First tube amp build: Fender Deluxe AB763 style amp (no reverb).

[mikepukmel]

Progress!

Is still very slow for this first build. maybe some other newbie will get something out of this:

- Got most of the components bent up and mounted onto the main board. The Vishay Dale's I got for the 1.5K that go around the 25uf/25v cathode capacitors are just a little too short. After bending, there is only less than 1/8 of an inch that sits in the eyelet. So, the slightest bump on the board, and these resistors roll right off. Too hard to solder, so I ordered CMF60's According to the data sheet, they're about 3.5" long, similar to the other more standard resistors.

- Got the thin brass plate, drilled and mounted behind the amp chassis face. Found that the threaded section of the nice CTS pots I got:

https://www.tubesandmore.com/product...io-solid-shaft

are too short to bolt through the brass plate, chassis and faceplate with the lock washer in place. <grumble, grumble> There is just barely enough thread sticking out to thread the nut on, with the pot directly on the back of the brass plate, with the faceplate on, but no lock washer. Im not sure of the ramifications of mounting the pot with no lock washer on back. <big frown> But I may be forced to do that. Or make this a real proto with no faceplate, and use some cheesy printed label. The amp is only for me anyway, but I hate for it to look awful. And I don't want to wreck my $3.5 pots!

*Finally* got the cap board soldered together, and wired up. Doesn't look really top notch like you guys build. But OK.

The wire I bought from a guy on Ebay is way hard to strip. Some kind of thin but really slippery insulation. I bought a cool wire stripper that is supposed to not nick the wire, but the stripper just slides over the insulation, its so tough. I ended up using one of my evil old style wire strippers. There is a notch that is just a teeny bit too big to strip the insulation off in one shot, but I rotate it around and score the insulation, not cutting all the way in, so it doesn't nick the wire, then grit my teeth and pull the piece of insulation off. So far, I can't see any damaged wire in a few inspections.

I don't think the big metal mounting standoffs and metal screws are going to do the job. They are really very strong and hold the fiberglass boards in place very well, and there is at least 1/4" distance from any electrical connection. But I'm thinking that with the amp powered up, if I have to probe some places in the circuit with it live, Im worried about a bumping the probes into a screw and shorting something out. Thinking about switching to nylon screws, but leaving the steel mounting standoffs on the chassis.

Nothing insurmountable, but ... its not turning out as neat as I would like. And, still no sound!!

[Justin Thomas]

Okay, not sure how much of a tickler you are, but...

Later in the Silverface years Fender would drill some of the holes in the faceplates larger than others, so that the pot mounting washer and nut rested on the chassis, instead of the faceplate. Usual
Ally the input Jack holes would be normal, and then the right-most knob. Then you had the pilot light helping keep the faceplate on at the far right. The oversized holes would still be covered by the knobs.

Just mentioning it as a possible option; maybe you only have buy one or two new pots insteadof all of them, if any if you rely on the input jacks & pilot light to hold the faceplate on.

Justin

Edit: regarding too-short resistors, you could also just mount the resistors on the cap leads. Some of us here do that... also frees up some extra space in the eyelets, which is always nice in a non-tweed Fender. Ampeg, not so much of a concern - I think they used boot eyelets!

Edit II: that opening line should say "stickler," but "tickler" was just too damn funny to take out!

[bob p]

Later in the Silverface years Fender would drill some of the holes in the faceplates larger than others, so that the pot mounting washer and nut rested on the chassis, instead of the faceplate. Usual
Ally the input Jack holes would be normal, and then the right-most knob. Then you had the pilot light helping keep the faceplate on at the far right. The oversized holes would still be covered by the knobs.

Excellent post, Justin!

As it turns out I have a photo in my projects folder that illustrates the solution to this sort of problem.

Example: 1975 Bassman 10:



^^^This case happens to be the opposite condition of what Justin was describing (oversized exposed holes for the jacks rather than oversized concealed holes for the pots), but ultimately they're doing the same thing -- As you can see here Fender had some jacks with short threads -- so short that they wouldn't have been able to use them while going through both the chassis and faceplate. So they enlarged the faceplate holes to allow chassis-only mounting of the jacks, and they used the surrounding pots and pilot lamp to hold the faceplate in place.

