Mods for Fender Champion 30

[Steve A.]

Effect of capacitor in NFB loop of IC?

I know how it works with resistors in the NFB loop of an IC stage (lower value resistor for less gain/higher value resistor for more gain) but how would that work with capacitors?
Specifically the chip is a TL072, the capacitor is 10pF and the resistor is 220k. This is from the clean channel of a ~2000 Fender Champion 30 which I think is too bright and I would like to mellow it down a little bit. Capacitors pass low frequencies better than high frequencies so I would think that I would want to increase the capacitance for a less bright tone. For coupling capacitors in tube circuits you often want to go up or down by a factor of 10 for a definite change in tone so for starters I guess I might try a 50pF or 100pF cap. Any thoughts on that?

Here is the link to the schematic:

http://support.fender.com/schematics..._schematic.pdf

Thanks

Steve Ahola

P.S. This amp is very loud and I would like to add a master volume to be able to cut the output level a bit. I was thinking of a adding a pot at the negative input of U4A. Perhaps a 100k pot wired as variable resistor to ground... Any thoughts on that?

The reason for the MV would be to reduce the always-present background sh-h-h-h-h noise if I wanted to record the amp. It just occurred to me that it might be better to use a SPST switch for attenuation, with a trim pot to set the amount of attenuation. That way with the switch off the circuit would be stock. I think it might be better if I added the switch on the output of U4A so that it would not affect the mixing circuit of the reverb signal...

[52 Bill]

I know how it works with resistors in the NFB loop of an IC stage (lower value resistor for less gain/higher value resistor for more gain) but how would that work with capacitors?

It works the same way, you can reduce the high frequency by increasing the value of the 10pf cap. To tailor the response it may require a cap with a resistor in series. I don't remember the formula for the frequency cutoff point, but I guess that you can find that online somewhwere. Of course all of this is only true when the cap connects the output with the inverting input.

Adding a master volume before the power amp chip will work, but how much of the hiss is being generated in the power amp circuit itself?

[tedmich]

a few of the more common NFB filter circuits common to op amps (with scary math!) can be found here
Introduction to Filters

[Enzo]

It is only 10pf, I think it is there for stability rather than tone. It would be easy enough to shape the tone a bit less bright without messing with that.

And I may be wrong, but I think increasing that cap will just move the whole effet to lower freqs. In other words it won;t make the 10kHz reduction greater, it will add 5kHz to its range. I made up the freq numbers.

JJust an opinion, but instead of changing those 10pf caps, look at the last op amp in the preamp, it has a 1000pf cap. Now that one does affect tone. But you could also pur a rolloff right at the entry to the power amp or a couple other spots, you could mess with the tone stack, All that stuff between U1a and U2a is avaliable for tweaking. And the 470pf cap on U2a.

[Steve A.]

It is only 10pf, I think it is there for stability rather than tone. It would be easy enough to shape the tone a bit less bright without messing with that.

And I may be wrong, but I think increasing that cap will just move the whole effet to lower freqs. In other words it won;t make the 10kHz reduction greater, it will add 5kHz to its range. I made up the freq numbers.

Just an opinion, but instead of changing those 10pf caps, look at the last op amp in the preamp, it has a 1000pf cap. Now that one does affect tone. But you could also pur a rolloff right at the entry to the power amp or a couple other spots, you could mess with the tone stack, All that stuff between U1a and U2a is avaliable for tweaking. And the 470pf cap on U2a.

Thanks for your suggestions! However the 1000pf cap you mention is after the tone stack; the OD channel is not too bright, just the Normal channel. So I'm thinking that the point to add a cap and trim pot to ground might be the output terminal of the Normal volume control.

Hmmm... I just checked the schematic for the Champion 30's big brother- the Princeton 65. It has the same basic circuit for the Normal channel volume control but it uses a 100pf cap instead of a 10pf. The Normal channel has its own bass and treble controls ahead of the volume control so that might make a difference but I think I will go ahead and first try tacking a 100pf cap onto the legs of the 10pf cap.

