Bottlehead Crack OTL AMP (Kit/DIY) - Official Thread

BHC Build Notes: August to September 2021

I ordered the BHC and Speedball upgrade kit during an August 2021 Bottlehead sale. I’d been on the fence about getting a Crack for a long time, but it has a strong reputation. It may be the closest thing to a benchmark OTL headphone amplifier on the market.

This post is intended for others similar to me: capable amateurs interested in building a Crack but uncertain of whether it’s right for them or how to do it. This post is not intended for experienced kit builders, nor for those uninterested in a kit.

This kit requires moderate journeyman craft skills and is NOT FOR EVERYONE. You MUST have solid attention to detail and the discipline to do things the right way and fix problems. The kit requires working with hundreds of volts, so if it doesn’t kill you it could surely ruin your day. There are some horrifying Crack builds around (timestamp 4:30 to 5:15). Don’t create another time bomb. At the time of writing one can pay a professional $200 for a basic functional assembly or $400 for assembly plus finished metal and wood. If you don’t have tools and finishing materials, the cost of DIY will certainly be similar. With any major error you will spend even more to have a functional amp. It took me about 40 hours for research, wood and metal finishing, mechanical assembly, and soldering. The soldering part was quick.

There’s no way to avoid rookie mistakes if you don’t have experience, as building involves hand-eye coordination plus real skills. Version 2.0 is always better than version 1.0. Do not proceed unless you are careful, patient, and willing to learn. Practice before attempting the dangerous portions.

The Crack demands space and attention. It must sit on an open flat surface, while the tubes and cables stick up high. As such, you will spend a lot of time looking at it even if not using it. I set out to create something that could blend into a living room or den, and look professional too. I didn’t want to think of a high school project that screamed “lazy” every time I saw it. So, I put as much thought into the cosmetics as the functional build.

Wood: My wood was nice (heh, heh), and ideal for first time wood project. The dimensions were perfect, seriously easy to assemble by woodworking standards, and took stain well. The finish is a personal choice, and from the start I knew I’d use natural black walnut dye. This follows from the retro look of the Crack and my disappointment with some modern stains. They can function as half paint and half stain, resulting in an artificial appearance. Natural walnut dye and linseed oil have been used for centuries, and they literally define authentic and traditional wood finishing. My design concept: a new old stock product from 1921, as built in 2021. I researched the look of 1920s products for plausible colors and finishes. Many electronics of the era had dark wood finishes, and were probably finished exactly as mine (walnut dye and boiled linseed oil).

A good result follows from patience and layers. Sand. Apply wood dye, dry, repeat and repeat again. The black walnut dye went on pale with the first coat, but became nearly black (burned looking) by the third coat. As water based, it required full surface saturation with each coat. The water raises the wood grain, requiring long drying periods followed by rubbing or fine sanding between coats.

Boiled linseed oil is easy to apply, but doesn’t dry for at least a day or two (i.e., sticky and then lumpy to the touch). It must be rubbed/smoothed between coats. Patience. The final finish has a satin sheen that’s hard to capture with a camera. Additional coats and rubbing (or a different oil finish) may increase the shine if desired.

Metal: Electronics in the 1920s often used black plastic, dials, and faceplates. However, that was too boring. I chose paint from a vendor founded in 1921 and a classic dark green. This is plausible new old stock. [Also, I may have been influenced by another excellent green amp.]

Painting the metal took much longer than planned. I used spray paint, and this may have been a mistake for the small size of the project. Next time I’d likely use a brush and sand between coats. The main top plate took paint well per four light coats in one session. It fully met expectations. Due to my stupidity in not putting down a big enough drop cloth (i.e., old grocery bags), overspray ended up on the floor and the bottoms of my bare feet. I had a darkly funny session of removing green paint with rubbing alcohol. Fortunately, it cleaned up easily because of its slow drying (24 hour) period.

The transformer bell housing is shipped as bare metal, and it’s prone to rusting. Mine had faint polka-dots from the bubble wrap packing material. It required three tries to get the paint right, with 72 hours of cure time between coats. The flat top looked great after the first try, but the edges had gaps from the aggressive aerosol blasts. I roughed the surface with sandpaper so the second coat would stick, but medium grit sandpaper was too coarse and the scratches showed through. I got it right on the third try with fine sanding and six very light coats applied every two minutes. Spray paint always leans toward orange peel too. The final look matched the intended new old stock theme.

In researching finishes, I found that some otherwise superb builds overlooked hardware tint. A dark top plate should have dark hardware unless one wants to draw attention to the screws.

One can paint screws, but paint gums up the slots and can often be scratched off. I used a metal bluing product. Squirt paste or pour liquid into a shallow dish. Cover the visible portions of the hardware for a few minutes. Wash it off. Repeat until the metal is dark, and be gentle with it as it sets. Wash and dry fully, but leave the screws bare and bright to maintain electrical conductivity.