On this particular specimen the left channel jacks (not shown) were long enough but the right channel jacks (shown) were not. It looks like fender received a shipment of short threaded jacks and decided to use them on the right side channel in some of the SF amps, where the faceplate would be secured by flanking the "loose" jacks with pots that were properly threaded to the faceplate. They used the short threaded jacks in the middle of the faceplate, reserving the longer threaded jacks for the left channel, so they could pin down the faceplate at the ends.

This kind approach might work for you -- IMO it would be better than the option of not having a faceplate and scribbling labels onto a blank chassis with marker or stick-on labels.

[bob p]

Edit: regarding too-short resistors, you could also just mount the resistors on the cap leads. Some of us here do that... !

That is a common technique for true point-to-point wiring, where the device with the longest/most robust lead is used as the anchor line across two points, and a parallel device with a more flimsy lead is hung across it.

I was going to suggest this but Justin beat me to it: Mount your short-leaded device to the long-leaded device. When I'm forced to do this I bend the leads at 90* angles to create space between the cap and resistor, to prevent resistor heat from drying out the cap. (this more important on power tube cathodes than anywhere else.) Then I'll spiral the short-leads around the long leads as if I were wrapping them around a turret before soldering them in place. The idea is to make the physical connection as strong as possible before soldering.

sorry, no photo handy.

[mikepukmel]

Thanks Justin, and Bob, awesome news!! The jacks I got are longer, so they fit fine, and I haven't tried the pilot light yet, but I think it has twice the thread depth as the pots. Easy to drill the face plate since its fairly thick and easy to handle.

Not one or two pots, but every one. <frown>

Thanks for the tip on wrapping the resistor leads around the cap leads. Ive seen that done in a few amp photos, didn't know whether it would cause problems, thanks!

Wouldn't call myself a stickler, but would really like this amp to come out presentable. Not like Im going to have a plexiglass plate on top of the thing where friends and neighbors could look inside at the wiring job, but . . . would be nice if it was neater rather than not.

[bob p]

Not one or two pots, but every one. <frown>

In that situation you might want to consider replacing a couple of pots in strategic positions with pots that have longer shafts. Use those pots to ensure that there are several good anchor points spaced along the faceplate to keep it tight. that way you get the benefit of the long threaded pots without the cost of replacing all of them.

[bob p]

Heat Transfer

Thanks for the tip on wrapping the resistor leads around the cap leads. Ive seen that done in a few amp photos, didn't know whether it would cause problems, thanks!

IMO the biggest problem comes along when people snug cathode resistors right up against cathode caps in cathode biased output stages. Those cathode resistors generate heat, and when the cap is snugged right up against them the heat is transferred to the cap. The cap responds by cooling itself through electrolyte evaporation, which eventually causes the electrolyte to dry out and causes the cap to fail.

Unfortunately I've seen a lot of builds where the two devices are snugged right up against one another on the board. Although that won't cause an immediate failure it will shorten lifespan for the cap. When mounting resistors that will dissipate any significant amount of heat I try to stand them up off the board a bit, and to position the cap to minimize conductive, convective and radiant heat transfer. By not snugging the cap up to the resistor you can almost eliminate conductive heat transfer from the body of the resistor to the body of the cap (though some conductive transfer will still occur through the leads). then you're left with convective and radiant heating. Radiant heat transfer diminishes with distance, and even a small amount of spacing helps a lot. The key consideration is never to snug the R and C right up against one another for what looks like a nice tight fit.

Heat transfer was a major consideration in the old mil-spec wiring methods, where components were elevated on turrets and the rules required no more than one component to be located on a pair of turrets. Elevation of the components facilitated some convective cooling of the parts. The single component rule made part exchanges faster, and it also helped to minimize heat transfer between adjacent parts.

Unfortunately these rules are routinely violated in the Fender and Marshall type board layouts, where everything is mounted in a single plane, flush up against a 2-dimensional board that doesn't facilitate cooling. Today most builders just copy these old designs without giving it any thought to the compromises that they made, or how the designs could be improved. The F/M layouts were done to facilitate ease of assembly, not for ease of service in the field and not to maximize component life. When you get to the point that you start designing your own boards then it'll be worth looking at the old layouts and improving some of their weak points.

[mikepukmel]

IMO the biggest problem comes along when people snug cathode resistors right up against cathode caps in cathode biased output stages. Those cathode resistors generate heat, and when the cap is snugged right up against them the heat is transferred to the cap. The cap responds by cooling itself through electrolyte evaporation, which eventually causes the electrolyte to dry out and causes the cap to fail.

