Custom Chassis?

While this is not Dynaco, it is a custom chassis build I made several years ago for a headphone amplifier. The top unit is a source selector / cross-feed circuit. The middle unit is the amplifier which uses (3) 6N1P NOS Soviet military tubes and is based on a circuit design by Bruce Bender. The bottom unit is the power supply. The chassis' were made from aluminum extrusions obtained from ProCo Audio and cut to dimension sheet steel from a local vendor. I did the drilling and punching in my workshop as well as building the shelf to hold all three pieces. Thought you might enjoy seeing this.

Best Regards,
Charlie

Nice job Charlie looks very nice indeed!

Thanks sKIZo you asked the questions I was thinking on the chokes I was thinking replace the two originals kit supplied units with two upgrades,
I'm still not sure I understand the benefits though, glad to hear your going the diy route heck whats the worst that can happen........... never mind LOL
Looking forward to hearing about your success.
Regards

I'm looking forward to not telling about my failure.

The better ... er, strike that (I stand corrected) BIGGER chokes came in today - that was fast! For anyone expecting a subtle difference in size, think again ...



Definitely want to have those handy when you're laying out parts as they take up a bit more real estate. Heavy little rascals too as you might imagine. That's a whole lotta iron compared to the originals.

Benefits? They ARE rumored to be quieter, but as quiet as the ST-120 is already, it might not be all that noticable. I'd think the biggest benefit would be reduced heat. Maybe lower saturation and more predictable performance under any load. I expect I'll be running the amp fairly hard, and that could come in handy. One other potential benefit is a larger choke tends to tone down the high frequencies some - for near field listening, that can be a good thing, depending on your speakers. My McIntosh XR16 big boxes have some serious tweet to them and I do currently flat shelf eq them down quite a bit over 8k. I also read that the choke works in tandem with the filter capacitor but is a LOT faster. Stands to reason that the larger iron would offload some of the cap's duties and make the system more responsive.

And ya ... all that is just what the innernet tells me, and half the lies it tells you are not true.

Good thread. Here is my latest 160 watt custom amp under construction during the testing of the primary wiring. First shot shows the two C-17_X chokes. The other shot is the top side with the 800DCV toroid power trannie and the 2 Edcor output trannies. I used a rotary switch for bias selection and a digital meter. All I have to do is select the tube and adjust the pot. Switch has an off position when done. This amp has 2 separated bias boards, the square holes on top are the bias pots. It also has 2 power supplies and 2 rectifiers. The holes on each side of the lettering are LED's that are red until the warm up is done then they turn green. Circuit has a lot of Dynaco DNA.

Serious hardware there ... my amp wants to grow up to be just like yours ...

Where'd ya get the rotary switch? I was looking for one in case I decide to go with just one meter. If I stick with the two, I'll just do the front/rear switches for each bank.

One other minor change on mine - a couple nice bat handle toggle switches instead of the sliders. Tried cutting a square hole today with the dremel and although it worked, it wasn't as clean as I'd like. Round holes are much easier. Grainger has a nice punch ... for only $80. I'll pass ...

sKiZo wrote:I'd think the biggest benefit would be reduced heat.

Why do you say there would be less heat? Aren't the chokes, no matter what the size doing exactly the same work? So their heat output essentially should be the same? True with one much larger there will be more mass to dissapate that heat, hence a perception of it 'running cooler', but won't the heat output be exactly the same?

OK ... semantics really ... let's call it less warm as opposed to less heat? The increased surface area and greater mass should allow whatever heat is created to percolate to air faster, which technically means the choke should run cooler.

Can't really comment on the efficiency of the larger choke in the circuit, but that could also work to reduce waste heat.

No comment on the pros or cons of moving the bias trimmer pots off the driver board, so I'll plan to do that also using the B50k linear pots I mentioned earlier. Got those, some knobs, and a neat little pushbutton power switch coming, and I think that's the last of the parts I'll need. Should be able to get started when the metal shows ...

