Reducing filtering in a Ceriatone Rocket

[Jennings]

The Ceriatone Rock It has increased filtering over the original Vox AC30...I'd like to try reducing it to AC30 specs.  Which filter caps should I change to which values in the attached layout?

[kagliostro]

The Vox AC30 in his history has changed the values of the filter caps

to which model/era amp do you want to refer ?

---

One of the latest solution was a switchable set for vintage and modern sound

give a look to the attached schematic

K

[Jennings]

I'm after the original low filtering set up of the early 60s JMI era AC30...I'm hoping that gives me the more organic and less stiff feel I'm after.  So I wondered which of those cap values needs to be lowered, and what value to make it. 

So am I right in saying, looking at the Ceriatone layout of filter caps, left to right then top to bottom the values need to be: 8 (instead of 22), 32 (instead of 22), 8 (instead of 22), 16 (instead of the 47) and both 32 caps replaced by a single 16?

[overtone]

not quite, 16uF is way too low - both those 32uF caps straight after the GZ34 are in parallel, so they add up to 64uF.
The original Rocket circuit called for 80uF in that location, which some feel strains the GZ34. To my mind that is only the case if you are hot switching on and off in quick succession and not running a Mullard. I think the 80uF was intended to keep it tight when at full tilt. Fully cranked this is a fantastic amp. This circuit improved my technique enormously, especially on the cleans - it does not let you hide.

You know that if you reduce the filtering you may find that you get some hum or ghost notes, right?

If I was dead set on this I would just take one of those first 32uF caps out of circuit at first and see how that works out.
Take it from there gradually.

[kagliostro]

Here you can see the values on the AC30/6

those used in the 1960 model seems to be the same

K

[jjasilli]

The first preamp stage tends to set the tone of the amp.  Vintage amps tended to have a low value cap there.  This is often considered to make a tonal improvement over larger caps in that position.  8uF as you state seems to work well.  When I rebuild a Fender or Traynor, I automatically put an 8uF there, in place of 16uf or whatever else was present in later year amps.  This seems to make an objective improvement in tone. You may be happy by first replacing just that one filter cap that feeds the first preamp stage.  Also it's a good idea to replace components one at a time, then use the amp for awhile.  So if a new issue arises, like ghost notes, etc., you have a better idea of the specific culprit.  Also, the manufacture may have settled upon component values for reasons other than tone, such as to eliminate side issues (like ghost notes mentioned above) in manufacturing runs.  However, there likely is plenty enough filtering before the last cap (feeding the 1st stage) so that no side issues should arise by changing it to 8uF.

For regular guitar (not bass), different values in the other filter cap positions, seem to alter bass response and may make overall tone somewhat different, but not necessarily better one way or the other.  

[Jennings]

Thanks folks for the good tips...I've always been an AC30 user, and love my 62 with rear topboost.  I'm trying to use the Rock It clone as my hack around version built into a small 1x12 cab with a 50w Weber uprated alnico blue to keep it compact.  Like the result, but I've always found the sound to be too "tight" and not have enough of the Vox's raggy and organic character...sounds like my tastes would be to reduce the filtering.  I know there're all sorts of other factors involved too, but I was hoping the filtering might be enough for me rather than any more drastic changes like tranny choice etc etc.  Most people seem to like more filtering and a harder tone than I do...I guess I just dig the lower filtered vintage tones and response.  I've never minded the original AC30s ghost notes etc...part of the charm for me, I guess...must have got used to them!

Good point on replacing each cap one at a time and trying for a while.  I think I'll try the idea with the first 22 becoming an 8uF and see where that takes me.  If it doesn't do it for me I'll continue swapping to the values I quoted above then...I do run a Mullard GZ34 in there, so I think I'm good to vary the 64uF filtering on that if I need to, but if I do alter it I'll take the advice and maybe just try cutting one 32uF cap out to halve it first and see what that does. 

[Jennings]

Quick update...switching that first preamp 22uF down to an 8uF is nice!  Obviously I've only just done it, so I have to play it in and get used to it...but I don't think it's quite there, so I might have to try swapping another cap.

[Jennings]

Well I did a fair bit of "suck it and see" testing and cap switching last night...the end result I'm fairly happy with at the moment. 

Currently the three 22uF caps I have now read 8uF, 22uF and 8uF.  The 47uF is now a 16uF, and the combine 64uF on the GZ34 is now halved to 32uF.

