All wired in

I populated the circuit board and made all remaining connections (... except one resistor which I just noticed from the picture is missing in the bias circuit). 

(board populated and wired in - front view)

(board populated and wired in - rear view)

(input section detail)

Next, will be a lot of continuity testing, tightening up mechanical connections of the pots, switches, and other components. Then, I'll follow a power-up sequence and do some testing to ensure everything is in the right place. Rob Robinette has a sensible procedure I planned to review, but I also just found this one by one of the amp builders I like (retired FAA technician who claims to be a hobbyist but is clearly a master of his craft), and am relieved to discover that I have already followed all of his steps up to this point!

You may notice some of my components have a black mark on one end. These are capacitors, which are designed sort of like ... Drake's Cakes Yodels or home made jelly rolls: two sheets of paper or plastic stuff rolled up together with electrolytic cream filling between them. The two terminals sticking out are each connected to one of the two sheets. The sheet that ends up as the outer most layer acts as a shield in the same way cable shielding works, so you always want the terminal that is connected to the outer layer closer to ground in the circuit to bleed off any interfering radio signals. The thing is, not all capacitor manufacturers mark which terminal is connected to the outer layer, so one has to just listen to them to figure it out. By connecting test leads to each terminal of a capacitor and plugging them into an amp, you can almost always hear a slight hum just from ambient RF and nearby A/C circuits working. By reversing the connections of the test leads - switching them back and forth - you can hear which way is quieter and identify the terminal connected to the outer sheet. and mark it so it's easy to orient correctly when placing in the circuit.

Sockets wired

The plan for socket wiring looks like this:

 (pre-amp socket wiring plan)

The idea is to keep signal wires away from the high voltage and heater wires, or to cross them at 90 degree angles. 

Here's how they look, mostly complete:

(Pre-amp and phase inverter sockets wired)

The pre-amp sockets are closer together in reality than as shown on the plan, so a couple wires on V2 (a plate load wire and a cathode wire) are actually closer together than I'd like.  I may be able to separate them a bit, but will wait until I complete the wiring of the leads, then try to bend things around one final time.

(Power tube sockets wired)

Speaking of leads, the thick looking wires in the layout  figure at top are the ones carrying the low voltage signal to the preamp tubes, or are otherwise connected to the sensitive tube grids (my guitar pickups typically generate up to about 300 millivolts AC). For these, it's good to encase the signal wire in a mesh wire shield, pull a sleeve of heat-shrinkable insulation over it (hence the added thickness), and leave the twisted mesh sticking out at one end so it can be connected to ground. You can buy shielded cable already made up, but since I have the materials, why not just make it. The net shield deflects radio frequency signals to ground before they can find their way into the amp via the signal wire. One could, and many do, add shielding to many more of the wires, but in this amp these are the ones most vulnerable to such interference since they are long in length and connect directly to the control grids of the tubes. Plus it's a pain to do, so I just did the four. Here are three of them:

(Pre measured Shielded leads)

These will be placed last, since they run in the space above the circuit board, and would get in the way of populating the board if I do them now.

Front panel wired

After one last validation of the schematic to the layout and the layout to what I've actually done so far, everything seems to be in order. I labeled the leads hanging off the circuit board and laid the board in where it goes, then sorted out pathways of the wires underneath board to untangle them, minimize crossing, and to make sure there was enough length in each wire before starting to solder the connections. 

(checking board alignment and organizing lead pathways)

The first connections to solder were the grounds running from each ground rail (see star grounding on Aiken's website), and the B+ leads to each filter capacitor for each power supply stage. Once I wired in the first stage filter caps (the big blue things), I tested them to confirm I had the full "unloaded" DC voltage I should have at the top of that circuit - a few weeks/steps ago, I had tested the rectifiers and found that without the rest of the filter circuit, they produced well under 400 volts, so was relieved to confirm that with the filter caps and choke in place, close to the right voltage was present. I also tested the bleeder resistor (you see it nestled between the two caps) and confirmed it drains the circuit of stored high voltage down to a safe level, about a minute after I switch the power off.

(testing the B+ "A" node)

("doghouse" wiring complete)

(power supply and heater circuit points tested to date)

Then I completed all the front panel connections I could. The junctions and wires you see still unconnected also involve wires that connect to the tube sockets or elsewhere, so I'll complete those when I have those wires ready to go. 

(front panel connections soldered)

I'm finding it tricky to get the soldering iron tip, the solder, AND a visual line of sight to all meet at the right spot, and do a good job. Some of my joints are messy, but they all feel solid and I tested them for expected continuity and resistances and all seem good.