Skizo:
Here is what the raw plate looked like before it was bent at a metal shop in a sheet metal brake. This was also back cut on the bend lines and notice there are no stress lines in the earlier picture in this thread.
I strongly recommend bending in a shop as 2.5 mm. is a lot stiffer than you might think. Cost me $15
I had a couple of failures with stress lines bending another one at home because the bending pressure needs to be evenly distributed across the entire bend length or you WILL! have stress lines.
The rotary switch is a "Lorlin" CK series from Mouser. I used a "Shurter IEC power input module with both the on/off switch and fuse cartridge onboard . No slider or other power switches are needed. Less holes, easier and/or cheaper.
Good luck. I like your enthusiasm





FYI:
Toroid transformer:Custom 800VDC "Sumner"

It's my enthusiasm that usually gets me into trouble. This started out as a simple kit build. dOH!

You've obviously got the bug too, and a lot more practice than me.

*****

Naked toroid!!



Quality stuff. The efficiency is really quite amazing. And they're much purtier than a standard stack. I've got em in my Sansui QRX-9001 and 8080db and they never really seem to be working all that hard ...

*****

I did notice the power setup in one of your earlier picks, but my amp is gonna be fairly high up and the back will be kinda hard to get to, so I like the idea of a front panel switch. Here's what I got ...



Another advantage of a bigger chassis ... thing takes up a lot of real estate! It's got a blue annular ring LED light - not sure whether I'll use that or not as LEDs can be a bit noisy. And it might detract from the glow of the tubes. Maybe hide a mini toggle to make the light optional.

< < < DIY METAL BRAKE > > >

As said I'm gonna go with a separate back panel, so the only bend will be the angle for the front panel. That should be around 30 degrees at most. Found this on YouTube ...



https://www.youtube.com/watch?v=xQlHqdudPvU

Might be just the ticket. Only thing I'd do is roll the edge of the top angle iron for a softer bend, or maybe try it with some scrap using a piece of steel pipe instead. Not much of an investment - zero really - as I have all the parts kicking around the shop anyway. I did order two blanks, so if I screw one up, I can always beat it flat for the chassis bottom and take the other one down the road to a real metal shop.

Switch - shorting or non-shorting. Got a number?

One thing on that purdy toroid you show: Take a look at mine. I had all the wires clustered together on each side so I only need two holes in the plate. That one you show they are strung out and you will have to make a lot of holes to mount it on top of the plate. Mine only needs 2 and they are hidden underneath. Otherwise reach for your wallet and buy a can for it that covers them. that makes it bigger as well. Hope you have the room.

Progress report ...

Not much, really. Got caught up in spring gardening and a couple outdoor construction projects.

Got the metal for the chassis plate, and that didn't go all that well. Got some good "accurate" sockets, bu t they just didn't have that "finished" look I was aiming for. Didn't like the look or tone of the unfinished metal either. Could have lived with that, but killer was that my drill press didn't have a deep enough throat to do the rectifier socket, so ... ordered a simplified version of the panel from FPE - it should be here next week. Here's the final version as ordered.



This does the CNC work for all the holes with exposed edges, and the two up front for the meters which are bigger than my bits. Keeps the costs down enough I was able to add back the "purty" around the pentode/triod switches. Any holes that will have covered edges after assembly, I'll do myself. The back plate will be machined separately and inserted into a cutout in the wood case. That I can do too.

I'm also planning to leave the transformers naked, with just a couple coats of Never Dull to polish them up. I may regret that, but they do look right nice at the moment. Never Dull is a good metal protectant (made for motorcycle chrome) so I don't expect any rust issues, and I can always touch them up later if needed. The bells surprised me - turned out a nice silvery gray semi metallic finish just rubbing them smooth with some 00 3M pads and the Never Dull wadding. Moderation is the key as too much and you'll rub the finish off.

Got all the parts ... I'm running out of excuses for not gittin er done!

PS ... got the switch ... it's got a real solid feel to it. Still debating on whether to wire it for the annular light ... cross that connection when I get to it. Hopefully not literally.

Something ya gotta love about Panel Express ... they certainly do keep you informed on progress.



In case anyone is wondering about the "infill" part, that's the ink fill on any printing. I did get one gratuitous bit of bling in the design.



