit the bill and since I was ordering parts from Tube Depot, I just added these as well. While they look and feel very nice, the nine-pin socket has some issues that could cause serious problems during the build. The eight-pin needed some enlarging of the mounting hole, but the nine-pin threw me quite a curve ball. It is smaller than the hole in the top plate and no amount of filing would fix that. I was going to insert a plastic washer in there, but in the end decided against it and just left it as is. The nine pin socket also has two other issues, namely:

a) there is no center pin in the tube socket, thus the LEDs have no natural mounting spot. Considering the fragile nature of LEDs, I consider this a major drawback. Searching this forum I found that Lior Amsalem also used these sockets (refer to http://bottlehead.com/smf/index.php?topic=2349.0) and I contacted him via PM. He advised that he used a bolt and nut though the center hole of the socket (yes, it has a hole you can use for this) and used some wire to create a hook. I used similar method to mount my LEDs. Lior's build was definitely quite inspirational to me as well.

b) the socket is already affixed to its mounting plate, therefore the orientation is already determined. It meant that none of the A1-A9 notes on my top plate were accurate. Luckily the Belton sockets have numbers on the sockets itself which helped avoid serious errors. It does however mean that none of the wire lenghts / routings at the nine-pin socket agree to the manual.


[*]VA Labs stepped attentuator - it was the first upgrade I did, and it helped me gain a lot of confidence, but I am not sure if I can notice the difference between this and the stock pot (my stock pot had no channel imbalance and sounded quite good). It is also physically very large. If I were to build another Crack I would use the Alps Blue Velvet just to save on space.


[*]Switchcraft output jack - I melted some of the plastic on the original jack and was not quite happy with the look so when PB recommended the Switchcraft 152B, I got one. I did not want to have a crowded jack like the Neutrik locking jack therefore it is a little ironic that I chose this one. Definitely more difficult to work on than the stock jack and the sleeve (to which ground wires connect to) is quite bendable. Based on the size and the difficulty in mounting the resistors, I would say that the stock jack is quite good.
[/list]

I got a lot of inspiration from the gallery section in the forums and, other than the names mentioned above, I prefer not to mention any names as I definitely can't recall everyone. However, I have reviewed every Crack gallery that still has pictures in it for inspiration.

Herewith pictures of the completed build:

bthWPAI.jpg


Jmk1NJK.jpg


5hAMe0s.jpg


y5WCWXx.jpg


63hNS1T.jpg


6e4flma.jpg


Important note - The below picture has L&R swapped on the Switchcraft jack - refer to the next post for the correction.

EEdiPM7.jpg


7GWgvl0.jpg


kC7HIjq.jpg


Still have several upgrades planned and I have to thank those who posted their upgrades here as it enables me to also upgrade my Crack in ways I never thought possible before.
 
Flat sounding amp and loose solder joints

With the first listening impression after the rebuild, I was quite disappointed. Volume through the iPhones headphone out was quite low - I could comfortably listen at 75% on the volume knob - and music just seemed flat. Best way I can describe this is that the bass was lacking, and I actually started preferring my Shure SRH440s straight out of the iPhone. Listening via my USB DAC solved the volume problem, but the amp still sounded flat and I started noticing a scratchiness that was not present before. I attributed this to the fact that the outlet has 1 pc, a laptop and 4 external harddrives connected. However, on performing a listening test before installing the PSU upgrades, the scratchiness was still present via the iPhone and running from a clean outlet. I reasoned it had to be the tubes.

As it turned out, I had some flaky solder joints on 2 spots. While the larger holes in the terminal strips are nice, it makes get proper joints really difficult once a joint gets crowded. In any case, on the photos in the prior post, I had already inserted a 5-pin terminal strip on the 8-pin socket's mounting screw. While inserting the terminal strip, I managed to dislodge the red wire going from terminal 13 to the 8-pin socket. Turned out this was the first problem and I resoldered the red wire properly to terminal 13. The amp still sounded the same and I just assumed this was how it should sound.

On removing the remaining two power supply capacitors from the terminals, I noted that the second capacitor had one loose leg. I only noticed that while performing the final upgrades (of this stage at least), but I can say that with the capacitors now properly soldered the Crack now again sounds glorious. The scratchiness is gone, the amp no longer sounds flat, and I now prefer to listen at about 25% of the volume knob. Unfortunately I only fixed this after installing both the Cree diodes and the choke so the previous listening tests are definitely not accurate and I can't comment on whether the upgrades led to an improvement in sound quality.

Left and right switched on Switchcraft B152

Based on the discussion here (http://bottlehead.com/smf/index.php?topic=7954.0), I started suspecting that I had swapped L&R on the output jack. Did a quick test in Windows and realised they were swapped around. While I was under the hood I decided to fix this and it enabled me to shift the wires out of the way, which creates more space for the cable tie mounts. Picture of the correct wiring of the output jack.

EQAdIFf.jpg


Cree diodes installed

richmi's Crack provided the inspiration and information for the installation of the Cree diodes (link here - http://bottlehead.com/smf/index.php?topic=7215.0). Without the detailed explanation he gave there, I would not have been able to do this upgrade.

Cree diodes were installed under the choke as I wanted maximum space next to the terminal strip. At this point I had not yet decided to install the 1uf K75-10s, but I knew that space would be tight. The protruding wires on the terminal strip were subsequently trimmed down properly. Resistance and voltages checks were fine, and it played music although it was still soft and flat as described above.