I tried swapping caps one by one, starting from left to right on those 22uF ones, and then playing the amp for a while on a variety of settings to get the feel for the change.  I have to say, jjasilli was right about that first 8uF making the biggest single noticable difference.  If I were to recommend just one change to anyone with a Rock It who wants to move it from the modern tight feel towards more of a vintage vibe then that really does feel like you've opened it up a lot.  The other changes I went on to make only added small amounts of flavour and took the edge off the remaining sterility really.

I have got to say, I'd previously already swapped some of the coupling caps out from Sprague Orange Drops to Mallory types, and also changed the values to mirror the JMI era AC30 "Normal" circuit values (as opposed to the Bass or Treble circuit values).  That had improved the bass response and midrange to my ears, but hadn't loosened up the response and feel of the amp...which is why I'd moved onto the filter caps.  I opted for the "Normal" AC30 circuit values, as this is what my own JMI AC30 values are...and the circuit version that the TopBoost unit works best with IMHO.

Verdict: positive results (to my ears!)...especially from that first filter change (22 - 8 UF) so thanks folks!  Doesn't seem to be an increase in either hum or ghost notes to my ears...or at least I haven't detected any yet.  Whole amp seems stable and sounding good!  I really think at this stage any other "improvements" really would be subjective...and would probably require swapping out OTs to see the subtle differences.

OK...here's my final dumb question for today (I hope!)...what's the purpose of the filter cap on the GZ34 there?  And why did JMI opt for a very low 16uF, Ceriatone a 64uF and Trainwreck 80uF?

[jjasilli]

what's the purpose of the filter cap on the GZ34 there?

http://www.el34world.com/Forum/index.php?topic=13736.0

http://forum.metroamp.com/viewtopic.php?f=2&t=38508

[Jennings]

Thanks for the links...makes sense to me...so the caps do the filtering and noise reduction (not least from the added diodes), and the diodes offer a protection should the GZ34 fail and do a bit of rectification on the way?  Won't this in turn reduce the supposed "sonic effect" of the GZ34 tube sag when the amp's pushed?  I've not had any trouble with rectifiers blowing to date in any amp, so could this be overkill for the rate times it happens?  I'm wondering whether removing them will add a bit more of an organic characteristic to the driven tone of the amp...I like to crank AC30s between half way and full when playing as I love the rich drive they produce.

OK...one last question (I say this a lot!)  On the Ceriatone/Trainwreck circuit pin 9 (grid 2) of all the EL84 has a 1K 5W resistor, whereas the AC30 spec looks to be 100r.  What's the purpose of the the beefed up resistance there, and what effect should it have if I revert down to JMI specs?

[Jennings]

Anyone had and good or bad experiences with lowering the grid resitors?  Will it actually do much tone-wise?  I run modern EL84s, so would this not be wise for them?

And also, I have noticed some recent threads discussing "hash" or a sort of fizzy high frequency noise that diodes can add to tone when used as they are in my amp.  When I play this amp I do notice more of that than with my actual AC30, which has always irritated me mildly.  I've never found anything that reduces this, and in part it's one thing I was hoping would disappear with "opening up" the amp with lowering the filtering.  Although the amp now has a looser feel and wider depth to my ears with the reduced filtering, the fizz is still there.  Should I remove the diodes and go back to JMI spec there, or should I play about with some caps to attempt to reduce the fizz if possible around the PI?  Has anyone tried with positive or negative results.  Currently I run Mullard GZ34s as I'm lucky enough to have a small stock of them.

[super&plexi]

Greetings, Jennings. I'm in same boat, or at least close to. I'm finishing up el84 build_18watter, and put in 10uf X`22uf switch in V1 position.

I wasn't sure from post if you mentioned actually checking values of caps, or just installing/swapping. I bring it up because I've noticed quite large fluctuations, or rather variations in tolerance, I guess. Illinois, in particular, but definitely quite a bit of manufactures.

I think 20% is normal, + or -, or higher one way, lower the other, or vice verse. Either way it can be substantial, and figured I'd mention it.

Sorry if I'm bringing up something you're hip to already.

