trinityamps.com

Hi All. I've really enjoyed reading build threads on this forum. I've learned a lot from -- and been inspired by -- threads like the ones from No457 Snowy (viewtopic.php?t=3188), recently Moodivarius (viewtopic.php?f=16&t=6267&start=0), and Stephan's first production build (viewtopic.php?t=2062).

I'm very slowly working my way through building a Tramp combo amp, so I'm going to post a build thread of my own. I mostly play mostly older rock and a few originals through a Fender Blues Jr III (the tweed version with the Jensen speaker), and my goal is to expand my range of tones without sacrificing portability.

This is my first ever amp build, and I'm excited. I'm about half-way through the build now, but I'll start at the beginning and catch up over several posts.

[Edit: I ran into clearance problems with this choice of cabinet/speaker. Please see my post on Mon Jan 06, 2020 9:23 am in this thread for details.]

I started with the cabinet. I wanted a Deluxe cabinet. Next build I'll try to make my own, but this time I shopped around a bit and wound up ordering Mojotone's "Narrow Panel Tweed Deluxe Style Cabinet" (https://www.mojotone.com/Cabinets_x/Cab ... er-Cabinet). Hopefully it's OK to post links to vendor pages in this forum. I didn't see any forum rules prohibiting it, but please let me know if I'm out of bounds.

I was confused by all the options, so I went with the stock cosmetics (unlaquered tweed, strap handle, chrome glides for feet). If I were doing it again I'd order it with tinted laquer, which costs an extra $40.

After I got the cabinet I did some reading online and decided I wanted the tweed to be laquered (for durability, stain resistance, and good looks). There are a ton of howto posts and videos with different recipes for laquering tweed, but I liked the very brief writeup on Carl's Custom Amps page (https://carlscustomamps.com/what-is-lacquered-tweed), so that's what I did. Next post will be about laquering -- or polyurethaning, I guess -- the cabinet.

For laquering the cabinet I used Minwax PolyShades Satin polyurethane with "Honey Pine" tint. Buying a quart of the poly, some mineral spirits, a decent paintbrush, and a package of 200 grit sandpaper cost me around $40. So ordering the cabinet already laquered would have cost about the same, and wouldn't have left most of a can of poly going to waste. Live and learn.

If I were doing it again, and if I had some clear polyurethane handy, I would use clear poly for the first coat, then switch to the tinted stuff. Using the tinted poly, it was really hard to keep the first coat even because it wanted to soak into the tweed immediately. This made things a little blotchy. Here's a picture after the first coat. You can see blotches on the side of the cabinet.

OTOH, it looked better as I added more coats. If I were doing this again and I _didn't_ have clear poly on hand, I'd have to think hard about whether it made sense to buy and open a whole second can of finish just for a single coat and an improvement that nobody else will notice.

Carl's recipe called for 4 coats, but I wound up putting on 5 or 6. I thought it looked better with each coat as the poly filled up the texture of the tweed and leveled out the surface. Overall, I'm really happy with how it came out.

OK it's late. I'll post more on another day.

Keep it coming. Like learning from others.

_________________
"The only problem with being retired is you never get a day off"

Thanks for the encouragement. I found some more photos of the finished cabinet. Here are some closeups that show how it came out. The picture of the front makes me think I should have done a couple more coats to make it look more like the bottom.


I ordered the cabinet without mounting holes. I measured, marked, and drilled pilot holes from the inside because the sides of the cabinet give a good references to measure against. I drilled to final diameter from the outside to avoid tearout on the visible part of the cabinet -- although the tweed might have protected from that as well.


[Edit: I ran into clearance problems with this choice of cabinet/speaker. Please see my post on Mon Jan 06, 2020 9:23 am in this thread for details.]

I spoke with Stephan on the phone about what speaker to install, and went with his suggestion of a Jenson Blackbird. This amp project is a splurge anyway, so I held my breath on the price and just ordered it. But wow, that's an expensive speaker. I like to support local business, so went to the Jensen website and found an independent dealer close to me (http://www.huberbreese.com/).

This speaker has a 29oz Alnico magnet, which explains its price. I only have a vague understanding of why Alnico speakers sound good (mostly learned here: https://www.tedweber.com/lets-talk-speakers/), but I wonder if the little Tramp has enough grunt to drive such a big (100W) speaker into compression? I guess I'll find out how it sounds later.

OK, next post will be about building the actual amp.

Awsome looking work.

I have the WGS Blackhawk HP, 100watt, sounds great on my Tramp.

