Harmony 415 - New (first) recon project

[SwamperGene]

Hi all 

New guy to this great forum, and I need a little help. I'm not an electronics pro by any means but can handle most of what's thrown at me. Have built a few pedals (that all worked lol) and did a few amp mods here and there, not afraid to "dive in" to things.

Anyways, picked up an old Harmony 415 a couple weeks age, thought I did pretty good at $40. It's definitely a "project"...moderate tolex/grillcloth wear, faceplate kinda ugly, but plug it in and it does make noise. To me this is a great starting point for a complete electronic and cosmetic restoration.

A few notes as to things I noticed right away that were "wrong" with the amp:

- Has one 8-ohm Jensen C12R that is probably original wired in parallel with a '70 Fisher 4-ohm speaker, so there's an impedance mismatch that's been there for ???

- Was obviously used with an extension cab at some point, has a 1/4" jack soldered onto the speaker output.

- Internally, the "MIC" input was split and going to both channels, with a horrible soldering job to boot :


- Also, as you can see in the pic, a new jack was added to channel 2 but was not a closed jack, leaving the input ungrounded when nothing was in the jack.

- C9 was visibly cracked.

So, I attacked all that stuff first....

- Disconnected the Jensen so I'm working with a 4-ohm load.
- Removed the extension jack.
- Removed the homemade channel jumper and corrected the MIC input per the schematic.
- Installed a new closed-type jack and correctly ground it (immediately fixed a white noise problem).
- Replaced C9 - with a NOS Mallory Plascap if it matters.

Also added a 3-wire power cord, and removed C20 in the process. Also replaced a couple resistors that were in my opinion too far out of spec, most notably R22 and R23, all replaced with spec'd NOS carbon comps. Also cleaned and retensioned all tube sockets.

So, now it sounds pretty good, but I know it can be better. The bottom end is a bit loose, a little "farty" when the volume is dimed. Overall clean/crunch tones aren't bad, but volume doesn't come close to my little Bugera V5. It's fairly quiet overall, but there is a random low-volume scratchy or crackling sound that is not affected by either volume pot and is there with nothing plugged in. So, on to test, and here's where my questions start. First off, I need to be sure I'm doing it correctly....I have all volume and tone pots turned down, tremolo pots turned up, no input, and 4-ohm speaker is hooked up. Here's my test voltages under those conditions, amp was on for a good hour prior to the numbers, there was a little deviation at times all over.



To me it looks like low voltage is an issue, and maybe as John did in his great 415 thread, I need to start out with a new 5Y3? Also, for what it's worth, the tremolo circuit doesn't work...maybe something there? I'm all ears at this point, looking forward to any suggestions.   

[John]

Hello Gene! If you do replace the 5y3, make sure it's NOS, not the Sovtek or similar. Otherwise, your voltage will be much too high-at least mine was. Tremelo: I remember PRR telling me to be sure the wiper on R36 had not lifted. That would cause wide fluctuations in voltage, as mine was doing. Of course, mine was doing that because I had taken some of that stuff out, not knowing that *everything* matters in these things. Put it back, of course.

Also, not that this is related to anything, but check the solder joint going to your speakers. At that little junction thing on the chassis, mine was loose/cracked. Of course, I found that AFTER I had replaced all the caps and half the resistors.  But hey, I needed the practice!

Anyway, when it's right it should sound pretty good. Mine's got nice tight bass and lots of headroom. Have fun!

[jjasilli]

random low-volume scratchy or crackling sound: likely due to "worn" resistors in the B+ power supply.  My standard practice is to replace all B+ filter caps and resistors in the B= rail, including plate resistors. This should get rid of that noise.  Then re-check your voltages. 

[eleventeen]

"random low-volume scratchy or crackling sound: likely due to "worn" resistors in the B+ power supply.  My standard practice is to replace all B+ filter caps and resistors in the B= rail, including plate resistors. This should get rid of that noise.  Then re-check your voltages."

I'll second that motion. I just replaced the 100K plate resistors in a Deluxe Reverb (silverface but AB763) I've had sitting around for over a decade but never got around to. (Thanks to the suggestions on this forum!) I modded this Deluxe up to a Twin (yes, 4x 6L6) 30 years ago for a friend and I got the amp back. I downshifted to 2 qty 6L6 and now it's more like a Pro Reverb or a Super. I used the 3 watt ones from Doug. Just 6 qty 100K's, first and second preamp tubes and the reverb recovery tube...and I didn't do the phase splitter because I don't have any 82Ks. Could probably use 100K there with no issue. The tubes in it are whatever I grabbed from what its laying around. 6 stupid resistors.

WOW. All the "random low-volume scratchy or crackling sound" is GONE. That was EXACTLY what was troubling the amp. The replacement knocked that crap back by 85% or more.

(And by the way, I had already re-capped it (original physically leaking B+ caps) and it became massively more stable but the sizzling noise remained.) 

