Harmony 415

[John]

Hi everyone. I'm working on my old Harmony 415 (Valco made) and replacing all those old brown caps. They all have .022mf printed on them. I ordered the Mallory .022's from Doug, and they are much smaller - maybe 1/8 the size of the original brown ones. My question is, are the new caps the correct ones, or did I order the wrong ones? It seems from what I've been reading that .022mf would be 22uf, and those look much bigger in Doug's store.

I did search all over and couldn't find anyone as dim as me that asked this question already. :) Thanks for any info!

Oh, the original .022's are roughly 1.5 inches long and maybe 3/8 inch diameter. They look like plastic or bakelite to me.

[Dave]

Different material used to make different kinds of caps requires more or less bulk to get the desired rating. So, size doesn't necesarily tell you much.

MF and UF are both used to refer to microfarads. They are the same.

If your old caps say .022mf and you bought .022mf/uf, then you stand a very good chance of having bought the right capacitor.
The only possible exception being that the old caps were mislabeled.

Dave

[FYL]

Pretty normal: modern components are much smaller.

Valco used 22 nF caps for coupling. Some were film caps (noted II on the schemo), others were elcos (noted III) - dunno why. Using film caps everywhere is a Good Thing™.



In the US, mF meant µF (uF for the gyros-challenged), so .022mf = 0.022 µF = 22 nF = 223 (value code on modern caps).

[John]

Thanks everyone for the fast replies! I was thinking the new caps should be right, but wanted to make sure before I started throwing globs of solder everywhere. lol

[tubeswell]

I have quite a few of those brown turd caps in different values and they are all on the small side, but they lack nothin' in delivery. I built a 5G9 using them and its the sweetest sounding amp

[John]

I'm recapping my Harmony 415, along with new resistors as I go along, and I'm replacing most of the wiring as well. I figure if I'm de-soldering anyway, might as well replace all that cheap stuff.
My first question is, would it make sense to go ahead and install new pots and jacks as well, or are they generally not a weak spot?

Second question is, would this be a good time to install new tube holders? AFAIK there's nothing wrong with the ones that are there. I also thought I might install a few grounding bolts to replace the places where the wires have been soldered to the chassis to make sure I have a solid ground. Make sense, or don't fix what ain't broke?

Also! (taking breath) Where it's easy I'm thinking about slipping shrink tubing over the leads on the caps and resistors? Is this a good or bad idea? And although this is strictly a point to point wire job from the factory, since the newer items tend to be much smaller I'm trying to organize the innards so the caps aren't strewn all over the place. Which come to think of it; are there any sort of holders that you could secure the caps in to clean up the innards a bit?

Any ideas, opinions, and "wtf are you doing" comments are appreciated!

[Structo]

You may want to be careful rearranging a point to point amp.
In a lot of instances they had a plan in the way things are arranged as far as lead dress goes.
You could neaten it up and end up with oscillation problems.

I would not replace any pots unless they are very noisy. The older pots many times have
a better taper than modern pots and as such cans sound better.
You can clean them with Caig DeOxit.

Tube sockets can be cleaned with DeOxit and retensioned if need be.

I think you have the right idea with, if it ain't broke don't fix it.

[RicharD]

>if it ain't broke don't fix it
Agreed.

[John]

You could neaten it up and end up with oscillation problems.

Yes, after posting that I was reading more old threads and found the one where the guy neatened up his amp with pretty wiring and ran into problems. I think (especially with my lack of knowledge) I'll just replace things as close to original layout as I can. Thanks!

[plexi50]

Same opinion here: If it's pretty much all there then only replace what is needed. Keep it vintage. You will not really gain anything by gutting and putting all new parts in it unless they are known to be bad.

[HotBluePlates]

You also don't need to unsolder grounds from the chassis, unless you're doing something like adding a 1 ohm resistor in between the cathode and ground of power tubes. The reason it is not advocated to solder to the chassis these days is that there may be a "better place" to locate the ground, and even a 100w solder gun will have difficulty making a chassis hot enough to melt the solder. So if you use a puny normal soldering iron, you're pretty much guaranteed a cold solder joint, and almost certainly one where the solder will pop right off the chassis cleanly if you tug at it!

