trinityamps.com

Ok Maybe it was me (bad tubes) or me starting to loose my mind. Cause mine sounded nothing like this when i first built it Maybe it was some of mods I did in the power sections. Maybe I had really crappy tubes in when I first built it And was modding around that
. I need another opinion.

Seems that It used to be much darker and boomer. I had the tone much higher to get this sort grit and brilliance out of it.

This is still the original Plexi 1 normal Chanel.
Process sm57 -stein berg c12- garage band with uniminds EQish and a bit of track verb and echo
Picture shows knobs and at about 3:10 in the clip i moved the input from low gain to high gain
After listening back I rushed some of the cuts and played a little sloppy sorry


http://www.soundclick.com/player/single_player.cfm?songid=12086133&q=hi&newref=1

This is what i'm getting out of the tmb side (still a little bassy have not mastered the sm57 thing. I was not getting this much gain when i first modded it.



http://www.soundclick.com/player/single_player.cfm?songid=12087165&q=hi&newref=1

what were the mods?

Sounds pretty good.
What guitar are you using?

What mods do you have now?

_________________
http://pdfelectronics.com

That's what I thought at first (more on that latter)

Same as I used in the clip in the brown thread 91 Les Paul into a pair of late 70's black back's 55 HZ

I went too far every were at once trying to get what I wanted out of the TMB in the mod's and didn't sit back and analyze it enough. (my bad). too busy chasing tone

On the normal clip remember I still had the 220 uf on the cathode no mods there. As for the TMB I had the mk 2 mods no bright cap slightly different boost on v1 2k7 and then a switch in another parallel 2k7. As for the V2 I had 1k5 and kick in .68uf and another 1k8 resistor. As for the PI had no changes. Pa had grid leaks down to 220k ,2k2 screen dropping resistor a shared 1k screen with two 100R's going to each screen. Also had 180r cathode resistor and 1000uf cap around it.

My issues with the sound on the normal channel there was a total loss in bass response and a little more light clipping from what I remember from when I first built it. I thought the bass response change was coming from the screen changes but man was I wrong (latter modded things back to stock and bass response was back to what I recalled ,(So I realized I was not nutz).
TMB clip had both boost's on still had the same bass loss and what seemed like my JCM 900 clipping (own a 4500). I had also installed the Lindz /brown note / Ruby fix across the grid leaks I was using 6.3v zener's and IN4007 well below what I have for a bias voltage. I then realized I must be getting too much diode clipping. So went back and re read Paul's article on the subject of fizz. He did warn about clipping with lower value zener's. So I said I'll try this as stated and slapped in some 10v zener's and lowered the vrm so I had 9.2v across the cathode resistor. At this point still pretty much the same bass response as above (no fizz) and said this is not for me

Now I'm back to square one almost , PA set back to stock except for screen resistor is at 1k as above and screen dropping resistor back to 1k5 and original cathode cap and resistor . Then d-mode the TMB v2 input resistor back to 100 k to ground. Now lets go back to the normal channel (this was my biggest mistake in the other mods i should have been going back to this as a guide the whole time ) max out the vrm and crank it up like in my settings above and behold the bass I had in the beginning and a much smother grit. ah but there's the little bit of fizz lying in the back ground. That drives me and some others nuts. (thought I had gotten rid of this way before with the moving of the screen resistors WRONG again) At this point I try the new plex mk2 normal channel cathode cap( as close as I can get about 2.9 uf with 2 caps together) start changing screen dropping resistors 1k8 then 2k my fizz issue still there. OK I try changing the PI input resistor from 470R to 820R low and behold Its gone. I swear (on the normal channel) And I stopped in the middle of typing this to check and managed to dial both tone and vol to eleven. This is the first head slap cause if what I am hearing and describing as fizz is actually PI input over load (this could also be aided from my guitar pushing the front end a little more than most) I still have to go back and try moving the screen dropping resistors back down as I know it changes the voltage down stream and if what I have is actually PI overload it may rear its ugly head again.