Thanks again for all of your help in my projects!

Steve Ahola

[Steve Conner]

A capacitor across the feedback resistor of an op-amp forms a low-pass filter. It kicks in at a frequency equal to 1/(2*pi*RC) and rolls off 6dB/octave above that.

So making the capacitor bigger will lower the frequency at which the rolloff starts, but it'll also increase the amount of rolloff at any given frequency above that.

Small values like 10pF are probably just to prevent the op-amp from oscillating, rather than shape the tone. But if you feel the tone is too bright, go ahead and make it bigger. You can use a trimpot in series with the capacitor, of roughly the same value as the feedback resistor, to adjust it.

[J M Fahey]

Dear Steve A. , glad to be of some use to you, after all you did for us during so many years.

1) As of your problem, if you find the normal channel very bright, it's easy to correct.
It has an "always on" active bright circuit, attached to the volume pot.
Just lift 1 leg of C13 and you're done.
You can add a "bright" switch to put it back at will, or, even further, add a small rotary switch (or a toggle, multi way type) to play with, say, OFF/470/1000/2200/4700pF

2) If you want to let the amp power crunch *a little* at less than full power , cut the track between R62 and C41 , wire R62 now free pad to a master volume pot and from its wiper come back to C41.
Use a 25K or 50K Log/Audio Pot (linear will work but not be as smoooth) and solder across it (from hot to ground terminals) a couple opposite Leds, your choice of colour.

Before tubeheads cut me to ribbons, SS power amp crunch *does* have some uses.
It's not bright and thumpy as tubey one, rather the opposite, flat and middly ... but sometimes this is good, to tame a buzzy distorter, be it external or the preamp itself.

Which in a way will solve part of Steve's problem.

Please post results.

[Mark Hammer]

Op-amp 101

Op-amps want to go full tilt....all the time. While we like to think of the various external components as adding gain, in reality they are adjusting the gain by applying negative feedback and "stepping on the brakes". The more negative feedback is applied from the output back to the input, the less of that maximum gain ends up being applied.

The most obvious way of considering gain and negative feedback is the value of the feedback resistor. Larger values of feedback resistance will diminish how much negative feedback is applied and allow the op-amp to inch closer to full tilt.

When the op-amp is in non-inverting mode, gain is set by the feedback resistance AND the ground-leg resistance from the inverting pin. Those two resistances act like a voltage divider would, with the inverting pin as the "wiper" and the op-amp's output as the input to that voltage divider. So, like any good volume pot, if the input-to-wiper leg resistance is much smaller than the wiper-to-ground leg, more of the signal (in this case, negative feedback) will be preserved. And, if the input-to-wiper resistance is noticeably higher than that of the groubd leg, you bleed off most of the signal (negative feedback).

How do you make a unity-gain op-amp? You bridge the output and inverting input with a piece of wire (zero ohms), and you provide NO ground leg resistor (functionally infinite resistance, though set by the chip's innards). It's like having a 50meg pot turned up full - no loss of any negative feedback, hence least gain the chip can provide: 1x.

So far, so good. Caps in the feedback path OR the ground leg will serve as frequency-sensitive resistances, and end up applying more negative feedback for some frequencies than others. If I stick a 1nf cap in a feedback loop, I have provided what amounts to a zero-ohm feedback path for higher frequency content. And, as mentioned above, if the input-to-wiper leg is a much lower resistance than the ground leg, more of the signal is retained. The result is that, with more negative feedback for top end, one steps on the brakes harder for top end, and gain is pushed farther away from maximum. Where that happens in the spectrum is a function of the feedback resistance and cap value.

I hope this makes sense. The key construct here is to think of op-amp gain as selectively diminished, rather than added.

[Steve A.]

1) As of your problem, if you find the normal channel very bright, it's easy to correct. It has an "always on" active bright circuit, attached to the volume pot.
Just lift 1 leg of C13 and you're done.