As for heat shrink on component leads... there's a better way. Unless you actually need the shrink tubing to seal something, it's easier to use some pieces of insulation stripped off of normal hookup wire. Problems with that method: if you use PVC wire, your insulation added to the component legs may melt as you solder the component. If you use high-temp insulation, you'll be fine. In the old days, the stuff was called "spaghetti" and was a black fiberglass cloth in a tube shape and about the size of insulation stripped off 18-22 ga wire (naturally it really came in a large variety of sizes).

There may be value in not adding insulation to the component leads, as it gives more space to probe for voltage/resistance later on if you have to troubleshoot.

[PRR]

> the stuff was called "spaghetti" and was a black fiberglass cloth

FIBERGLASS?? You run with a rich crowd.

Spaghetti is varnished cambric. Lightweight plain weave cotton cloth.

"tell her to make me a cambric shirt... "

[rafe]

don't replace the resistors unless they are out of spec IMHO it's not necessarily an improvement

[HotBluePlates]

Spaghetti is varnished cambric. Lightweight plain weave cotton cloth.

Duh! That makes sense. I think I assumed it was stiffer stuff because the varnish gives it some stiffness.

I've also had much larger versions of this stuff that was some form of synthetic material (yellow, not nylon, but also obviously not-organic).

[John]

Where can I get some of that braided insulation? I haven't had much luck googling it.

Thanks,
John

[sluckey]

AES has it.

http://www.tubesandmore.com/scripts/foxweb.dll/catalog@d:/dfs/elevclients/cemirror/ELEVATOR.FXP?PAGE=SUBCAT&SEARCH_TREE01=WIRE&SEARCH_TREE02=17_SPAGHETTI

[HotBluePlates]

I was gonna say McMaster-Carr, but other places probably have it as well. McMaster is cheaper, but you're looking at buying 10-100ft minimum quantities depending on what you get. I didn't check all of their sleeving, but the smallest inside diameter looks to be 1/8" which may be too big for use on component leads.

But my point was that you can simply use any high-temp insulation off of hook-up wire. You can use stuff you normally throw in the garbage instead of spending more money. It works just as well.

[John]

But my point was that you can simply use any high-temp insulation off of hook-up wire. You can use stuff you normally throw in the garbage instead of spending more money. It works just as well.

Sorry, I misunderstood. You meant, what works better than heat shrink. :) I think that's what I'll do; there's only a few places where the leads are pretty close to each other (from original) and I figured while I'm doing the time I'll insulate them.

[John]

Okay, I have yet another question. I'm pretty much done except for a few caps/resistors I just ordered, and replacing the 2 prong cord with the 3 prong grounded cord.

The schematic shows the caps in a nice aluminum can from original. However, a few years back when I had it to the amp tech in town, he disconnected that cap can because it was leaking, and replaced it with new caps. However, the values are very different. C13 is a 30uf, and the other 3 are all 47uf.

I've read other threads where it's said that the 5Y3 rectifier doesn't like high cap values. I was already considering replacing the old can with the nice JJ one that Doug sells; it's a 40/20/20/20 IIRC. Does this sound like a worthwhile thing to do? As far as I'm concerned, any parts I buy from here are free; I'm just paying for the information. :)

One MORE thing; as far as I've been able to tell reading about it, when I replace the 2 with the 3 prong plug, I basically don't drop any caps out or anything like that, correct? Simply make sure I have the polarity correct when I wire to the switch and fuse?

Thanks,
John

[sluckey]

I'd leave the power supply caps alone, unless the amp is humming.

Referring to the schematic above, remove C20 when you change to a 3 prong cord. Connect the black wire to the fuse.