Now on to the TMB the fizzer of the two. Set the tone controls at top dead center and start moving the gain an volume up in unison and you get the fizz at about 2- 3 o'clock I'm thinking this is really PI over load. This is where I have the second head slap cause the tmb could always get louder and fizzier then the normal. Looking back at if I can crank up the normal channel now (pinned) that would be max volume ( if what I was getting was PI overload) . To me there's no way to get the tmb any louder without the PI overloading. (correct me if I'm wrong) Sure you can get more gain from the front end but at output going into the PI has to be closer to the output level of the normal channel. Re read the mod instructions in the kit. You may get PI over load when adding gain to the tmb you may need to lower the master volume pot. So I first maxed out the gain pot and played with the master until I could find the spot without fizz and maximum gain (its extremely touchy and very narrow). I've started playing with lowering the volume pot down to 250 k but its not there yet. This is all based on that some guitar players like me have a habit of once and while or more often than not maxing things out.

If I get the TMB max volume without fizz to that of the normal channel I would be more than thrilled cause the normal channel totally cranked now sounds F'n great and has more than enough volume !. Will update this more as I've as I continue on back from square one

That sounds awsome! What speakers do you have?

oops! NM. Blackbacks. Helps to read the rest of the thread before responding?

_________________
Les

http://www.soundclick.com/unimind

I really think the "fizz" that you and I are hearing IS in the PI. Is it overloading the input? Or, what I now think, is it because of the way it is designed for EL84s? Because EL84s clip a lot earlier than other tubes, the 18 watt PI is designed to clip a lot earlier than the output so as to NOT drive the EL84s too hard. This is done with the 56K tail resistor. If you lower this resistor you will increase headroom in the PI but you will then be driving the EL84s too hard (overloading them). When I switched to 6CM7s, I lowered the PI tail to 15K (the 6CM7s can be driven a lot harder) and the "fizz" is gone. The pre amp is still the same so it is still hitting the front of the PI the same but now there is a lot more headroom in the PI so it can take it. Like I posted in my 6CM7 thread, the amp now has more of that big Plexi crunch with no fizz. Now I do not have 6V6s but I know that for them you should have the PI tail around 22K so I would guess you would get similar results. More headroom and more Plexi like crunch and no fizz.

My conclusions are that the 18 watt with EL84s is designed to have less clipping in the pre amp and more clipping in the PI and power section. This is what gives it that unique "creamy" sound. The big amps were designed to have more clipping in the pre amp but with a "clean" PI that would drive the power tubes when cranked. This is what gives them that nice tight crunchy sound. Putting the big Plexi pre amp in front of an EL84 based PI and power amp sounded okay tone wise but it was just too fizzy and not crunchy enough. I spent months trying to get rid of the fizz. Tweaking this and that, reducing gain here and there, Ruby mod etc. Put in 6CM7s, changed PI tail to 15K and I have not touched the inside of my amp since. I really think that if you want that Plexi like crunch without the fizz, you will want to go with 6V6s or 6CM7s and change that PI tail.

_________________
Les

http://www.soundclick.com/unimind

FWIW, We played the MkII design as posted in the forum, for quite a while and didnt hear a fizz per se. It's now in the studio for extended testing, so if a player notices it, I will get feedback on it very soon. Using Mullard RI power tubes and Mullard RI / Tung Sol pre tubes.

That said, I'd still really like to try those 6CM7s

_________________
Stephen
Web: www.trinityamps.com. Facebook: facebook.com/trinityamps. Twitter: @trinityamps

@Unimind
I Hear where your coming from cause I really thought about 6V6's myself. But I can be a real stubborn bugger at times. As for the Tail resistor (after my dumbass got scholled in another thread ) I did some reading and so forth on LTPI. And I had more of the understanding the lowering the tail resistor actuall causes more voltage drive at the output. Which at this point makes me also wonder if changes things on the cathode resistor (I just don't know that part) Where as having the higher tail resistor lowers the the voltage drive to the output tubes. Which makes sense with the input sensitivity of 6v6's vs el84's.