Don't overthink it. Don't measure 'em, just replace those plate load resistors. I daresay there's no cheaper anything you can do to an old amp that produces such a dramatic improvement.  

I know the (tweed) Deluxe fans are mentally modding that amp as soon as they see the schemo! 

[jjasilli]

 82K is necessary tobalance some PI's.  Unbalanced PI is not the end of the world.    Also 33K + 47K = 80K; or 100K in parallel with 47K = 82.5K

[SwamperGene]

Thanks guys for the ideas.

John I have a printout of your thread out in the garage, that is a great thread for this amp, so I have pretty much gone over all that stuff. Using my meter, the tremolo pot seems to be functioning correctly.

I just replaced all the plate resistors and so far, the scratchies seem to be gone. 
 There is a little bit of a steady hiss, not affected by volume changes, it's not too bad though.

Voltage out of the 5Y3 is still showing low, newest test is at about 276-280 (it drifts a bit). When I first cleaned and retensioned all the sockets, I initially had 305 but it dropped after about a half hour...maybe gotta repeat that step again.

[SwamperGene]

Hmmm....

The schematic is vague about setting the tremolo controls when testing voltages. Specifically, there's reference made to every pot except R36, the Intensity pot. If I roll this all the way CCW, B+ is 307. Could it simply be that R36 has to be rolled back for testing, which looks like it will put all my voltages within specs? Or have I inadvertenly stumbled onto a clue for the non-working tremolo issue in that turning up the Intensity is dropping my supply voltage? J5 is open throughout all of this.

 

[eleventeen]

?? 307 measured relative to 310 spec'ed is "dead on the money" in my book. I don't see why you find that alarming.

Phase shift oscillators (characterized by the three caps in series) of the kind in your trem oscillator can be a tad finicky when it comes to working, especially starting. With leaky caps, it might have trouble starting up, and those caps you got there don't exactly look so great.

[John]

If I roll this all the way CCW, B+ is 307. Could it simply be that R36 has to be rolled back for testing, which looks like it will put all my voltages within specs?

May be.

[PRR]

The stand-out WHAT?? on your voltage notes is the 6.3V at power tube grids.

This must be zero. Less than one volt.

Voltages are checked with tremolo, at minimum, off, or even (if possible) with trem tube yanked (easy in this amp). The amp MUST work trem-less. Things get a little wonky with trem, but if the sound wobbles and nothing melts, that's good enough.

Positive voltage on grids mean leaky coupling caps or no working path through grid-to-ground resistors. Since both grids are same voltage, and it is unlikely two caps will leak equally, I'd followthe resistor path from grid to ground.

If you get this 6V with trem pot up, then the trem coupling cap is leaking or the power tubes are very gassy.

[SwamperGene]

The stand-out WHAT?? on your voltage notes is the 6.3V at power tube grids.

This must be zero. Less than one volt.

Voltages are checked with tremolo, at minimum, off, or even (if possible) with trem tube yanked (easy in this amp). The amp MUST work trem-less. Things get a little wonky with trem, but if the sound wobbles and nothing melts, that's good enough.

Positive voltage on grids mean leaky coupling caps or no working path through grid-to-ground resistors. Since both grids are same voltage, and it is unlikely two caps will leak equally, I'd followthe resistor path from grid to ground.

If you get this 6V with trem pot up, then the trem coupling cap is leaking or the power tubes are very gassy.

Yeah I thought that 6V was a problem too. Gonna recheck later today with trem tube pulled but I do believe that it went down to like a half a mV with trem pots turned down, so might be the trem coupling cap...I think the whole thing should be recapped really. I've replaced more that half the resistors in the amp and saw improvements each time, I did notice a lot of ground issues centered around all the jacks which were not getting a good ground to the chassis and fixed all of that.

Played through it for a good hour last night, sounding much better though I still think a bit low on volume and "punch". I've been comparing it to my Bugera V5 which does have a master volume setup, I know percieved volume between the two won't seem too different, but I'd think the 415 should at least sound as loud as the V5, which it doesn't.

[SwamperGene]

Update:

Hi folks...work kept me pretty busy the last couple weeks so it's been a little slow going here 


Got a few things accomplished but some issues still remain. First was I recapped the tremolo section and that now works. Then I recapped everything else except for the multicap can and C17, C18, and C19 ( http://www.el34world.com/Forum/index.php?topic=9335.msg84724#msg84724 ). Voltages all looked good and steady, within +/- 10V of specs...even the grids were now showing only a few millivolts. But, after leaving it sit "on" for about an hour, I noticed all my voltages had dropped, B+ went from 320 to about 270.

Volume, to me anyway, still seems to be an issue. My little 5W Bugera still kills the 415, which I'd think should be close to a Blues Jr. volume-wise?

One odd thing I noticed...the input jacks are closed so the inputs are grounded when nothing is inserted. When playing through either channel, if I open the other (unused) jack on that channel, there is a noticable volume boost. A little confused by that 

[sluckey]

But, after leaving it sit "on" for about an hour, I noticed all my voltages had dropped, B+ went from 320 to about 270.