Next: secure the board (which has been unfastened so I could peek under it and adjust leads as I went) and connect the leads from board to sockets and back panel...

Wiring the circuit board

I started attaching the wires that run from the circuit board to elsewhere in the amp. There are ongoing debates about the best wire to use for different situations. For example, some wires carry high voltage DC to the tubes' anodes (plates) while other wires carry low voltage AC signal to the tubes' grids, and one can understand how the type of wire could ultimately make some timbral difference. Some folks say they can actually hear the differences among wires of varying gauge, composition and construction (stranded vs. solid), and others say they can't. Some say selection choice is more about ease of laying the wires and making good quality connections, or whatever Leo or Jim's engineers decided back in the 50's and 60's. Having no real interest in historical accuracy nor any useful experience doing this, I lean toward the "ease of use" approach (lighter gauge solid wire), thinking I'll probably just do a better job overall since it's easier to put things just where I want... but in other projects I've done (from carpentry to coding) this approach has often worked against me in the end, coming also with limitations in durability or flexibility. So, I go with Michael's recommendation and use silver stranded #18 definitely for the strictly-signal wires and also for as much of the others as I can. I have a lot wire to choose from, but I don't think I have quite enough to be consistent throughout. so I'm using other #18 stranded, as well as some #20 solid in cases where I have multiple wires connecting on one terminal and need to conserve space. And, my predisposition to conserve and repurpose materials means I'll also use up scrap wire bits trimmed from the transformer (which are so pretty!). So, it's a hodgepodge with an aim to place the highest quality wire where it probably matters most, and my whims, within generally acceptable bounds, for the rest.

I use the scale drawing of the chassis floor to measure the length of each wire lead before soldering it on. Grey #18 stranded for the plates and cathodes, black mostly stranded some solid for grounds, red stranded mostly #18 some #20 for DC supply, white/blue #18 stranded for the signal wire)

(Circuit board with about half the wires in place.) 

I've completed the front panel controls as much as possible prior to making the connections with the board and mounted everything, though some front panel pots remain only lightly attached since they'll have to be pulled back to access the front terminals when I make the final connections to the board. Throughout the process of adding components, I was doing testing to make sure my switch and pot connections were good, and at one point got unexpected results. The connections were good, but I realized I had soldered the wrong value resistors into place from the jack (tip connector) to ground - should be 1M ohm, and I had 100 ohm resistors in there for some reason. I suppose it's partly because I haven't fully sorted and binned the gajillion resistors that got mixed up in transit, and some of them ended up in the wrong drawers. But mostly, it's because I have never bothered to learn how to interpret the resistor codes printed right on them (just laziness on my part).  So I test them before placement to confirm the values, but must have forgotten to do that in this case. So, I pulled the jacks once more, and put in the right resistors. 

(The mostly-completed front panel controls. (Left to right) presence (high cut), master volume, mid, bass, tone stack select switch, treble, normal channel gain, lead channel gain, (bottom 3 in a row) gain boost/cut switch, V1a cathode resistor select switch, normal channel impedance Hi/Lo switch, (above the three just noted) channel mode select switch, lead and normal input jacks.)

On the rear panel, I had trouble making a good solder joint with with the output transformer leads to the speaker jacks, the solder just kept balling up and looking wrong. Turns out, these leads, since they come directly from and are probably literally a continuation of the secondary winding, are coated with the same thin insulation material (think thin coat of polyurethane) that keeps the wires of the winding coil from shorting each other. Since I clipped these leads down in length, I clipped off the ends that were left as clean uninsulated wire, so I as was trying to solder insulated wire. I discovered that one must burn off the insulation using molten solder in order to clean the wire. Immersing the wire end in a glob of hot solder for 15 or 20 seconds seemed to do the trick. Normally I'd be able to put my ohmmeter on each end of the two wires soldered to the joint to test continuity, but in this case, the end of the transformer wire is buried somewhere inside the transformer, and I just haven't taken it apart to test it yet, so we'll see.

 

(rear panel connections complete: (left to right) speaker outputs, negative feedback selector switch, power tube bias test points,  rectifier select switch)

I've also updated my layout (below) to better reflect actual placement of things... updating actual wire colors will be the last thing I do since that' s a work in progress. 

