New trainwreckish Plexi build
Moderators: pompeiisneaks, Colossal
New trainwreckish Plexi build
I just finished a new plexi build with the same format and general layout as a Trainwreck amp.   The circuit is a classic channel linked plexi, to which I have incorporated distributed filtering,  paralleled inputs (4 parallel triodes counting both channels), power scaling and GMX, bootstrapped master volume....   Actually, there are too many little details in this build to mention, feel free to ask if you see something you don't understand on the photos.
Initially it was meant to be an AX84 October, but it evolved to something else. I'm using a Hammond 372jx PT and a Pacific 25525 OT. I used a higher voltage PT (+20v) than the October to account for the added preamp tube. With 6v6 all the relevant voltages were within a 3% of expected, but I preferred the body of KT66 and the preamp voltages came about 25v down. I adjusted the resistor chain to compensate.
This was my fist layout from scratch. I "prototyped" the amp by drawing very detailed layouts in Visio. Many of the components I used are huge (PEC pots, AMRG resistors, Sozos). Early on I realized that all components had to be drawn to scale to avoid bad surprises.
I had no relevant issues with stability or noises (it is dead quiet) but it took me some time to figure out that I needed a very abundant bias current source: full wave rectified from B+ with 33k coupling resistors! The bias raw voltage is limited to 100v by a zener across the smoothing caps.
One thing that surprised me is that the presence control is not very effective, no matter how much NFB I let in. I think it is related to power scale, although it did not happen to me before, even with the same bootstrapped master volume.
The tone is just awesome. It is my second build with this preamp (the fist one had 4xel84). I love the "oily" fat tone I get with it.
Regarding GMX, I like the effect, but I keep getting distortion artifacts in many settings. This made it almost unusable with 6V6, less so with KT66. Anyhow, for the time being, I took it out. Once the voicing is finished and I've had a few playing hours with it, I'll put it back in and see if I can iron out the problems. I think I'll have to buy a scope for this one.
I made the layout for Vintage Sozos. Since it took 4 months to get them in, I installed TAD mustards first. As a result, now I know that the Sozos do sound MUCH better. Better high end, feels like there is more detail. Pity they treat their clients so bad.
Pablo
			
			
						Initially it was meant to be an AX84 October, but it evolved to something else. I'm using a Hammond 372jx PT and a Pacific 25525 OT. I used a higher voltage PT (+20v) than the October to account for the added preamp tube. With 6v6 all the relevant voltages were within a 3% of expected, but I preferred the body of KT66 and the preamp voltages came about 25v down. I adjusted the resistor chain to compensate.
This was my fist layout from scratch. I "prototyped" the amp by drawing very detailed layouts in Visio. Many of the components I used are huge (PEC pots, AMRG resistors, Sozos). Early on I realized that all components had to be drawn to scale to avoid bad surprises.
I had no relevant issues with stability or noises (it is dead quiet) but it took me some time to figure out that I needed a very abundant bias current source: full wave rectified from B+ with 33k coupling resistors! The bias raw voltage is limited to 100v by a zener across the smoothing caps.
One thing that surprised me is that the presence control is not very effective, no matter how much NFB I let in. I think it is related to power scale, although it did not happen to me before, even with the same bootstrapped master volume.
The tone is just awesome. It is my second build with this preamp (the fist one had 4xel84). I love the "oily" fat tone I get with it.
Regarding GMX, I like the effect, but I keep getting distortion artifacts in many settings. This made it almost unusable with 6V6, less so with KT66. Anyhow, for the time being, I took it out. Once the voicing is finished and I've had a few playing hours with it, I'll put it back in and see if I can iron out the problems. I think I'll have to buy a scope for this one.
I made the layout for Vintage Sozos. Since it took 4 months to get them in, I installed TAD mustards first. As a result, now I know that the Sozos do sound MUCH better. Better high end, feels like there is more detail. Pity they treat their clients so bad.
Pablo
You do not have the required permissions to view the files attached to this post.
			
									
						Re: New trainwreckish Plexi build
A beautiful build.  I guess i don't know what GMX is though.  Do you find the bright channel and normal channel interact due to the 220pf cap around the mix resistors ?  How do you like the AMRG resistors ? 
 
			
			
									
									
						
- 
				Marioverkill
- Posts: 61
- Joined: Thu Jan 05, 2012 2:23 pm
- Location: NB,Canada
Re: New trainwreckish Plexi build
NICE !!!!
Coicidentally,I was experimenting with bootstrapping a DC cathode follower last night.What sonic effecet does bootstrapping the master volume have ?
Do we get some soundclips?
Nice work
Mario
			
			
									
									
						Coicidentally,I was experimenting with bootstrapping a DC cathode follower last night.What sonic effecet does bootstrapping the master volume have ?
Do we get some soundclips?
Nice work
Mario
Re: New trainwreckish Plexi build
The GMX is a term coined by Kevin O'Connor in his books.    It's a "trans-conductance multiplier".   It makes the amp feel like a much more powerful amp.  Not louder, if the OT does not allow it.   Just like switching the half power switch in a 18/36watt but without the increased 3dbs.
Not sure to understand your question on the interaction between the channels. They share everything but the first stage, and they are paralleled so you could say they are very interactive. In Vintage mode they share the cathode's resistor and cap too.
I like the AMRGs. Very little noise, but it's difficult to tell without comparing it with another similar amp, with different resistors. I know that the distortion is smooth and completely free of "artifacts". They are huge, though, I had to change the layout a lot to make space for them.
Pablo
			
