NOS Mallory TC-72 a bad idea for my La2a?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
looks pretty cool to me, Mallory made good stuff, if the caps in a 1964 twin reverb test ok and the amp sounds killer and the pwr trans ain't drippin then i leave them in, especially since sprague atoms are 15 bucks a pop, who wants to pay 125 bucks for a measly ol cap job? (75 parts, 50 labor) especially if they need a quad of Tungsol 6L6GC;s.
they got any more of those caps on evilbay? i tried poly and lytic in the LA2a and could hear no difference.

i use a meter from an old voltmeter for leakage, no series resistance, 1 ma is full scale, pre charge the caps so you don't smoke the meter, or turn the supply up reaaaaaaaaalllllllllllllll slllllllllloowwwww.

oh yeah, and one other thing,

mr putin,
























































Your Gonna Die!
 
Last edited:
Why on earth would you use an old capacitor when these days you can get better caps? Would you fit your 2020 car with 1950 tyres?
Leo Fender was confounded as to why anyone would want to make a guitar amp distort on purpose. Musicians and design engineers just think differently from one another. For some musical objectives, “better” isn’t better.

That said, I don’t attribute much magic to this cap in this circuit. But I do think it’s sensible enough for someone to want to evaluate and learn what, if any, effect could result from using historical vs. modern components in a historic device.
 
That said, I don’t attribute much magic to this cap in this circuit. But I do think it’s sensible enough for someone to want to evaluate and learn what, if any, effect could result from using historical vs. modern components in a historic device.
That’s exactly it - my understanding, from reading the exhaustive ‘all things La2a‘ thread and others is that ‘the sound is primarily affected by: the transformers, 12ax7 in V1, 12bh7 in V2, and this Mallory / Sprague cap in C5. So I’m curious, does the mallory make any/much difference - but obviously wanted to check I wasn’t going to screw anything up because it’s an old electrolytic

it’s not a 2020 car with 1950 tyres - it’s a 1960 car with 1960 tyres (if they’re safe). The 2020’s car is what I’m going to build with the next chassis.

(thanks for all the other cap info Soapfoot - lots to go through there. And to everyone else who has weighed in).

Time to test it for leakage...
 
ok, so I’ve tested it for leakage following abbeyroad’s instructions earlier in this thread - just in case anyone’s curious how it turned out...

I tacked a couple of wires to V2 pin 1 and one of the ground points in the chassis, then the cap and 1k carbon comp (didn’t have any metal film 1k) to the wires. Just giving methodology in case any of it has a bearing on the results...

888AF992-9045-403E-8A10-0A4088F53B48.jpeg

Firing it up, initially it very briefly read just over 1V DC before quickly going down. After about 5-10 minutes it was 0.097 vDC. it continued going down. After an hour or so 0.037. Another couple of hours 0.025. It’s been 5 hours now and seems to have pretty much settled around 0.010 v DC possibly still going down. I‘ve also tried turning the La2a off, discharged the cap, turned it back on and it very quickly settles back down at that value

A modern electrolytic cap should have less than about 100uA leakage. OTOH a film cap would probably have many times less.

if I understand abbeyroad correctly that gives me leakage of around 10 uA (that’s from 98v on pin 1) - would that be considered acceptable to make it worth installing? am I testing incorrectly?
 
I am not a fan of using old electrolytic capacitors for all the obvious reasons.

If your bench testing confirms that it is useable have at it, the worst case is you have to replace it again.

Measuring +50% capacitance sounds questionable.

JR
 
I think you’re fine to evaluate it

just bear in mind that much of any distinction you’ll hear will be due to the high out of spec value
 
mainly because I hadn’t seen any Mallory TC-72s that looked like the one I have, only blue metal can ones, I went googling...

probably just trying to talk myself into it at this point and, as someone said on page 1, making a way bigger deal than is no doubt warranted but found mine in a 1968 catalogue

5583045A-9D9F-4E45-9429-CE986140778B.png2E211F63-184A-400B-9582-B45031EC692A.jpegFD201A65-1B1B-4ABE-8340-0127F849088B.jpeg

Measuring +50% capacitance sounds questionable.

probably still not a brilliant idea, but if I’m understanding this catalogue correctly at 14.5uF, this is actually still in spec!
 
