Enlight HPC-250-101 is aging

[ckrtech]

I just rebuilt a Pentium III machine. I use my bench ATX power supply whenever I work on a system, and then I turn my attention to the build's PSU before integrating it. This is an Enlight HPC-250-101 as mentioned in the build quality pictorial thread here - https://www.badcaps.net/forum/showpo...postcount=1644, but it appears to be a different revision.

I have some interesting findings here. I have only tested with the tester. I have a scope, but I haven't use it with a PSU. I also don't own an ESR meter.

I know that the PSU tester is a bit of a "well, at least I've got something to use for testing" device - but the +12V reads 11.5 with basically/almost no load. -12V only showed LL. Adding a peripheral molex connector got 11 to show under the -12V (pic below). I don't know how trustworthy the PG is on this device. I will say that this power supply is basically an AT power supply with an ATX connector. -5V is available, and there is no on/off switch on the back.

My knee-jerk reaction upon seeing these numbers on the tester and the TEAPO and G-Luxon caps inside is "aah. whatever. Re-cap it."

What is more interesting is the glue. You can see instances of the glue starting to turn brown in places - check out the primary side as well as the coil in the center of "capacitor city" on the output side.

I am tempted to re-cap and DE-glue (and re-glue?). I would probably leave the two primary caps (Rubycon CE 470uf) but still remove the glue around them. What are your thoughts? Also - any other components worth replacing/upgrading? Tough to get numbers on the components - there is glue everywhere.

[Quaddro]

Re: Enlight HPC-250-101 is aging

Old Enlight usually use delta as the oem.
and old delta, somehow, is always low in 12v when tested with no load.
As a group regulated characteristic, you need give a load to 5v rail, to raise the 12v rail.

what about -12v rail?
Ignore it.
Any modern system not use this rail.

Glue after very long time can be a conductor.
Well, but i never found any dead psu because of it.

Upgrade suggestion.
Upgrade the secondary rectifiers. that listed only provide 10A.
use at least 20A, or you can double (2x20A) it to double capacity.

[ckrtech]

Re: Enlight HPC-250-101 is aging

Hey Quaddro. Thanks for your input.

Most of the PSUs that I restore are used in vintage systems like 286/386/486/Pentium. Although this one is going in a Pentium III, I would like to have the -12v still going.

I suppose I could add a component part I have around here in order to get it to draw from +5V and see if 12V comes up. Aside from the glue (and ignoring that -12v for the sake of this question) - would you say fresh caps aren't really necessary?

[Quaddro]

Re: Enlight HPC-250-101 is aging

Ow..

lets check the picture.
12v1, 11.5
12v2, LL
-12, 11. (that's mean the value is -11v)

Nothing wrong, i think.
There's a single rail psu, that's why 12v2 only display LL.

I never replace capacitor if it's not necessary, even in 10+ years old psu.
Well, esr meter is very useful here.
As long as it fit the criteria, then it's will be fine.

Yes, capacitor, will degrade. but how much it will affect your system?
That depend on your load.
If it low enough, then a non bulged caps is still good to go.
If it will run at full capacity, then fresh caps is must

[ckrtech]

Re: Enlight HPC-250-101 is aging

Enlight is now in circuit.

[momaka]

Re: Enlight HPC-250-101 is aging

What is more interesting is the glue. You can see instances of the glue starting to turn brown in places - check out the primary side as well as the coil in the center of "capacitor city" on the output side.

That's the infamous organic tan glue that turns conductive when it turns brown. Unlike Quaddro, I actually have seen it make power supplies blow up - though it is worth noting that the blown PSUs I saw it in, it had turned almost black from heat.

Nevertheless, it's always a good idea to remove that tan glue, especially from the primary side or any high-voltage components. I also like to remove it from the secondary-side inductors, as sometimes it can eat through the varnish.

I am tempted to re-cap and DE-glue (and re-glue?).

Re-cap? - YES, highly recommended. Even a partial recap would be good.
De-glue? - YES, absolutely.
Re-glue? - No.

The glue is put there only during manufacturing to hold down the components as they get soldered. Once soldered, it is no longer needed. So you do not need to reapply the glue.

