January 12, 2006

Tweed Deluxe Reverb? Never heard of it! If Leo had wanted us to have one, he would have made it for us. Who'd want it anyway? 

........Well, I would!

The project took life by first taking the Deluxe Reverb AB763 schematic and stripping out the normal channel and tremolo circuits. Then I added a bright switch and middle control. After some fine tuning in Visio, I cranked out my final working schematic.

The aluminum chassis is a Hammond 1444-20, 17x4x3. It's really thinner than I like. If I were to do this again, I'd use the steel version of this chassis. It's not that much harder to work with. However, it would have been a special order from Mouser, so I opted for aluminum.

The PT cutout was done with a jig saw and some hand filing. A router guide template would prove useful if doing several of these, but I had a nice hole cut faster than I could have built the template. All other holes were done in a drill press using HSS bits or a step bit. The octal socket holes were punched with a 1 1/8" chassis punch.

The control panel layout was done using Visio. The Tweed Deluxe cabinet has a 2" x 12 1/8" cutout, so I had to do some careful layout work to get it all to fit correctly. Layout from left to right is fuse, power, standby, indicator, reverb, master, bass, treble, volume, bright, inputs. That's BUSY!

I had one design layout flaw that I caught while wiring the filament string. I had placed the reverb in/out/footswitch jacks below the reverb transformer for a nice looking and easy wiring job. But my filament string was gonna be tucked away under the bottom lip of the chassis and would pass very close to these jacks. So, to minimize hum, I relocated the jacks and put a small cover plate over the original holes.

All caps came from Mouser. Materials to build the turret board came from Hoffman. The safety shield for the caps was cut from a sheet of 1/8" acrylic that was sandwiched between two wood blocks in a vise. A heat gun was used to make the bend. Hopefully this will keep careless fingers off the business end of those caps!

The transformer set is from a '65 DRRI and was provided by MikeyMann. All wiring (except power cord and transformer leads) is either 18 or 22 gauge teflon insulated 19 strand silver plated copper. I had some on hand and bought the rest on eBay. (There really isn't anything that can't be had from eBay!)

I already had all the miscellaneous stuff, tube sockets, switches, indicator, fuse holder, standoff turrets, etc., from my last build in 1976. Those gold plated sockets looked brand new. Some of the other stuff got a good cleaning.

The next phase of this project is going to get crowded, so this is a good time for some preliminary checks.

All external components have been installed but not necessarily connected. Note that the HV center tap (red/yellow)  and bias tap (red/blue) are not connected at this time. At this point I did a first power on check to verify PT operation and rectifier and filament string wiring. All was well so I put all tubes in for a quick check for lights. And finally, a voltage check between rectifier cathode and HV center tap showed the expected unfiltered 120 hz. DC voltage. All checks at this point passed so it's time to move on.

Building the main board was actually several challenges rolled into one. Divide and conquer. The board needed to be about 2" wide to provide adequate wiring space inside the 4" wide chassis. I spent several hours working with Visio to generate my layout template. Then several more hours were spent going over and over, refining and verifying that everything was correct. It's the old carpenter's rule, measure twice, cut once. I didn't want to find out later that I needed one more turret lug, or a jumper here, etc.

The final template measured 2.2"x10.8". Now the fun part. The G10 board was cut to size and the full size template was printed and attached to the board with double sided carpet tape. All holes were carefully center punched and then drilled at the drill press. (This method works fine for a one of a kind board) Then I used Hoffman's little flaring tool set in the drill press to install the turrets.

Back to the soldering station to install jumpers, wires and load the components. BTW, all resistors are carbon comp, 5% tolerance, high reliability (yellow band = 25ppm, whatever that means!) mil spec.
 

Here's the board layout if you're curious.

The board is fully loaded and ready to be installed. Actually, this was to be the most challenging part of the entire project. I really could have used a swivel jointed curved needle nose pliers with a retractable cutter on the tip. (and one more quart of nerves!) By the time I finished the tube side wiring, I was exhausted. The control panel would have to wait. And there ain't gonna be no rematch, so don't ask.

Now that's a science all by itself! Not just with amp building, but with any complicated electronic system. I work for FAA at an enroute radar station and believe me when I say that a lot of R&D goes into a grounding and bonding system at such a facility.

So I read up on amp grounding schemes. Aiken has some excellent reading. And I pondered what to do....Star, multi-point star, buss. They all have good points and grounding does not need to be taken lightly, especially with an aluminum chassis. My ground system was the very last thing I wired. I was hoping to come upon some magic solution.

But in the end, I went with my tried and true method from 30 years ago. Single Point Ground is what I called it back then. I guess some would call it a Star today, but hey, it could even be a 2" long buss!

The chassis is 90% complete at this point. I say 90% because you can never quite be sure that some gremlin is not gonna rise up and piss in your coffee during the first burn.  I got lucky. The components held up and I only had one minor wiring error. The coax cables that feed the grids for V1 run underneath the main board. Yep, they got reversed. An easy fix that I caught just prior to first burn. I dummy loaded the output and set the bias for 25 ma idle current. All plate, screen, grid, and cathode voltages were checked and found to be acceptable. I just let it cook that way for the remainder of the day. I think it's always good to just let the juices flow at idle for a while before applying an input signal. I made a point to stop, look, listen, and smell every few minutes during the first hour. By the end of the day confidence was high.

