Question about small caps on preamp tube pins

[Steve A.]

A friend brought up a question about using a low value 1kv ceramic cap from the plate to the grid on the initial preamp stage: if the cap was to short out wouldn't that create a safety hazard? Like if someone was to touch the tip of a cable plugged into the amp? Or could it damage the pickup or guitar circuitry?

Would a mica cap be less prone to short out? FWIW I am more familiar with caps failing open. I guess it would be safest to use a short length of shielded cable as the capacitor, or little maze like Peavey uses on their C30.

Thanks!

Steve Ahola

[soundguruman]

That's why we use 1000V silver mica, or a 600 volt shielded wire with the grid connected to the center and the plate connected to the shield.
Boogie amps have been using 1000V coupling and snub caps for some time, and that is not what fails.
Between the center and the shield, with the insulator in between, is the wire capacitance necessary. The longer the wire the more capacitance.
But actually in a tube amp anything could short, including the tube. And it might damage the pickup, etc,..."might" is a pretty far fetched possibility of odds.
In fact pretty much all technology "might" short and burn. But day after day the vast majority does not.
We all use good quality parts, wire and tubes that we envision as "short circuit proof" or failsafe. And we know they do fail, but very rarely when assembled correctly.

[jrfrond]

A cap across the plate resistor with shape the stage response as a low-pass amp, with a fairly-sharp cutoff frequency. It limits the upper bandwidth.

A small cap at the grid to ground is used to choke RF at the stage input. In places with high RF, like here in NYC, it is always a good idea. I place it right at the tube socket.

[Chuck H]

On my last design I have the first preamp tube VERY close to the input jack. It's a very high gain amp and with just a resistor and it's leads from the jack to the grid pin the amp is unstable at the highest gain settings. It's also glassy bright. Rather than try to incluse a 1-1/4" length of shielded lead I went with the grid to plate cap. As described above I ended up at 4.7pf. Amp is stable and it just trims the glass off the top end. But as mentioned in other posts it doesn't detriment the clean tones as badly as the ground shunt cap. I'm happy.

[loudthud]

On my last design I have the first preamp tube VERY close to the input jack. It's a very high gain amp and with just a resistor and it's leads from the jack to the grid pin the amp is unstable at the highest gain settings.

I want to try this on a TW Express type amp, and a few other layout tricks. No sense running signals any farther than necessary just to use an eyelet board.

[Steve A.]

That's why we use 1000V silver mica, or a 600 volt shielded wire with the grid connected to the center and the plate connected to the shield.
Boogie amps have been using 1000V coupling and snub caps for some time, and that is not what fails.
Between the center and the shield, with the insulator in between, is the wire capacitance necessary. The longer the wire the more capacitance.
But actually in a tube amp anything could short, including the tube. And it might damage the pickup, etc,..."might" is a pretty far fetched possibility of odds.
In fact pretty much all technology "might" short and burn. But day after day the vast majority does not.
We all use good quality parts, wire and tubes that we envision as "short circuit proof" or failsafe. And we know they do fail, but very rarely when assembled correctly.

I agree although I do like to stack the deck in my favor. How about running two 1kv ceramic caps in series, which will lower the capacitance and bring it down to those low values which might be all you need.

Steve Ahola

P.S. Probably not a good idea to play poker with me!

[Gingertube]

This is where experience counts - The pure design engineer (like mugins here) will tell you that the B+ rail is at AC 0V and so putting a cap across the load resistor or putting a cap from anode to the real 0V should do the same thing - right?
Well NO - the B+ is only at AC ground because we tie it to 0V (AC wise) with the bypass cap. So when your electrolytic bypass cap becomes inductive at higher frequencies then B+ is no longer equivalent to the real 0V - make sense?

You really do want B+ to look like an AC ground to much higher than audio frequencies and that is why I (and many others) add a film cap (100nF to 470nF, 630V ) across each elctrolytic B+ bypass cap in the preamp. Once again there will be some who object to that on the basis that adding a film cap across an electrolytic which will look inductive at high frequencies will cause that pair to act like a tuned circuit and it can cause high frequency (super sonic) ringing. That is possible but I've never seen it happen. If in doubt put a 2,2 Ohm resistor in series with the film cap. That will kill any ringing.

This is also why an old amp will sound brighter when you do a recap job even if you stick to exact original values.

Cheers,
Ian

[Chuck H]

One interesting thing about the TW type amps is that looong instrument cables are suggested as a way to tame the overly glassy top end at lower distortion settings. Glen kuykendall (if you know who he is) uses fourty feet of standard instrument cable in front of his Express. I didn't see the sense in using extra cable as a needed compliment to the amp. So my little 4.7pf cap solves for that too.