[rzenc]

One MORE thing; as far as I've been able to tell reading about it, when I replace the 2 with the 3 prong plug, I basically don't drop any caps out or anything like that, correct? Simply make sure I have the polarity correct when I wire to the switch and fuse?

I believe you should remove C20 - 0.3 - cap from switch...

Hope this helps
Best Regards

Rzenc

[John]

Thanks for the replies. I was thinking as I soldered in the new .03 I might have to take it back out.  ;) It's okay, I need the practise!

[John]

Okay, yet another question. C17 is on the tremolo circuit, and I don't have a .25 cap. I do have .047, .01, .1, and .033. I have never used the tremolo in the amp and don't plan on starting (in fact, I had  considered taking out the trem circuit altogether, but know I don't know enough to do it properly yet). Anyway, can I just pop in a .1 and be okay? Or should I hook 2 of the .1's together? And if so, what's the proper sequence, end to end or parallel? I did try to search but couldn't come up with the right phrase I guess. With all the reading I've done here I should have stumbled across it. :)

[John]

Oh for petey's sake. Just saw the "capacitor math" thread. On this page. Where's the duhhhh emoticon???

[John]

Okay, got everything installed that I switched out. After the slow startup with limiter, etc. plugged in and played. Sounded awesome, really sweet, lots of sustain. However, after 2 minutes it started to die out again, although not nearly as bad as the original problem that prompted me to replace the caps and out of spec resistors, frayed wire, etc.   So, hooked up the meter and got the following voltages that are in red on the picture. Things that puzzle me in my noobidity are the slightly higher heater voltages, lower voltages nearly everywhere else, and the wide disparity in output tran. voltages on each side.

I do know I have to replace R24 since it is visibly leaking and getting hot. The leakage was already occurring prior to my starting the project, (edited to add: the clear stuff on the outside didn't dawn on me what it was till I fired it up and it turned to liquid... another duh moment) so I guess that was one of the many problems. I am also going to replace the tangled up mess of filter caps that are much higher in value than the schematic calls for with the JJ can; it'll get the values back to where they should be and do away with the non-secured cap issue that I don't like.



Anything obvious Ive missed, and anything I should triple check myself one would be appreciated. Thanks!
Oh, and I know it's just a cheap old amp, but when it's running right it sound fantastic with my Epi.

[PRR]

How can you have 284V supply and 350V at a tube? Tubes only waste power, don't create power.

It is unlikely to have the two ends of the OT at very different DC voltages.

At the two power tubes known-good?

If you swap them in their sockets, do the wrong voltages stay with the tubes or stay with the sockets?

Replace the cathode bias resistor under the power tubes. While you have money out, get 200 ohm and 250 ohm 10 Watt parts. Use the 250 for now. We'll want to know the cathode voltage, cathode resistor, and plate voltage to verify safe plate heat.

Ohm-meter the R22 R23 grid resistors, be sure they conduct to ground. Failed resistor will lat that tube float-up to uncertain bias. There is also the R36 trem pot wiper: if that loses contact both power tubes get uncontrolled.

With power tubes out and J5 Trem jack shorted, what voltage at power tube grids pins 2? Should be less than 1V. If higher, replace C8 C9 coupling caps.

Don't start a new thread for a continuing discussion. I can't remember lunch, much less what all the names and amps and threads are about. And clicking between two treads is confusing. If I can figure out this moderator wand I may try to link this new thread into your previous thread.

[John]

Thanks PRR. Sorry about starting the new thread. :)

The 2 power tubes are new, although I guess that doesn't guarantee they're good. I just got them from tubestore.com before I found this place.

I'll swap them in the sockets and see what happens. Thanks for all the suggestions, I'll do all that and post the update (in this thread! :) )

Thanks again, this place is great!

[John]

Okay, latest update.

Got my package from Doug and replaced R24 cathode resistor with 10W 150 ohm. The 250 got broken in shipping so I couldn't test that one, but here's the voltages. ( I don't suppose I can just super glue the outside back on? )

http://www.el34world.com/forms/valvedata/mdstihl415.htm

The cathode resistor is reading 14.9 at the +.

