New project : vajra 36W

clobo · Post by **clobo** » Wed 09/17/08 7:36 pm

Hi,

2 years ago I originally built version 1 to standard Ritchie Plexi specs, using Trinity chassis and custom ordered transformers. Then last year I rebuilt version 2 to sIII plexi specs. Both versions sound good, but too close to my other amps (18W and 50W).

This time I feel I need something completely different, and wanted to try EF86 for normal channel, and cleaner TMB channel. So, I'm actually rebuilding my 36w version 3 to vajra specs.

Board is already finalised and ready to drop in chassis. Just in case, I also planned the board for easy fine tuning, more towards Z/YGM specs (version 4 ?)

I just ordered EF86 and 12DW7, should be in friday. Hope to finalise during the weekend, I'll post the results.

Claude

Post by **zaphod_phil** » Thu 09/18/08 10:32 am

Good luck with the build, we'll keep a look out for your posts!

BTW my own experience with playing an SIII TMB was that I found I preferred the character of its Normal channel.

clobo · Post by **clobo** » Fri 09/19/08 5:56 am

Thanks Phil,

To my ear, comparing Ritchie and sIII (plexi voicing) sounds in one sentence...

Normal channel
Ritchie : original 18W trem voicing, with more punch
sIII : early Zep (telecaster)

TMB channel
Ritchie : lots of options, but basically more agressive (modern) sound
sIII : more vintage (early plexi vibe) sound

In both versions, I prefer the normal channel.

Now I still have to hear the (cleaner) vajra TMB, and (hotter) YGM variant.

Claude

Post by **zaphod_phil** » Fri 09/19/08 8:15 am

It's interesting that you say you prefered the Normal channel of both TMB amps. AFAIK the tone control is the only difference between the Normal channel of Richie's original TMB design and the SIII.

I'm also fine with the Plexi-ish voicing of the SIII's TMB channel. I just felt I had to push it awfully hard to get it to break up, while the Normal channel seemed much more dynamic and touch responsive. Anyway, these are my own subjectives opinions and what matters is that everyone has fun with whatever version of 18W/36W takes your fancy.

Now what is the YGM variant, some kind of Traynor hybrid?

clobo · Post by **clobo** » Sat 09/20/08 5:11 pm

Hi Phil,

There are slight differences between various versions of Ritchie and sIII (normal channel). However I find the major difference comes from the 500pF bright cap bypassing sIII (plexi voicing version) normal volume. At low to medium setting, I find this cap really nails that early Zep sound with a Telecaster. I guess I probably play slightly cleaner sound than average member

You're also right about Traynor. I'm actually aiming at YGM-2 (Guitar Mate) as it is close to actual Vajra TMB channel, but YBA-2 or YGM3-4 would also work here. Understand I have limited choices, because I only have one dual triode left for building the whole TMB preamp.

But for now I'm still focusing on EF86 channel, lots of interesting options to try here, including one very strange (boutique) mod I would like to understand better, but my maths are kind of rusted.

Normal (V) implementation we see everywhere has supply rail directly feeding plate (220K) and screen (1M). However a modern (Z) implementation adds another 68K between the supply rail and the previously mentioned circuit.

I would really like someone to explain that mod.

Thanks

Claude

Edit : added drawing

klingo · Post by **klingo** » Sun 09/21/08 8:23 am

Hi claude,
i'm curious about the modern implementation. is the screen grid decoupled with a capacitor? is the value still 0.1uf?
is the Z letter refering to the so called doctor?
sorry for all these questions and not being able to help

Post by **zaphod_phil** » Sun 09/21/08 10:31 am

clobo wrote:There are slight differences between various versions of Ritchie and sIII (normal channel). However I find the major difference comes from the 500pF bright cap bypassing sIII (plexi voicing version) normal volume. At low to medium setting, I find this cap really nails that early Zep sound with a Telecaster.

That's fascinating. I'm going to have to check that out.

clobo wrote:Normal (V) implementation we see everywhere has supply rail directly feeding plate (220K) and screen (1M). However a modern (Z) implementation adds another 68K between the supply rail and the previously mentioned circuit.

I'm not sure what that was for - possibly trying to fix a hum problem in the amp.
However, there's a whole bunch of interesting things you can do with an EF86 in a guitar amp. For example, reduce the screen grid resistor value - try 470k , 330k or 220k, to make the stage more non-linear and less hi-fi. You can also put a 250k pot between the screen cap and ground, to vary the stage's natural compression. All good fun stuff to play around with.

DanLurie · Post by **DanLurie** » Sun 09/21/08 12:17 pm

The new VOX AC15 (classic or something like that) the EF86 has a pentode/triode switch.

clobo · Post by **clobo** » Sun 09/21/08 3:00 pm

Hi,

Yes, I saw that EF86 implementation while repairing a so called doctor amp. The other components around EF86 were pretty much the same we see everywhere.
- supressor and shield all tied to cathode (pins 3-8-2-7).
- .1 uF screen decoupler, to ground
- 2.2K/25uF : cathode (+ others), to ground.
- there was a very strange tone control, worked backwards, had lots of range, but also had quite strange voicing.

So far the amp works, both channels sound fabulous, overall... very loud, clean, compressed, responsive, quiet

(...well sometimes..)

However I still have to deal with severe oscillation, occuring only when chassis is open (no shield), and no guitar actually plugged in. Problem seems originates in the second 12DW7 plate circuit, 150K-50K split feeding one side of PI. There is no resistor (or pot) to ground, so this whole section is floating at B+ DC voltage, and tends to act like an antenna.