You may notice there are some new resistors on the pots that are not shown on the schematic (which I'll update one of these days) - this is due to my not having the right value pots in some cases, and really wanting to use what I've got. The presence pot is 25k ohms (should be 5k), the TMB side of the treble pot is 500k ohms (should be 250k) and the TMB side of the mid pot is 250k ohms (should be 25k). It's mainly that I'm thrifty - I just can't see spending $10 - $20 on special duplex pots that have two different values on one spindle, as my dual tone stacks require. And in the case of the presence control, I've already got a PEC 25k pot (military grade high quality) so why not use it instead of doling out another $4 + shipping?  The extra resistors work in parallel with the pot resistances to make them function as if they were the spec'd values, so the electrons moving through the controls "feel" as if they're experiencing what's shown in the schematic. 

... And, here's the fully wired board. 

I still need to clean up the flux and solder splatter and and go over all the connections to be sure they're sound. Next will be going over the final schematic to be sure it's right, and final layout once more to be sure layout matches schematic matches what I've done to date, before I shoe-horn the board into the chassis, install components, and complete all the connections.

Input Jacks and front panel controls

I'm in the midst of pre-soldering a bunch of components and leads onto the input jacks and switches before installing them, so I don't have to do as much work in a tight space. This includes the "top" of the pre-amp ground buss - the thick bare wire running over top of the jacks, that will connect to the pre-amp ground on the circuit board. 

Unfortunately, after taking this picture and admiring/examining it, I realized I soldered the wrong terminals of the two input jacks onto the ground, so will have to un-solder and re-do it. Oh, well... it isn't the first time I've had to do that, and won't be the last I'm sure.

To prevent more of that, and just to be extra sure I wasn't missing anything critical in my thinking about how to set up the Mode switch, I re-checked my map of signal paths for all cases of channel used and switch position. The diagram excludes two of them: selecting the Cascade mode when plugged into the Normal Ch is not something I'll ever do since (1) the signal flow is designed purposely to go the other way, with tone shaping, voltage dividers and bias options on V1B that assume it will be the second gain stage in this mode, and (2) though it will work technically, it will probably result in a very low-power signal and an ungrounded grid on the Lead Ch, which may introduce noise. And only one Jumpered mode is shown - the second is identical to the one shown, just with the signal coming in via a different jack, so will be electronically identical.

Although I did this a few months back, it might be interesting for you to see where I worked out how the mechanical switch would address the various logical combinations/requirements of the mode switching. I wanted a 3 position toggle switch to do the job, and the one I found has twelve terminals (the round dots) with blades that connect them (the thick dark vertical lines) in various combinations. When the switch down, four blades connect each pair of terminals in the top and middle rows. When the switch is centered, the first and last blades slide down to connect first and last terminals of the bottom row with their middle row partners. And when the switch is up, The four blades connect each pair of terminals in the bottom and middle rows. Given that system, and knowing the logical requirements of my circuit, it was fairly straight forward to figure out whether and how the switch could work. 


 

Rectifier and heaters

I wired up the rectifier socket, complete with safety diodes to protect the PT in case of shorted tube, and a 2-phase solid state rectifier (which is just a pair of additional diodes), connected to a switch so that I can easily switch between tube-rectified and diode-rectified voltage. The rectifier takes high voltage AC current and converts it to high voltage DC that can be used by the power tubes and pre-amp tubes to amplify the signal at various stages in the signal path.  In my case, the PT is generating 350VAC, which will be converted to something between 400 and 450VDC for tube rectification (with different tube types generating different output volts) and 455 to 465VDC for solid state rectification. The layout snippet will put the second picture in context. 

(Rectifier layout)

(Rectifier all wired, and power tubes with heater wires complete.)

I also wired up the heaters to all the tube sockets (tubes use filament-type heaters, like an incandescent lightbulb, to free up electrons that then flow from cathode to plate). For a number of reasons:

• because we don't want heavy AC current traveling in wires near to the delicate low voltage signal 
• wires, which can pick up the noise
• and because of the closeness of the pre-amp tubes to each other and to the circuit board, 
• and the way the signal wires come from underneath the board, right down against the chassis 
• and also because I just wanted to try different things

...  I used two different methods to wire the heaters, both common practice though some folks claim one or the other method to be superior. In the case of the power tubes, I laid the twisted heater red and black wires (stranded #18) down on the chassis at the back and came forward to the tube sockets in linear fashion. For the pre-amp tube heaters I used blue solid core wire (which keeps its shape and stays where you put it space), following the same route at first, but then went skyward and arced up and down into each socket. Both methods keep the heater wires as far away from signal wires as possible, and/or on dissimilar axes, to prevent noise induction from the electromagnetic field created in the wires.

(Pre-twisting the power tube heater wires.)

(Final position of pre-amp tube heater wires.)

(It's a lot wire. But all tucked back as far as possible away from signal circuits.)