I would probably leave the two primary caps (Rubycon CE 470uf) but still remove the glue around them. What are your thoughts?

The glue around the primary caps is usually the only place I leave it, unless it is also touching other components.

Also - any other components worth replacing/upgrading?

There may or may not be a critical cap for the 5VSB on this PSU, depending on if it's a 2-transistor 5VSB circuit or not. But if it is, you will DEFINITELY want to change that. On that note, also always recap the 5VSB on any PSU you intend to use. If the 5VSB caps go bad, sometimes they can cause a lot of damage to other components in the PSU - particularly the PWM controller.

Old Enlight usually use delta as the oem.
and old delta, somehow, is always low in 12v when tested with no load.
As a group regulated characteristic, you need give a load to 5v rail, to raise the 12v rail.

Exactly!

Upgrade suggestion.
Upgrade the secondary rectifiers. that listed only provide 10A.
use at least 20A, or you can double (2x20A) it to double capacity.

If you are talking about the 12V rectifier, that may not be a good idea. Old PSUs like this usually use a fast-recovery rectifier on the 12V rail with a typical voltage drop of at least 0.8V. Replacing it with a Schottky (typical voltage drop of 0.5V or less) can sometimes make the 12V rail go way too high, especially with a high 5V load for which the PSU is made.

I never replace capacitor if it's not necessary, even in 10+ years old psu.
Well, esr meter is very useful here.
As long as it fit the criteria, then it's will be fine.

What you are doing there is quite risky.

First off when dealing with old PSUs, always check if they have a 2-transistor 5VSB circuit. 2-transistor 5VSB circuits do NOT play well with bad caps on the 5VSB output. I've seen burned PWM controllers many many times due to bad output caps on the 5VSB rail.

Furthermore, 2-transistor 5VSB circuits with a "critical" cap (typically 10-47 uF cap on the primary side next to the 5VSB transformer) CAN damage not only the PSU, but also even your motherboard and attached peripherals. Look up Bestec ATX-250-12E if would like to see a good example. A typical failure mode of such circuit is extreme over-voltage on the 5VSB rail (8-15 Volts not uncommon at all).

Given the vintage of the O/P's PSU, I am almost certain its 5VSB circuit is a 2-transistor design.

[ckrtech]

Re: Enlight HPC-250-101 is aging

momaka - I am late for thanking you for your response. Thank you.

I am currently removing/documenting caps as well as making a cap map. On that subject, the large TEAPO output caps (2200uf, SX series) are (according to SX Datasheet) ripple current: 1450mA /100KHz and ESR: 0.095 (Max 25C /100KHz). Nichicon UPJ seems like a possible replacement with ripple: 1470mA (10kHz to 200kHz) and ESR: 0.066 (20 ̊C / 100kHz). Slightly lower temperature for the specified ESR, but still a notable lower ESR. Any objections? I saw other caps that were closer to an ESR of .040 or even .022ish, but I didn't know how far was safe to stray from the original 0.095 - provided that datasheet is applicable for my TEAPO caps.

Concern -

I've noted your commentary on 5VSB. I am doing a full recap, so any electrolytic in there is going to be replaced. I assume (wonderful word) that I should be in the clear as far as 5vsb is concerned in regard to caps. However...

All non primary input caps seem to be TEAPO and G-Luxon. The only exception is a single, 35V 100uf Rubycon YXF (C16 near the TL494CN in the attached photo)

In the 5vsb "area" (middle third), I found something rather disturbing. Check out R25, glue (partially removed), and the orange area of the board - all next to TL494CN - PWM control, is it not? What the heck is going on/has happened?

[stj]

Re: Enlight HPC-250-101 is aging

oh nice.
that is a classic example of why you remove the glue.

the glue hardened and went conductive, then current passed through it and burned the area!!!
the glue is acidic btw, that's why the tin plating has gone from the copper resistor leads!

