Tweaking my Sceptre.

[Boy Howdy]

Okay, so don't kill me, I'm modding my Sceptre. It wasn't 100% original when I got it and it's not in great shape anyway. And hey, let's face it, these amps can certainly be made better for guitar. They were, after all, designed to be bass amps. The only change they made was to add reverb and trem.

http://www.freeinfosociety.com/elect...unnsceptre.pdf

First thing I did was to change out the 1st stage cathode bypass cap (250mF became 2.2mF). Big improvement already. Goodbye woofiness.

Next I got rid of all that mess at the inputs and replaced with the standard 68k/1M. Goodbye brittle crunchiness. Huge improvement. I don't hate the stock Sunn sound, actually I really like it, but it is very much a specialized sound. Super bright treb, super fat bass and no mids (unless the mid boost is engaged and then it just sounds bad). Don't even think of trying to use a distortion pedal in front of a stock Sunn, sounds awful.

I'm tweaking with the second stage cathode bypass cap and the treble cap values at the moment.

Two things that still need to be addressed (two that I can think of anyway): The midboost switch and the .1 caps just before the power tubes. I have questions about these caps ... in a minute.

I haven't seen a schematic with the midboost circuit anywhere online. Anybody got one? There is one in the Aspen Pitman book that I've been studying.

The midboost switches in or out three different changes. One, it switches out the stage two .2mF cathode bypass cap when the boost is switched on. I disabled this so the brightness stays in both positions. I like this better. BUT, I'm not stopping there. I think that boost sounds awful so....

There are two 100K resistors that moderate two 12pF caps when the switch is in the off position. So there is some midboost at all times, regardless of switch position. When the switch is on, the 100Ks are bypassed altogether for a huge freq boost. I plan to make the 100K setting the maximum boost (I may tweak beyond this. TBD) and to eliminate the current maxed out setting, altogether. Once I taste test the sound with the 12pFs out of the circuit I will decide on a plan.

About those .1 caps before the power tubes though, can I change these to, probably, .02s to reduce the bass? Will this require rebiasing? I don't know that much about this end of an amp. It won't require any other components be changed, will it?

Do those Black Cat caps have any monetary value, like with mustard caps for instance?

I'm looking to get the best possible clean sound out of the amp, and it already sounds better than my other Sunn. I realize that you're not going to get great distortion sounds with just one preamp tube, but I don't plan to repurpose the reverb tube at this time. I'm just looking for great clean sounds out of this amp.

Anybody have any suggestions for what I'm doing? Or for other mods for Sunns? Just for curiosities sake.

Lastly, remember opinions about tone are subjective. Your mileage may vary.

[tboy]

About those .1 caps before the power tubes though, can I change these to, probably, .02s to reduce the bass? Will this require rebiasing? I don't know that much about this end of an amp. It won't require any other components be changed, will it?

Yes, you can do that without any need to rebias or change other parts.

[Boy Howdy]

Thanks, tboy.

[Boy Howdy]

Okay, about that midboost switch, disregard what I wrote above. I got some values wrong, etc. Faulty memory, you know.

Here's a schematic with the mid boost switch:
Sunn SCEPTRE Amplifier Schematic

I had a little time to play with the midbooster last night. I isolated the three changes that the switch makes and listened to them individually. Basically, the switch gives two options: the .2mF cathode bypass cap on the second stage in or out, or a 1.5mF cathode bypass cap on the PI tube in or out. This second option also engages a 12pF cap to blunt the giant treble boost afforded by that bypass cap. I like the sound better with that 12pF disconnected. But I haven't settled on anything yet, still have some more options to try. I may put a resistor in place of that 12pF just to tame the volume increase when the 1.5mF is engaged. I also thought I could put a 25mF in place of the 1.5mF and just have an overall volume booster. Could come in handy. There have been plenty of times when I've wanted more volume during a solo when the drummer suddenly decides to go apeshit. A switch would be faster than adjusting the volume control. But then I thought that would be much better on a footswitch. So now I'm thinking about those possibilities. And others. Stay tuned.

Are those Black Cat caps any good?

[Boy Howdy]

The tweaking continues.