The rectifier tube is reading 418 vdc, which is about 100 higher than it should. Also, the heaters are still running about 6.8 - 6.85.

At least the voltages are now holding steady everywhere and not wandering all over the place; I suppose re-installing the trem pots might have had a little something to do with that *hanging head* I had taken them out to give me more room to get all those caps and resistors in place where the caps are now. I figured I might as well fess up that I didn't realize they were part of the bias circuit until PRR said to "check that R36 wiper"  . So that was why the voltages were all over.

I haven't installed the cap can yet as that's going to be pretty involved. But the chassis was not laid out to have those caps laying all over. I did manage to neaten them up slightly. The voltages going to them are also high of course. I think I posted before that the original can was a 40x20x20x20, and someone replaced that with individual caps, mostly all 47uf except for one 30uf.

The only thing I can think of is that the PT is putting out too much voltage. I wish I had thought ahead before I started the project to check the voltages, but it does make sense as to why it started going through power and/or preamp tubes.

Thanks,
John

[PRR]

> PT is putting out too much voltage

Heaters are high by the ratio of 110-118, the difference in wall-outlet voltages over the decades. That alone, I would ignore.

But 320-413 is 30%, way more than wall-volt drift.

Please, someone jump in and explain the different things sold as "5Y3". I recall that some "5Y3"s were really some "better" ex-Russian tube, leading to higher B+ voltages than expected.

398V is over rating for 6BQ7, though I would run new-production EL84 in that zone.

15V over 150 ohms is 100 mA cathode current. Taking ~~10% off for G2 current, 90mA in plates. 90mA times 398V is 36 Watts in two plates, or 18 Watts per plate. The 6BQ5 is rated 12W Design-Center, we can assume maybe 14W max safe. EL84 is similar. The amp is running 28% hot in the plates, which will give short life.

Taking 150 common cathode bias resistor (a good value) the plate voltage should not be over 360V.

A "correct" 5Y3 may be the answer.

How is the hum with Volume full down?

[John]

Hum with volume full down is almost non-existent.

I'll look around for info on the 5y3 tubes.

Thank you!
John

[Geezer]

EDIT: I merged the 2x separate threads on this amp into one single unit. 

Geezer

[John]

I'll be ordering a 5Y3 from Doug (along with another 250 ohm 10W to test) because when I pulled the tube to take a closer look it was cracked near the pin end. Even I was that observant   . Before I pulled it, I measured the AC supply to the tube and it was 358VAC, still 418DC output. This *should* fix the high voltage problem (?). Thanks again for all your ideas, inputs and comments, and I'll keep you updated cause I just know you're dying to hear. :D

thanks,
John

*edit* Before I order, there's not a tube I should be trying instead? The schematic does call for the 5Y3, and I see Doug sells the Sovtek.... seems he sells what 99% of the people need.

[John]

After much googling, it seems the Sovtek will let through higher voltages than the old Sylvania brown base style. I think if I get the Sovtek and up the  value of the 5W cathode resistor, say at least a 250 instead of the 150 that should help with dropping the voltage to the plates. Or am I better off spending twice as much for the Sylvania? It seems odd that one tube over another would give 100 vdc more!

[HotBluePlates]

I don't know the actual russian type number, but almost any new production "5Y3" is really a russian type that has less voltage drop in the rectifier. If you're having voltage issues, getting a new production rectifier is the wrong way to go.

The Sovtek 5Y3 from Doug is $18. You can buy old-stock US 5Y3's from Antique Electronic Supply for $9-18.

Normally, the Sovtek is not a problem, and I always support buying from Doug. In this case, you would benefit from buying old production, and the price is the same or less.

[John]

Thanks. I just got done more research (believe or not I do read way more than I ask lol) and the other solution seems to be either install a higher ohm cathode resistor, or run a resistor with bypass cap for each tube. I think the NOS tube is a lot easier!