One check mark I'd like to see is "Buyer didn't make any really stupid mistakes during the design phase."

Machining a groove to thin the sheet aluminum where the bend is going to be is is a good idea. Use a ball nose cutter and perhaps round off any sharp edges the cutter may create before bending to reduce stress risers in the material. Another possibility to help hide stress marks is to perhaps anodize the aluminum after forming. Just a couple of suggestions. Blake

I decided against the groove ... that silly little back cut added a LOT to the cost of the panel for whatever reason. Multiple passes, flip and new CNC program, that kind of stuff. I figure to do a more gradual bend with a fairly moderate curve to it and keep my fingers crossed. Here's a quick and dirty concept drawing I did a while back ...



Lots of changes to the design since then, and excuse the funny looking stuff on the front panel. That all got some serious parallax issues when I "bent" the panel in photoshop.

PS ... now that it's too late, I'm thinking what I SHOULD have done was add a few closely spaced shallow relief cuts on the front of the panel right at the bend. That would have covered up any stress marks (which knock on wood, I won't get anyway) and also tend to act as guides to keep the bend where it belongs. Also, I think it would have added a nice touch to the looks. Ah, well ... next time. If I DO get some stress marks, I can always add some pinstripe tape ... waddaya think ... flames or chains ...



I think I got some skulls here somewhere too.

OK then ... got the panel back last week and drilled all the mounting holes (well, most of them anyway). Here's the first dry fit from the top.



Quite happy with it ... definitely different. One of my major concerns was adding space between the bottles, and that worked out right nice. There's even plenty of breathing room between the big bottle rectifier and the transformer.



Excuse the wandering knobs ... the small ones next to the driver tubes in this pic are for the bias pots. Once I got the tubes in, I can see where that would have made for a lot of tender fingers. Just as well to put them on the front panel. Only change there is now the power switch and light will move to the wood front of the chassis. There will also be a couple mini-switches between the meters so I can switch from front to back on each bank when reading or adjusting bias.

Speaking of driver tubes ...

I'm sure I posted these up earlier, but what's a few more pixels ...

My chassis plate was designed to recess the driver board with only the glass of the tubes showing. That required relocating the big caps (C13-C16) to the bottom of the board. That done, the tallest part of the board is the sockets themselves. Here's the finished product ...



One nice feature of the board is that both sides have solder tabs so you can flow thru the board and get real solid contact. This was a BIG advantage with swapping the caps to the other side. I just put a temporary riser under each so I got a bit of air space and could also inspect the joints for good flow both sides. Here's what it looks like from the bottom ...



I also had some fun with the PIO coupling caps, adding a heatshrink cover and custom printed labels to pretty them up. There's always a chance of shorting the standard bare metal cases to ground or worse, so there's actually a practical side to the mod.

Unfortunately, I already had the board built before I decided to replace the original board mount bias adjusters with standard pots. I'll have to pop those pretty PIO's to remove those and add twisted wire long enough to reach the new 50k pots on the front panel. Not a perfect world, and I'm here to keep it that way!

Speaking of not perfect. Somehow or other I missed the target when drilling the mounting holes for the VTA driver board. That put the glass envelopes right up against the back of the holes in the chassis plate. dOH! Easy fix though ... The board is mounted about an inch below the chassis plate and I was able to bend the bolts just a bit to center the tubes. Shhhh ... I won't tell anybody if you don't ...

More later ...

The rectifier is a NOS 1972 Philips 5R4GYS built in Holland. Same factory that did the Amperex Bugle Boys, so I have real high hopes for the sound. Parts Connexion was running a special on them for $40 - they usually go for around $55.

Only real difference in the spec is the voltage drop is a lot higher than the usual 5AR4 varieties. I'm told I may lose a couple watts output, but I don't see that as a problem. The KT-120 power tubes have loads of reserve in this amp and I may even come out ahead. Plus, I get the big bottle look and reserve of the GZ-37, at half the price.