[momaka]

Re: Enlight HPC-250-101 is aging

On that subject, the large TEAPO output caps (2200uf, SX series) are (according to SX Datasheet) ripple current: 1450mA /100KHz and ESR: 0.095 (Max 25C /100KHz). Nichicon UPJ seems like a possible replacement with ripple: 1470mA (10kHz to 200kHz) and ESR: 0.066 (20 ̊C / 100kHz). Slightly lower temperature for the specified ESR, but still a notable lower ESR. Any objections? I saw other caps that were closer to an ESR of .040 or even .022ish, but I didn't know how far was safe to stray from the original 0.095 - provided that datasheet is applicable for my TEAPO caps.

Check the physical dimensions of your Teapo caps (i.e. diameter and height). Within any cap series, the ESR and ripple current depend only on the can size of the capacitor. So if your Teapo SX capacitors match the size of the Teapo SX in the datasheet, then that's the ESR and ripple current your look for.

That said, for those 2200 uF Teapo SX caps of yours (I'm guessing 6.3V), Nichicon PJ would indeed be the closest match in terms of ESR and ripple current. Rubycon YXJ would be second closest, though it does have slightly lower ripple current.

Generally, it is best to stay within 20% of all specifications, including ESR. But often times with crap cap brands, you will find that to be impossible so it's okay to deviate a little. In my experience, up to 50% lower ESR has never caused any problems with these older PSUs that use caps rated just a tad bit better than general purpose.

So with that said, you could probably use Nichicon PW and PM, Panasonic FC, or even United Chemicon LXZ without problem - all of these are entry-level low ESR caps that have been known to do really well in PSUs (after all, they were developed for PSU use).

Thus, if you can't find Nichicon PJ in stock or Rubycon YXJ, any of the above should be fine too. On some of these old PSUs, I've gone as low as motherboard-grade capacitors and still had good results. So most of these older PSUs are quite tolerant of ESR deviations. In many cases, it takes up to 10x lower or higher ESR for the PSU to start misbehaving.

Concern -

I've noted your commentary on 5VSB. I am doing a full recap, so any electrolytic in there is going to be replaced. I assume (wonderful word) that I should be in the clear as far as 5vsb is concerned in regard to caps.

Yes.
5VSB circuits use flyback topology, so even caps with very low ESR (i.e. motherboard grade) will work with it. In fact, I highly suggest using better caps on the output of the 5VSB. Chemicon KY, KYB, and KZE, as well as Nichicon HE usually work very well there. The other series listed above will also work well.

In the 5vsb "area" (middle third), I found something rather disturbing. Check out R25, glue (partially removed), and the orange area of the board - all next to TL494CN - PWM control, is it not? What the heck is going on/has happened?

Can't say for sure why that has happened. If you provide a picture of the bottom (solder) side of the PSU, I may be able to trace what that resistor is connected to.

Most likely, it is a minimum-load resistor of some sort - either for the 5VSB, 3.3V, or 5V rail. PSU manufacturers sometimes use resistance values that are too low, and the resistors end up burning like you see. If that's the case, I typically remove/replace them, depending on which rail they are on.

However, that burned resistor could also be a series resistor for the PWM controller's Vcc supply. If that's the case, then that's an indication that your 5VSB or its auxiliary secondary rail have been going over-voltage. Like I said, I've seen this happen very often when either the 5VSB's output caps or "critical" cap go bad. It doesn't always result in over-voltage in the 5VSB, but it almost always results in over-voltage to the PWM controller (in which case, the series resistor that goes to the PWM's Vcc "eats" all of that extra voltage and burns up). That said, this resistor looks a bit too big to be PWM controller's series resistor on Vcc, so I'm still thinking it is a load-resistor. But I guess we will see if you provide photos.

oh nice.
that is a classic example of why you remove the glue.

the glue hardened and went conductive, then current passed through it and burned the area!!!

I doubt that's what happened.

Even when fully cooked to a crisp, that glue is hardly conductive to pass more than a few mA of current, if even that much. I've measure it a few times, and the worst I've seen it is go down to a few 10's of kiloOhms - enough to pull up a Gate on a MOSFETs to make it fully turned ON (thus causing a major blow-out) or mess up a PWM controller's logic.

the glue is acidic btw, that's why the tin plating has gone from the copper resistor leads!