Don't worry, when I get around to installing that cap can, I'll be spending more money here. :)

Thanks again to everyone for putting up with all my newbie, don'tknownuttin questions!

John

[FYL]

I don't know the actual russian type number

Sovtek currently ships a modified version of the Russian 5ц4м (5C4M), halfway between a GZ30 and a 6106.

 

[HotBluePlates]

That's probably the number! I had found data sheets for the 5C2M and 5C3M, but they were obviously higher current and bigger rectifiers.

How does the GZ30 compare? I'm not familiar with that type number, although I think I recall the 6106 as an industrial 5Y3GT.

[John]

http://www.tubesandmore.com/cemirror/inv/T-5Y3GT_GE.JPG

This is the one I ordered; it looked the oldest.  Just a side note, the 5Y3GT I have now was not cracked before I started the project; I had pulled all the tubes and inspected them before deciding to re-cap everything. I could have cracked it when pulling it out, or maybe it was just it's time; I have no clue why, how or when it got damaged.

[FYL]

How does the GZ30 compare? I'm not familiar with that type number, although I think I recall the 6106 as an industrial 5Y3GT.

The GZ30 was a compact version of the 5Z4G. It was superceded by the GZ34, which offered better specs at nearly the same price.

http://www.r-type.org/pdfs/gz30.pdf

The 6106 was developed by Bendix as a hardened 5Y3GT for mil or aerospace applications. Lower plate R, much longer lifetime.

http://tubedata.itchurch.org/sheets/131/6/6106.pdf

[John]

Update: Installed the NOS 5y3. Plate voltage is now 366, voltage at cathode resistor is now down to 13. IF my math is correct by cipherin' out PRR's equation above, I'm now running 14.42 watts per EL84, or 28.84 total (that's taking off the ~10% for G2 current (which I have no idea what that actually is)). So while this will be at the edge of "okay", would I be better served upping the cathode resistor to 200 ohms?

I'm going to go run my guitar through it whilst I wait on replies. ;)

[PRR]

Fine.

[John]

The only issue I hear is some break-up when I go past 6 on the volume, sort of a dt-dt-dt sound, which could be speaker maybe? It's been on my workbench in the loafing shed for the past month, so it might just need some dry air.  I think I'll swap in the 200 ohm just to see if there's a tonal difference or not, and double check all the ground connections.

Thanks again to everyone for their help!

[John]

Final post on this amp. I promise. :)

The "dt dt dt" sound was due to the JJ power tubes. Swapped them out for the old tubes with Harmony still stenciled on them, no noise. Put in the Sovteks from Doug, and got a richer, cleaner sound. So, easy fix for that problem. I read somewhere that JJ's can have a structural flaw if they get too hot, and something inside rattles. (?) Dunno, but taking them out fixed it.

Put in a 250 ohm cathode resistor, and took voltages again. The plate voltages are still high, but with the 250 I'm getting around 12.5 watts dissipation (if my math is correct.) So I'm not going to worry about it unless I start eating tubes.

Did the chopstick thing and found one old joint that was just a little crackly, so reflowed that.

So, put everything back together, brought it in the house and played for about an hour, at low and high volume, pushing it with footpedals and without. It sounds great, sweeter than it ever did. Hum and or hiss is almost non existent. Now I get a nice smooth breakup around 7, dimed it's got a nice crunch but still sustains nicely. My wife (the drummer) was quite impressed with the new tones, and admitted that all my late nights haven't been wasted.

Also, I can see now why my amp tech didn't jump right in and replace the caps (although I never could get him to put on the 3 prong cord, either). De-soldering was a huge pain. Every lead from factory was actually wrapped around the lugs, usually more than one wrap. I was faster at the end than the beginning, so I'm sure experience would have helped too.

Finally, thanks again for the comments, thanks to PRR for his equations that allowed me to figure out stuff for myself. I've learned far more than I thought I would with this project. I think I have 20 hours of reading time in alone.  Now I'm noodling around figuring out how to afford justify building a little 5 watter. :D