Cap on the input to an amp, does capacitance matter much?

Blue Shadow said:
many amps great the incoming signal with a capacitor, something from a few microfarads to maybe ten or so. Usually a bipolar cap. I'm wondering the the capacitance matters? The Haflers seem to have a 10 uF there but many units have 2.2. If it does not matter, the cap there to take DC out of the signal (and what else I don't know) wouldn't it make sense to go with a sweet film cap instead of the cheap bipolar electrolytic used at the factory?
Click to expand...
For the input, anything between 1uf to 10uf will be fine in *most* designs. I use 2.2uf film most of the time and have the same in my own Hafler amps that are 630v rated, though a 250v is plenty. BiPolar is used because if you get a strong enough DC "spike" (polarized) in the wrong polarity, the cap would pop, and in doing so the spike will enter the amp stage causing damage. Electrolytic caps work "alright", but after a while drift high in capacitance and ESR, and leak from drying out attenuating the signal more and more. For the "cost", film is there "forever" without change through time.
 
I want to thank Blue Shadow for bringing this topic up (capacitance of coupling caps between preamp and amp), as it recently entered my consciousness and I was glad to see it was recently discussed.

I found the discussion here to be very useful, and particularly appreciate the comments provided by Skywatcher about “…anything between 1µF and 10µF will be fine…”, but was left somewhat uncertain if I could sub in a cap different from what the original design had. I was able to get more insight into that in a contemporary, recent thread (https://audiokarma.org/forums/index.php?threads/audio-grade-capacitors-catharsis.907641/) (mentioned above too) that gave guidance in how to compare (and substitute) film vs. electrolytic caps used in this particular location (preamp-amp coupling). In that thread, an exchange between Binkman and leestereo on the 3rd page brought up the nugget that if an electrolytic cap is used as a coupler instead of a film cap, the electrolytic should be 10x larger capacitance than the film cap (the capacitance of the latter being what the design specifies to insure a proper roll-off frequency).

Armed with this info, I approached a 30-odd-year-old NAD 7225PE in the dining room, which I use a few hours several days a week, driving a pair of OLAs. This receiver has a 10µF/16V polar electrolytic caps in series at the inputs to the main amp. I removed and measured these the other day, and found them in questionable shape: 11.6 and 11.1 µF, 2.6% and 2.4% Vloss, and 2.8Ω and 2.2Ω ESR. I have never recapped a device other than spot replacement of failed components or, in the case of speakers, old crossover electrolytic caps, but given the above, I was interested in changing these particular caps even though the unit was performing well.

Here is the relevant part of the main-in circuit. The caps in question are C401 and C402.

7225PE main-in.png


I do have some 10µF/50V electrolytic handy, but they are unbranded and have higher Vloss and ESR than the originals, and therefore not appropriate for this location. Next I thought I might replace them with some 10µF Panasonic ECQ-series metallized polyester caps I also had on hand, but they were much too large to fit in the space available (without daddy-long-legs type leads). But by reviewing the info in this thread and the other thread mentioned above, it was apparent that using a 1µF metallized polyester cap from the same Panasonic series should fit the bill fine, and it would fit perfectly in the chassis as well. And I have a bunch of them.

Here is a photo with the old 10µF/16V ones and the new 1µF polys. The former are among the smallest electrolytics I have handled, undoubtedly due to the low spec on the voltage tolerance.

IMG_0337.JPG

I installed them today and listened a bit. The NAD sounded quite good before, and they sound very good if not excellent now (only 1 hour in; clarity and low-end extension are excellent, so I didn't lose anything on the low end). My aural memory isn’t good enough to do a useful comparison before-and-after, but I’m certain I have not degraded the sound. So I am happy to confirm that replacing a 10µF polar electrolytic with a 1µF non-polar film cap, for a location where it is in series in an amp input, works well. I am also happy to not have to worry about further degradation of the original caps.

I’ll probably be looking over the schematics of some of my other older equipment to see if they are candidates for this treatment. The 1µF/250V Panasonics are only about $0.40 a pop in multiples of 10.
 
northpaw said:
The 1µF/250V Panasonics are only about $0.40 a pop in multiples of 10.
Click to expand...
and even though the manufacturer would get a lower price for higher volume it would still be well above the penny they paid for the lytic that was in there originally and worked just fine. Seems to be a nice upgrade. Glad it worked out for you.
 