Will have to wait until next weekend, in the meantime I would apreciate tips from other builders. So far my plans are to add a separate supply section for the 12DW7 preamp, actually shared with PI (more that 2 stages..)

Thanks

Claude

Thanks

Claude

DanLurie · Post by **DanLurie** » Sun 09/21/08 7:22 pm

However I still have to deal with severe oscillation, occurring only when chassis is open (no shield), and no guitar actually plugged in.

Why is that a problem? With the amp buttoned up, does it oscillate? If not, don't worry about it. Many amps need their shield to be stable.

clobo · Post by **clobo** » Sun 09/21/08 9:43 pm

Hi,

I played the amp for a couple of hours, in its (shielded bottom) cab this time.

The amp remains stable until TMB inputs are unplugged, then oscillation cycle starts, along with hum and crackling (on all jacks and pots). The funny thing is, just barely touching input jack sleeve with a guitar cord clear all problems at once

Might be an intermittent Cliff jacks problem (happened before), or a ground loop problem, or DC on jacks. Unfortunately I will be on business trip this week, so debug will have to wait.

BTW this amp can be pushed extremely loud, while staying so articulate, never boomy or shrill. 12DW7 TMB channel is a mix between early Blackface and JTM-45, while EF86 channel is a mix between early Vox and modern boutique amps.

Here's an external picture, I'm actually recycling chassis, transformers, and cabs from a previous built (4x10 Celestion G10S-50). I'll show internal guts later, as I need to solve the problem then cleanup the mess.

Claude

DanLurie · Post by **DanLurie** » Mon 09/22/08 5:32 am

Hey man, the amp looks just like a Marshall

I bet that noise problem is the cliff jacks... I've said it a bunch of times.. I don't like cliff jacks!! Get some switchcraft jacks (L12A) and some Insulated fiber shoulder washer and do it right.

clobo · Post by **clobo** » Fri 09/26/08 8:44 pm

Hi,

I'm just back from my business trip, I will look at this issue more closely over the weekend.

I know it's a Voxish amp now, if I decide to keep it, I will eventually build another 2x12 open back with Alnico. I built the actual cabs 2 years ago, went for the Marshall look cause I was initially building a Ritchie plexi.

However this 4x10 cab sounds pretty good, and is much easier to carry around than my other full size 4x12.

Claude

clobo · Post by **clobo** » Sat 09/27/08 2:30 pm

Hi,

Main problem was my 12DW7 second section plate circuit layout. This high level signal was too close to its own TMB input jacks (both signals being in phase). I revised my layout and moved signal bearing components on the tube side of the board, while B+ is now being fed from the control side of the board (DC supply = signal ground).

I also moved the 68K grid resistors from jacks, to tube socket. This might not have the proper vintage look, but it is technically correct and also improves shielding around the jacks and leaves fewer bare wires (possible antennas).

Finally one of the Clif jack had a contact problem.

Amp is now stable at all settings, however the TMB channel still generates slightly more noise and hum than my other amps. Adding a master volume (at PI input) would solve the problem, however this would defeat the essence of that extra-clean channel.

Claude

clobo · Post by **clobo** » Mon 09/29/08 10:11 pm

Hi,

Here's my revised Vajra-36 board, this one is pretty quiet. Curiously layout has not changed much from official RP.

Master-Vol is not used, seems like a good place for PP/SE control.

Also the EF86 can support Z option, replace red jumper by 68K and also lower the last voltage dropping resistor (47K > 27K). Hard to explain the mod, but I hear slightly less gain but more solid drive (partial triode mode

Claude

Fredaxis · Post by **Fredaxis** » Tue 09/30/08 12:42 pm

Bonsoir Claude,
About the PP/SE pot in a 36W....
Will it still be a PP/SE pot or a 2 tubes/4 tubes???
I have no schem on hand but wondering???

Fred.

Alexo · Post by **Alexo** » Tue 09/30/08 1:34 pm

Heya, try grounding the suppressor grid on the EF86, instead of tying it to the cathode like everybody else does, I find it works better that way and the voltages fall into place where they should, curious to hear other's thoughts....

clobo · Post by **clobo** » Wed 10/01/08 8:09 pm

Fredaxis wrote:Bonsoir Claude,
About the PP/SE pot in a 36W....
Will it still be a PP/SE pot or a 2 tubes/4 tubes???
I have no schem on hand but wondering???

Fred.

Bonsoir cousin, enfin, bonne nuit chez vous...

Yes it still is a PP/SE, this control acts like a post PI Master-volume, applied to only one side of the Phase Inverter. This option was part of the original Vajra 18 (found in the download section), but I have not implemented it (not the sound I'm looking for).

Claude

Post by **rjgtr** » Thu 10/02/08 8:03 am

Hey I'm glad to see you making this. I build a Vajra (see my gallery) and it is a very cool amp.

Have you solved all of your issues?

clobo · Post by **clobo** » Thu 10/02/08 10:35 pm

rjgtr wrote:Hey I'm glad to see you making this. I build a Vajra (see my gallery) and it is a very cool amp.

Have you solved all of your issues?

Hi Richard,
You built a very nice amp. In fact I did study that layout, but at the end did not follow it because it did not fit my chassis (Trinity) very well.

I solved all oscillation issues on my initial prototype, mainly by repositioning the 12DW7 second plate circuit farther away from TMB input jacks. However I still need to further reduce idle noise and hum coming from the TMB channel (remember, no master-volume).

For now (if I can get some spare time) I will be rewiring a clean amp as shown, and try to still improve from there.

Claude

Donations

Links

Welcome

New project : vajra 36W

New project : vajra 36W

Ocillation problem fixed

Board