I don't have a lot of free time this week but plan to do clips of both the 470 r vs 820 r on the normal channel. As well try to explain a little better what I was getting at with the voltage drive coming out of normal vs TMB. Think I could fix what I was talking about if I had a scope , a week to relearn how to use it, 6 beers just cause and all my fingers and toes to calculate the voltage divider to drive the tmb side of the PI

Yeah I don't fully understand all of it either that's why I ask "experts" to chime in and let me know if I am wrong. I know thru experimenting and by "ear" what seems to work as far as the sound and tone I want.

I did hear a difference between the 470r vs 820r and going back to the 820 seemed to help with the fizz. As far as the input from the TMB, I had put a 470k - 1M resister after the voltage divider on pin 2 of V2 to knock the signal down entering the tone stack. One of my "gain reducing" mods.

_________________
Les

http://www.soundclick.com/unimind

Kind of but not quite. The 18W PI is designed to have less gain, with the 820 ohm shared cathode resistor instead of the 470 you find in most Marshall/F*nder LTP PIs. Then the 56k tail resistor does mean the 18W PI operates with less output swing/headroom, but not with the intention that it should clip much, since it also has less stage gain.


Actually that's incorrect. The 18W PI is intended to have almost no clipping at all. The 820 ohm unbypassed cathode resistor both reduces gain(as I mentioned before), and also biases the 12AX7 triodes towards a neutral point in their characteristic, so they have to be driven a lot harder to clip.


The shared 470 ohm resistor in the PI of the bigger Marshalls biases the triodes more asymmetrically towards the "warm" side of their characteristic, as well as increasing stage gain. So these PIs will provide some great crunch tones when pushed, which is why that value is used in the Trinity Plexi 18W. By contrast the typical Vox/Matchless PI has a 1.2k shared cathode resistor (equivalent to 2.4k on a single triode), which biases the triodes asymmetrically on the cold side. When you push the Normal channel of an AC15, AC30, or a of Matchless Spitfire, the PI starts to hard clip with a distinctly fizzy tone. When you push past this fizzy zone and the power tubes start to break up, then the nice warm Brain May style power tube distortion becomes more dominant with these amps.


So the issue isn't mainly about the PI, but about the intrinsic tone of the power tubes. Not everyone likes EL84s being driven by a high-gain front end, although that combination has always sounded fine in cascaded 18W amps that I've played or worked on. If you change to lower gain power tubes then the PI has to be made hotter to compensate.


Yes, since that reduced the PI gain.

_________________
Great tones come in small glass jars!

Thanks Phil,

On the Original Plexi, we use 470. On the new Brown version, we are using 820.

_________________
Stephen
Web: www.trinityamps.com. Facebook: facebook.com/trinityamps. Twitter: @trinityamps

For me like I said the 820 fixes the clippy fizzy thing on the normal channel. Previously at bout' 8 I would get the fizzy and now 8 to 10 it doesn't get any louder just more sweet gain I assume what Zaphod is describing as power tube distortion.

Now the TMB is the other side of the coin. I can still get the fizzy I assuming from the overall output of the TMB vs normal.

At this point i've got a master 250 k pot and 470k dropping resistor coming of the treble wiper. Have to play around the values a bit and re adjust for the over all load. But the fizzy's seem to be gone and have the el84 sizzle . Still have to do some more listening

On a side note the VRM works more like I assumed it work before it always seemed to (enhance the fizzy) as you turned it down now with them gone I can turn it down and voila

Yes, I was referring to the original Trinity Plexi there, and not the Brown version.

_________________
Great tones come in small glass jars!

Thank you again Phil for clearing things up. I seem to mix up gain vs clipping vs headroom all the time as I still don't understand it all. But my ear tells me what sounds good and I just could not get things to sound the way I wanted until I went to the 6CM7s which, by the way, you had suggested to me a long time ago. So thanks again for that.

_________________
Les

http://www.soundclick.com/unimind

After re-reading this and thinking about it, the ONE thing that mostly got rid of the FIZZ with el84s was to lower my PPIMV to around 7-8 (2-3 o'clock). That would lower the input signal to the el84s. I remember setting it around 2-3 o'clock and just leaving it there. But because I had the 820R on the PI I was lacking that Plexi crunch.

_________________
Les

http://www.soundclick.com/unimind