Trainwreck power supply/filter cap board

[Raoul Duke]

Question:

Has anyone on TAG ever tried to design a power supply/filter cap board to replace the cap stack in an Express or Liverpool? The bias and pre-amp boards seem well established and pretty practical - but I ask about the cap configuration for two reasons:

1. I’ve started to research what’s available in snap-ins and radials to try and model something in DIYLC
2. The reason for this is that I’m starting to plan out a Liverpool build of my own - and this is probably the one change I’d want to make for practicality reasons (ease of build, component selection, future service etc).

I realize it’s not considered true to Ken Fischer’s original designs; but I think this might be a worthwhile pursuit - perhaps one that he might even have arrived at himself eventually given his pragmatic approach to designing and building. I’ve searched here and found nothing - but given all the innovative and wicked smaht folks here - I wanted to ask before I start to embarrass myself with some rudimentary ideas for this, lol.

I’ve got a few other practical questions for those who are experienced with this design - but will consolidate those into an RFI-type post once I try to find answers on my own first.

Thoughts?

[RockinRocket]

I don't see why you couldn't make a board or chassis mount them but really the Mallory or modern equivalent Spurge Atom cap stack fit pretty snug in a 2" tall chassis. Glue them in there and wire them up with solid core wire they aint going anywhere. I really don't know what all the fuss is about when I read about how unsafe it is. Its not.

[M Fowler]

I used to build Rockets with two of the filter caps incorporated in the circuit on the tone board and punched chassis hole for 50/50 can cap or put remaing caps on small board.

Mark

[Roe]

Like Larry, I've placed filter caps together with the plate resistors on the preamp board

[johnnyreece]

Like Larry, I've placed filter caps together with the plate resistors on the preamp board

Yeah, I tend to like this method more nowadays. I have also built a small cap board for the stack, as well.

[Raoul Duke]

Understood and thanks for the replies all.

I’m not concerned about the safety of it really. I agree with Rocket that the glue and solid core would keep them still. I was more looking for something that was more practical/efficient/simpler from a service/component selection view.

I’ve seen a number of builds where some or all of the caps end up on the main board and haven’t ruled that out yet either. Just trying to figure out the simplest, most intuitive way to approach this.

I happened to discover Martin’s Express design from a number of years ago and that’s got me thinking about “form follows function” as well. Really interesting.

[solderhead]

Question:

Has anyone on TAG ever tried to design a power supply/filter cap board to replace the cap stack in an Express or Liverpool? The bias and pre-amp boards seem well established and pretty practical - but I ask about the cap configuration for two reasons:

Some people want to build amps in exactly the same fashion that KF built them as an homage to his designs -- others do it in the belief that there's something magic about his layouts and that the amp won't work properly unless every minute detail is copied from the original. The truth is that cloning his layouts is not strictly necessary if what you want is to have an amp that works like it should.

Although many people like to build their amps in the way that KF would have built them 40 years ago, I think it's a wise decision on your part to look toward the future in adapting the designs of the past -- especially if you're going to embrace the snap-in capacitor format as snap-in caps really call for PCB designs. The writing has been on the wall for new designs favoring snap-in caps for decades now. Axial caps are getting harder and harder to find with good specs (if you can even get detailed spec sheets on them) and they are priced at a premium to the levels of performance they offer. I've given up on them and I've been designing my own boards around snap-ins for decades now. The added benefit of designing your own PSU board for your amp is that there isn't a lot of variation in PSU design layouts if you're using cap input pi filters -- all amps that use that type of PSU topology are similar enough that they could use the same board. The benefit of that is that when you do the design work for a PSU board, you can leverage that design in all of your builds that follow. If you're willing to break free of convention it makes sense to do the things that will make your life easier over the long haul.

[Raoul Duke]

Ok Folks,
Here's my first stab at using DIYLC to invent something. I used actual part #s and specs/dimensions and tried to pay attention to wire runs and other parts' locations. I realize there's still some opportunity to combine the PS/Bias board into this design as well; but was trying to fall in the middle of "traditional build" and completely reworked. All caps are 500v.