Just to be on the safe side, I do plan to run a Weber WZ68 Copper Cap solid state rectifier when I first fire it up and until I verify the build. I've also got a Mullard 5AR4, an Amperex 5AR4, and a Shuguang 5AR4 here, so I'll be able to roll some and take notes. I'll also be doing the diode mod on the rectifier socket - better safe than sorry.

The TungSol KT-120's are also some real beasts - IF your transformer can handle the load - not a problem here. Reviews are that the bass is exceptionally crisp and strong, the midrange is real sweet, and the highs are detailed yet not shrill. Very comfortable listen across the range ... Who could ask for more? They can do 150 watts in a standard push/pull configuration if the amp is designed around them. Won't get anywhere near that with the ST-120 due to limits in the amp itself, but ... I'm thinking that's a good thing. Won't be driving them hard at all - at those levels they should have long happy lives and be impervious to anything I can throw at them. Bonus - they're actually cheaper than a premium tube like the ever popular SED Winged C ...

Yup ... I was reading your initial evaluation of the tubes and the results of your testing. I also remember you mentioning there was no discernible difference in sound quality with the higher settings, so ... back down to .55vdc ...

Speaking of bias ... I found some real nice meters. Ya, I know ... who needs em ... but ... Bling!

Most of the ones I found were either the wrong scale or hard to read. Found these at JameCo's eBay store and thought they looked perfect for the job.



Didn't expect much for $10 a pop, but they feel solid. I'll be putting test jacks in too, so I'll be able to verify their accuracy over time. I'll use one per bank and just add a mini SPDT for each to select the front or back tube on that bank.

That looks pretty slick! Nice job!

Nice job! You actually finished it! My problem is that I keep getting into mods and now I am re-designing a new improved power supply. Good thing I have lots of perf board to experiment with before I order the PCB's. Make sure you put up some pics of the wood base and finished amp.

I got a long ways to go before it's "finished", but I'll post up on any major changes.

Also ordered a back plate from FPE today. I was gonna cut that myself but they do a much nicer job.



The back of the case will be a wood panel with a cutout for this plate. Here's hoping I sized the holes right for the fuse and plug holders.

PS ... I'll be using RCA gold panel jacks for the input. If anyone's wondering why two sets, I currently drive my amps off a Sansui quad and use a loop adapter on the back of the amp.



I figure as long as it's a scratch build, put the loopback right in the chassis. I'll just bridge each set of jacks internally.

Oh. And I know what you mean about changing things. My job's easier as there's not many circuit changes to deal with, but as for the rest, I'm pretty much making it up as I go along. Got a good idea of what I want, just not exactly how it's gonna get done. That's the fun part!

OK then ... rainy day, so good time to dry fit the bottom. Let me know if I forgot anything.



Nice to have some more room to play with. Those big chokes look right at home there, and I won't even need to add any extra holes to mount them as they line up quite close to the output xformer mounting points. Not quite - I'll have to ream the holes a bit to handle the slight offset, and maybe add some washers under the nuts.

Also plenty of room for the can cap. I'll make a couple strap mounts for that and attach to the chassis using the output xformer mounting hardware.

Haven't drilled the holes for the controls on the front panel yet, but I'm pretty sure that's how it's gonna work. Two stepped attenuators wired as mono so I can adjust balance, and the annular power switch on the left. The quad of 50k pots is for bias, and I'll also have standard pin test jacks between the meters for a VOM hookup. I know I don't need two ground pin jacks - it's a symmetry kinda thang. Oh, and the black line is where the panel will bend for the front control surface. I'll get all the holes drilled before I tackle that.

While I'm at it ... the SCM's are a tight fit on the standard chassis, but they actually get sorta lost in this one.



Lots of room to grow them up ... any advantages to going larger than the standard pair of 330μF 400v? The chokes and large caps facing down on the driver board already dictate the final depth of the wooden base.

Re; >"any advantages to going larger than the standard pair of 330μF 400v? The chokes and large caps facing down on the driver board already dictate the final depth of the wooden base.

Those caps are in series so what you really have is half the capacitance (165 uF) at double the voltage rating (800 volts). I have added more capacitance to the VTA ST-120 and it makes no difference in the sound of the amp or it's ability to play louder at higher volume levels.

Bob