I think you'll be happy. Very defined crisp sound.


Scott

[Edit: This post describes some issues with lead-free solder, but I think I've resolved them by using a different flux and a hotter soldering iron tip temperature. Please see my post dated "Tue Dec 31, 2019 9:06 am" later in this thread]

Thank you! I can't wait to hear how it sounds.

I had some confusion right at the beginning of building up the amp chassis. The Tramp Builder's Guide says "Align the sockets so that the Octal (8) pin #1 points toward youand Noval (9) pin #1 points away from you," and I wasn't sure what that meant. The suggested orientation is clear from the layout diagram, but it took me a while to figure that out. In any event, here's how I installed the tube sockets.

[Second edit, Feb 2020: After a ton of reading about the environmental costs I mention in the paragraph below, I'm convinced they're not an issue, and lead-free is the right choice for me. Just circling back so as not to leave this loose end dangling.]

The Guide recommends 60/40 rosin core solder, and soldering lead-free is supposed to be more difficult. Lead-free solder is also supposed to have environmental costs that I don't yet understand. I decided to try this build using lead-free solder anyway. This is another first for me: first amp build; first experiment with lead-free solder. I'm very open to advice/feedback/criticism on this choice, so please don't hesitate to post.

After some reading (https://www.kester.com/Portals/0/Docume ... .19.06.pdf , https://www.assemblymag.com/articles/84 ... ng-by-hand , and others), I settled on SAC305 solder. "SAC" is an abbreviation of the atomic symbols for tin (Sn), silver (Ag), and copper (Cu). "305" describes the proportions of the alloy, which is 96.5% tin, 3% silver, and 0.5% copper by weight. From what I can tell, it melts at a lower temperature than most other lead-free wire solders, and flows pretty well, athough not as well as tin/lead solder. I chose 1mm solder wire as a nice compromise between having enough control for smaller connections and having enough volume to fill up the eyelets on the board.

Here's the solder I started with, but don't buy this for your build. I ran into some problems that I'm still figuring out, so I can't yet recommend it yet.

I have a nice Weller 80W soldering station, so I ordered pretty big chisel tip to try and get heat into the big eyelets quickly. I got both the solder and the soldering tips from Digikey (https://digikey.com).

The flux core on the solder I ordered is Kester 268 "No Clean" flux. There's a lot of it (3.3% by weight), and I was hoping this would mitigate some of the challenges of going lead free. It works fine on smaller components, although it definitely doesn't flow as fast as 60/40 tin/lead, and doesn't look as pretty.

On the other hand, I couldn't get it to flow onto the tag strips at all. This doesn't seem to be a heat issue. The solder would melt on the tag strip, and I could see the flux covering the metal, but there was no adhesion between the solder and the tag strip terminal. I tried sanding lightly to break through the oxide layer, but got nowhere.

At this point I decided I'd solder at least the tag strips and mains connections with leaded solder. I had this spool on hand, and it works quite a bit better on the tag strips.

Stay tuned for more exciting adventures...

Suggestions: 1) Rotate the tube sockets 180 degrees (so key way on the octal socket points down and to the left, viewed from inside and the noval socket gap between pins 1 and 9 points down and to the left); also for easy identification mark pin 1 on the octal socket. 2) The 63/37 solder works great for printed circuits, but is a PITA on eyelets and lugs because it goes from wet to solid rapidly (works great on the small small red switches and small lugs of push-pull controls); tried to use on last Tramp build and it globbed out on the backside of the board (as expensive as solder is, I still went back and got the good old 60/40). YMMV.

Hank

_________________
"The only problem with being retired is you never get a day off"

Thank you for posting this. I wondered why the Guide called for 60/40 rosin core specifically, and your post explains it. This is why it's great having a community... to beat sense into people like me who have trouble following directions.


This is also good advice, but I think I'm already doing as you suggest. In the photo above, the chassis opening is toward the bottom of the picture, so when you lay the chassis down and look from inside, the keyway and noval socket gap both point down and to the left. I'll post a photo from inside later on today to double check.

Thank you again. I really appreciate the feedback.

LH, don't change the sockets, rechecked myself and you got it right the first time. My apologies. I still suggest marking the octal pin 1 on the inside w/ a Sharpie or something. Makes it a lot easier to find the right pins since it's hard to see the keyway from inside

_________________
"The only problem with being retired is you never get a day off"

Thanks @Hankules. That's a great idea to mark the pins on the inside of the chassis.