Blue Shadow said:
many amps great the incoming signal with a capacitor, something from a few microfarads to maybe ten or so. Usually a bipolar cap. I'm wondering the the capacitance matters? The Haflers seem to have a 10 uF there but many units have 2.2. If it does not matter, the cap there to take DC out of the signal (and what else I don't know) wouldn't it make sense to go with a sweet film cap instead of the cheap bipolar electrolytic used at the factory?
Click to expand...
Hey Blue Shadow,

What made you bring this topic? Did you also have a specific unit in mind? If so, let us know.

Manfred gave you the best answer so far...if you ask me...:)

Best

Soundork
 
I started it because there are a number of amplifiers that have caps on the input. Sony TA-3200F has a ceramic disc cap hidden back on the RCA jacks that are switch selected in the normal position vs the direct connection when in the test mode. Value unknown but since it is ceramic (based on recollection not looking it up) it was probably a low value. Sony indicated it dropped frequency response down in the low single digit range and without it the amp would dig deeper.

Then there is the Hafler DH-200 which has a 10µF cap on the input to the board. Recommended to bypass this one, use a bipolar for sound quality but what size is really needed? These two amps ain't that much different, put in a signal get out 100 watts, what happens inside who knows. So let's ask on ak there are a few folks that are knowledgeable on this. Seems using a few µF film caps would be a small charge for us to absorb as we correct all these old amplifier. Could be an add on for amps that are truly direct coupled to avoid DC in the output if there is DC in the input.

Just trying to learn.
 
northpaw said:
I want to thank Blue Shadow for bringing this topic up (capacitance of coupling caps between preamp and amp), as it recently entered my consciousness and I was glad to see it was recently discussed.

I found the discussion here to be very useful, and particularly appreciate the comments provided by Skywatcher about “…anything between 1µF and 10µF will be fine…”, but was left somewhat uncertain if I could sub in a cap different from what the original design had. I was able to get more insight into that in a contemporary, recent thread (https://audiokarma.org/forums/index.php?threads/audio-grade-capacitors-catharsis.907641/) (mentioned above too) that gave guidance in how to compare (and substitute) film vs. electrolytic caps used in this particular location (preamp-amp coupling). In that thread, an exchange between Binkman and leestereo on the 3rd page brought up the nugget that if an electrolytic cap is used as a coupler instead of a film cap, the electrolytic should be 10x larger capacitance than the film cap (the capacitance of the latter being what the design specifies to insure a proper roll-off frequency).

Armed with this info, I approached a 30-odd-year-old NAD 7225PE in the dining room, which I use a few hours several days a week, driving a pair of OLAs. This receiver has a 10µF/16V polar electrolytic caps in series at the inputs to the main amp. I removed and measured these the other day, and found them in questionable shape: 11.6 and 11.1 µF, 2.6% and 2.4% Vloss, and 2.8Ω and 2.2Ω ESR. I have never recapped a device other than spot replacement of failed components or, in the case of speakers, old crossover electrolytic caps, but given the above, I was interested in changing these particular caps even though the unit was performing well.

Here is the relevant part of the main-in circuit. The caps in question are C401 and C402.

View attachment 1864215


I do have some 10µF/50V electrolytic handy, but they are unbranded and have higher Vloss and ESR than the originals, and therefore not appropriate for this location. Next I thought I might replace them with some 10µF Panasonic ECQ-series metallized polyester caps I also had on hand, but they were much too large to fit in the space available (without daddy-long-legs type leads). But by reviewing the info in this thread and the other thread mentioned above, it was apparent that using a 1µF metallized polyester cap from the same Panasonic series should fit the bill fine, and it would fit perfectly in the chassis as well. And I have a bunch of them.

Here is a photo with the old 10µF/16V ones and the new 1µF polys. The former are among the smallest electrolytics I have handled, undoubtedly due to the low spec on the voltage tolerance.

View attachment 1864223

I installed them today and listened a bit. The NAD sounded quite good before, and they sound very good if not excellent now (only 1 hour in; clarity and low-end extension are excellent, so I didn't lose anything on the low end). My aural memory isn’t good enough to do a useful comparison before-and-after, but I’m certain I have not degraded the sound. So I am happy to confirm that replacing a 10µF polar electrolytic with a 1µF non-polar film cap, for a location where it is in series in an amp input, works well. I am also happy to not have to worry about further degradation of the original caps.