Could be the 5Y3 but replacing all 4 caps in the can will likely fix that. The cap can should be very high on the list of things to replace.

When playing through either channel, if I open the other (unused) jack on that channel, there is a noticable volume boost. A little confused by that 

That's normal operation for the way those jacks are wired. Look at the schematic. When plugged into only one jack, the signal goes thru one 68K to grid. But there is another 68K to ground thru the closed switch contacts of the unused jack. The 2 68Ks make up a 2:1 voltage divider. IOW, only half the guitar signal gets to the grid.

When you open the unused jack you remove the 2:1 voltage divider and ALL the guitar signal gets to the grid.

It's easy to change to the typical Fender hi/lo input jacks. Just look in "the tube amp library of information".

[SwamperGene]

Thanks sluckey, yes the can and the 5Y3 are both next on the list. Found another bad resistor too (random scratchiness is back).

Thanks for the info on the inputs, just did it on one channel gonna try it out in a bit. 

[SwamperGene]

Well...time for another update. As usual that thing called work keeps me from moving along as fast as I'd like.

Was still having some random scratchiness going on in last post. I've since finally found the remaining resistors I needed and that has cleared up. Basically this thing has had all resistors and caps replaced now, and has a nice NOS Sylvania 5Y3 in it which seems to have stabilized the voltages a little. Also, using a wooden dowel, I found a few wires that were acting up and replaced those as well, everything has been dressed as close as possible to how it was removed. Also did the input mod to both channels, great stuff!

Next it was on to cosmetics. This amp was no beauty...major tolex issues, stained/sagging grill cloth, and the worst...the bottom board of the cabinet had been wet at some point and all of the plys had separated completely. I wish I had taken more pics along the way but have a few to show the amp's condition.





So...on with the woodworking cap. Using a piece of 3/4 pine I had, I made a quick and dirty router jig to transfer the box joint pattern and replaced the bottom panel. This went great, even rounded the edges properly and all. (really wish I tookj pics here) When all was said and done, I had a nice rattle-free shell. Went to my local JoAnn fabric store and grabbed 3 yards of pleather and a cool-looking black glittery nylon that looked like it would make a nice grillecloth. Total materials were less than $15. This was my first attempt at doing this, but the results aren't too shabby :)





So now I have a decent looking vintage amp :)

Back to "the amp". Everything seems to be working great. Threw in 2 new EH 12AX7's I had, using good pulls for the EL84's. I get a little hum with volume but nothing out of the ordinary. It is now very loud, when I first lit it up in my detached garage my sound meter (wife) went ballistic and I was only at "3" on the volume 

I still find it to be a bit "farty" on the bottom end when it starts to crunch. Still running the 4-ohm Fisher speaker - probably from a stereo, has a whizzer on it. The only electronics that haven't been replaced are the cap can, pots, and transformers at this point. My voltages are all 5%-7% over specs which I'd say is a direct result of 118V house voltage which is about 7% over the schematic. I'm wondering if a little bias adjustment would help? I was thinking of replacing the 130-ohm bias resistor with 120 in series with a 25-ohm rheostat to give me a little on either side of stock. Any thoughts on this...might it help the loose low end?


You guys here and this site have been a lot of help and I thank you all. This has been a great learning experience, already getting ideas together for a ground-up build 
My

[John]

Gene, nice restoration!

With my very little experience, I think your farty bass will be fixed by replacing the filter caps. You can either get the can from Doug - or do what my amp guy did to mine, which was disconnect the cap, and fit in the individual caps. IIRC, he put a 32uf on the rect. tube, a 47uf on the el84 cathodes, and then 22uf for the others. I *think* that just the simple cathode resistor is the way to go, not getting fancy with the rheostat or whatever.
Don't worry, if I've got any of this wrong someone will be along shortly to slap me down! 

I keep meaning to dig back into mine to redo a bit of the wiring and get rid of some of the hum, which maybe just play time would take care of; it's been sitting a lot since I got my other one done. But again, nice work!

[SwamperGene]

Thanks John, and thanks again for the cap! 

Which brings up my next question. The stock can is 50/40/20/20 and the JJ is 40/20/20/20. I know two 20's are direct swaps, but can anyone tell me which arangement would be best as far as the 40/20 replacing 50/40? I'm thinking I could use the 3 correct values (40/20/20) then add a cap to get the last 20 where I need it for the rectifier...but been reading that the stock value of 50 may actually be a bit high? If I take it to around 30, will that affect anything downstream?

[John]

I *think* you could use the 40uf at the cathode resistor for the power tubes, and then use the 20's everywhere else. I know the schematic says 40uf at the 5y3, but it's pretty accepted that 40 is a little high for that, especially if you're using a "real" NOS. If it seems you have ripple with only the 20 at the 5y3, maybe then parallel another 20 with it?