After trying a ton of different values, I decided to change the second stage cathode bypass cap from the .22 to .44. Sounds similar - belltones aplenty - but a little less harsh. Sunns can sound a bit cold and sterile in the treble, the .44 softens the edge a bit, warms it up slightly. Makes it slightly less scooped sounding too. I also took the treble cap from 270pF to 390pF.

I tried every value cap I had on hand from 250mF to the stock 1.5mF in that midboost circuit (with that awful 12pF disconnected) and came up with nothing I liked much. I would have liked to try something in the 2 to 5mF range but didn't have any.

I decided to just keep the midboost circuit as it is when the switch is in the off position. I removed everything from the switch itself and will use it to add a cap (probably between 10 and 22mF) in parallel with the second stage cathode bypass cap. I might use the other side of the switch to cut the bright cap on the volume control to 100pF to cut the upper mid content a bit too. These changes will give a nice fat sound for jazzy things.

More to do, but gotta order parts.

[imaradiostar]

Have you tried disconnecting the negative feedback from the transformer secondary? Between the ultralinear output and the nfb at the phase inverter I bet it's pretty tight and clean, probably to a fault. The ultralinear output alone should provide more than enough feedback to keep things from falling apart when pushed a little bit.

rather than a bright cap I'd be tempted to add a 470k/500pf combo between V1a and the volume pot for a nice treble shelving effect.

Based on designs I've built with similar tubes I'd be tempted to get rid of the negative feedback into pentode and add a bias network and "master volume" between the pentode and the phase splitter triode. At the same time I'd play with lower pentode plate resistors to see how it effects the tone- I'd probably land between 15k and 82k- 6an8 can take a lot of current! You'll have to alter the screen and cathode resistors accordingly. You may find you can get a "sweeter" more pleasant tone at lower volumes and have a more responsive amplifier overall. The amp could take on a little of that pentode harmonic enhancement like a matchless clubman or old ac15- not quite as clean as before be really much more pleasant sounding.

glad to hear someone messing with an old sunn- I've often thought these things were a few tweaks away from being a much better guitar amp!

jamie

[Boy Howdy]

Well, that went right over my head.

I got my learner's permit, but Moms won't take me driving much. It may be a while before I get my license. Gulp.

I'll study what you've written, Ima, and get back.

I've been playing these Sunns for 10 years stock so I've gotten pretty used to them. I like it bright, but yeah, they are pretty clinical. I don't think I'd want to put in a master though. I can't imagine there would be enough gain with just the one tube to make it a desirable option. I did think about the possibility of converting the reverb tube though at some point. http://www.schematicheaven.com/dumbl...7fullschem.pdf

I also plan to convert two of the input jacks to a preamp out/power amp in loop. I did this to my Traynor and have had lots of fun putting tube preamps in the loop (among other weird experiments). I think Dumble is onto something putting the distortion pre after the clean pre. And of course I can always put a volume pedal in the loop (or construct an outboard master box) instead of installing an extra pot.

[imaradiostar]

Sorry if I said too much. I have a way of doing that. I'll try to give my opinion in a way that makes sense...maybe others will agree or disagree and we can all learn something.

The output circuit of the Sunn is pretty similar to a hi-fi amp- in fact I think it's lifted in its entirety from one, I just don't recall which. It has its strengths and its weaknesses. Part of this is that the pentode (first half of the 6an8) is direct-coupled to the phase inverter (second half of the 6an8). The plate or anode of the pentode is connected directly to the grid of the triode and as a result sets the operating voltage and current of the phase inverter. It's not a terrible way of doing things if you like for the amp to stay clean but it has two problems for guitar amp use- it can't put out as much clean signal to the power tubes as other designs and the pentode, used with the large plate resistor, isn't in its musical "sweet spot." Fine for a clean hifi amp, not as good for a loud guitar amp.

I was recommending a master volume and changes to the pentode not so much to create "preamp distortion" as to add (or take advantage of) the harmonic sweetening effects of a pentode. It's a really neat sound if it's done correctly and it can be complementary to a lot of different pedals and guitars. It's also really useful if you decide to add a preamp in the way you described- you can send the output of the preamp out to a pedal or second preamp or effects processor and still have a volume control between the effects and the power amp.