Assembly: The kit requires the physical installation of the jacks, sockets, pot, and transformer. This is mostly simple, with the transformer presenting a slight challenge. One must layer the transformer, bell housing, screws, several washers, and threaded nuts in the correct order. The sandwich must be kept in place as the screws are tightened, so one could use three hands. Ignore the counterproductive instructions to tape the screws and instead create a jig to support the main plate vertically and hold the transformer half way up. Also note that one screw is included to cut threads into four mounting bracket holes. Use a lubricant and clean the screw between cuts to ease the process.

Wiring: A wire stripper and needle-nose pliers are your best friends. Most of my electrical construction time went into twisting, braiding, measuring, cutting, and stripping wires. Bending every stripped wire tip into “L” or “C” shape maintains mechanical stability. Keep the wires straight near the ends, as curved insulation will split if heated during soldering. The manual does not address this.

Soldering: I began by watching a couple Youtube videos, soldering short pieces of wire (not from the BHC kit), and checking for conductivity with a multimeter. I bought two school project kits ($10 each) to refresh my skills. I hadn’t soldered in ages, and also needed to determine the characteristics of my soldering iron. The practice kits were very small, PCB based, and much more challenging to solder than most of the BHC. Both kits were helpful for learning temperature settings, the amount of solder required, and how to avoid burning yourself or overheating parts. I built and “upgraded” a tiny mono amplifier (it works…but it’s a p.o.s. that sounds worse than a phone speaker), and an electric Christmas tree decoration (it doesn’t work). Several Christmas tree reviews reported high failure rates, criticized the parts as junk, or received bad parts out of the box. I fried the power switch before the leads were hot enough for the solder to stick, and watched black plastic melt all over the PCB. The battery holder had bad dimensions and wouldn’t hold batteries either. Set your expectations accordingly.

For the most part the BHC has large parts, great wires, and is very easy to solder. At least 95% of it was easier to solder than my practice kits. I switched from a small pinpoint iron tip to a mid sized flat tip, and everything was quick and easy. Until I encountered the infamous BHC LEDs.

The LEDs: While the BHC kit mostly requires “mid-level” soldering skills, two LEDs crank up the difficultly. The LEDs are (1) very, very, very small, (2) extremely fragile with long and soft leads, (3) can be overheated/melted during soldering, and (4) must be installed in the most cramped portion of the unit. The instruction manual is defensive about the LEDs and misses the point. Before the LED step, all hardware and wiring is large and easy to handle, but it suddenly changes. The manual doesn’t prepare one to leave the region as open as possible, whereby prior wiring choices can make the LED soldering process very hard. If I’d cut a couple of the 9-pin socket wires only 1/2” longer this step would have been much easier. Other experienced builders have issues with the LEDs, and Bottlehead sells extras on their website. There must be a better way!?

Grab some tweezers and a magnifying glass for the LEDs, because you’ll likely need them. I strongly recommend that Bottlehead add a full page explaining soldering techniques and routing options. I spent perhaps 30-60 minutes studying and then fitting just two parts. Upon completion, I was only 2/3rds confident in my work and shopped for replacement LEDs in case I’d fried them. They were fine.

See the problem area below (my blue circle and text added to Bottlehead’s stock image.

Wrap up: My Crack works fine after about 20 hours of run time over several days. It’s a fine decorator item too. I like its performance with a Svetlana 6N13S (6AS7G) output/power tube and the HD-6XX, but not as much with the factory GE 6080 (way too bright) or a similar RCA 6080 (rumble and excess bass). I have quite a few 12AU7 input tubes, and am still experimenting with those. As expected, the HD-600 is hissy and noisy so the Speedball may improve its performance (it really wants a balanced amp). Somehow the HD-6XX works “right” as is, without perceived noise and with copious gooey-tubey character. I don’t think I’ll install the Speedball on this build, as it misses the point of a retro amp and as the HD-6XX works fully as desired. I can choose a more technical amp if I want technical performance. Finally, the factory pot isn’t very good so I’d upgrade other hardware if adding the speedball.

The green appearance changes a lot per the lens and lighting.