Last edited:
I am not a fan of using old electrolytic capacitors for all the obvious reasons.

If your bench testing confirms that it is useable have at it, the worst case is you have to replace it again.

Measuring +50% capacitance sounds questionable.

JR
Agree-I am very suspicious of ancient capacitors measuring significantly higher than high range of tolerance (after ruling out component doesn't have +75/-10% tolerance). I have found ancient electrolytics in power supplies that obviously had a ripple problem, but measured much higher than marked. A 2nd capacitance 'meter' and a 3rd had drastically different measurements. One said it was shorted, but an ohmmeter didn't, the other read almost as high as the 1st.

I put it on an impedance analyzer at 120 Hz and it measured some inexplicable positive phase angle that made no sense...so I declared it to no longer be a capacitor. Then I double-checked that on another impedance analyzer that agreed with the previous one.

Handheld digital C-meters seem to be fine for healthy capacitors (that are not still on circuit boards). I argued with someone once about checking a capacitor supplier's parts in-circuit because they thought they were out of tolerance. I understood not wanting to lift leads from several boards to measure them, but I showed them three different digital C-meters gave three significantly different results, because of auto-frequency decisions in their software.

I don't know what the resolution of the problem was, but someone got stung by submitting data from out-of-calibration equipment (for a non-standard test, anyway). A case of arguing with idiots...the measurements were more meaningless than calibration certification.
 
don't be suspicious, the caps i replace on Marantz receivers are always about 50% over their spec, most from around 1975. rarely have i found a bad lytic and i have a garbage can full of pulls. not a wastepaper basket, i mean a small garbage can. Thousands.
Elna is the best as far as beating spec, I even chop them in half with the dikes and there is always a ton of juice left. there is a lot of lytic paranoia out there. why? because everybody knows there is liquid inside, so everybody thinks that liquid surely must be dehydrated.
the few dried out caps that i have seen were always installed next to a power resistor or power transistor by some genius engineer.


the only thing worse than lytic paranoia is the "be sure to bypass every chip with a 0.1 cap on the supply rails" God only knows how much time and money this has cost the industry.
one of those "just in case" things for which i have yet to see a single case where that was necessary. add up those tants. billions and billions. and they all need thru holes or pads, and they all need a "pick and place" cycle of action. And, and don't forget the touch up or replacement of parts that floated out of the board during wave solder. rant off.

and be it advised that under the cardboard of the Mallory LA2a cap is a metal canister which probably contains the same amount of juice it started with.
 
Last edited:
I am just a design engineer so when in doubt I read the data sheet.... I found a crusty old image of a Mallory spec sheet,
2020022219.pdf

if I read it correctly (not guaranteed) it appears to say -10% +75% , so +50% is well within spec.

Carry on...

JR
 
don't be suspicious, the caps i replace on Marantz receivers are always about 50% over their spec, most from around 1975. rarely have i found a bad lytic and i have a garbage can full of pulls. not a wastepaper basket, i mean a small garbage can. Thousands.
Elna is the best as far as beating spec, I even chop them in half with the dikes and there is always a ton of juice left. there is a lot of lytic paranoia out there. why? because everybody knows there is liquid inside, so everybody thinks that liquid surely must be dehydrated.
the few dried out caps that i have seen were always installed next to a power resistor or power transistor by some genius engineer.


the only thing worse than lytic paranoia is the "be sure to bypass every chip with a 0.1 cap on the supply rails" God only knows how much time and money this has cost the industry.
one of those "just in case" things for which i have yet to see a single case where that was necessary. add up those tants. billions and billions. and they all need thru holes or pads, and they all need a "pick and place" cycle of action. And, and don't forget the touch up or replacement of parts that floated out of the board during wave solder. rant off.

and be it advised that under the cardboard of the Mallory LA2a cap is a metal canister which probably contains the same amount of juice it started with.
Tiny SMD collateral damage (loss) by hot air removal of large target components... :O(. Somebody has to check because the rework folks often don't realize what else was or wasn't there.
 