Yup, that's another reason to remove the glue. It seems it's actually many times more acidic than regular silicone. Makes one wonder why the manufacturer didn't use regular silicone. I'm guess this is probably the cheapest thing they can find... hence the results.

[ckrtech]

Re: Enlight HPC-250-101 is aging

Oh yes! I did not include the dimensions, but they are on the spot for height, width, and lead spacing. The replacement UPJ may actually be 1.5mm taller - not much at all, and height is certainly no big deal for fit.

Thanks for the general guide on the percentage difference. When I first started working the filters at digikey for low ESR for 2200uf, the large percentage drop in ESR was quite noticeable.

I've included a photo of the solder side. It looks like you can easily locate the burned area in the middle of the photo. I suppose when I order replacement caps, I can get a replacement resistor as well. Gonna take extraction and cleaning to see the stripes - or an Ohm reading and hope it is still at the appropriate level if I can't make out the colors on the casing.

[ckrtech]

Re: Enlight HPC-250-101 is aging

However, that burned resistor could also be a series resistor for the PWM controller's Vcc supply. If that's the case, then that's an indication that your 5VSB or its auxiliary secondary rail have been going over-voltage.

Hrm. Seems like all items are linked to PWM pin 12: Vcc.

I have enclosed the following:

1: A "monster" combination of photos that shows a horizontally flipped solder side image, the component side, and a side view of the TL494. It should show the two resistors in line with Vcc (blue path) as well as a path traveled that ties Vcc to a jumper that then connects to the C16 positive side (red path). C16 happens to be the only non-Teapo, non-G-Luxon cap on the board. It is a Rubycon 35v 100uf YXF series cap.

2: A close up showing the connection between one leg of each of the two resistors and TL494 Vcc.

3: The TL494 datasheet for the pinout. It doesn't specifically mention the "CN" package, but I assume I am OK. Link - https://cdn.badcaps-static.com/pdfs/...88cd5e8891.pdf

[momaka]

Re: Enlight HPC-250-101 is aging

Wow, what a crappy circuit layout. And the output design isn't that great either.

The 5VSB rail starts with rectification from the diode right next to C53, gets filtered from C53, through L7, filtered by C54, through L10, and out through the purple 5VSB wire.

Meanwhile, the secondary auxiliary rail (sec. aux. for short) is rectified by D29(???) - or whatever that diode is next to L10 and R70, and goes straight to pin 12 (Vcc) of the TL494 PWM controller. Filtering is done by that Rubycon 35v/100u cap (C16), which is connected to Vcc through a jumper wire (what a terrible idea!)

There are no series resistors with the PWM's Vcc, like I expected. This means an over-voltage on the sec. aux. can damage the PWM controller quite easily.

That said, the burned resistor R25 appears to connect to diode D15 (a regular glass 1n4148) and from there appears connected to the secondary-side BJT drive circuit, which consists of two little TO-92 transistors, a bunch of resistors, and a few diodes.

With that said, if the PSU seems to work fine, definitely recap the 5VSB circuit output and check the voltage on the sec. aux. rail, especially with a 0.5-1.5 Amp load on the 5VSB rail. As long as the voltage is well below 30V, the PWM controller should be safe. As for R25, that could have started burning either due to over-voltage on the sec. aux. rail or if it wasn't selected properly (or if the PSU isn't working, the two small TO-92 driver transistors could have shorted or the two glass diodes). In any case, measure R25 out of circuit to see what resistance you get on it. Resistors typically go open-circuit or high-resistance. So if your resistor is reading several hundred Ohms or less, it is probably still okay (though do try to compare it to its burned color code). Let us know what you find.

Thanks for the general guide on the percentage difference. When I first started working the filters at digikey for low ESR for 2200uf, the large percentage drop in ESR was quite noticeable.

Interesting. I never tried sorting by ESR on there, as in the past, they used to not always put the info there.

