Changed a cap and lost some bass...?

[hnb2907]

Hi folks,

I've been tinkering around with my Chinese amp, and have replaced a 100nF 400V MKP1841 Metallized Polypropylene Film Capacitor on the 6J8P frontend. I've dropped in a 100nF 630V of these as a replacement http://wduk.worldomain.net/acatalog/SONIQS_SAX_range.pdf

Now, the mids have more clarity, the treble is bright (maybe too much), but the bass seems almost none existent.

I've put the old cap on a component tester at work, which claims it is 98pF, 4ohms@100hz, 0.25ohms at 1khz, 0ohms at 10khz.
The new cap measured the same way claims it is <0.25ohms at all those test frequencies. Maybe the different internal resistance of the cap at 100Hz is the reason?


As I'm a beginner here, the question is....

Looking at my circuit diagram, the cap changed is C1. Would replacing both R1+R2 with, for instance, 15Kohm each change the time constant of the R/C network to roll off at a lower frequency? I'm thinking that 1K5 to 15K could produce a noticable change of 1 decade. Then through trial and error find the correct resistance? Maybe I should be looking at the 8K2 resistor?

I'm basing this idea on the idea that the cap is somehow causing AC negative feeback onto pin 6 of the 6J8P, and that C1+R2 or C1+(R1//R2) are creating the low frequency cut off.

Maybe someone could give me a quick explanation of what's going on around the 6J8P and how to tweak it...?


Cheers,
ChrisC.

[ray_moth]

If you increase R1 and R2 to 15k, you will have 15k in series with the cathode and the bias will be so high that the tube will probably not be able to conduct. Your problem is almost certainly C1, which bypasses grid 2 to the cathode and is fairly critical. If its value is too high, it could cause LF instability because of the negative FB loop. If it's too low, the symptom would indeed be weak bass and the middle and high frequencies would be emphasized. The replacement cap you've fitted may be defective in some way.

[hnb2907]

Originally posted by ray_moth
If you increase R1 and R2 to 15k, you will have 15k in series with the cathode and the bias will be so high that the tube will probably not be able to conduct. Your problem is almost certainly C1, which bypasses grid 2 to the cathode and is fairly critical. If its value is too high, it could cause LF instability because of the negative FB loop. If it's too low, the symptom would indeed be weak bass and the middle and high frequencies would be emphasized. The replacement cap you've fitted may be defective in some way.

Thanks for the sanity check - I'd completely missed thinking about the bias!

I'll have a check of the new caps on the meter. Is it more likely that the difference in internal resistance is causing this, rather than them being faulty?


Cheers,
Chris.

[Yvesm]

Originally posted by hnb2907

. . .Is it more likely that the difference in internal resistance is causing this, rather than them being faulty?


Cheers,
Chris.

I can't beleive the internal resistance could have such effect !
In your first post, you say having measured 98pF ? Typo ?

Yves.

[kevinkr]

Try a different 100nF (0.1uF) cap in that location, it could be that you don't like the subjective change in tonal balance that the new capacitor gave you. (Other brands may sound different depending on construction.)

Even (good) cheap amplifiers are often built with combinations of components chosen through careful listening to work synergistically with other choices made in the design. It is quite possible that the substitution of a better cap (or one that at least measures better) has unmasked a design compromise somewhere else in the circuit. Or it could be a bad sounding cap...

I would also measure the frequency response of the entire amplifier, and you will probably see it is reasonably flat down to at least 50Hz.
Do this with an 8 ohm resistive load, with an output voltage at 1kHz of say 1.0Vrms - sweep the frequency down to 20Hz and up to 20kHz.
Note that 707mVrms represents the -3dB points in the response curve. (You need to use a meter with better than 20kHz bandwidth or a scope.) Note that this is the small signal response, but should not be far off the mark for typical listening levels with efficient speakers. (125mW)

Finally I assume you meant 98nF, not 98pF - this would certainly result in rather poor low frequency performance..

[hnb2907]

Originally posted by kevinkr


I would also measure the frequency response of the entire amplifier, and you will probably see it is reasonably flat down to at least 50Hz.

I did this a few weeks ago before I made any changes at all to the amp, and it was flat from around 55Hz to about 27kHz. Had planned to pick up my signal generator from my workshop on my way home tonight, went in to collect my snailmail, and still forgot to bring it home. Doh.

Finally I assume you meant 98nF, not 98pF - this would certainly result in rather poor low frequency performance..

indeed - 98nF - another one of those days - as you can probably tell?!

My current plan is to check the frequency response and then put the original caps back in, to make sure the sound returns to how it was. I've got a collection NOS 6SJ7's of various ages/manufacturers on the way to replace the Chinese 6J8P with. maybe I'll go back to the caps another day....

Anyway, I'm still having fun learning a bit more about all this - thanks for your help eveyone...

C.