Now... My design is super similar to a TW design and uses a similar layout ideology, BUT, in my defense I did come up with it around 1999 when no one knew what was inside a TW amp. However, my third gain stage was a typical 1.5k cathode but with no bypass cap. After the TW design became known I tried the "cold clipping" thing and ended up liking 6.8k the best. So now my design is just a little more Wrecked than before. But I swear I didn't clone a Wreck when I designed it. My amp is a little lower gain than a TW too. But even "KF built" Wrecks are often reported to go into oscillation if the gain is advanced too far. That's unacceptible in a market design. I would consider my amp like a lay players Wreck in behavior. A little more user friendly and can actually do a clean tone without the need to adjust the guitar volume. You get a ton of great OD without the need to tame the beast with dark speakers, long input cables and limited yet fussy adjustments. Not that I think the TW design is a bad amp. Quite the contrary. I just don't think it's the kind of amp that the average guy walking into Guitar Center can even begin to understand and use. So in that regard my amp is sort of like "TW for Dummies". I'm still (and ever) making refinements. But if your interested I;ll send you a layout and schem.

[Steve Conner]

But even "KF built" Wrecks are often reported to go into oscillation if the gain is advanced too far. That's unacceptible in a market design.

Urban legend has it that KF would voice his amps by moving the internal wiring around.

That would indeed be unacceptable to anyone with a hint of EE training, but in practice it just seems to feed the hype and make the original Trainwrecks sell for much more than amps that actually work.

Some boutique amp builders will take this argument a step further, and claim that an EE can never build a truly great sounding amp because they're fixated on making it work reliably.

[Chuck H]

+++
My "limited" head can't get past thoughts like "Well it shouldn't oscillate at ANY settings. Because someone WILL set it there." BUT...

If you listen to TW clips you'll hear something special happening. My favorite is to compare the real TW clips and the KF colaborate Komet amp on the Komet site. You hear the mojo with the TW but not the Komet. KF did voice amps in unusual ways. Every real express I've heard is smooth yet has that edgy chip on the pick attack, cleans up bright on the guitar volume control without being harsh on OD and has a generous amount of useful and good sounding swirl. These are properties that other amp designs have failed to accomplish consistently. Ken did it with every amp he sold. Because of his excellent results and consistent performance Ken was the only "guru" I've ever been willing to respect. I, for one, would love to know more about his process.

[Steve Conner]

Ken did it with every amp he sold. Because of his excellent results and consistent performance Ken was the only "guru" I've ever been willing to respect. I, for one, would love to know more about his process.

I hope it involved a lot of hot glue to keep the wires in place once he was done voicing! I can't see how an amp that oscillates could ever be called "consistent". All that positive feedback will amplify the effects of any small changes in the circuit. You could change tubes and it would sound completely different.

I think one of the two things must be BS. Either the edge-of-oscillation claim, or KF's reputation.

[soundguruman]

If the amp is designed properly, it won't oscillate.
If it does oscillate, the builder is just too lazy to finish designing it, or probably does not know how...
But there are plenty of builders who modify or build oscillating amps, and don't know better.

[Chuck H]

All that positive feedback will amplify the effects of any small changes in the circuit. You could change tubes and it would sound completely different.

I think one of the two things must be BS. Either the edge-of-oscillation claim, or KF's reputation.

A retube on a TW does alter the tone. Usually in a bad way. Ken stopped building and then eventually shopped out most of the service and the amps never sounded the same. There are original example that sound nothing like they're supposed to. Others seem to be more resiliant in the face of alterations. My point was just that even though his methods were different, they were repeatable, at least for him. That's significant. Just because you or I can't do it his way doesn't mean it doesn't work like that. It just means we can't do it. It would be nice to dismiss KF as just another in a long line of cracks that come and go. That would somehow validate our design ideologies over his. All I'm saying is that in light of the tones that "he" was able to consistently create and the high demand for his product I can't really do that.

[Steve Conner]

Well, my point is that if retubing a Trainwreck changed the tone in a bad way, then KF didn't create the tone consistently. Consistent means that the tone stays the way you want it over the working life of the amp. It means you can make a batch of 100 in a contract factory and they'll all sound the same, and any service shop can fix them given the service manual.

If you owned a Wreck, were you supposed to take it back to Ken for retubing followed by rejiggling of the wiring till the mojo came back?

Having said that, Philbrick made a highly successful op-amp module that only worked because it was unstable.

[diagrammatiks]

Well, my point is that if retubing a Trainwreck changed the tone in a bad way, then KF didn't create the tone consistently. Consistent means that the tone stays the way you want it over the working life of the amp. It means you can make a batch of 100 in a contract factory and they'll all sound the same, and any service shop can fix them given the service manual.

If you owned a Wreck, were you supposed to take it back to Ken for retubing followed by rejiggling of the wiring till the mojo came back?

Having said that, Philbrick made a highly successful op-amp module that only worked because it was unstable.

you ever see the schematic for wreck steve?

it's a wonder they sound the same between power ons.