Bassman 10 -- Where's the Love?

[bob p]

Can't say I've ever pushed marginal filter caps to breaking with a low frequency test signal

Sometimes it's not the filter caps that fail in the stress test. Sometimes the stress test kills off the cathode bypass caps that people leave behind when recapping!

I've been on a Fender amp binge lately. Maybe I've just been lucky in stumbling across several good finds, but I can't pass up a good deal and these amps tend to follow me home. If good things like this really do come in groups of three, then one more Fender amp is likely to be in my immediate future...

A couple of weeks ago I had a 135W Twin Reverb follow me home. I found it at a Mom and Pop shop. It was above average cosmetically, it played loud and clean, it was a great sounding amp. My interest was piqued by it's pair of EV SRO 12 speakers with the 6 lb alnico magnets. The amp sounded great until I engaged the tremolo... then it started crackling and sputtering and making all sorts of ghastly noises like it was intermittently passing DC. I accentuated these problems during my demo to drive the amp's price way down.

When I got it home the first thing that noticed was someone else's half-assed re-cap job. Some "guru" had re-capped the PSU caps with Atoms, going down to the smaller BF values on the first node, but they skipped the bias caps and never worked on the bias circuit. They had also left all of the old cathode bypass caps in the preamp, either because some of them are hard to get to on this amp or because they still tested OK. Someone had obviously tried to do a tremolo slow-down mod because one of the oscillator low pass caps had grown to a larger value and changed from blue to orange. And the opto bug had been swapped out, but the trem still wasn't working. Obviously, we know where this is headed -- somebody had done hack job on the amp and had cut several corners and fixing this amp was going to be an exercise in fixing someone else's mistakes.

Though the PSU caps had been replaced it only took me a minute to kill it with a simple sine wave signal trace as I worked through the stages. A test signal was enough to take those marginal bypass caps in the preamp and push them over the edge. Those caps still tested fine but they were obviously bad. Replacing them fixed the audio amplifier stages and then I moved to the oscillator.



Someone had definitely tampered with the tremolo oscillator, as they had removed one of the blue drops on the oscillator's low pass filters and replaced it with a larger value OD, presumably to increase the time constant and slow down the oscillator. It looked like they were surprised to find the oscillator not working when they were done, so then they replaced the opto bug. But the tremolo still didn't work! All of the caps tested OK at low voltage but the circuit still wasn't working and it was making horrible noises. Whoever had been working on the amp threw in the towel and sold off the amp as "unfixable."

I know better than to trust a cap tester so I just replaced oscillator caps one at a time, starting with the one that gets exposed to B+. What amazed me was that all 3 caps in the low pass filters were bad. WTH? Then It dawned on me -- ALL THREE of the low pass filter caps all tested good at low voltage, but were all leaking at working voltage. But why? Did the cap that normally gets exposed to B+ fail? When it started passing DC did it then subject the remaining caps in series to B+, causing them to fail like falling dominoes? Or was there a more plausible explanation?

If I had to venture to guess, I'd say that whoever tried to do the oscillator slow-down had overheated the new 0.2uF OD that he was putting in, which sat right in the middle of the series string. When soldering-in the cap, he must have also overheated the 0.1uF blue drop on it's left and the 0.2uF blue drop on it's right, because they shared common solder joints. The result was that when he finished the mod, the caps all tested good on a low-volt tester, but failed to function in circuit, and started passing DC which led to all sorts of noise and inconsistent oscillator behavior. Then he replaced the bug, apparently looking for a fix to the problem anywhere but where he had been soldering, not willing to admit that his poor technique caused the problem. It took replacing all 3 of the low-pass filter caps to bring the oscillator back to life. IMO this had to be a case of heat damage.

The end result is another happy ending: I got a great deal on a malfunctioning Twin Reverb because it was obviously damaged, and I fixed it for only a few dollars in parts.

I guess the moral of the story is that the tried and true rules of thumb still apply:

a. don't do a half-assed cap job by replacing only some of the electrolytics
b. don't trust low voltage cap testers
c. never trust someone else's work.

This isn't the first time I've had to fix someone else's cooking because of oddball damage that was hiding inside of an amp. In this case the tremolo appears to have been broken because of poor soldering technique that ruined the low pass filter caps, and nobody could figure out why the circuit wasn't working. Nobody was willing to doubt their soldering technique and everyone trusted their cap tester. The result was that they could not see the cause of the problem so they gave up on fixing the amp, and it got sold-off as a lemon.

This is the first time I've seen the blue drops cause a tremolo failure, but to be fair I think it was because someone had overheated them.

