Tek 501 power supply

[Grandchuck]

Two Tek 501 power supplies were purchased at a Hamfest a while back.  One worked and one did not.  The one that didn't had almost no output voltage and the current limit LED was on.  C55 was shorted (see the schematic).  This was easy to find with the help of Shannon tweezers.  The original electrolytic was of a style I have not encountered before.  It is in a shiny, hard case.  The new capacitor is of the familiar style.

Now it is time to clean the other supply as it's rather grungy.  Both units easily meet all of their specifications.

[Gyro]

That's a Tantalum bead capacitor. They're pretty notorious for failing in older gear, particularly when it hasn't been powered up for a while or the capacitor is run too close to it's voltage rating.

You're lucky that the current limit cut in, because in the wrong situation, they can make a nasty burnt mess of the PCB. They can normally be replaced by decent quality electrolytics these days, as seen on the other unit.

[Nikola Tesla Junior IV]

Tantalum capacitors are electrolytic capacitors. Some Wet, Some dry, using Tantalum. 

The Tech's at TEK weren't stupid. They used those capacitors for their frequency characteristics, stability, and longevity.

They also provide lower ESR (equivalent series resistance) than AL electrolytics as well as a higher operating temperature capability and lower leakage.

AL electrolytics are good for HIGH Capacitance, some say high Voltage - It's debatable, other then that they are used as an economic version that just does enough to get the job done.

If your Smoothing, Cleaning a Power supply, or dealing with Audio, stick with the Tantalum.

My 2 ¢

[Grandchuck]

Tantalum capacitors are electrolytic capacitors. Some Wet, Some dry, using Tantalum. 

The Tech's at TEK weren't stupid. They used those capacitors for their frequency characteristics, stability, and longevity.
[end quote]


But, they used a part rated at 20 volts in circuit that is expected to normally reach 20 volts?

[bdunham7]

The Tech's at TEK weren't stupid. They used those capacitors for their frequency characteristics, stability, and longevity.

In some cases tantalum beads were used specifically for their unique characteristics.  In many more they were used back then because they were small, readily available and thought of as longer lasting.  It was also thought--and there was research to 'prove' it--that they could be run at 100% rated voltage with virtually infinite lifetime.  Unfortunately their most common failure mode is a black-hole short, unlike electrocaps that fade away slowly.  In many cases--not all--you can sub in electrolytics and in some of those cases they may actually be better. 

Also, based on some previous posts here and other research, apparently the weak point of tantalum beads is that they can be damaged by more than ~2% of rated voltage in reverse.  So if you have turn-on surges and things that are limited to one diode junction (as in this case) then you don't want to ever use a tantalum bead capacitor with less than a 35V rating.  Now when I'm doing any sort of repairs, I more or less heed this.  I either use a 35V or 50V part if I can.  If it is a large value, I will switch to an electrolytic if I think it is reasonable based on the circuit.  For power supply decoupling with something like a 100uF capacitor, I'm pretty sure there's an electrolytic that will work.  I haven't had any substitutions turn out poorly yet.

[Grandchuck]

The replacement part used is a 100 uF electroytic (35 volts) and the unit exceeds all of its specifictions.

[Nikola Tesla Junior IV]

Well, without writing a novel about theory of operation, after departure of OP-Amp U40, the energy is stable expressing little to no spikes. C55 100uf20V should not see more then 19.99xxV. In theory it can run forever at 20V as stated above, and has been tested - This is correct.

However-

Bad quality Caps - If rated voltage of tantalum cap, is too low,  if high current applied to the capacitor for a long time, if it is NOT of quality, and the insulation resistance becomes low, they can begin to leak at an increasing rate and with high current, the actual withstand voltage capacity will diminish, effectively halving or more, the voltage rating/capacity. Low quality tantalum capacitors have no real ability above room temperature to actually withstand the rated voltage and when they reach higher temps, the voltage rating will rapidly drop - thermal derating.

IF that isn't the #1 reason they fail/failed, then #2 would be:

High AC ripples. If there is a large enough AC ripple, it CAN and WILL cause a breakdown within the Capacitor even at Voltages well BELOW the cap's rated value. IF the ESR of the cap is not the proper value to deal with AC Ripple, or Ripple is to high, this will occur.

As with most electrolytic caps, the tant cap has polarities.. which generate heat from AC Ripple, which can cause thermal derating / thermal breakdown.

Most people don't know this.