I may actually build this onto the pre-amp board rather than a separate board.

TW-L Cap Board.JPG

Constructive criticism is welcome.
Thanks!

[solderhead]

I am not familiar with DIYLC, so I don't know what kind of features if offers you. Does it only allow you to draw pictures of layouts, or does it allow you to include traces? I'm guessing it's more for organizing parts on a layout rather than formal PCB design.

I've noticed that you've kept the grounds separate for each B+ node, and that you've made separate eylets (or pads) for each node's power and ground connections. Is it your intent to physically put all of the caps on one PSU board, while maintaining electrical isolation (keeping each cap's ground isolated) so that you can run separate B+ and grounding wires to the individual locations for each stage on your preamp board?

Since it looks like you're planning on using snap-in caps with the 10mm lead spacing, pay attention to the PCB thickness when you finally get around to finalizing your board and mounting things. Snap-in caps work best with some PCB thicknesses, and not at all well with others. On those occasions where I've tried to use snap-ins on boards that are too thick, I've had to counterbore a mounting spot for them to reduce the net thickness to a point where the leads were able to establish a good "snap" fit. IME 1/8" G-10 is too thick and 1/16" is what you want.

[Raoul Duke]

I’m new to DIYLC; but I think you’re correct that it’s mostly for organizing layouts. I’m still learning though.

Yes, it is my intention to keep the caps on one board and I separated the grounds for the purpose you’ve identified mainly because it’s worked well for me in previous builds - but I’m always open to suggestions and learning new ways to skin old cats.

Thanks for the tip on board thickness - excellent practical info.

This will be my second project using snap-ins. I made a power supply board with snap-ins and radials last year for a Dumble BM/HRM using Martin’s excellent plans and have to say it worked out perfectly. His “precision power supply” as it’s known is my inspiration for trying it with this build.

Thanks again for the feedback!

[johnnyreece]

I'm a big fan of DIYLC. Yes, as the name indicates, it's mostly for layouts, however, I also now use it for schematics (there is a schematic symbol library).

[dorrisant]

I designed a board for this using axials. Not the snap-ins you mentioned, but still easy enough to install and repair...

https://ampgarage.com/forum/viewtopic.p ... ss#p454738

I could find the drawings in my files and post if needed.

[Raoul Duke]

Ahhh, yes; I see what you did there. Very nice, tidy combination of using the axials in a more organized and serviceable manner. Makes perfect sense and refines the original implementation very nicely.

Thanks for the pics, I think I can extrapolate the dimensions close enough from those. I’m still on the fence about which way to go. Seems both MOD and JJ make the values in axial and are easy enough to find…

I’m going to go read the thread right now. Very much appreciated!

[M Fowler]

I did various caps boards too, nice job Tony

The two main filter caps are under the cap board.

Mark

Express with Rocking Class Pacific transformers 010.JPG

[solderhead]

I really appreciate the craftsmanship that goes into some of the boards/mounts. The attention to detail on many of these builds is very impressive.

as an engineer I'm going to ask a silly question:

Why does everyone use the ridiculously oversized cement resistor, and wirewound vitreous enamel resistors? Can anyone explain the reason for these design choices, or is that just part of the mystical secret sauce that gets copied over from the original amp builds just because that's the way that KF built a specific amp? I'm not saying that doing any of that is wrong and I'm not trying to belittle the decisions -- I'm just trying to understand why some of the non-conventional choices were made.

It's interesting that the amp was supposedly designed with the specific intent to avoid the inductance provided by a choke, but then inductive resistors get added into the mix -- big ones -- with ridiculous power ratings.

I'd like to be clear that I'm not criticizing -- I'm just trying to understand some of the design elements and choices that aren't blatantly obvious. If anyone can explain some of these odd choices, I'd really appreciate it. If anyone's got math to support these choices, that would be even better.