I was wondering about that cap- why it was there and what it would do. Being on the negative input side of the IC stage I see that it would add brightness rather than cut it. It should be easy enough to desolder one lead to see if I want any kind of brightness cap there. Wiring it in series with a pot would make it like the bright side of a 5E3 tone control.

2) If you want to let the amp power crunch *a little* at less than full power , cut the track between R62 and C41 , wire R62 now free pad to a master volume pot and from its wiper come back to C41. Use a 25K or 50K Log/Audio Pot (linear will work but not be as smoooth) and solder across it (from hot to ground terminals) a couple opposite Leds, your choice of colour.

Thanks for the tip!

Before tubeheads cut me to ribbons, SS power amp crunch *does* have some uses.
It's not bright and thumpy as tubey one, rather the opposite, flat and middly ... but sometimes this is good, to tame a buzzy distorter, be it external or the preamp itself.

As much as I like power tube distortion it is not that easy to get playing live since you usually can't crank up your amp that loud. (And if you use a smaller one it might not be powerful enough for the venue.) Is power scaling a solution to that? Longtime AMPAGE contributor Richie Hall and his brother sell VVT kits to add to amps- will that help with power amp distortion? IMO the stomp boxes available today can offer a reasonable fascimile of preamp tube distortion. If a VVT kit will allow you to get power tube distortion at lower volumes that would make a great combination.

I hope this makes sense. The key construct here is to think of op-amp gain as selectively diminished, rather than added.

Thanks! So what is the gain of a TL072 stage with no negative feedback? Just wondering...

Steve Ahola

P.S. I have a few books on IC audio circuits from my days building all of those Craig Anderton kits that I should study!

[J M Fahey]

So what is the gain of a TL072 stage with no negative feedback? Just wondering..

.
HUGE , although it falls 6dB/oct towards high frequencies.
Yet it still is around 100X (40dB) up to about 30 KHz, not bad at all for Audio circuits.

[Mark Hammer]

ISo what is the gain of a TL072 stage with no negative feedback? Just wondering...

What you look for is the graph you will likely find in every op-amp datasheet that is generally titled either "open loop gain", "gain bandwidth product", "open loop frequency response", "Gv-F", and a couple of other similar names that will summarize a graph that has db gain on the vertical axis and frequency on the horizontal axis. The typical picture will show something like 100db "open loop gain" (100,000x) at low (subaudio) frequencies that stays high until maybe 1khz or so (higher in some chips) and then starts to drop dramatically as frequency increases. The pic that Juan showed in one example.

Slew rate is related to, but not identical to, this graph. Slew rate depicts the extent to which the op-amp will "respond" to input signal. If something slews at 9v/us, then it is not going to be impaired in providing an output of 1V at 10khz, because either the output amplitude or the frequency would have to be MUCH higher for that slew rate to pose any limitation.

The gain-bandwidth product also depicts what the chip can do, as you move up the frequency spectrum. The general principle is that if you applied NO negative feedback, the chip would happily provide up to 100db gain at low frequencies, but noticeably less at higher frequencies. Clearly it won't be able to actually do so with a 100mv input and 9v supply, because it will quickly run out of headroom, but assuming you had a 10uv input signal, and a +/-18v supply you ought to have little problem knocking that input signal up to +/-10V....the hum and hiss might drwon it out, but we'll conveniently ignore that part.

While modest open-loop gain may be seen as a shortcoming, in some contexts, it can provide added value. The venerable Proco Rat fuzz used an LM308 op-amp, and at full tilt would request a gain of just under 70db from the chip for content above 1500hz or so. That's interesting, but more interesting was the fact that with no negative feedback whatsoever, the LM308 could only yield about 50db gain at 1khz, and about 40db gain at 6khz or so, using the compensating cap Proco did. Obviously the gain was even more limited, going up from there (about 30db at 12khz). The net effect was to filter out a lot of the "fizz" one might expect from a distortion device by deploying its gain-bandwidth limitations to advantage.. Anyne attempting to make a Rat with a CA3140 would be irritated by the tone produced with a chip capable of 70db gain at 1khz, or a TL081 still rockin 60db gain at 5-6khz.