So what I do instead is select only the brands I want to look at (Rubycon, Nichicon, Panasonic, and Chemicon), the capacitance(s), the voltage(s), "Through Hole" mounting, and check "In Stock" box. After that, apply filters and go for a second round, where I select the temperature range (all that list 105C) and diameter I need. That will show all 105C-rated caps, some of which are not low ESR. But I have folder with a list of low ESR PSU-suitable caps' datasheets, so I know which series to look for only. Then I enter the quantity I need and sort by price. That about does it.

[ckrtech]

Re: Enlight HPC-250-101 is aging

Thanks again for your feedback. The PSU seemed to run just fine when I had it hooked up to the Pentium III prior to pulling caps.

In any case, measure R25 out of circuit to see what resistance you get on it.

I had to do a bit of scraping with the multimeter probe to get a reading. Finally got ~140 Ohms to hold steady. Size-wise... 2 watt seem correct?

The band is more of a continuous loop? Hmm. Found a site on the web that basically says it is a burned out resistor (in case there was any doubt of that). I was wondering if perhaps the casing was basically gone, and that was why I couldn't see any normal bands. Google search suggests that the line I am seeing is a spiral of conductive carbon?

I'd like to replace this little burned out guy but definitely need a sanity check as far as what replacement I should use.

check the voltage on the sec. aux. rail, especially with a 0.5-1.5 Amp load on the 5VSB rail.

So to double check - I have a TL494, and the Vcc can handle between 7V and 40V. The oomph (higher voltage) coming into PWM Vcc is from the secondary rail. The secondary rail comes from "pin" 7 of that transformer, goes over that diode D29 (I'll double check the number) and hops right into pin 12 of the PWM - the light blue path I drew in my earlier image.

How should I (safely) load the 5VSB and test the sec. aux rail? I've got a scope - so that should let me clamp easily. I suppose I could grab the leg of R17 closest to pin 12 of the PWM & alligator ground clamp the PSU...case. Not exactly close to where I would be getting the signal. I guess I could put a second probe on TL494 pin 7 (GND) and MATH the two signals. Never done that. Do I need to worry about how I ground though, really, since I am not concerned with noise so much as just trying to get a DC < 30V check? Just thinking out loud here.

I need to dig a little grave for that resistor.

[momaka]

Re: Enlight HPC-250-101 is aging

I had to do a bit of scraping with the multimeter probe to get a reading. Finally got ~140 Ohms to hold steady. Size-wise... 2 watt seem correct?

Yeah, 1-2 Watts seems correct, depending on if this was a carbon film or metal film resistor. But I'd go with 2W just to be safe.

I need to dig a little grave for that resistor.

Well, it doesn't appear quite dead yet. While 140 Ohms is not a valid resistance for standard E12/E24 4-band resistors, it's still probably somewhat close to what it was originally. I would *guess* it most likely drifted up from either 100, 120, or 130 Ohms.

The fact that it hasn't gone open-circuit yet with the case this badly burned suggests that the conductive glue could have caused most of the damage actually. Personally, I've seen that conductive glue become corrosive enough to eat the casing of resistors. Combined with heat, their casings peel over time just like yours.

That said, I looked at my collection of half-bridge ATX PSU schematics, and it appears most PSUs typically use 1.5 KOhms for that resistor. So why the low resistance reading you are getting, I have no clue. Double-check the resistance again, if possible. It also might help to sand the leads of the resistor to get a better reading. And make sure your multimeter shows no more than 1 to 2 Ohms resistance offset. (I sometimes use my cheap meters, and with their cheap dials, sometimes they will show a 30-50 Ohm resistance offset! I have to rotate the dial several times to clear the contacts. Only then I get less than 1 Ohm offset on the 200 Ohm range.)

The band is more of a continuous loop? Hmm. Found a site on the web that basically says it is a burned out resistor (in case there was any doubt of that). I was wondering if perhaps the casing was basically gone, and that was why I couldn't see any normal bands. Google search suggests that the line I am seeing is a spiral of conductive carbon?

Yeah, that spiral is the "innards" of the resistor. A spiral color code would be quite deceiving to the reader (and funny too!)

I'd like to replace this little burned out guy but definitely need a sanity check as far as what replacement I should use.

Well, since everything appears to work with that "140-Ohm" resistor, the probably it might be best to get 130 or 150-Ohm resistor and call it a day.

