Can I increase capacitance on these TONE BOARD caps?

[roger2]

Hi all

This is not a DIY amp project so I hope it is OK to post this question here. I am restoring and doing modest mods to a Kenwood KA-4006 integrated.

My question is in regards to caps Ci15 and Ci16 shown in the schematic attached. They are currently 100uF/10v. It would be just as easy to replace these with 220uF/10v as they are close to the same physical size as the original caps. I will be using Panasonic FM.

Could it be detrimental to this circuit to increase capacitance of these two caps? If so why? Could it be beneficial? Better power available to the opamp or better decoupling?

Thank in advance...

[Jeff R]

They are power supply filter caps, so increasing the value should be fine. However, is there additional filter capacitance for the B+ and B- lines? If so, if might make better sonic sense to use a metalized polyproplyene of 10 uF. It is not so much the absolute value of capacitance, but the quality of the capacitance.

[Zero D]

nvm

[roger2]

@ Jeff R

I believe (I am a bit noobish) that there is additional filter capacitance for the B+/B- lines on this board if you count the lines that supply this board...they come directly from a simple shunt regulator (14v zeners) with 470uF/16v electrolytics on the downstream side of the zeners. But there are 680R resistors, Ri26 and Ri27, between the other voltage source and the opamp. Does that mean anything? Is it easier for the opamp to draw from Ci15 and Ci16 than from the other caps? There has to be a reason they were put there by the designer, what was that reason?

I don't think 10uF of any type of poly will fit there. The 220uF Panasonic FM will fit just like the original 100uF. I am not looking for drastic improvements or putting in major effort. Like I said, physically, the new 220uF is about the same size as the original 100uF. Easy change. Just don't want to throw something out of whack.

Beside the role of power supply, do the caps in question, Ci15 and Ci16, also decouple the opamp?

[Jeff R]

The resistors decouple the op amp power supply. Theoretically, the resistors form a low pass filter (Ri28 with Ci16), reducing any voltage fluctuations on the input supply lines. In this case, you want lots of capacitance (i.e. Ci16) to supply transient power, as you can't get a lot of instantaneous current through the series resistors. That implies you need a good quality cap to supply high frequency current, since it can't come from the main power supply due to the resistors.

Using 220 uF won't cause a problem, though.

[roger2]

thank you Jeff R

After posting here I also posted this question, and some other related questions, at AudioKarma to get some additional opinions. Several members there agreed with you that it would be OK to increase from 100uF to 220uF on these two caps

[indianajo]

At >20 years I would worry also about Ci1 and Ci2. These show to be electrolytic, and if they dry up they can distort the input sound. I've had this happen in my Sony 250 taperecorder I was using as a preamp, as well as my dynakit st120 power amp. You should force film caps in there even if you have to fly them above the pcb and lengthen the leads a little bit. You should also confirm Ci5 and Ci5 are not electrolytic (plus on one end or minus in balls pointing at one lead, or "np" after voltage) and if they are , change them too. These need to pretty well match the old ones. That is if you don't want to take the amp apart again and do it later. Electrolytics are like a "time fuse" that blows eventually.
On my ra88a disco mixer, the resistors in the local input section power rails were to reduce the hum from the power supply. The op amps were capable of taking the higher ps voltage, but hummed worse without the resistors, even with bigger caps at the op amps. Sloppy but cheap technique to reduce hum.

[roger2]

Thanks indianajo, lots of good thoughts there...

(BTW, Anderson IN here)

RE: Ci1 and Ci2, these have already been replaced with Silmic II. In fact all of the electrolytics have been replaced, but I will be going back into this board again to exchange the new PW caps I used at Ci15 and Ci16 with FM, most likely of higher capacitance as was the original question in this thread.

Having had more time to think, and on your suggestion as well, I would like to put film at positions Ci1 and Ci2. I think a Wima MKP would actually fit there no problem. But I have some issues with the PCB traces on one of these and, because of that, will probably opt to leave the Silmic II's in place. The pad is very narrow and part of it simply flaked off and is gone. I would be afraid to do any more soldering there, as the current Silmic II cap appears to be solidly in place. Film would be nice, but I think I better leave well enough alone in this case.

Ci5 and Ci6 are indeed electrolytic. They are NP and have already been replaced with Nichicon Muse NP. Replacing those with film, although again, a good idea, may be impossible due to the layout of the board and space limitations. But I will double check this.

As far as mounting caps up off the board and extending the leads, I would be hesitant to do this as I have read warnings against this at Rod Elliot's site.

regards

[indianajo]

Your new NP electrolytic caps should be good for another ten to twenty years, I wouldn't worry about redoing it now. Where'd you get them? Newark has no NP caps, Mouser lists Lelon as NP caps in the index, but nothing about that on the catalog page.
I flew a lot of film caps above the board on my Hammond organ, but those only go up to 8000 hz so I can't say it was successful at a full audio 20 khz. I also left the old leads in the board, cut the leads and hooked over the leads from the new caps. I don't like lifting traces off the pcb, do it more than necessary I'm sure.
Modern parts can oscillate with excess lead length. I had to stomp out a 1mhz oscillation on my Ra88a mixer when I replaced 4558 op amps (slow slew rate) with 33078 op amps (7 v/microsec slew rate).
The hammond does not have modern transistors.

[roger2]

Mouser has Nichicon ES series NP electros that are marketed as "Audio" type. I have read opinions in threads that these are good to use wherever a NP is required (unless you go to poly of course).

Link: Mouser Electronics - Electronic Component Distributor Aluminum Electrolytic Capacitors - Leaded

I just started at the full Mouser product list screen and worked my way down. You can see the path near the top of the screen in this link. This show all their NP electros. From there I would choose Nichicon for the manufacturer to narrow it down further, then look for the size needed in ES or "UES" to match the naming convention used.

See pic attached of Nichicon Audio cap line.

Regarding flying caps: The warning at the ESP site had something to do with increased inductance caused by the longer leads. It is recommended there to always keep cap leads as short as possible. But in your case...up to 8kHz only...you may be right and maybe this wouldn't be an issue. I do not know honestly.

Interesting that a mere 7v/uS slew rate caused oscillations. Although that is higher than the 4558's slew rate, it is still much lower than other current op amps (I am learning this practically as we speak...I am engaged in a similar thread at AK)

So how did you become aware of the oscillation in your mixer and how did you fix it?

.