Wednesday, December 29, 2010

Recapped the B+ power supply of my 1973/4 Fender Princeton last night. This is the amp that had sat in a shed for a while and sounded anemic as hell in a raspy farting through a cork wall sort of way. The main problem was not a failing speaker as the seller thought, but a lifted ground reference in the phase inverter section that proved a simple fix. The amp was far from done, needing the logical next step: removal and replacement of 37 year old electrolytics.


Since the overall amp is so compact, I faceplanted the cabinet and used the upper rear baffle railing as my chassis stand.


Here's the original equipment, a Mallory axial 20MF hailing from the 43rd week of 1973 and a Mallory 40-20-20 multi-section can cap built in the 29th week of 1973 (you'll just have to trust me on that one).


Can caps that tie to the chassis require a nice solid structural solder joint that completes ground and secures the cap to the amp. For this, liberal application of heat is a thousand times better than trying to remove and replace a chassis pool with an underpowered or small iron. Here's the 100 watt Weller I use for chassis work in comparison to the variable temp iron I use for all other component grade work.


Of course, with a behemoth hot iron, all of its benefits in this single application are detrimental in every other way. Shifting of existing components that would otherwise be untouched is crucial; I've smelled enough melting plastic already, thanks.


This tab got the desoldering treatment, I'm about to bend the tab up and free this joint. The joint on the opposite end proved easier to just heat and rotate the can, shifting that tab out of the pool. Then it's just a matter of bending the remaining tabs up and dropping the old can out.


Ancient (remember, electronic eras pass faster than those of humans) solder flux reminds me of cerumen, and it too is generally undesirable when built up and can take multiple passes to get clean. Above shot is after a single scrub/soak pass of isopropyl, and yes the good people at Fender put that flux there in the first place. I have a few cotton swab heads full of the stuff if any collectors are interested.


Now it's ready for a new cap. I'm not concerned over the existing solder there and think of it as tinning for the new joint.


The replacement multisection is built by CE Manufacturing in the US, purportedly off the original Mallory tooling. In comparing parts side by side I find that very easy to believe, fitment was EXACT.


Somewhat hard to see, but the pull job here has a bulge in the center vent.

I bought the caps in reference to the schematic as opposed to referencing the actual amp, which is why the replacement multisection is a 20-20-20 as opposed to the 40-20-20 that was original. Frankly, I'm not overly concerned at the 20% drop in total capacitance of the power supply. It seems as if manufacturing tolerances of electrolytics tend to over build as opposed to under build, and I've heard people bellyaching about adverse effects that over capacitance has on tone. Seriously though, retrofitting a lower capacitance to follow the mid 1960s circuit isn't a life or death matter.


Here it is, ready to burn in slowly on a variac. I'm not in love with the recreation of the original joint at the head of the axial cap, but it seems solid enough and the other did last for 37 years just fine.. I only addressed the B+ caps for now as this is my amp and I expect to be in and out of it for a while; tinkering with the guts and getting it just right to my ear, so COMPLETE overhaul is less of a pressing issue. Once I've insured a solid electrical base, next step is to rotate through a pile of 10 inch speakers to find the best match for my practice and recording needs.

2 comments:

pearshapedhuman said...

It looks like that internal Mallory was added later to beef up filtering in the pre-amp gain and recovery tube. Harmless other than the 6 bucks it probably cost you.

crochambeau said...

Hmmm, that didn't occur to me, I've never been inside another Princeton. The similar date code led me to believe it was original, but the set up does seem a little funky. Thanks for the input!