layout philosophy question

[AcePepper]

Hi All,
I've built lots of amps, but am a newbie to the "Dumble Culture";-) I'm working on a layout for a HRM. I've done many amps with the power supply/filter caps laid over the main board and these have worked well for me. See attached drawing. Neater layout, shorter wire runs, distributed filtering, easy grounding.

Is there any GOOD reason why all the Dumbles and clones have a seperate filter cap board? Does keeping the filters away from the signal parts keep noise down? Or does it matter? Is inductive capacitance a problem? Or something?;-) I'd like to do something a little different, layout-wise, unless there's a good reason not to or to do it the old way.

Also, what is the advantage of series 47uf filter caps vs. single 20uf caps? It seems like one could get better quality axial caps than the small radial caps. It also seems like lotsa little caps is more that could go wrong and harder to replace than larger axial types. Whatta y'all think?
Thanx!
Ace!8-)>

[heisthl]

Looks beautiful and I bet it would work great. Even "the master" would put a filter cap next to cl1 in some builds. I have always liked single board builds with minimal outboard components for ease of construction. You are probably right about the radial vs. axial caps but it sure is nice having those bleeder resistors when you have your hands in there.

[d95err]

Hi All,
Also, what is the advantage of series 47uf filter caps vs. single 20uf caps?

The main advantage is that you get twice the voltage rating, so you can use smaller and cheaper caps. The bleeder resistors ensure each cap sees exactly half the B+ voltage.

I guess it was more of an advantage in the old days when high-voltage, high-value, electrolytic caps were really large.

[swt]

maybe other guys will chime, but i've built clones with distributed capacitance, and they hum a lot...and then change it to the original gorunding schem and the hum was gone. Don't know why but happened lots of times.

[Pete]

Also, what is the advantage of series 47uf filter caps vs. single 20uf caps? It seems like one could get better quality axial caps than the small radial caps. It also seems like lotsa little caps is more that could go wrong and harder to replace than larger axial types.

I think we shouldn't overlook the cost factor. Even with double the number of caps used, with the lower voltage radial caps, wouldn't the cost be a bit less over all?

[Bob-I]

Personally I like the filters on a seperate board but I see no reason why it cannot be all on one board. I also like to use a single 22uF 450V cap instead of the 2 47uF caps shown on most Dumbles, fewer parts to deal with.

I don't like distributed filters. IMHO it allows more noise to be generated by sending less filtered signals across the board.

[ampdoc1]

Why not the best of both? Put all your caps on the end of the board and add an additional row of "holes" to your board width for the DC distribution. I built a couple of TWrecks like this several years ago using a VERY high quality PCB, but it could be done easily enough with turrets or eyelets.

ampdoc

['67_Plexi]

maybe other guys will chime, but i've built clones with distributed capacitance, and they hum a lot...and then change it to the original gorunding schem and the hum was gone. Don't know why but happened lots of times.

Using series caps has never caused a hum in any amp I've ever built. There's no legitimate technical reason it should.

The reason they are used was stated previously, basically because of the physical size of the larger value caps. It goes back to 60's.

Be very careful of the voltage rating of caps you use:
Let's say a cap in a normal single cap filtering stage goes bad and shorts (they can do) the worst that would happen is the fuse would blow. Now lets say you have 2 caps in series, but you choose a lower voltage rating to keep the size down. If one of the caps goes short, you then end up with an overvoltage situation on the other cap, up to 2x. An electrolytic with a serious overvoltage will get hot and possibly self destruct.
As an apprentice (many moons ago) I was taught to always voltage spec. capacitors so the working voltage spec. exceeded the normal terminal voltage to the lowest potential in the equipment (generally ground). You have have full HT to ground on the positive terminal of first cap.
Standards and guidances such as UL6500 are very stringent in areas of fault tolerance and reducing potential fire and personal injury hazards.

If you are builing amps for yourself, you carry the risk of hurting yourself, but if you sell them, you need to be aware that there's way more to this than just building a good sounding clone. Most of us live in the US, the land of the lawsuits

Alan.

[drz400]

maybe other guys will chime, but i've built clones with distributed capacitance, and they hum a lot...and then change it to the original gorunding schem and the hum was gone. Don't know why but happened lots of times.

There are advantages to using multiple parallel caps but not series caps except for increasing voltage rating

The ground path must be looked at as a whole, Grounds are usually the last thing amp designers learn about and it should be the first, hums wouldnt be such a mystery. Absolutely nothing wrong with local filter caps to the stages they supply. In fact Technically this is the correct way to do it. It keeps the flow of the current at the stage it is filtering.

Distributed capacitance is quieter if we are talking the same thing, one filtered supply feeds the next so they are in series instead of all parrallel off one point. Just put your scope on the supplys at each point and see the ripple drop. By the time you are at the most important stage #1 the DC is nice and pure where it really counts. The only problem with having distributed supply is you can run out of voltage. This might not be what you mean though since the Dumbles are distributed, He just clumps all the caps together, Why?? Because the size of the caps used back then made it difficult to do it any other way, The same reason the used caps in series to get the higher voltages. These days the caps are so small you can put them where they belong to keep wire runs short. They belong next to the stage they are decoupling. There are other ways it is done but usually from a "that is the way it was done before" mentality.

['67_Plexi]

I think the term 'distributed capacitance' was used incorretly in the beginning of this thread. I think they were talking about using two series to 'distribute' the voltage across two caps in a stage.

To me distributed capacitance refers to the distribution of filtering throughout all stages of the amp.

I totally agree on your comments regarding grounding and keeping the HT loops as short as possible is totally on the mark, especially the higher current power stages.

['67_Plexi]

There are advantages to using multiple parallel caps but not series caps except for increasing voltage rating

There's no positive advantage to using series caps to increase the voltage rating. As I stated in my message above it's basically regarded as a bad design practice these days. It was a little different back in the 60's.....safety wasn't at the top of the list, you know with the two prong cords and a polarity switch with a single cap protecting you from getting zapped and all that.

[drz400]

There are advantages to using multiple parallel caps but not series caps except for increasing voltage rating

There's no positive advantage to using series caps to increase the voltage rating. As I stated in my message above it's basically regarded as a bad design practice these days. It was a little different back in the 60's.....safety wasn't at the top of the list, you know with the two prong cords and a polarity switch with a single cap protecting you from getting zapped and all that.

Yes I agree it isnt great practice but when someone cant get a hold of a 500V+ cap it can be the only way. If a filter cap shorts to ground sure it will take out the other cap too but even if you dont have them in series and have a 500V cap when needed sparks are going to fly if a cap shorts out, hopefully your fuse will protect the PT. There are advantages to using parrallel caps to get the same uf like 2-47uf instead of 1-100uf

[KT66]

FWIW, increasing the voltage rating by placing caps in series will extend their life. From Aiken's site: Capacitor life is directly proportional to the ratio of rated voltage/operating voltage. If you run a 500V cap that has a specified lifetime of 10,000 hours at 250V, it will instead last 20,000 hours. If you run it at 500V as specified, however, it will only last 10,000 hours, which is what it is rated for. So, as you see, there is no benefit to running a capacitor right at it's rated voltage, it is better to run them conservatively with respect to voltage and temperature for maximum lifespan.