Though the experimental side of me also says you could try getting a 1-KOhm resistor too and see if the PSU works with that. Again, all H-bridge ATX PSU schematics I looked at in my collection (about 10 of them) showed a 1.5 KOhm resistor. I think only one used something like 820 Ohms. The secondary-side BJT drive circuit is pretty standard otherwise: two NPN transistors (typically 2N2222A, C945, or 2SC1815), 4-5 resistors (all in the 1 to 3 KOhm range), 4-5 1n4148 diodes, and a 1-2.2uF cap.

It also makes sense for that resistor to be around or more than 1 KOhms. After all, typical voltages for the sec. aux. rail range between 12 to 25V, with the TL494 being able to take up to 40V. If that resistor in your PSU (R25) really is 140 Ohms, you are looking at over 1 Watt of power dissipation at 12V and over 2 Watts at 17V. Personally, I've never seen anything like that before either, so I think it might be a good idea to also get a 1 or 1.5 KOhm resistor.

So to double check - I have a TL494, and the Vcc can handle between 7V and 40V. The oomph (higher voltage) coming into PWM Vcc is from the secondary rail. The secondary rail comes from "pin" 7 of that transformer, goes over that diode D29 (I'll double check the number) and hops right into pin 12 of the PWM - the light blue path I drew in my earlier image.

Yes, that is correct.

How should I (safely) load the 5VSB and test the sec. aux rail? I've got a scope - so that should let me clamp easily. I suppose I could grab the leg of R17 closest to pin 12 of the PWM & alligator ground clamp the PSU...case.

That would be fine. For simple DC voltage checking, you can grab the voltages from anywhere pretty much.

That said, make sure all of the filter capacitors are back in the PSU before powering ON anything, especially the 5VSB filter caps (otherwise, be prepared for magic smoke!)

I also suggest to take the voltage readings both with R25 in the circuit and with R25 removed. R25 is only really needed for the primary power supply section. Without it, only the 5VSB will work - which is okay for our purposes.

That said, to load the 5VSB, a 12V 10-20 Watt car bulb and/or some resistors might do the trick. I personally use a 20W halogen bulb from Ikea (rated for 12V.... so at 5V, it draws around 1 Amp). If you don't have any of these, a hair dryer, heat gun, or toaster oven rated for around 1400-1600 Watts will have approximately 10 Ohms of resistance (for the US 120V market, that is). So with that, you will load the 5VSB with approximately 500 mA, which is still good enough. Or four to five 12V PC fans should also provide a reasonable load on the 5VSB.

[ckrtech]

Re: Enlight HPC-250-101 is aging

I did a bit of sanding, and the resistor is now "646 Ohm." It is possible that I could do a bit more sanding - but it seems like I got two good points for it.

Multimeter is a pretty good one - Fluke 87 V. I'm unfamiliar with offsets.

What's the worst that could happen if I just stick a 1.5KOhm 2 watt resistor in there right from the get-go?

[ckrtech]

Re: Enlight HPC-250-101 is aging

OK. Some side notes and questions about caps.

I've recapped a PSU before, but I am trying to read into things a bit more with this one. I've got some questions.

First - Thanks again for the Nichicon UPJ thumbs up. I feel pretty good about that one.

So then comes the subject of what appear to be general purpose caps and their ripple frequency.

TEAPO actually has a catalog (attached image) with competitor series equivalents - a short cheat sheet for upgrades here -

One thing interesting is that they specify competitor series that have a lower ripple frequency @ 120 Hz than their own line of caps. For instance - TEAPO SEK 50V 10uf is ripple current of 50 mA @ 120 Hz. It recommends (for example) Chemi-con KMG as an equivalent replacement, yet the same ratings and case size for KMG has a ripple current of 40 mA @ 120 Hz.

Using the coefficients in the tables for the TEAPO SEK and the Nippon Chemi-con KMG, the high frequency numbers balance out a bit more -

TEAPO SEK is 72.5mA @ 10~100K
Chemi-con KMG is 70mA @ 10K, 72mA @ 100K

So it still comes up just short in the high frequency.