Sorry for the rambling, but sometimes fixing someone else's work can be an amusing and entertaining experience. I enjoy it when the fix involves solving a problem that somebody else wasn't able to solve, and solving the problem provides a bonus in the form of a great deal on an amp.

[bob p]

Now for a quick tale of premature failure. I just fixed a DRRI this week, built 2008. Its main filter is a 47 uF 500V with a 220 uF 100V stacked in series. There's a 470K resistor across the 47, and a 100K across the 220.... why Fender chose this combination is beyond my imagination. In any case the amp was buzzing ominously, and it turned out the Lelon brand 47 uF was just plain kaput. Whaz going on here? Cheap part doomed to fail in spite of being under-challenged? Go figure.

I have been ranting about cheap quality chinese caps for a long time. but an 8 year old amp isn't premature failure if you're talking about a chinese cap. to me that's old age failure as I rarely see chinese caps that last that long. Where are Lelon caps made?

I seem to remember seeing mostly IC in the recent production Fenders.

When I read your story I thought that you had to be dealing with someone else's hack repair, but looking at the DRRI schem you're absolutely right -- the first node totem pole is a 220uF bypassed by 100k sitting on top of a 47uf bypassed by 470k. That's really odd!

[Leo_Gnardo]

I have been ranting about cheap quality chinese caps for a long time. but an 8 year old amp isn't premature failure if you're talking about a chinese cap. to me that's old age failure as I rarely see chinese caps that last that long. Where are Lelon caps made?

I seem to remember seeing mostly IC in the recent production Fenders.

When I read your story I thought that you had to be dealing with someone else's hack repair, but looking at the DRRI schem you're absolutely right -- the first node totem pole is a 220uF bypassed by 100k sitting on top of a 47uf bypassed by 470k. That's really odd!

The previous record holder was a failed 18 year old cap in a Mesa. That one was parked atop the power transformer EZ-bake oven, the way they do in Mesa land, so I didn't blame it much for quitting. I guess we'll be looking forward to a spate of DRRI & maybe other model power surprise repairs. FWIW the only other amp I've seen such an odd combo of power supply main filters was a late 80's Engl 100W "Straight" head. Strange that Fender does this, a plain 47/500 or a pair of 22/500 made by any competent company would work fine. Nothing is run off the junction between the series caps. DRRI filter caps live in the bathtub/doghouse too, not subject to heat.

[The Dude]

......I seem to remember seeing mostly IC in the recent production Fenders.......

Yep, Illinois Caps- now made in China. And, I'm noticing a higher than normal failure rate of these in the newer Fenders. In fact, just this week, I fixed a relatively new Hot Rod Deluxe with one cap totally dead and 2 others oozing.

Edit: I think I was wrong. I believe it was the DRRI. The customer had two amps.

[g1]

looking at the DRRI schem you're absolutely right -- the first node totem pole is a 220uF bypassed by 100k sitting on top of a 47uf bypassed by 470k. That's really odd!

On a bit of a tangent, this faintly rang some bells, a bit of searching and I see we did discuss this before. Some interesting speculation and I guess I never got an email reply back from Fender as to why they did this.
Enzo phoned someone at Fender and their response was "I have no idea".
http://music-electronics-forum.com/t36373/

[bob p]

On a bit of a tangent, this faintly rang some bells, a bit of searching and I see we did discuss this before. Some interesting speculation and I guess I never got an email reply back from Fender as to why they did this.
Enzo phoned someone at Fender and their response was "I have no idea".
http://music-electronics-forum.com/t36373/

Interesting thread. I posted my opinion there. Thanks for the link.

[bob p]

I got sidetracked a bit, needing to fix someone else's amp, so the Bassman 10 got put on the back burner for a while. Steve and Leo, you guys brought up some interesting ideas about the idiosyncracies of the B-10 circuit, and now that I've had a chance to focus my attention back onto this amp I'm going to give my opinions on them, based solely on the schematic. (I haven't attacked it with a soldering iron yet).

Looking at one schematic if you were going to use this for guitar you'd probably want to remove the .0047uf cap between the treble and volume control on the Studio channel for starters.

Steve, I agree that my first impression was that I should rip out that anomalous 0.0047uF cap, because if it doesn't exist in other amps there's no reason to have it in this one. But I honestly don't think that removing that 0.0047uF cap that goes from the Treble wiper to the Volume pot is going to help much. I actually think that removing it would end up doing more harm than good.

Here's why:

Looking at the magnitude of the treble pot (250k) and the volume pot (1 Meg) resistances, I think that the majority of the effect that will be produced by that 0.0047uF cap will be through the formation of a high-pass filter by the cap and the volume pot. I'm going to ignore the resistance of the treble pot for the time being, just because the impact of the 1 Meg pot is bigger. It would make an interesting simulation, though, to also factor in the effects of the treble cap, but for now I'm going to keep the analysis simple.