Sometimes I replace all the electrolytics of the same age if I find one bloated or leaking in an SMPS (post-hf circuitry). Sometimes I do it if there is no visible degradation but there is major malfunction and I have no docs. More often than not (but admittedly not always) SMPS and motor drivers have returned to life.

Or I remove mains rectifier filters (snap-ins) to check and have found occasionally essentially-open parts (470 uF was 50 nF). Two in particular (same kind of SMPS) had 2000 hr/85C parts, with only 2 of 5 pins used by (poor?) design (less heat conduction capability with 2 vs all pins). The datasheet (Hyundai) also showed that value was larger than standard offerings in that can size...apparently a custom for size issues. That one (twice) was obvious, but everything else was rated 105 C at much higher hours, and more work to deal with (SMT and stacked boards, & no docs). I could not find equal or better replacements, so I rationalized the better quality components should have much longer lifespan on the basis of high hour and temp. ratings and looked OK...so no PM replacement was done on the other SMD parts.

Lucky hunch, and while I was initially offended that two systems failed the same week, I estimated they were operating 24/7 for 9 years and changed my mind I was impressed they lasted that many hours (with no idea how high internal temperature got). That kind of anecdotal observation isn't statistically meaningful, but contributes to forming opinions about age vs. labor and reality of components not being intended to last forever. That SMPS was part of a modular subassembly that wasn't serviced by the mfr. - they recommended replacement of the PS & fan ass'y (almost US$3000) to speed return-to-service. There was no spare purchased. It was mechanically very complex and I had to make my own notes to get it apart & back together. The original capacitor wasn't available. I found one option for replacement in stock and it was 3000 Hr /105C with higher ripple current rating and only had 3 snap-in pins by design. I'm hoping I'll be retired before I ever see it again.

A third unit failed later but that dept. didn't want downtime and I told them my reliability gambling logic for the repair...so they ordered a replacement unit.

There could be other component parameters that degrade (ESD, for one), so just capacitance and leakage checking aren't the whole picture.

At home, labor cost or downtime are just annoyances, not expenses. There are so many other situations where money is poorly spent I don't worry about the cost of low-dollar parts sometimes. I guess that falls under just-in-case or for confidence category. Sometimes something doesn't seem to merit the effort but there's a desire to give something a 2nd chance.

I used to have a co-worker who would break open (solvent-welded without screws) plastic SMPS power cubes that no longer functioned, to replace electrolytics. I thought that was crazy but did a couple myself for the challenge. Sometimes there's a weird one with multiple outputs and a special connector and you can't find a replacement. That thankless task sometimes saves the day but I sure try to talk my way out of some jobs.

When it's your money, you can rationalize what to put the effort into. I don't know how many dealer car repairs included things that were 'also needed'...that industry doesn't seem to like admitting incorrect diagnostics or repair attempts...customer usually pays :O(.

When I do decide to replace something and the original is no longer mfd., I match any critical mechanical dimensions (like PCB hole diameter, spacing and can height), then look for equal or better ESR, ripple current rating and lifetime/temperature ratings. Since I can't predict what parameter isn't important in something I'm just repairing, and frequently technology has improved, I feel like there may be benefits.

Those old Computer-Grade screw-terminal capacitors do last an amazingly long time and can be pretty hard to find or have minimum order/lead-time challenges. Someone (physicist/engineer) gave me advice to also not (when possible) substitute a physically substantially smaller electrolytic in high current situations because the thermal mass of the larger device plays a role in its dissipation of internal heat from losses (what causes electrolytic aging).
 
Last edited:
Back
Top