I could obviously skip TEAPO's own equivalent catalog and just walk datasheets from digi-key as I assume TEAPO naturally set up their catalog recommendations so they provide higher values than the caps you would be replacing with theirs.

But this brings up the question regarding the importance of ripple current low/high frequency for general purpose caps. Is their wiggle room, or is it best to always look for higher values & therefore jump a series (20 cents more or whatever) to get that extra oomph in mA?

[momaka]

Re: Enlight HPC-250-101 is aging

I did a bit of sanding, and the resistor is now "646 Ohm." It is possible that I could do a bit more sanding - but it seems like I got two good points for it.

Interesting. I wonder why the resistance is higher now. Did you try cleaning the resistor well from the conductive glue prior to the previous measurements. If there is some stuck between the conductive spiral, I think that could mess up the measurements.

What's the worst that could happen if I just stick a 1.5KOhm 2 watt resistor in there right from the get-go?

Nothing. The main power supply just won't be able to turn on if the resistance is too high.

At worst, the BJTs on the primary may not get fully turned ON, which can cause them to heat up more than usual. But in any case, once you put the PSU under load, this will show up rather quickly, as the PSU won't be able to keep up under load and probably shut down. So it's unlikely that anything would get damaged.

Multimeter is a pretty good one - Fluke 87 V. I'm unfamiliar with offsets.

Then you probably have nothing to worry.
I was talking about the cheap multimeters that have really crappy probes and dial contacts. I have a decent (high-end) Radio Shack multimeter and several really cheap meters. With the cheap ones, I always check the resistance offset by shorting the probes together on the resistance setting I want to use the multimeter. Like I said, bad dial contacts and also cheap probes can give a serious resistance offset (especially in the low resistance range) - enough to make you conclude something is much higher resistance than it should be (or shouldn't be).

But as a general rule, regardless of what kind of multimeter I am using, I always check the probe resistance first to make sure the probes are making proper contact to the meter and are functional. There's nothing more frustrating than trying to measure a voltage (or current) and seeing nothing on the multimeter only to find out that it's not working properly. I've wasted a lot of time that way trying to troubleshoot problems that never existed with the device under test.

TEAPO SEK is 72.5mA @ 10~100K
Chemi-con KMG is 70mA @ 10K, 72mA @ 100K

So it still comes up just short in the high frequency.
...
But this brings up the question regarding the importance of ripple current low/high frequency for general purpose caps. Is their wiggle room, or is it best to always look for higher values & therefore jump a series (20 cents more or whatever) to get that extra oomph in mA?

Put it this way: electrolytic caps have pretty lousy tolerances. Capacitance can vary by as much as 20%, so most circuits are usually designed to handle even bigger variations (half the capacitance or two times the capacitance is not uncommon - though I never suggest going lower than the original capacitance). ESR? It can vary quite a bit with temperature - sometimes as much as 2 to 3 times higher at -10°C compared to normal room temperature of 20°C. (Of course, equipment is hardly ran at that temperature, but it gives you a realistic idea of what you should expect of the caps if you take a piece of electronic equipment that's been outside in the winter and you bring it inside.) That said, the ESR of caps also decreases quite a bit as the temperature increases. In a mildly warm PSU, the temperatures can easily reach 40-50°C and in hotter ones, 60-70°C is not uncommon either. So you can imagine what happens to the ESR - it drops quite a bit.

So where does the ripple current stand? Well, ripple current is the amount of AC current the cap can take before it overheats internally (which over time will cause the cap to fail). Thus if you use caps with lower ripple current ratings, you could technically make your new caps overheat. However, a good circuit design should never be made to stress caps near their maximum ripple current. Think of the ripple current rating just like the voltage rating of cap: if your circuit is only running at 2V, for example, you can use a 4V, 6.3V, or even a 10V cap - all would be safe. If the original caps were 10V for that circuit and you replaced them with 6.3V, nothing bad will happen.