			
									
									
						Not sure to understand your question on the interaction between the channels. They share everything but the first stage, and they are paralleled so you could say they are very interactive. In Vintage mode they share the cathode's resistor and cap too.
I like the AMRGs. Very little noise, but it's difficult to tell without comparing it with another similar amp, with different resistors. I know that the distortion is smooth and completely free of "artifacts". They are huge, though, I had to change the layout a lot to make space for them.
Pablo
- 
				Marioverkill
- Posts: 61
- Joined: Thu Jan 05, 2012 2:23 pm
- Location: NB,Canada
Re: New trainwreckish Plexi build
Stunning! Your inner Virgo is showing.
			
			
									
									
						Re: New trainwreckish Plexi build
Bootstrapped MV is the most transparent MV I've tried.  My problem with power scaling was that the master volume is used in much lower settings than with, say an attenuator.  So, if your master sucks tone, you lost it. 
I once implemented 5 different master volume types (traditional, tandem pot and crossline PPIMV, PI cathode pot and bootstrapped) in the same amp, just to see which one I liked most. Putting a pot on the PI's cathode works fine, but you can't get a clean sound out of it. PPIMV is good too, but drive increases as you dial down power scaling, because of the lost NFB. The traditional lost all treble at higher settings. Bootstrapped was best overall. For non NFB amps like the rocket I would rather use PPIMV.
Pablo
			
			
									
									
						I once implemented 5 different master volume types (traditional, tandem pot and crossline PPIMV, PI cathode pot and bootstrapped) in the same amp, just to see which one I liked most. Putting a pot on the PI's cathode works fine, but you can't get a clean sound out of it. PPIMV is good too, but drive increases as you dial down power scaling, because of the lost NFB. The traditional lost all treble at higher settings. Bootstrapped was best overall. For non NFB amps like the rocket I would rather use PPIMV.
Pablo
- Reeltarded
- Posts: 10189
- Joined: Sat Feb 14, 2009 4:38 am
- Location: GA USA
Re: New trainwreckish Plexi build
Nice! We think alike, but you just think a lot more!

			
			
									
									
Signatures have a 255 character limit that I could abuse, but I am not Cecil B. DeMille.
						Re: New trainwreckish Plexi build
That is nice! Question: what is the 10k and diode at the cathode follower?
			
			
									
									
						Re: New trainwreckish Plexi build
The real reason is that I copied AX84 October's design, witch has them.  
After investigating a little (it's mentioned on Merlin's book on preamps), I found out that it is used for protection against arcing at switch on.
The preceding gain stage (DC connected to the cathode follower) is pretty hot with a 100k/470R anode/cathode, which is different from other amps I've seen that usually have a pretty cold stage in that position. This might be the reason why this protection is needed.
Maybe someone that really knows what he is talking about could chime in here...
Pablo
Edit: I uploaded an schematic corrected for KT66
			
			
						After investigating a little (it's mentioned on Merlin's book on preamps), I found out that it is used for protection against arcing at switch on.
The preceding gain stage (DC connected to the cathode follower) is pretty hot with a 100k/470R anode/cathode, which is different from other amps I've seen that usually have a pretty cold stage in that position. This might be the reason why this protection is needed.
Maybe someone that really knows what he is talking about could chime in here...
Pablo
Edit: I uploaded an schematic corrected for KT66
You do not have the required permissions to view the files attached to this post.
							
					Last edited by pablogt on Thu Mar 29, 2012 9:17 am, edited 1 time in total.
									
			
									
						Re: New trainwreckish Plexi build
Pablo,
I have a similar build in the works but it is not up and running yet. I do remember KOC saying that powering the GMX from the Vs node provided cleaner power - I see in your scheme you are powering it from Va.
My design uses a cathode biased power amp, we'll see how the SB power scaling with both an improved bootstrapped master, a PPIMV and GMX works out.
John
			
			
									
									
						I have a similar build in the works but it is not up and running yet. I do remember KOC saying that powering the GMX from the Vs node provided cleaner power - I see in your scheme you are powering it from Va.
My design uses a cathode biased power amp, we'll see how the SB power scaling with both an improved bootstrapped master, a PPIMV and GMX works out.
John
Re: New trainwreckish Plexi build
Thank you for your input, John.   Actually, GMX kit's documentation states Va, but Vs is certainly worth a try.  
Pablo
			
			
									
									
						Pablo
Re: New trainwreckish Plexi build
That's a Merlin Blencowe special and prevents arcing at startup. Cheap insurance and tonally transparent.jelle wrote:Question: what is the 10k and diode at the cathode follower?
Re: New trainwreckish Plexi build
Thanks guys! I did buy the valvewizard books but I'm afraid that they are out on permanent loan......  
			
			
									
									
						