Unfortunately I've seen a lot of builds where the two devices are snugged right up against one another on the board. Although that won't cause an immediate failure it will shorten lifespan for the cap. When mounting resistors that will dissipate any significant amount of heat I try to stand them up off the board a bit, and to position the cap to minimize conductive, convective and radiant heat transfer. By not snugging the cap up to the resistor you can almost eliminate conductive heat transfer from the body of the resistor to the body of the cap (though some conductive transfer will still occur through the leads). then you're left with convective and radiant heating. Radiant heat transfer diminishes with distance, and even a small amount of spacing helps a lot. The key consideration is never to snug the R and C right up against one another for what looks like a nice tight fit.

Heat transfer was a major consideration in the old mil-spec wiring methods, where components were elevated on turrets and the rules required no more than one component to be located on a pair of turrets. Elevation of the components facilitated some convective cooling of the parts. The single component rule made part exchanges faster, and it also helped to minimize heat transfer between adjacent parts.

Unfortunately these rules are routinely violated in the Fender and Marshall type board layouts, where everything is mounted in a single plane, flush up against a 2-dimensional board that doesn't facilitate cooling. Today most builders just copy these old designs without giving it any thought to the compromises that they made, or how the designs could be improved. The F/M layouts were done to facilitate ease of assembly, not for ease of service in the field and not to maximize component life. When you get to the point that you start designing your own boards then it'll be worth looking at the old layouts and improving some of their weak points.

Thanks again for the comments and tips here. On my dry fit of the main board last night FINALLY! (part leads bent and parts sitting but not soldered in), the signal caps and electrolytics are all mashed right up on the resistors. These audio signal caps are a lot bigger than the orange caps, and have a lot less leeway. Going to take a lot more effort to re-bend the leads and put the resistors up above other resistors and caps with a good size air gap in there. Wish I had started with a turret board rather than an eyelet board. Next time!!!

The reason I wanted to just "clone" an old design was that I was naive to think I would not run into these kinds of problems.

[bob p]

Don't worry too much -- my post was intended to be guidance for a layout, not an expectation that you'd rip apart an existing layout and start over.

Not all resistors generate enough heat to be troublemakers, so you don't have to reposition every component if your layout is already done. Just look at the resistors that dissipate power (generate heat) and be careful not to snug the electrolytic caps up against them.

In a cathode biased amp the 6V6 cathodes would be important. That's not a concern with fixed bias 6V6. It's good practice to think about resistor heat when you do the layout for your noval tubes, but not as critical since they draw less power, generate less heat and aren't likely to cause immediate cap failures.

[mikepukmel]

Having a hell of a fight with the heater wiring. The sockets I got (supposed to be good ones, from a good source), have really tiny slots in the wiring tabs. They are very flimsy and Im afraid to cause metal fatigue by bending them too much. One wire fits in fine, then I can't get the second one in without one or two of the strands bending back. End up clipping off the end, re-stripping. Took a sabbatacal after about 3 hours last night. Got a few wired, then one or two just would not play ball. Maybe I need to find better 9 pin sockets and go from there. Any suggestions on really *good* 9 pin sockets? The 8 pins I got are really solid, they are Belton, with really thick tabs. The 9 pins are also belton but aakk really cheesy.

[mikepukmel]

Just scanned Antique Electronic Supply, and Hoffman. All of the 9 pin sockets look about the same at least in the photos.

[Justin Thomas]

What are you using (gauge) for heater wire? And do you tin the ends first?

Justin

[Enzo]

They will all look about the same. OK, so one wire fits? Great. First, you might be using oversize wire, no need for #12 wire in an amp. But if two wires won't, fit try this. Bare the wire to go into it a little extra longer. Now strip the second wire and wrap it around the first up next to the insulation. The extra length you stripped will now still poke into the socket tab. Solder all.

But really, I'd just knock the wire size down a notch. 20 is plenty.

[mikepukmel]

I think I got 20 for the bulk of the wiring and 18 for the heater. Have to check (its not written on the wire). I did try to tin the wire but I ended up with too much solder on the end of the wire, and the fit was worse. My soldering and wire stripping skills are in need of a boost here!

Stripped properly (which is tough for me to do at this point) the wire slides right into the slot. Then I get the second one all screwed up trying to fit it above or below the first one, in the slot.

[mikepukmel]

with my skills the way they are, I can easily see how the post CBS fender wiring looked so hideous.