I have tested the ESR on the filter caps in my Quad 405-2. They are the original blue BHC10,000uf 63v (making them 35-40 years old.) One reads around .08-.111 ohm and the other .025-.029 on my ESR meter. They both measure close to their values in capacitance. The ESR seems relatively low, are they still ok?
I find that a little difficult to believe. A 100 watt amp with 8 ohms draws 5 amps peak. A .14 ohm resistance in series will only cause the supply to lose seven tenths of a volt. Yeah, it’s high, but not enough to cause THAT much trouble. If you slip the decimal point, THAT would noticeably drop the power.
Was a 4 ohm rated amp with 4 ohm speaker. AC voltage on speaker was converted to watts. ESR meter was Peak ESR meter from Newark. You can come over and measure the old caps if you want. Burglar stole the meter but not cap marked defective. All new one way caps brought power to 40 watts, but a difficult to find bipolar 75 uf cap restored to 100 watts.
I see manufacturers have started giving us low frequency ESR numbers, which IMHO is a step backwards, but let's say a new cap is under 0.04 ohms at 120 Hz. End of life would be 2X or .08 ohms, though it would probably still be functioning well. I'd rather see dissipation factor, which would be 0.3 new and 0.6 at end of life. It would be unusual to measure the correct capacitance value and yet have the cap be no good. Note that at this level you can't get an accurate measurement without a 4-wire connection to the cap.
One cap has gone a little astray, but it is still Ok, and will probably stay so for the next 30 years. If you think you are going to outlive it, you can replace it, but before you attempt anything, I recommend you browse through all the recapping horror stories on this site (and others).
It should be a sobering and calming read.
Living with something that is potentially problematic vs. actual problems is not the same.
The mantra is: if it ain't broke, don't fix it
It should be a sobering and calming read.
Living with something that is potentially problematic vs. actual problems is not the same.
The mantra is: if it ain't broke, don't fix it
I have successfully re-e-capped for cause a Fomoco AM radio 1956, Dynaco ST70 amp (4 times), PAS2 preamp (2X), ST120 amp, Hammond A100, H182(2X) organs, Shober Recital Organ amp, Peavey PV-4C amp, CS800s (2x) MMA-875t (2X), MMA-8150t, M-2600 amp,Allen S100 amp (2x) several DC supplies from the Allen 300 organ, Allen switcher supply from a nineties organ, Reader's Digest AM/FM radio 197?, Lenoxx Sound AM/FM radio, Newark Stellar Lsbs HDTV converter (switcher supply), Herald Electronics RA-88a disco mixer.
I function test my work after every couple of caps. Points my eyes right at any error I made. The H182 took 177 e-caps. After wearing out the ST70 caps 3 times, I quite buying TV parts store shelf caps and started buying long service life (>3000 hours) caps from newark & digikey.
Last edited:
I think the rule of thumb here is to compare the RATIO of the best sample to the other samples.
The BEST measurement is your bogey & tells you about your meter
Anybody measuring 10x bogey --> trash
Anybody measuring 5x bogey --> trash
Anybody measuring 2x bogey --> trash
...pick whichever one of those lines you like ":^)
Also at some point you gotta test your bogey against a different batch of the same value parts to see if the bogey is trash as well.
The BEST measurement is your bogey & tells you about your meter
Anybody measuring 10x bogey --> trash
Anybody measuring 5x bogey --> trash
Anybody measuring 2x bogey --> trash
...pick whichever one of those lines you like ":^)
Also at some point you gotta test your bogey against a different batch of the same value parts to see if the bogey is trash as well.
As an aside, I am using a sound card for analysis, it needs an attenuator to avoid overloading it. I found this diagram for a simple circuit:
https://www.diyaudio.com/community/...ut-and-audio-interface-balanced-input.394417/
However, I found that I need to dramatically increase R2 to 180k (from 9.1k for a claimed 20db reduction) to get a decent attenuation around 23db (550mv down to 3mv). Can someone post the calculation for the DB attenuation for a given R2 resistance?
https://www.diyaudio.com/community/...ut-and-audio-interface-balanced-input.394417/
However, I found that I need to dramatically increase R2 to 180k (from 9.1k for a claimed 20db reduction) to get a decent attenuation around 23db (550mv down to 3mv). Can someone post the calculation for the DB attenuation for a given R2 resistance?
You also have to take into account the load you have on the attenuator. The values given are for open circuit loading.
It appears you have somewhere around 1k of actual loading. Check the specs of the sound card.
Hi, it's the same as the earlier DIY article I linked to, the only differences being the change in R2 to 180k, and use of a .25" phone plug instead of xlr (goes into the same jack). https://www.diyaudio.com/community/attachments/mes_setup-png.1129894/
So for 23dB attenuation ( with R2 in k ohms) you need:
-23dB = 20 x log ( 10//60 / (R2 + 10//60) )
10^(-23/20) = (10//60) / (R2 + (10//60) )
0.0708 = 8.57 / ( R2 + 8.57 )
R2 + 8.57 = 8.57 / 0.0708
R2 + 8.57 = 121
R2 = ( 121 - 8.57 ) = 112.5k
If the math is right, the difference is probably in the measurement of the 3mV.
-23dB = 20 x log ( 10//60 / (R2 + 10//60) )
10^(-23/20) = (10//60) / (R2 + (10//60) )
0.0708 = 8.57 / ( R2 + 8.57 )
R2 + 8.57 = 8.57 / 0.0708
R2 + 8.57 = 121
R2 = ( 121 - 8.57 ) = 112.5k
If the math is right, the difference is probably in the measurement of the 3mV.
I have built my attenuator using the 180k ohm resistor mentioned above.
Voltage drop (using a 100hz tone to drive the amp) shows 21.4vac (rms) output on the amp drops to .114vac with the attenuator.
Using the formula RV = 20*log(V2 / V1), I get a 45.5db reduction.
I have always thought +10db = 10x the wattage (or +3db = 2x watts), but this calculation doubles that: 10db=20x. Most amplifier meters also show +10db = +10x watts.
What am I missing? Something to do with peak to peak measurement?
Voltage drop (using a 100hz tone to drive the amp) shows 21.4vac (rms) output on the amp drops to .114vac with the attenuator.
Using the formula RV = 20*log(V2 / V1), I get a 45.5db reduction.
I have always thought +10db = 10x the wattage (or +3db = 2x watts), but this calculation doubles that: 10db=20x. Most amplifier meters also show +10db = +10x watts.
What am I missing? Something to do with peak to peak measurement?