The second part of this involves converting the phase inverter (triode half of the 6an8) to fixed bias operation. This helps the phase inverter to develop maximum possible clean voltage, something more difficult to control when the phase inverter is directly coupled to the pentode in front of it. This means that you create a voltage divider (one resistor to ground, one to B+) that establishes a portion of B+, in this case we want 1/4 of B+. These resistors are either large values and connected directly to the grid of the phase inverter triode or they're smaller values to establish the voltage and a traditional 1 meg grid leak resistor connects from that 1/4 B+ point to the grid of the triode. This should give maximum clean output from both sides of the phase inverter.

I've built this design in one of my own amps with 6L6's instead of 6550's and had really good results. It allows you to hit the pentode with just enough signal to compress and add some harmonic content without crunching or distorting in the traditional sense. By getting rid of the negative feedback around the output section into the cathode of the pentode you have a situation that requires far less input voltage and should, in theory, sound more interesting for guitar use. Set the master higher and the preamp volume lower and it's still capable of its good-old super clean sounds but perhaps a bit less restricted sounding.

Two other thoughts-

I don't think I'd bother modding the reverb return to use a tube. It's just reverb- it's not in the signal path and there is little to be gained by changing it. If the transistor circuit quit working I'd build a new one based on an LND150 mosfet and roll on knowing that I'm not wasting any filament current to have more reverb than I could ever use. I would, however, add a 1 meg "dwell" pot to the input of the reverb driver triode so I could scale back the drive for a slightly cleaner verb tone.

If you decided to add preamp out/power in jacks I highly recommend adding an LND150 setup up as a source follower to buffer the output so that any remote effects or cables don't hurt the tone of the amp. I know that's mouthful but when the time comes it's pretty simple- less than 10 parts and they'll all fit on a smallish perfboard, giving you a buffered output that won't load up the tone stack. If you decide to do that just ask and I'm sure someone can help you.

jamie

[Boy Howdy]

Dog, Jamie! You've just given a man who can't cook all the ingredients for a five star meal! Mmm, boy that raw meat looks tasty! Kinda frustrating. LOL. I do realize, BTW, that this isn't just for my benefit, that others will learn from this. That said, I may have to put you on the hook for more detailed info.

Again, I'll study this. I've gotten a little busy here lately, but will spend some time with this over the weekend.

The thing about the reverb, what I was proposing was converting the reverb tube to a distortion circuit like in that dumble. I don't much care for spring reverb and the pan is defective in this particular unit anyway.

[Boy Howdy]

Well, I finally got back to my Sceptre.

Per Jamie's suggestion, I tried disconnecting the negative feedback lead from the output jack. This did indeed improve things! It's louder, sounds more detailed and feels much better: less constipated, more dynamic, more touch responsive. Shoulda done this years ago!

I also removed everything from the mid boost switch and then ran a couple of leads to it from the V1b cathode to engage/disengage a 2.2mF cap in parallel with the .44 cap that I already had in place. This makes for a nice "fat boost". It's the perfect complement to the regular sound. I'm not sure which "channel" I like better, they're both great!

I replaced the .1mF post PI caps (what are those called anyway?) with .02mFs. Yes, much better. Full, but not bloated.

I rigged a temporary post PI master ala the Matchless Clubman 35. http://www.freeinfosociety.com/elect...essclubman.pdf . Didn't like it. Too much gain, for starters. Too touchy, too easily loaded down, too crunchy. Not good.

[imaradiostar]

Sounds like you're making some good progress. The phase inverter in the Sceptre isn't really conducive of a master volume after the PI- the one I was recommending between the Pentode and the split load triode was only so that you could hit the pentode a little harder without overloading the triode. In my experience the split load phase inverter is not at its best when overdriven heavily.

By removing the feedback loop you've increased gain and "opened" the sound, so to speak. I'm glad you like it. I think this is the best thing you can do for old ultralinear amps. It seems to me that ultralinear is enough to keep the bass a little tighter and sound a little purer without the more abrupt clean to dirty transition of most amps with feedback around the power section.