  As to the diode inquiry, I'll definitely leave that question to the more informed folks around here. I'm wondering the exact same thing, and was contemplating asking the same question. (I know the safety issue, not sure of sag, but I guess if they're B4 rect., then they probably don't change much, sagwise)

 I think I'm going to go put them in right now, so if I can record a B4-aft, I will post, or just my own 2bits.  But going with lower V.rect. will. I changed out my GZ for 5ar, or 5u, whver, the lower voltage one is, and that makes noticeable difference.

Speaker change is HUGE, and while I'm not familiar w/yours, I'll agree with anyone who says the response/feel,(sag) of an amp,   is highly influenced by speaker choice, whether or not the amp is SS, or tube rectified.

Have you played the old amp through this amps speaker, and does the 'hash' come through, or subside, or?...

either way, i'm confident that the folks on this forum will be able to get everything 'squared away'. Good luck, and hope I haven't wasted your time.

[Jennings]

Hi super&plexi!  Thanks for the reply...that's an interesting point you raise about cap tolerance variations...I must admit years ago I used to use a lot of Orange Drops, and measured those.  Found them to be very close tolerance wise...and since then I've never measured unless I suspect a duff cap etc.  This Ceriatone I've changed the orange drops out to Mallory types, and changed the values to JMI AC30 "Normal" values.  This was to get closer to the circuit in my early 60s AC30, which is a Normal version with added rear topboost.  The Mallory types did give a slightly less sterile sound, but actually not much.  The lowered filtering I've just tried really did open it up...however I'm still left with a fizz that I don't on other amps, and a slightly sterile tone compared to the AC30.  Yeah, I know it'll never be spot on to a 50yr old original, but it's still nowhere near as rich sounding/feeling as my 18w builds end up.  I might measure some of the Mallorys I have laying about just to get an idea of how much they wander spec-wise.  I haven't tried the filters, but when I lowered them I did at first try with some caps I had laying about, then ordered new cans to neaten it up...same result, so I have a feeling any variation there isn't hitting my ears at all tolerance wise.

Well, since the original AC30s did have the diodes, and I've never had a recto go yet...and this is only my personal build...I'm thinking today I'll simply try unsoldering the diodes and giving the amp a play without them.  I'm hoping that it'll either cure the problem, or do nothing (to my ears).  If I get no positive result I'll probably pop them back in just for the protection.  I'll post back with results.  I'm also toying with lowering those 1k resistors on the EL84 pin 9 to 100r...no one's lept in to shout "noooo", and I figure again that I've not had any problems with amps I run with low values there, bar one AC40 I had about 15yrs ago with a burned out carbon comp.

Speaker-wise I've tried a few in this amp...Celestion Gold was the obvious choise, but I found it too treble heavy and nothing like the Blue is was supposed to ape.  Original Blues, contrary to popular belief, actually have a lovely midrange sound...yes, a big clear high, but also a lot of smooth sounding mids too...to my ears.  In the end I settled on a Weber 50w Blue Dog type.  This isn't exactly G12 Blue like wither, but does have the added smoothness and mid/bass I was after in addition to clear highs.  Each speaker I've tried sounds very different to the next, but the fizz and so on hasn't altered.  If the last reversions to JMI values don't cure it I'll probably put it down to OT.

[Ed_Chambley]

You have noticed a fizz?  What range is it most prominent?  B string 10 to 12th fret?  Is it immediate upon attack, or does it appear as the note is held?

Also, I use a few Blue Dog Webers and have noticed the lower the wattage, the better.  I have a 30 watter, no dope in a bassman type of build, easily 40 watts and it has no problem handling the wattage when the amp is dimed.  Weber doesn't use the same specs and say eminence.  They say they have a different way, but I am not sure what it is. I do believe their speakers are over-spec in wattage.

[jjasilli]

I have found that NFB eliminates fizz.

[Jennings]

Well originally the amp sounded a bit sterile (brittle maybe?) to my ears, and I found the cranked sound was a little "fizzy", in that as you held notes or chords and the sound fades away you could sort of hear a kind of tatty overtone or raggedness to the tone.  Not a problem with a band as such, but it always annoyed me at home or recording.  I've called it the "kazoo" effect!  It reminds me of that ragged fizz that a kazoo has to its tone!  I hoped that opening the amp up with reduced filtering would cure this as well as the sterility, but it's only really fixed the latter.  So now I'm trying to work out whether I can pinpoint and fix it with component changes, or whether it could be more to do with the OT etc.