BTW, take a close look at that old brown tube box just north of the chassis. Catch the packing date on that dude? That's the same year I got started in all this stuff!

First burn is always a rush. It's usually either Miller Time or a very humbling experience.

Well, it's Friday morning, I'm home alone, what shall I do?

I think I'll try to make some noise. I connected into one of the TRRI speakers, strung out the reverb cables and turned it on. It made some noise all right. SQUEALED LIKE A STUCK PIG! Power off immediately and fire up the soldering iron. My bet is the OT plate connections are reversed. Hey, it's a 50/50 thing and I got it wrong the first time.

The Strat is plugged in and I'm having fun. Something is not quite right though. Sounds like the gain is turned way way up even at a minimal setting. This one took a couple hours to find, mainly because I had to strongly suspect a wiring error or bad component. I danced all around the problem for some time before realizing that the error lay with my rendition of the original AB763 schematic. The vibrato channel has an extra gain stage due to the loading effect of the reverb and tremolo circuit. The tremolo pot puts a 50K resistance on the signal path just prior to the mixing resistors that feed the PI. I had totally omitted this resistor! Gator clipped a resistor in and everything quieted down. The final fix is a 100K pot for a master volume control. I'll just give up the 'Middle' tone control.

After debugging, I spent 4 hours with my strat tied to this thing! I couldn't leave. It felt good. It sounded good. By the time I realized I needed to move on, I'd celebrated too much to do any more work. The cab work could wait 'til tomorrow. Anyway, I just thought of another tune to try!

It's tomorrow! Imagine my joy when this chassis slipped snuggly between the side cleats in this cab!

Actually, this point was my biggest research challenge. I posted the question "what is the exact distance between cleats" in several different places.... Mission Amps, Mo Jo, Weber, and on AGA. I got a variety of responses, but no one could give me a measurment. Well, actually, one guy on AGA nailed my question with real numbers, but by this time, I had charged ahead and bought the cab from Marsh Amps on eBay. I'm pretty sure he's sourcing from Mo Jo. I knew it would be within 1/2' and was prepared to modify the cleats to make my chassis fit. I was pleasantly surprised.

There are only a few things left to do now, 1) Shellac the cab, 2) make a faceplate, and 3) get a speaker.

Shellac is done. I applied 4 coats of Zinsser's Amber cut 50/50 with denatured alcohol.

As it turns out, the speaker selection is going to be limited. There's not much room in that cab. I'll either have to stay with the one I borrowed from the TRRI or find a suitable low profile model.

Almost done. It's been a challenge to fit everything from the onset. As you can see, putting it all in the box was no exception. But my friends, you're looking at the genuine original, one of a kind, AB763  Tweed Deluxe Reverb. Who cares?  Probably no one. And certainly not me. Hey, I got one! What a trip!

As you can see, the speaker selection is done. I chose the Jensen P12N reissue. It has a much brighter sound than the borrowed C12-K from my TRRI.

Well, I guess you knew I couldn't let it be. And I knew it too! I really enjoy playing through this amp, but I love tinkering under the hood.

So I made some major changes. All the iron has been upgraded to Mercury Magnetics and I've switched to 6L6s for the output. I'm currently running on borrowed Groove Tubes from my TRRI but will be changing to JJ Teslas before the end of the week. With the bias set for 46 milliamps and the plates at 394 volts, the static plate dissipation is 18 watts. This amp truly is a prime example of a wolf in sheep's clothing!

The sound is fuller on the bottom and the sparkle is still there on the high end. I don't think I'll be switching back to the 6V6s anytime soon.

While I was under the hood, I removed the Pentode/Triode switch. That was a nice toy to play with, but after a month I found I was just leaving it in the Pentode position. An external speaker jack now occupies the vacated switch hole.

I'm still waiting for a faceplate. A local trophy shop says they can do it but with spring sports season in full swing, they're too busy for now. Meanwhile, I've used Visio again to create and print my own. It looks OK printed on photo paper but is not very durable.
 

(June 28, 2006)...

The 6L6 JJs have been running fine for over a month now. I finally dropped the idea of a faceplate. I polished the front panel to a 'near' mirror finish and used white dry transfer lettering. Looks good!

Oh, and I've replaced the Jensen P12N with a Weber ceramic California.

The 'Tweed Deluxe Reverb' is done!

I hope I get as much enjoyment from playing this amp as I did from building it.
 

Epilog (July 1, 2006)...

I had so many extra parts left over from this project (transformers, speaker, tubes, etc.) that it seemed a shame to just stick'em in a box. But that's exactly what I did....sorta!

Take a look ...

Lil Sister

The wolf in sheep's clothing! Still looks innocent enough. The Weber shines through.

Popping the hood reveals the super charged drive train.... All Mercury Magnetics iron driving JJ 6L6s which in turn drive a Weber California!

Tight control panel.

The wolf! MMs and JJs pump out 28-30 watts of unclipped power into the new Weber Cali.

Still an AB-763, although I now hesitate to call it a Deluxe Reverb.