However, the only reason why it's always suggested to have new caps meet or exceed the ripple current rating of the originals is because there is no easy way to measure ripple current in a circuit without expensive equipment. Thus, to avoid risking of picking caps that are not suitable for the circuit, it's just easier to go with ones than can handle more ripple current. That said, going down to even 2/3 of the ripple current of the original caps would still be safe in most cases. Just refrain from doing that, *if* possible.

That said, the example you gave above with the ripple current of the Teapo SEK vs. Chemicon KMG is less than 5% difference between the two. You can be sure as heck it won't matter in a real circuit - or it better not. If it does, then that circuit was designed very badly to begin with. A good design (or even a merely decent one, for that matter) will never have components run near their maximum ratings.

[ckrtech]

Re: Enlight HPC-250-101 is aging

Thanks once again, momaka.

I did some sanding on the resistor. There was definitely some glue around one edge of the spiral - there still is, actually. The results? 2.2 MOhm! It looks like it would take quite a bit of work to get the thing completely clean - and even then, I am not sure that I wouldn't do a bit of damage to it. I may need to up the grit if I proceed. Probably will just order a 1k and 1.5k resistor...

That said, the example you gave above with the ripple current of the Teapo SEK vs. Chemicon KMG is less than 5% difference between the two. You can be sure as heck it won't matter in a real circuit - or it better not. If it does, then that circuit was designed very badly to begin with. A good design (or even a merely decent one, for that matter) will never have components run near their maximum ratings.

What about the difference between cap brands/series in the low (120Hz) range vs the high (10k or 100k) range? There is a much higher difference between the two series for 120Hz.

Also - any harm subbing a low ESR cap for general purpose?

[momaka]

Re: Enlight HPC-250-101 is aging

I did some sanding on the resistor. There was definitely some glue around one edge of the spiral - there still is, actually. The results? 2.2 MOhm! It looks like it would take quite a bit of work to get the thing completely clean - and even then, I am not sure that I wouldn't do a bit of damage to it.

Oh man, I hope you weren't sanding/cleaning the conductive spiral on the resistor all this time, because now it looks like your resistor is pretty much open-circuit.

Probably will just order a 1k and 1.5k resistor...

That should do. They won't even need to be rated for 2W. 1W would be perfectly okay.

What about the difference between cap brands/series in the low (120Hz) range vs the high (10k or 100k) range? There is a much higher difference between the two series for 120Hz.

Caps that only show ripple current for 120 Hz are usually just general purpose. The only time you need to worry about ripple at 120 Hz is when the cap is used in a linear power supply filtering some rectified line voltage. In SMPSes, you can be sure the caps will be running under higher frequency. And with that, as you see between the Teapo SEK vs Chemicon KMG, the difference in ripple current becomes smaller.

Also - any harm subbing a low ESR cap for general purpose?

Almost never a problem. In fact, it's often recommended when recapping SMPSes, as low ESR caps tend to have longer endurance life too.

[ckrtech]

Re: Enlight HPC-250-101 is aging

Oh man, I hope you weren't sanding/cleaning the conductive spiral on the resistor all this time, because now it looks like your resistor is pretty much open-circuit.

Just the leads the first time. The second time I shot for just the ends and tried to leave the middle cylinder with the spiral alone. I wouldn't be surprised if I caught it a few times while sanding, but I suppose I already made up my mind when it came to the replacement value resistor prior to that second go-around with sanding.

($0.24 metal oxide Yageo OK? Haha)

That should do. They won't even need to be rated for 2W. 1W would be perfectly okay.

Excellent.

In fact, it's often recommended when recapping SMPSes, as low ESR caps tend to have longer endurance life too.

Awesome. This influenced my choices a bit - made things easier.

Attached my spreadsheet of previous caps and new caps in the BOM at digikey in case people need a reference in the future (or if you have any suggestions. I probably won't commit it until later in the day on Friday the 13th.)

Originals on the left column and replacements on the right. Lead spacing was accounted for for each cap but not included in the spreadsheet. I also couldn't find an entry for G-Luxon SM series 0.68uf in any datasheet, but I feel the Nichicon UPJ should be fine. The G-Luxon ripple value for that cap is a guess. A few of the bolded numbers demonstrate a noticeable difference between old and new (i.e. Ripple in a few spots)