Speakers I've ruled out, as I've tried a range of speakers with the amp, and the Weber Blue alnico 50w sounds the best by far yet.  Although I do also like their 12A125A, but that really alters the "vox-like" sounds I'm after.  The "kazoo effect" is still audible with any speaker combination with my Rocket, but strangely when I try the cab and speakers with other amps it's not there. 

I'd like to convert my Rocket to be as close circuit-wise to my JMI AC30 topboost, simply because I like the tone of that amp and AC30s have always been my amp of choice and I'm used to how they respond to my playing.  I'd like this amp to be my smaller (dimensionally) 1x12" AC30-a-like I can throw in the back of the car rather than treat with kid gloves like I do with the JMI amp.  I think adding NFB might change the feel of the amp...but if it does smooth out the kazoo effect I might well try it.  So you folks don't think having the diodes on the GZ34 would do anything to alter the sound/feel of the amp then?  Or lowering the 1k grid resistors on the EL84s to 100r?  Currently, choke and transformers excepted, the circuit is now at JMI values for all but these last two things.

[Glennjeff]

Remove the diodes on the rectifier, rewire rectifier appropriately then test for fizz. If it doesn't go away put them back in. If it is the diodes you could put them back in and try some of those 0.01uF 2kV caps from GZ34 / Diode anodes to ground. It will take 10 minutes. Thought you were going to try that. Some Audio DIY's go to a lot of trouble to get rid "quantum noise" associated with SS rectifiers,  little 0.01uF 2kV caps scattered all over the place like confetti, even on every heater connection to ground. Don't think Nick would have put the diodes on his designs if they fizzed though.

Fizz IS very annoying.

[Jennings]

Good point...I'd been side tracked with the soldering iron and a DAC90a! I'll try tonight sans diodes. I'm not keen on spending weeks sprinkling tiny caps like glitter though so I hope I find a result. I'm guessing Nik added the diodes as most folks hear no difference and it helps with modern valves perhaps?!

[Glennjeff]

Nick's pretty thorough in most aspects of doing stuff in my experience. Don't think it would be in there if it wasn't up to "good ears" standard. I've got that rectifier setup in my D'Mars , by choice, no FIZZ. But eliminate each uncertainty one at a time, they scratch at you from the inside of your head you know, much worse than FIZZ.

[jjasilli]

Hard to rule out speakers.  They're all different and interact differently with different amps, tubes, OT's etc.  (I have a cab that fizzes with many amps, but fortunately not the one I built it for. Lucked-out for once!)  I'm still thinking parasitic oscillation is the problem.  Possibly, lead dress is an issue.  Other than global NFB, another solution is a suppression cap usually 47pF across the outputs of the PI.  Do you have one?  IMHO it's worth trying the suppression cap.  Then if still needed, you can temporarily try global NFB to see of it makes the fizz go away.  If global NFB does kill the fizz, but you don't like the tone further refinement might be tried. 

[Jennings]

OK...report back time!

I read up a little on the opinions and experiences of people who had varied the rectifier set up, and figured I'd take the plunge.  I can see the diodes being useful as a protection, especially when considering the apparently lower robustness of modern GZ34 valves.  I think generally speaking the tone with diode is fine...but I guess a few of us, namely us who gravitate to places like here, like to tweak just to see if their ears either notice at all, or like the result.

I wouldn't say that removing the diodes and the bridging cap are like night and day, but I do like the sounds of the amp, and I would say that it's maybe a little more like the JMI as a result...which my ears like.  Obviously I wouldn't notice in a full band set up etc etc, but otherwise I do actually think it's added to the organic feel.  Maybe the diodes do move the tone a little to the slightly more robust and rigid feel of a solid state recto amp...although I know that whole debate is subjective!  

I'm not using the stand by on my amp either...and I'm running a Mullard GZ34.  So far it's performing and sounding great.  I really do think the slight fizzy kazoo element to the tone is reduced in this amp with those changes.  I'll have to see if this is the case over time and in the playing environment.  Currently I'm happy with the reduced filtering and the additional recto alterations.  I'd say the amp before was punchy and tight, and now it's looser and more woody.  A bit like the difference between the RI AC30s and JMI ones, if you compare the two.

[roseblood11]

How could protection diodes for the gz34 influence the sound? Is this even possible?

[MFowler]

Only if the rectifier tube goes bad and the diodes are now rectifying then the B+ might be higher then what the tube rectifier put out.


Mark