[guitician]

I remember making a fuzz with an LF357 that has a 50v/us slew. It was great at getting nice square waves and made the guitar sound like a synthesizer.

[Steve A.]

Okay I removed C13 which as Juan Manuel Fahey pointed out acts like a bright cap in the clean channel. Much better but it was still a bit too trebly so I added a 100pf cap across C10 (I actually mounted it on the legs of R19 since there was more room there.) That got the clean channel working great but the OD channel sounded kinda fizzy so I screwed around with the NFB loop on U2B. I put a 100pF cap across the 470K NFB resistor R15 which smoothed it out a bit and then added a 560k resistor across it to bring the net resistance down to 255k. Heck the OD channel is sounding so good I might even use it once in awhile.

A big thanks to everybody for their suggestions and explanations of the IC circuitry in the amp.

Steve Ahola

EDIT: One more change. I soldered a 1500pF cap in parallel with the .0047uF cap for C22. On the treble control this is the capacitor going to ground.

I also noticed buzzing notes from the reverb tank (mounted in cardboard and isolated from chassis vibrations by two pieces of thin white foam- not very pliable after ~14 years) I used a piece of sponge rubber weatherstripping to provide more isolation. No more buzzing notes. Hooray!

BTW it is really tricky taking out the chassis and putting it back in. Unless it is lined up exactly the corners will dig themselves into the sides of the cabinet and trash the foil shielding trying to get the chassis unstuck. So I got two pieces of .008" tin flashing from Ace Hardware, cut them to the right shape and glued them to the cab with Pliobond 25 contact cement ($7.99 for 8 oz vs $5.99 for 3 oz for DAP Weldwood.)

My father worked in a sweatshop making formica countertops back in the 50's and 60's and they used Pliobond for their contact cement. This was back before low vapor standards and it was quite toxic to breathe if you didn't have fans blowing in fresh air. In any case they bought it in 5 gallon buckets (or even bigger) and my father would take some home in a coffee can for projects and repairs around the house. The plastic liner in our refrigerator got cracked so my father fixed it with Pliobond- I don't think that we ever got rid of that smell! So I had developed quite a dislike to that shit but I figured I would give the modern Pliobond 25 a chance to get back in my good graces. Yep- it is a good adhesive. There is a warning to be sure to clean up any that you spill because when it hardens it will not be easy to remove.

Getting back to the sweatshop in Lafayette they had a showroom on the ground floor and the shop was in the basement with no windows as I recall- it smelled of sawdust and pliobond. My father developed cardiopulmonary problems and I am SURE that the sweatshop had contributed to much of that. Blodgett's Linoleum eventually dropped the formica countertops so my father went to work for Andco on Adeline in Berkeley not far from Ashby. This shop was on the ground level with big doors and plenty of ventilation- the way you would expect it to be! The idea of putting a shop like that in a basement was sheer idiocy!

[Steve A.]

Crap! I think I burned it up!

I had just put the finishing touches on the amp- it was working beautifully- but I had to take the chassis back out because the plastic bushing for the treble control went in crooked. When I put it back together I forget to put the speaker leads back on the speaker- something that I had already done several times but I'd shut it right off because I didn't hear the background noise from the circuit. Only this time when I reconnected the speaker leads there was still no output.

I checked the voltage on the PT and that was normal (20vac to CT on both hot leads/40vac lead to lead). I thought I better check here before proceeding any further.

Will the power chip burn up if run with no load? Will it burn up if the speaker leads were shorted together? The number on the power IC is TDA1514A which I see is available on eBay from China.

I guess I better find my 8 ohm dummy load and plug it into the ext speaker jack before doing any further tests.

Thanks for all of your help on this!

Steve Ahola

P.S. I just dl'd the spec sheet for the TDA1514A and see that there is thermal protection which might have saved my ass.

http://www.datasheetcatalog.org/data...ps/TDA1514.pdf

[Steve Conner]

Running a SS power amp into an open circuit should do it no harm at all. Now if the speaker wires had been touching each other, that would be bad.