Materials for Constructing and Finishing the BHC:

  • Natural black walnut dye (i.e., walnuts and water)
  • Boiled linseed oil (Danish oil is similar and offers additional color options)
  • Rust-Oleum Satin Hunter Green spray paint
  • Birchwood Casey Perma Blue Paste for visible screws and 9-pin tube clamping ring
  • Wood glue
  • Replacement volume knob for visual level confirmation and to match the design theme
  • Silicone spray to protect the bluing/patina applied to visible screws and top hardware
  • Blue thread locker (i.e., Loctite) to prevent loosening of non-electrical hardware over time
  • Electrical rosin core solder (60/40 or 63/37 tin and lead)

Tools Used:

  • Soldering iron, spring rest/stand, sponge, solder removal wick
  • “Helping hands” soldering system (4 arms) for practice but not the BHC itself
  • Wire stripper and wire cutter tool (20 AWG stripper)
  • Wire flush cutter (small tip allows tight space access)
  • Multimeter with at least one probe alligator clip (black line)
  • Needle nose pliers (essential), and second set of pliers is useful too
  • Tweezers and a magnifying glass (LED installation)
  • Ruler or tape measure (12 inches or 30 cm)
  • Screwdrivers, wrenches, and sockets (various, from small to medium sized)
  • Sandpaper (220 grit most often used)
  • Sanding block to maintain flat wood surfaces and even edge breaks
  • Microfiber cloth to support transformer housing and protect finish from damage when soldering
  • Dial calipers to measure potentiometer shaft diameter for replacement (it is 0.25”)

Additional Supplies:

  • Spare wire for refreshing soldering skills
  • Soldering project kits to refresh soldering skills (e.g., $5 to $10)
  • Toothpicks to apply and remove metal bluing in nooks and crannies
  • Cotton swabs (aka Q-Tips) to apply and remove various liquids
  • Rags (cotton) for applying finishes and wiping between coats
  • Rags (polyester) for removing sawdust and rubbing wood smooth
  • Rubber or chemical gloves to keep walnut dye/stain off my fingers
  • Plastic and glass containers (e.g., cleaned yogurt cups, cookie boxes) to hold scraps and liquids
  • Isopropyl alcohol (rubbing alcohol) 91% or 99% for removing oils and wet paint
  • Paper grocery bags and empty cardboard boxes for disposable work surfaces
  • Painters tape / masking tape for assembly and liquid control
  • Packing tape to secure bags and cardboard pieces together
  • Plastic cling wrap to separate slow drying wood and paint finishes and facilitate handling
  • Kitchen detergent to remove surface oils after final assembly (diluted on a damp rag)

Hi @generic I really appreciate this guide. I keep wanting to build my own. I will get around to it eventually and with your great guide and walkthrough it’s going to be easier. I know that there are loads of very comprehensive guides about the Crack but I like yours also.


Well done @generic , both for the build and the extensive post!!


Built a BHC a few weeks ago. First tube amp and I understand the appeal of tubes now. The meme setup 6XX + BHC is really something else.

I’ll need to take more pics once I get around to finishing the wood… I have a pretty nice idea in my head I need to execute. Hopefully, I get some time to do that soon!

Also wish my pics would translate better to how it looks IRL… the transformer cover is a Galaxy Purple - has a really nice sparkle effect depending on the lighting. I wanted to experiment a bit with the chassis plate so instead of making it fully black - I mixed very tiny amounts of the galaxy purple in to give it the white specks. The goal here was to make it space-like and I think I achieved that.
Check it out: BHC Build - Chassis Plate


Nice custom color job. :+1:

1 Like

Thank ya! It turned out great. Wish you guys could see it in person


Has anyone tried to rebuild Crack using PCB instead of point to point?
Point to point just takes too much space.

The transformer, tubes, tube sockets, headphone jack, power jack, RCA jacks, and volume pot require a minimum amount of space too. The wiring isn’t very complicated and fits between those components. The Crack produces a lot of heat underneath (caps and resistors), so generous air gaps are probably a good idea. A PCB may allow a height reduction of the wood base, but the surface spacing facilitates airflow and stability. I swapped the 6080 for a 6AS7G (6N13S), and it stands about 6" over the metal plate.

1 Like

Hi Guys,

I come to a dilemma situation choosing a pure OTL amp for my ZMF Verite open and HD800S. I have down selected to 2.

  1. BHC with speedball + build services (since i have no skill and time to build) + Mods: Total: $1205
    Capacitor upgrades (Mundorf EVO Oil)
    Audio Note Potentiometer
    Cree Diodes
    Choke C-7X

  2. Woo Audio WA2 $1399.

Will 1 out perform 2 or vice vera? Which one will you pick? Of coz tube rolling also a determine factor.

Thanks guys


I doubt the BHC will hold as much of the value of the $ you spent, if you try to sell it in the future.

The woo will lose value too, but probably not as much.

Honestly, if you want a Crack with that much added to it, I think you’d be better going with a Crackatowa. Maybe have a cap upgrade done to it, then call it good. Better bang for the buck. Better power transformer (which in my book is more important than diodes and chokes). Tube rectification. Etc.

I don’t mind Woo much, but you’re better off saving $300+ buying pre-owned.


I concur with @ProfFalkin about the Crack stretching for the >$1k value range. Many people enjoy building kits, but that’s a different market than retail and reselling as used. People (rightly) doubt the quality of finished kits, as many are butchered during assembly. I watched a couple horror show videos in preparing to build mine. Factory finished products typically have a known degree of quality control, and vendors have service departments.