There are no benefits to using AL electrolytic cap, other then the cost effectiveness, easy to find/get, or high capacitance/size.  Electrolytic capacitors are just that. The difference remains within the materials used to create it.

Based on real data, If a high quality 20V Tant Cap was used without being exposed to AC Ripple (Properly working Mainframe,Scope,Lines) that had matching ESR characteristics for the tested 50Hz in house power, then it would in fact be ideal for the 501 Power supply.


 

[Nikola Tesla Junior IV]

The replacement part used is a 100 uF electroytic (35 volts) and the unit exceeds all of its specifictions.

Will do the job fine, I'm quite certain.   However there IS a rhyme and a reason to things, and there is a necessity for it to be understood.  Often times, it is "throw a higher rated AL-E cap at it", rather then understanding ESR values, Energy flow as it pertains to materials and voltage derating, and the characteristics of such as they work in different types of circuits from switching, rectification or coupling.

[bdunham7]

There are no benefits to using AL electrolytic cap, other then the cost effectiveness, easy to find/get, or high capacitance/size.

I think that both the ability to tolerate a moderate amount of transient reverse voltage and the lack of a common failure mode of a sudden dead short certainly qualify as 'benefits'.

[Nikola Tesla Junior IV]

There are no benefits to using AL electrolytic cap, other then the cost effectiveness, easy to find/get, or high capacitance/size.

I think that both the ability to tolerate a moderate amount of transient reverse voltage and the lack of a common failure mode of a sudden dead short certainly qualify as 'benefits'.

I think you misunderstand that Tantalum Capacitors ARE polarized electrolytic capacitors, that we are talking differences in material costs, size, and that both will behave much in the same way. AC ripple effects polarized capacitors, period. Transient reverse voltage? Typically a large size AL-cap will endure slightly more AC ripple purely due to the size and spacing of materials. Other then that they fail in the same way and Aluminum electrolytic capacitors can fail short as well, perhaps not as often due to the fact they shoot their acidic guts all over the PCB traces, MC's and everything else that is important and hard to find.

These things are why you hear everyone tell you to replace the AL-E caps, right off the bat, whether good or bad (I do not agree with this).

Of the Tantalum capacitors and Aluminum capacitors I have seen fail, the Aluminum ones are much more devastating despite the non-sense you read online about Tantalum capacitors starting fires, sure it can happen.

I have a definitively larger pile of ruins from acidic guts of AL-E caps then I do from Tant failures.

What is the current condition of C55? Shorted and fine (yes, not always the case)

[bdunham7]

I think you misunderstand that Tantalum Capacitors ARE polarized electrolytic capacitors

No, I didn't misunderstand that at all, otherwise I wouldn't be mentioning reverse voltage and such...

Of the Tantalum capacitors and Aluminum capacitors I have seen fail, the Aluminum ones are much more devastating despite the non-sense you read online about Tantalum capacitors starting fires, sure it can happen.

I have a definitively larger pile of ruins from acidic guts of AL-E caps then I do from Tant failures.

What is the current condition of C55? Shorted and fine (yes, not always the case)

I've repaired numerous burned traces and  a few really toasted power supply boards due to shorted tantalums.  Haven't burned down any buildings so far but there sure was smoke.   When they don't blow something up it is because the circuit can tolerate the short without damage, as in this case.  I can't recall repairing anything recently with leaking electrolytics, typically they seem to dry up if they fail.  And I've yet to see one fail short although sometimes they blow up nicely if a rectifier fails.  YMMV.

[PwrElectronics]

Looking at that schematic, red flags!!

1.  Cap is across power supply rail.  The output can go to 20V and its a 20V cap.  At my employer, design rules would require a 50% voltage margin on tantalum caps.  We do this no matter where they are.  Too many failures.  I had one on a internal regulator (not going off board) on a product that was 12V and it was something like a 25V rated cap and used to get units back from the field occasionally still.

2.  Circling back again to the voltage rating.   A type of cap known to not tolerate any overvoltage or surge is on a "naked" power supply output.  We lose lab supplies all the time at work from inductive transients in the load spiking the supply output and blowing the regulator.  Killed one myself a few years ago trying to duplicate some failures the customer was seeing on their vehicle that involved loose battery cable connections.

Right now, our design rules for things like 5v sensor supplies that go off board into the wiring harness cannot have a Tantalum cap as the regulator filter as any spikes, etc making it back and past the ESD or other transient protection have caused too many field failures.