I just replaced the filter caps in a tube crossover, and it now take the fuse out on power up. (yes, I put 'em in correctly)
I suspect the new caps have a much lower initial impedance and the inrush current is too much. I'm using Cornell-Dublier, and the values and voltages are correct.
Anyone out there have any suggestions?
I suspect the new caps have a much lower initial impedance and the inrush current is too much. I'm using Cornell-Dublier, and the values and voltages are correct.
Anyone out there have any suggestions?
yup
If they are the same value as the originals, and installed properly - polarity wise - then you're probably right in assuming they are drawing more charge current.
You could try using a small wattage light bulb in series with the AC line to see if current limiting helps.
If so, you can build a soft start ciruit, or simply add a thermistor in the AC line, which is what I do. I get mine from old PC power supplies.
If they are the same value as the originals, and installed properly - polarity wise - then you're probably right in assuming they are drawing more charge current.
You could try using a small wattage light bulb in series with the AC line to see if current limiting helps.
If so, you can build a soft start ciruit, or simply add a thermistor in the AC line, which is what I do. I get mine from old PC power supplies.
No- same value and voltage. The new caps are considerably smaller than the originals, but they read the correct capacity on a cap tester.
I rigged a supply to output the rated B+ (450 vdc) and hooked it up to a variac I borrowed. I'm raising the voltage 5 volts or so every couple of hours until I get to +450. Hopefully, that will form the caps and they'll behave after this treatment.
If you guys have and old HAM radio enthusiast in your neighborhood, get to know him. They know a LOT of tube and HV tricks.
I rigged a supply to output the rated B+ (450 vdc) and hooked it up to a variac I borrowed. I'm raising the voltage 5 volts or so every couple of hours until I get to +450. Hopefully, that will form the caps and they'll behave after this treatment.
If you guys have and old HAM radio enthusiast in your neighborhood, get to know him. They know a LOT of tube and HV tricks.
glorocks said:
By raising the voltage so slowly to the rating voltage of 450V, I believe that you are in effect forming the capacitor at each of the lower voltage(s). This may affect the performance of the capacitor. It's better to go straight in at the rated voltage and keep an eye on the leakage current over time. Once it steadies out, the cap is properly formed.
See this link for good general outline:
http://www.vcomp.co.uk/tech_tips/reform_caps/reform_caps.htm
Hi,
presumably the mains transformer is an EI. This should not need a soft start but a small benefit will accrue if you choose to use one.
The slow charge circuit can be a Power Thermistor in the secondary circuit to limit the peak charging current. This will be very beneficial. Fit it.
A light bulb will not do, it is PTC, you need NTC.
presumably the mains transformer is an EI. This should not need a soft start but a small benefit will accrue if you choose to use one.
The slow charge circuit can be a Power Thermistor in the secondary circuit to limit the peak charging current. This will be very beneficial. Fit it.
A light bulb will not do, it is PTC, you need NTC.
the high series resistance ensures that the capacitor is initially charged by a very low voltage and as the capacitor leakage reduces so the charge across the plates increases.MCampbell said:
It is wrong to suggest that staged charging is less effective than gradually increasing charging.
Ultimately the di-electric film will be the same thickness and it will suit the final voltage.
I use much higher resistor values to achieve much slower re-forming.
AndrewT said:
Andrew,
You may well be right that ultimately the dielectric layer will be the same thickness. What I am struggling with is the advantage of staging the voltage or of slow reforming. Certainly I have never seen it recommended by any manufacturer. Has your experience been that sow reaging will ultimately lead to a lower leakage current for the reaged capacitor. Or is there some other benefit that I'm missing?
Matthew
AndrewT said:
Andrew,
I do a fair bit of reforming, so I'm very interested in your finding.
I use a GR-1617A Capacitance Bridge for reforming (mainly because it only cost me about £50) which does a fine job, and should allow me to stage to voltage.
I'm certainly going to give this a try. I'll dig out an old pair of 500V lytics and try reforming one at 500V only and the other stagewise up to 500V.
To get the results you've indicated, what increment and period have you used? Are we talking 5V increase per hour?
OK- I raised the voltage to the +450 rating over a period of about 4 hours, bringing the voltage up 10 or so volts every 15 minutes, or when I remembered to do it.
One cap came up nicely, and when put back in the circuit worked as expected.
The second cap, although measured at the correct capacity on my cap tester, (low voltage), blew up when I got to around 100 volts. No fuse on the test rig to get in the way of catastrophy.
So it appears the second cap was the problem and was defective from the vendor. Haven't seen this before, but life is full of suprises.
Thank you all for your help and insight- all of which was most illuminating-
One cap came up nicely, and when put back in the circuit worked as expected.
The second cap, although measured at the correct capacity on my cap tester, (low voltage), blew up when I got to around 100 volts. No fuse on the test rig to get in the way of catastrophy.
So it appears the second cap was the problem and was defective from the vendor. Haven't seen this before, but life is full of suprises.
Thank you all for your help and insight- all of which was most illuminating-
i work at a company that does a lot of repair work. we recently got some caps in our "bench stock" that had one particular value (47uf/100V) that we use often, with about half of the batch mismarked for polarity. i think the clue (other than one tech blowing one up) was that the uncut leads had one long one(+) and one short one(-), with the shrink-on label reversed (i.e. the long lead was on the same side as the (-) stripe).
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.