I’ll probably be looking over the schematics of some of my other older equipment to see if they are candidates for this treatment. The 1µF/250V Panasonics are only about $0.40 a pop in multiples of 10.
Click to expand...
I am not sure that your original capacitors were bad, Northpaw. Usually capacitors have tolerance of 20% on each side. But, I also might not be reading right what you said.

Best

Soundork
 
I don’t think they were bad, at least not in the sense of causing noticeable (by me) sound degradation. The capacitance values are certainly within spec (they are labeled with code M, which is 20%), but from the discussion here and in the other thread linked above, this value isn’t critical anyway. What I was concerned about was the values for Vloss and ESR. I think they are higher than what is desirable in this location (in series, in the signal path), but that is just my opinion based on what I have read. I don’t know what the original spec was for Vloss and ESR for these particular caps, so I can’t say they drifted high over 30+ years, but my guess is they have. For 40 cents ea,, I can replace them and not worry about them going forward.

I might add that my primary impetus for opening this unit up was to check the PS caps. These NAD PE units have two sets of rails, the extra one being for the higher voltage PE circuit. In another PE unit I have, the PS caps on the upper rail failed, giving very noticeable hum. Those caps were located under the top tuner board, and that warm location may have hastened their failure, over several decades of use. The 7225PE unit here has that same configuration. So I pulled those caps: C515 and C516, 330µF/63V; the rail for them is ±50V. They checked out fine (295 and 320µF, with Vloss and ESR under 1). I noticed they were Japanese caps, not the local ones for this China-made receiver, and were 105°C rated. I put them back in, after cleaning up the old glue on the PCB and some corrosion it induced on nearby resistor leads.

On the overall issue of coupling caps, I am with Blue Shadow: it is something I wanted to learn more about. Further, we all know that some compromises had to be made in parts selection in this level of consumer gear for them to be offered at the prices they sold for; the best companies made wise choices that still performed well. I put NAD in that category, as I do almost anything Henry Kloss touched.

But the view from today, for me, some 30+ years on, is that some components may be degraded (certainly electrolytic caps are among those), and moreover, that some improvement over even original performance may be possible by informed replacement of limiting parts. Again, for me, low value electrolytic caps (less than a few µF) in critical locations such as coupling caps and speaker crossovers, are prime targets for replacement with film, given that very high quality film caps in this size are far more affordable now than they were when these units were manufactured.
 
Back in t he early 80's, I got a Hafler DH200 kit. The original boards had a 10 mfd, 16 volt non-polarized electrolytic cap for C1. Hafler offered an amp modification kit DH-203 for the "golden ear" crowd. The kit replaced the original input caps with larger-sized 2 mfd polypropylene that had to be mounted on the back of the boards. As with the vast majority on my modifications on gear over the years, my non-golden ears couldn't really tell a difference. Doesn't look like capacitance was all that significant.
 
We have a beta tester giving his feedback on our new Dh-220c design. This design has a dc servo so you can run it dc coupled. I used a 1uF Kemet stacked poly in parallel with a Nichicon muse bp 10uF. The tester is using Magi plainar type speaker, expensive ****. Well what would you know if he can hear the difference between different input caps and no cap at all. Imaging bass response etc. So that gets me thinking how come I can’t hear a difference. Wonders of the world This amp is ultra low distortion. There is no feedback cap with a dc servo design. I gave the tester a list of polypropylene types to try out, this should be interesting.
 
Blue Shadow said:
.... Then there is the Hafler DH-200 which has a 10µF cap on the input to the board. Recommended to bypass this one, use a bipolar for sound quality but what size is really needed? These two amps ain't that much different, put in a signal get out 100 watts, what happens inside who knows....
Click to expand...
re. the Hafler DH-200, w/early PC-6 circuit board input: These simulation plots show the response differences for 4 input coupling caps: 10uF, 4.7uF, 3.3uF & 2.2uF (numbers are for 2.2uF).

At 2.2uF the response loss at 20Hz is negligible, below the threshold of audibility. The increased phase & group delay may be more important, tho there isn't much recorded music that low anyway.

You'll get similar response curves for other amps, tho the rolloff frequencies will differ depending on follow-on resistance/impedances.

Consider the Wima MKS2 10uF PET (polyester) cap instead of electro. Small enough to fit comfortably in some boards. Will fit the DH-200 top side, tho not flush to the board.

2.2uF.jpg
 
You must log in or register to reply here.
Back
Top Bottom