Looking at the combination of the 0.0047uF cap and the 1 Meg pot, they form a high-pass filter with a -3dB frequency of 34 Hz. I think that the reason that they put that cap in place instead of using the traditional Fender jumper wire was to protect the speakers in the ported cabinet from damage due to over-excursion.

At the time that the amp was built Larry Graham was already slapping; slapping the strings places a tremendous LF strain on a speaker cabinet. Whether Fender was actually worried about slapping, or just LF over-excursion in general, it looks like they realized that over-excursion in this tightly vented box was a real threat to speaker life, so they made a deliberate decision to roll off any "useless" LF content that could damage the speakers. Realizing the the 4-string bass' low E had a 40 Hz fundamental, they designed a 34 Hz high-pass filter to get rid of all of the dangerous LF content.

Honestly speaking, I think the only reason that the 0.0047uF cap is there is to offer out-of-bandwidth protection for the speaker array. Guitars and Basses of that era never produced content below 80Hz and 40 Hz respectively, so the filter would never effect the reproduction of the instruments' sound. I think it's there only to protect the speakers -- just like the LF rumble suppression circuits that we used to have in our stereo preamps when everyone had turntables. Those rumble suppression circuits were there to protect the speakers from death by over-excursion in the event that a tonearm was dropped onto a record.

In the big picture, I don't think that replacing the 0.0047uF cap with a jumper wire will make any difference, because the LF content that gets removed by the filter is below the LF limits of the guitar and bass as signal generators. I'd wager to say that you put that 0.0047uF cap into any of the other vintage Fender amps you'd never even notice that it was there.

[bob p]

Another minor mod to wake up the "studio" channel: there's a treble-deadening cap in parallel with the first stage plate resistor, clip it out and hear the high frequencies come thru.

... There are also .002uF caps to ground right where the 100k bias resistors connect to the 1.5k grid stoppers.

What's interesting about this amp is that if you calculate the rolloff frequencies for both of those filters, it's evident that Fender was executing a specific plan for voicing this amp; Fender placed the same voicing constraints on the front and back ends of the amplifier.

Looking at Leo's suggestion of the Normal channel's plate resistor bypass cap: it's definitely there to limit the first stage's frequency response to midband signals. The combination of the 100k plate resistor and the 0.001uF bypass cap form a 1600 Hz low-pass filter. The result is decreased gain at this stage for that mid-frequency and above content. Although that's great if your intent is to create a traditional rounded bass tone, losing all of that 2k-5k content doesn't appeal much to guitarists. There's an obvious explanation as to why this channel wouldn't appeal to guitarists. It's got lots of low-mid twang and it's devoid of upper harmonic content.

But clipping out that 0.001uF cap won't be enough to bring the amp out of it's dark-sounding character. There's still that capacitor nonsense between the grids of the output tubes and ground that has to be dealt with.

Now take a look at the plate resistors on the Schmidt Inverter. Unlike the typical 82k/100k pairings that you find on all of the other Fender amps you'll see that the values have been reduced to 47k/47k. That accomplishes two different things: First, it makes the phase inverter operate on a different load line than other Fender amps, with lower gain. Second, the 47k resistors combine with the 0.002uF caps from grid to ground to form a low-pass filter at the input to the output stage. It has an f3 value of about 1700 Hz. Once again, Fender is deliberately voicing the amp to roll-off the upper midrange content.

It would seem then that Leo's suggestion of snipping out the 0.001 cap on the Normal channel's plate resistor would definitely restore the frequency response of the first stage in the normal channel, and that that will be necessary but not sufficient to completely "open up" the amp's midrange response. To completely fix the problem, those 0.002uF caps on the 6L6 also need to go.

I think that these circuit mods by Fender are interesting -- Fender made a conscientious effort to voice this amp, with a specific intent to make it a bandpass filter in the range of 34 Hz to 1600 Hz, rolling off frequencies below and above those limits. I can definitely confirm that in it's native form, the amp does seem to be very honky and midrangey in character and is definitely lacking the top-end sparkle that is stereotypical for Fender's SF amps. I think I've just figured out why this amp isn't particularly popular among guitarists, and exactly how to remedy that situation.

I'll report back after I've had a chance to snip out some parts that need to go into the trashcan.

[bob p]

I guess I should finally post back and update this thread -- the mods I described in the preceding two posts work just as the math predicted -- a few snips with a pair of dykes and the amp opened right up.