The Crack kit is in the $300 range for a reason. I’m not happy with its volume pot per channel imbalance. The 1/4" headphone socket is a bit too deep and the left channel is silent if longer length jacks are inserted all the way. (When pulled slightly out it produces stereo output, it’s a flimsy socket rather than a mono short.) Treat the standard Crack as an inexpensive product [if you own tools and have decent build skills] that performs well in the OTL genre. However, its output is not dramatically different than the <$300 factory new and finished DarkVoice I just acquired.

If you do choose a kit build service, I’d ask the builder for end user references. This is not for lack of trust but to get feedback on the upgrade’s performance and sound character. I doubt that more than a handful of the visitors to this forum (if any) have heard a similarly tweaked Crack.

The Woo will almost certainly be easier to sell, but I have no experience with it. No comment.


So, I know I’m slow AF, but it’s getting there. Will be starting the wiring/soldering part hopefully tomorrow.

Acquired the upgrades I wanted for the most part… Alps pot, better knob, upgraded tubes. Couldn’t make a decision on caps but I can add those later if I figure out what I want. Couldn’t figure out the size issues for upgrades and what exactly is needed to squeeze them in.

Wood and paint didn’t quite come out as I planned but I’m good with it. Staining was fine but I struggled with the polyurethane. I just couldn’t get it applied without blemishes. After doing it 3 times and sanding back to zero and restaining over and over… I’m leaving it. The wood got a nice look to it after being stained and sanded so much, lol, but the last poly coat is still f’d up.

Definitely enjoying this project once I finally got started. Already thinking about what’s next…maybe something in a metal prefab case :laughing:

I’ll share some more details and better pics when I get it all finished.


The wood I received (standard Northwestern alder) is more porous than most wood I’ve worked with previously. It’ll absorb whatever you put down and much of it will migrate to the mitered corners and/or fully disappear inside. It’ll look okay for a while but then revert to the natural fiber structure. The wood fibers suck up finishes as little straws, and after they are done the wood will take on a translucent and deep (but unstable) appearance.

I applied quite a few layers of black walnut dye, charcoal powder, and boiled linseed oil before reaching the look I wanted…and eventually plugging up the tubes/pores. Traditional products are never as stable as modern polyurethane products, but they can be easily tweaked over and over again. My final two photos above show how it looked 3-5 coats ago. The wood is darker now.

After letting the latest coat of oil cure for about 3 months I’m thinking of light sanding and adding another coat. Or, perhaps I’ll wait another 3 months, as linseed oil takes forever and a day to cure.

Swapping the pot is a great idea, as the factory pot is not good. Per my reading, be extremely careful soldering the Alps because it’s designed for (at most) a single quick soldering pass. Some builders solder the wires to a subassembly board.

I’d also swap the factory 1/4" socket for something better if I could do it again (probably a locking Neutrik). The factory socket doesn’t tolerate jack length variations well, with some inserting too far and missing contact on the tip. I have to pull those jacks out a couple mm for a stereo signal. I flipped the board to see what the problem was, and it’s caused by excessive flex in the plastic housing and (wobbly, flimsy) metal TRS connector bands.

Soldering was straightforward except for the LEDs (as above). If doing it again I’d bend the legs into an “A” shape, affix the leads with alligator clips, and solder underneath (to minimize heating the LED itself). I’d also have purchased the $5 set of 5 spare LEDs from Bottlehead. Finally, before you cut the wires to length inspect the routing requirements and what comes next (also where the Speedball goes if you plan to add it). Some wires (e.g., green heaters) would be much easier to route if only slightly longer than the instructions state. The cramped front tube socket area involves the bulk of the problems.

Good luck!


Thanks for the advice. I was debating whether to get the Neutrik socket but didn’t because I didn’t want to mess with drilling the holes to mount it.(sounds extra lazy when I say it out loud ) After what you’ve said I’ll go ahead and add it. I do like the look of it much better as well. What I’d really like is to have an XLR instead as all the cables I actually use are 3 meter and terminated in XLR. I guess that’s not possible though? Couldn’t find anything about it.

I did see the Alps pot on Amazon with the board already attached. Wasn’t sure it was compatible that way…wish I would’ve looked into it more. May have been easier. Oh, well. I am doing the speedball as well.

1 Like

I have four thoughts regarding the Speedball:

  1. Leave it off because I like the easy-going sound for some moods and have technical amps if that’s what I want.
  2. Leave it off because mine came out as retro through-and-through, per the old-school look, the cheap pot, and the cheap socket. Design purity.
  3. Put on a V2.0 Crack built with technical components from the start.
  4. Sell it as NIB in favor of something very different.

or you can get something like this, a board specifically to mount the alps pot on The ε27 volume control potentiometer mounting board