QkplknU.jpg


Triad C-7X choke installed

I had originally planned on just doing the Cree upgrade, listening for a week or two, and then installing the Choke. However, since I was in a groove, I just connected the choke. Initially connected the choke to terminal 14 & 15, but realised that I had made an error while reviewing the solder joints. In order to ensure the magic smoke stays inside the amp, I corrected my mistake and connected it to terminals 13 & 15. At this point I also decided to relocate the first two PSU caps to create more space for the final film cap. This is where I realised that I had a flaky joint on the PSU caps. Once this was corrected, resistance and voltage checks were fine and I had a quick listen to see what it sounded like. With the solder joints fixed, the amp sounded much better. In hindsight I should have spent more time listening to it, but I was already planning to replace the last PSU cap.

1rQLIBa.jpg


Using K75-10s as mounts for the final PSU cap

I had planned on using cable ties to secure the final PSU cap to the choke, but I was not happy with the sturdiness of this arrangement. I also had four 1uf K75-10s that were no longer going to be used (based on the post by Snarii here - http://bottlehead.com/smf/index.php?topic=4744.90 - I realised I would rather use 4.7ufs as output bypasses), and I was contemplating using them as bypasses on the final PSU cap. It did not take long to realise that I could mount the K75-10s to the top plate using cable tie mounts, and then fasten the last PSU cap to the K75-10s.

Potential fire hazard: One of the K75-10s touched the exposed terminal lugs on the switch. I only noticed this on the final review before I did my final voltage checks. Since this is where the power enters the amp, and the shell of the caps are very probably conductive, I assumed it would lead to fireworks. The temporary solution I did last night was cable-tying some folded paper to break the connection. At this stage I cannot remove the caps without disassembling the entire PSU again, so for the final solution, I will create a "bed" consisting of insulation taped cardboard that will span the entire bottom of the cap. This will then be affixed using cable ties. None of this is evident in the pictures I have posted.

Picture of the K75-10s mounted and awaiting the final PSU cap

Ux6ZbjO.jpg


Installation of the final PSU cap

The final PSU cap is a 270uf 400V Jantzen Cross Cap. I can get these for pretty cheap, thus it was a pretty easy choice to make. I was planning on buying a 220uf, but was just too eager to press the "Buy It Now" button and ended up buying the larger cap. The diameter is only slightly larger than the 220uf, so I thought I would see if I could get it installed rather than send it back. Installation was quite straightforward, but as can be seen, I have quite a few wires running off the 5-pin terminal strip and the PSU is definitely getting crowded.

The idea to use the additional 5-pin terminal strip comes from Mikey's build (refer to http://bottlehead.com/smf/index.php?topic=1632.0) and it really made wiring all the bypasses much more simpler. Even without the bypasses, I would still use this just for the pure ease of accessibility.

Since access to the PSU terminals is now severely limited, resistance checks became a little trickier. Getting the alligator clip attached to terminal 12 was tricky. Terminal 13 is no longer accessible, and I did not perform a resistance check there. With the voltage check extreme care needs to be taken as space is suddenly very cramped. I did not measure terminal 13, but the voltage measurement at B5 was good, so I assume that T13 is correct. Perhaps I could also measure the T13 resistance there, but I did not consider it at the time.

dr8Ro3m.jpg


View from the top

8LL13Cr.jpg


View from the rear

a9pfBCG.jpg


Future upgrades planned

[list type=decimal]
[*]Installing higher feet to get the undercarriage off the floor
[*]Replacing the remaining 220uf PSU caps with Rubycon 470uf caps
[*]Installation of Speedball
[*]Replacing the output caps with Mundorf 100uf MCaps
[*]Bypassing the Mundorfs with 4.7uf K75-10s
[*]Bypassing the K75-10s with 0.056uf K72P6s
[/list]

How does it sound?

Well, suffice to say I had a big smile on my face after the final upgrades were in place. Due to the points noted above, I can't really comment on how each incremental change affected the amp. Upgrading the final PSU cap to a film cap and bypassing it with the K75-10s makes me believe that more bass is present, but it is really tough to say - installing the upgrade between dinner, bath-time and putting kids to bed meant that 4 hours had passed between listening tests.

A sincere thank you to all the others before me who provided the inspiration and guidance that made this all possible.
 
Not to put too fine a point on it, but the Crack is not designed to be used with 44 ohm headphones.
 
Hi Doc,

I know and I can see how my post can lead you to believe I use the Crack with the SRH440s. My setups are:
[list type=decimal]
[*]Work - iPhone to SRH440s (tried the SRH440s in the Crack once, but bass was loose and flabby)
[*]Home1 - Computer to Sound Blaster X-Fi HD USB to Crack to 250 Ohm DT770s. Lots of IT equipment on the outlet so it gets pretty noisy.
[*]Home2 - iPhone to Crack to DT770s. I usually move the Crack to either my bedside table or the kitchen table where it is the only thing plugged into the outlet. Crack will be almost dead quiet when plugged in there with the exception of some extremely slight background noise at max volume.[/list]

Before I disassembled the Crack many months ago, it definitely had something magical. When I first fired her up after the rebuild, the magic just was not there and the sound was as described. I never thought that it could be my sloppy work that was the problem. Fixing the loose connection at T13 did not bring back the magic, but once the power supply caps were properly soldered the magic was back in spades. It was immediately noticeable in both the increase in volume and return of the bass.

Guess I am quite surprised that an oversight this small caused such a big difference in sound and I am wondering how many who complain of insufficient volume just have a flaky solder joint somewhere.

As a last note, properly soldering the power supply caps was the single biggest sonic improvement done to this amp to date.
 
As I have posted a few times over the years, in the Crack the power supply is completely inside the signal current loop. So any problem in the power supply is likely to show up in the signal.

People often use the term "the signal path" but I find that to be ill-defined, and the definitions I've seen have more magical thinking than science behind them. That's why I use the term "signal current loop" which is at least a well-defined term.
 
Back
Top