Layout adjustment 1

Doghouse for some of the filter caps

Recently, when I was dry-fitting the large components to get a sense of the space in the chassis, I realized my plan for laying out the three 22uF filter caps along the front edge of the circuit board would not work because the backsides of all the potentiometers (tone and volume dials, which in my design are duplex pots, twice the depth of the normal ones on the standard Bassman) would trap the relatively large caps in place, preventing easy maintenance. And filter caps are one of those things that sometimes need replacing. The advantage of putting them there would have been that they would be physically close to the portions of the circuit they serve, while not taking up space on the circuit board, which I wanted to keep as small as possible. This was an idea Doug Hoffman showed in some process photos of a 60's era Fender he was renovating.     

So, the current plan is to relocate just those three caps to the outside of the chassis, in their own compartment (doghouse) to prevent the high voltage leads from being accidentally shorted. The location will be very near to where they would have been, just on the other side of the chassis panel, maintaining the physical nearness to the circuit, yet getting them out of the way and in a location better for maintenance. 

I made and mounted a little circuit board just for them (circled in red).

No selectable speaker impedance 

When I was thinking of what I wanted in my design I thought it would be good to have flexibility in what speakers I could use with the amp. Other amps I've had offered multiple speaker jacks supporting speakers of various impedance, and I've found that useful. Many outputs transformers (OTs) have multiple secondary outputs taps to drive different speakers, often including 4, 8 and 16 ohm taps. I assumed mine offered such options as well, what with all those wires, so that's what I designed for. Actually the array of wires on my O45RS OT are on the primary side to better match various power tubes, and on the secondary side it has just one 16 ohm tap. Turns out, this is really just fine - it's slightly less efficient, but driving a single 8 ohm speaker works just fine with that tap, or I can drive two such speakers in series, both of which are options for me. The efficiency issue probably only produces subtle audible difference at high volumes.  Removal of the impedance selection switch actually creates a bit more distance between wires in that area of the chassis, which could help limit any interwire signal interference. It also requires a change in resistor values to drive the switchable negative feedback circuit, since that circuit will now be fed from a 16 ohm output tap instead an 8 ohm tap. 

That part of the layout plan now looks like this:

Early Wiring

I followed the advice of one of the amp gurus and hung paper copies of my layout and schematic behind the work table so I can easily check my work against design before I do it.

I worked on a number of controls I can do prior to their installation in the chassis - it's easier to solder components between the tiny lugs of a switch when the switch is held in a little clamp right at sternum-level, rather than mounted in the chassis, maybe at an odd angle or buried under other stuff. That included jumpers and resistors on some of the tube sockets, 

bypass capacitors on the volume pots, as well as resistors and common grounds on the set of 4 mode switches for the input channels. 

I installed the IEC connector (wall power input receptacle), main fuse, on/off switch and lamp, and went on to installed the main transformer, output transformer, and choke and pulled some wires around. This power transformer is truly a thing of beauty, and will power everything in the amp including a variety of heavier than usual (for the Bassman) power tubes - here's the wiring diagram: 

I grounded all the center taps (CTs) to the main PT ground (center of pic below), connected the line-in power to the PT and on/off switch, and did some testing to make sure all the voltages were as they should be. 

Then hooked up the lamp ...  the first bit of work my amp has done - yay! 

Chassis config

I did some testing with the power transformer, output transformer and choke, laid out on the chassis, to determine least noisy orientation. By design, the power transformer (PT) generates a 60 cycle hum that can be picked up by nearby wiring or the output transformer (OT, which ultimately sends amplified signal to the speakers). By connecting headphones to the otherwise disconnected OT  and powering up the PT, I can hear how they interact in close proximity and orient the OT to best avoid the unwanted hum. Same with the choke and OT. 

(This photo, taken later, shows the orientation of the transformers and choke, which seems to eliminate noise in the output transformer.)

Then I drilled holes and started populating the chassis with sockets, stand-offs, etc.

The tube spacing is tight, but I think oriented as best as can be for quiet wiring.

Tranny test

 I couldn't find a wiring diagram for the power transformer, so decided to do some testing to figure it out... 

I later wrote the manufacturer and the doc they sent based on the model # validated my findings and cleared up some confusion about a couple of leads from the secondary winding. This transformer is unusual  (at least for the size/class of amp I'm dealing with) in that it is tapped for two separate 6.3V circuits - I wasn't looking for that possibility as I tested, so having found one 6.3 V circuit, couldn't quite figure out what the additional leads were for.

The transformer is pretty beefy - easily meets my needs, and probably was meant to drive a 4-tube power section.