Take a look at an Ampeg B18n schematic- it shows how to bias the triode with separate resistors and an added cap. If you've done that you can add a pot and another cap between the pentode and triode (before the triode bias network) to allow you to turn down the signal into the phase inverter triode.

I built a similar amp with a 6au6 pentode out front then a 1 meg volume control with a bright cap and a 6u8 for an additional gain stage and phase inverter. I found lots of interesting sounds by overdriving the second pentode but keeping the level into the PI more reasonable.

jamie

[Boy Howdy]

Thanks, Jamie. I will study this and get back in a couple days. I'm sure I will have questions.

One concern after just a cursory glance at the schematic, is the issue of preserving the 1.5 mF/100K thing coming off the cathode of the pentode. I left that intact (when I got rid of the weird stuff on the midboost sw.) and I think it's a big part of the Sunn sound. Sort of in the way a speaker defines the overall voice of a given guitar sound. But whatever, maybe it should go. Nothing ventured, nothing gained.

And yeah, I understood that the master vol. you were talking about was intended to go between the pentode and triode. The post PI master was just something I wanted to try. It was simple enough. I guess the pentode has a lot more gain and that's why it was so shockingly grindy.

Anyway, I'll get back.

[Boy Howdy]

Anybody know of a simple, or even a not so simple, way to slow the rate on the tremolo in this amp? It can get so fast that the effect becomes inaudible for like a third of the pot rotation. But it doesn't get slow (enough) and sexy like that Fender Rhodes piano sound.

[Boy Howdy]

Questions for Jamie:

I'm not sure how to handle the B+ situation. There are two sources of power coming into the Sunn circuit and one with the Ampeg.

On the B-18 the .1mF cap attaches to the 16 ohm tap through a 22K cap on the OT, since we're not doing negative feedback, I should leave the .1mF in place but disregard the 22K and the feedback connection, correct?

Use the same value caps and resistors in the Sunn as are used in the Ampeg? All are 1/2 watt, I assume.

I don't have a 3.3M or 10M resistors on hand. By putting a couple resistors in series I can get:
3.33M (3M, 330K) or
3.27M (3M + 270K)
and
9.8M (6.8M + 3M) or
10.2M (8.2 + 2M).
Close enough? Which ones should I go with?

I assume I should just ignore the pin numbers on the Ampeg schematic and instead look at what pins on the Sunn correspond with plate, cathode, etc. on the Ampeg. In other words, match plate to plate, cathode to cathode between the two schematics regardless of pin number? But wait a minute, the 7199 pentode (?) half shows only two screen grids. Color me confused.

I will probably implement this after I make sure everything else is working properly, but I'll ask now since I'm thinking of it. The master volume will come between the .02 cap and the 10M resistor? Just a simple volume pot (1M?) and bright cap (what value?) configuration?

You said you would want to experiment with pentode plate resistor values (and consequently the cathode and screen resistor value as well), care to recommend some good starting point values for each?

The Sunn uses two terminal strips straddling the PI tube. My resistors aren't long enough to reach across. I've wondered about this before, are resistors with longer leads available?

Sorry for all the questions. Enquiring minds ... yada.

[imaradiostar]

When I get home from work tonight I'll dig into the schematics and post a proper reply but for starters I'd leave the existing pentode resistors in place- since you're lowering the values solder your new resistors over top of them, paralleling them for the lower value. It's a chance to practice your math (or cheat and use an internet based java calculator).

As for the grid of the triode- you want it to rest at between 1/4 and 1/3 of supply voltage, there are a number of ways to do that. If you'd like you can connect a 220k resistor from B+ to the top lug of a 100k pot and the other lug to ground. Connect a small value cap (5 uf 150v) from ground to the wiper of the pot. Connect a 1 meg resistor to the wiper of the pot (with the + lead of the cap) and the grid of the triode so it forms a grid leak to the voltage divider. This will allow you to dial in that voltage exactly and probably end up with a lower noise circuit than on the Ampeg. It's a set and forget so you can use a fixed resistor once that triode is resting with about 1/2 of B+ between the cathode and anode- give or take.

It's not a super exact science...that's the fun part. Sometimes things that don't work correctly sound awesome, like the traditional vox and marshall cathode follower. It's not very linear but it sure does sound good!

jamie