This post is about filing things. I had some trouble with the LED bezel and lock-ring. There's a groove inside the bezel that's supposed to snap around the skirt of the LED. I couldn't get the skirt to fit into the groove; it looked like the skirt was just a bit too thick (or the groove was too narrow). Here's a photo -- with the assembly removed from the chassis -- illustrating the issue. You can see that the "arms" of the bezel are splayed out just a bit to accommodate the LED. This prevented the lock ring from sliding over the splayed arms and securing the LED in the chassis. When the led snaps in properly, the arms close up just enough to let the lock ring slide over them.

I solved this by lightly filing the bottom face of the LED (the one that the leads come out of) with a small flat file until the skirt on the LED fit the groove.

While I had the file out, I used it to shorten the tabs on the potentiometers so they don't poke all the way through the chassis and interfere with the faceplate, as suggested in the builder's guide.

Finally, I couldn't get the fuse holder to slide into the cutout on the chassis without forcing it, so I dressed the cutout slightly with a round file on the curved face and a flat file on the flat face. This fixed the issue.

By the way, don't be fooled by the filing. This chassis is gorgeous. When I ordered the kit, I didn't apprecate how nice the parts were going to be. It's a real credit to Stephan, and I couldn't be happier. I'm also really impressed with the way things are layed out.

Interesting change: the schematic calls for an audio taper pot in the power control, but the supplied part is linear taper. I don't have enough context to know which is preferable... just noticed it in passing.

And then I was ready to run my first wires. Woo!

The LED rings are really tough. I have simply cut them so they spread open to fit the LED but they should slide on.

The change is OK. You will find the control acts a little differently but gets you to the same place in the end.

_________________
Stephen
Web: www.trinityamps.com. Facebook: facebook.com/trinityamps. Twitter: @trinityamps

Thanks Stephan. That's good to hear.

Moving along with the wiring, here's heater wiring and mains connections all completed. I felt a little silly heat-shrinking the mains connections, when right next door there are going to be uninsulated components over 300V. OTOH, the hot leg of the mains connection is the only thing in the chassis that isn't fused, so maybe it's justifiable.

I installed the green neutral wire from the power tag strip, and just left the other end flying. I'll solder it to the board once the board is available. I did it in this order (rather than installing it on the board, then soldering to the tag strip), so I could solder the tag strip once and be done with it. This also let me route the wire cleanly underneath the green and blue wires going to the neighboring lug on the tag strip, which took a little effort. I'm not proud of my soldering job on the tag strips in general. Even with the change in solder/flux, I was struggling to get the solder to flow without cooking the attached wires and melting insulation.

As you can tell, I didn't get the memo on twisting the transformer primary wiring and routing around corners of the chassis. I straight up forgot those instructions while I was wiring the mains. I'm optimistic that this won't have a negative impact, but if I get hum, I'll know where to look.

Here are the grid and screen resistors soldered to the octal tube socket. I copied the picture in the builder's manual when I mounted the 10k grid resistor, but left pin 6 of the socket unsoldered until the wire from the master volume is available. I kept the lead length between the resistor and pin 5 as short as possible. In retrospect, I like the way @Moodivarius wired his grid resistor better (viewtopic.php?f=16&t=6267). I'd probably copy that if I were doing it again. Note that I resoldered the tag strip later to make those joints a little better.

I routed the wires from the secondary winding of the output transformer around the mini switch holes, but I didn't install the switches. I imagine it's going to get crowded in there, so maybe I'll have an easier time wiring switches outside the chassis, and then installing them. Time will tell...

This picture shows the chassis wired up to the point where the builder's guide says to start on the eyelet board. I deviated from the instructions and installed the line to the octal tube screen resistors as a flying lead. This gave me an excuse to redo the solder connections I mentioned above.

At this point I insulated the ends of the all the loose wires, plugged the chassis in, and measured transformer voltages. I'm at 662VAC on the high-voltage secondary and 7.13VAC on the heater circuit. The heater voltage seems pretty high, but maybe it will come down when there's a load on it?

LH, I'm not saying to rewire the the heater wiring, but if you build another amp, try wiring the heaters from the endpoint socket back to the next. It took two AX84 and two Trinity builds for me to figure this out. Once I did I noticed Stephen recommends doing it his way in the Build Guide. It lets you keep the heater wires twisted tighter around the socket area. Your amp build still is looking great regardless.

_________________
"The only problem with being retired is you never get a day off"

Hi Hank. Thanks for pointing this out. I went back and took a look at the photo on page 32 of the Builder's Guide. I think I understand what you're saying.
Just to confirm: the difference is that I didn't keep the heater wires twisted between the tag strip and the octal tube... I have that big loop of red wire running from the power tag strip to the octal socket. If I had wired it as you suggest, the heater wires would have been twisted tightly the whole way.
Did I understand correctly?

Thanks again for commenting, and for the encouragement!

LH, what I used to do was wire from the transformer heater wires>tag board>octal socket>noval socket on the Tramp (and others). Figured out it was better to wire back from the last socket; noval>octal> tag board. That way you can trim, adjust and twist your red and black wire lengths. As I said, not suggesting you go back and redo what you've already done, but give it a try if you ever build another amp. Will post and attachment of second Tramp. You can enlarge the pic and see that I should have left a little more length on the black heater heater where it goes around the noval socket (a little close to some of the other pins). Anyway, it looks better and tighter than some of my previous heater wiring

_________________
"The only problem with being retired is you never get a day off"

That makes good sense. Thanks for the clarification. I will do that on my next (?) build.

Another update:

While I was getting ready to build the board, I stumbled on this video (https://hackaday.com/2015/05/20/how-to- ... backwards/). It's long, but you can watch it in fast-forward and skip the repetitive parts. It suggests that you can reduce noise and hum in your amp by figuring out which leg of each tubular cap is connected to the outermost layer of foil, and connecting that leg to the lower impedance part of the circuit so it acts like a shield around the capacitor. It seemed worth a shot!

In the end (see below) I decided the impact on this build will be minimal. It was still a fun exercise, though, so I'll write it up here.

The first thing I did was simply connect each capacitor directly to an oscilloscope probe & ground, then reverse the leads. I didn't take a photo, but I was trying to duplicate what's shown in post #19 of this TDPRI thread (https://www.tdpri.com/threads/lunchbox- ... ed.992175/). It didn't work well. The variation in signal strength just from moving things around was bigger than the effect of swapping the leads. I wasn't able to figure out which lead was connected to the outer foil.

So I decided to build a testing box like the one in the video, but with a mechanical switch instead of a PCB and solid state relays.

where all of the shields and the scope ground are tied to an aluminum case. I had this project box on hand. It's aluminum, and about the right size, so I used it.

I put the hole for the switch and the holes for wires on opposite ends of the box, so I can work the switch without getting too close to the capacitor and oscilloscope probes.

I didn't have any spare shielded cable to use. I thought about cutting up a guitar patch cord, but I didn't want to use cable that thick. In the end I found an old RCA-to-RCA video patch cord and used that. Here's the whole thing wired up.

The copper braid on each pair of shielded wires is twisted together and soldered to a piece of single-conductor stranded wire. Those two single conductor wires "Y" together into a single wire that runs to the project box. I connected it to the project box by soldering it to the locating tab on the switch mounting washer.

Here's what it looks like on the outside. To keep the wires from sliding in and out of the grommets I used zip-ties on the inside of the box and heat shrink on the outside.

Finally, I put ring lugs and alligator clips on the wire ends. I forgot to tie the scope ground line to the case when I was wiring up the inside, so I simply connected the shield and inner conductor at the ring lug.
If your scope has a low-pass filter or bandwidth limit, it helps a lot to turn that on. Here are pictures at full bandwidth (50MHz on this scope) and with a 20MHz bandwith limit turned on. You can see that the signal with the limit engaged is much cleaner. The toaster is there just as a good source of 60Hz interference.


Finally, here's the switch in the other position. With this cap (0.01uF), there's a noticeable difference. With larger caps the difference gets much smaller. I think this is because the larger caps have lower impedance at 60Hz, so the hum simply couples through to the grounded leg regardless of switch postion.

This was a lot of fun, but in the end I don't think it will have much impact on the tramp build. After doing this, I went back and had a good look at the schematic. Only one of the tubular capacitors (C3) is connected to a truly low-impedance part of the circuit, so this is the only tubular cap where I imagine getting the orientation "right" might make a real difference.

Wow, this post got long!

Cool device. This is discussed in some of the builders guides. Instead of the toaster though, just touch the body of the cap with thumb and forefinger. That should do the same things.

PS it wont make a difference in the Tramp though!

_________________
Stephen
Web: www.trinityamps.com. Facebook: facebook.com/trinityamps. Twitter: @trinityamps

Coco answered before I did. Go to Trinity Amps site (bottom of opening page), click on Support and look at the OSD Builders Guide, pg. 54. Apparently all you need is an O-scope . Still a cool device, like Coco said

_________________
"The only problem with being retired is you never get a day off"