Rebuilding my kit - shielded twisted pair options, and a film cap [resolved]

[twofires]

Hi Bottleheads,

So, my SEX amp has an intermittent popping issue (through headphones and speakers) that is maybe once every hour of use, and is concerning enough that I'm just going to strip the whole thing down and start again. There was a water-soluble flux issue during the build, brought about by my inexperience, that may be causing intermittent shorts, and so I just want to clean the whole thing out, try again, and hope the transformers and chokes are okay.

Anyway, I'm able to account for pretty much every part, except for the shielded twisted pair power wiring w/drain wire. The closest I've found that isn't AUD$400 for a massive reel is this stuff:

https://catalog.belden.com/techdata/EN/8450_techdata.pdf

It's about AUD$95 for a 15m cut shipped from the US, but I can live with that.

The tech data says it's rated to 300V, but as there's around 400V flying around inside the amp, I thought I'd check if this is an appropriate substitute.

Also it seems 1.5uF 630V+ PP film caps are hard to find outside of this general purpose Vishay:

https://au.mouser.com/ProductDetail/Vishay-Roederstein/MKP1839515084HQ?qs=5W%252BGeepUzoc9cdLA%252BylF3g%3D%3D

and a rather chunky and expensive Mundorf M-Cap Supreme:

https://www.soundlabsgroup.com.au/p/Mundorf-MCAP-Supreme-1u5-600V-Capacitor.html

Is there any leeway with capacitance with that cap? What would raising it to 2uF do to the circuit?

Thanks in advance,
Simon

 

[Paul Joppa]

...

Also it seems 1.5uF 630V+ PP film caps are hard to find outside of this general purpose Vishay:

https://au.mouser.com/ProductDetail/Vishay-Roederstein/MKP1839515084HQ?qs=5W%252BGeepUzoc9cdLA%252BylF3g%3D%3D

and a rather chunky and expensive Mundorf M-Cap Supreme:

https://www.soundlabsgroup.com.au/p/Mundorf-MCAP-Supreme-1u5-600V-Capacitor.html

Is there any leeway with capacitance with that cap? What would raising it to 2uF do to the circuit?

...

There is considerable leeway. The calculated value is 1.25uF, and anything between half that and twice that is acceptable. Larger values give slightly deeper bass with slightly  less power on the deepest bass - neither is likely to be audible in ordinary use.

I'll defer to PB on the wire. You can also ask at replacmentparts <at> bottlehead.com for the cable and capacitor. Mention this thread when you email.

 

[Paul Birkeland]

That wire will do the job, the STP in the SEX amp doesn't carry much voltage. 

Water soluble flux can ruin a build, but I have brought kits back to life by running them through a dishwasher cycle without soap. 

 

[twofires]

There is considerable leeway. The calculated value is 1.25uF, and anything between half that and twice that is acceptable. Larger values give slightly deeper bass with slightly  less power on the deepest bass - neither is likely to be audible in ordinary use.

I'll defer to PB on the wire. You can also ask at replacmentparts <at> bottlehead.com for the cable and capacitor. Mention this thread when you email.

Thank you for that, Paul. It seems that the options for caps increase a fair bit if 2uF is acceptable, so I might go with that. I'll keep the replacement parts thing in mind, too. It all comes down to shipping in the end.

That wire will do the job, the STP in the SEX amp doesn't carry much voltage. 

Water soluble flux can ruin a build, but I have brought kits back to life by running them through a dishwasher cycle without soap. 

Excellent. Thanks for that PB. Finding an equivalent wire down here in Oz is pretty much impossible - too small a market for many specialty products I guess.

Re the flux: yeah it was a lesson learned, for sure. The Keystone terminal strips were a bit oxidised and my wimpy no-clean solder didn't have strong enough flux to cut through, so I unwittingly added the water-soluble stuff. It worked in the short term, but then clagged everywhere, hid in places, and trapped forgotten lead cutoffs. The total immediate damage on first power on was one power resistor, a Dayton cap, and an RCA tube. It took out another tube before I got it clean enough to function, but it's not ideal and I'm looking forward to a clean slate. I have better RMA solder these days.

I've seen you mention the dishwasher thing before - how does that work exactly? Are we talking disassembling and desoldering and washing the hardware and passives? Or you just bung the whole thing in minus the wooden base and tubes? I always assumed that would destroy switches and pots and transformers and such.

I'm basically planning to bin everything but the transformers, chokes, tubes, base, top plate and mounting hardware, and replace the rest to rule out gremlins. This still leaves the possibility that I damaged the PT-10 when I (very gently) sanded the chickenwire marks & coating off the sides and repainted. I recall you folks telling someone this may damage the transformer - something about eddy currents? The transformer did get quite hot in operation, and there was a definite hum, but it also seems like these things are normal within certain limits. So maybe it's fine? I dunno if there's a test with a scope or meter or something to confirm.

Anyway, I can try working with the iron I have first - if I have to build it a third time I'm sure I'll have it down pat by then. 

 

[Paul Birkeland]

Or you just bung the whole thing in minus the wooden base and tubes?

Yes, that will do it, and definitely run a dry cycle. 

 

[twofires]

I opted for the full teardown in the end. That water soluble flux (ChipQuik) was crusted on everywhere, in spite of my best cleaning efforts while the amp was whole. It's a miracle the amp worked at all in that condition. When I was desoldering leads and removing excess solder from the chokes and transformers I could still smell the stuff cooking off - a distinctive black bean stir fry sort of smell.

See the attached for a an example of how the flux caked on to a tag strip, and the horrified look on my now depopulated top plate.

Other than having to bin a lot of components, the transformers and chokes came out okay. The only minor note is that terminal 18 on the PT-10 is a little bit wiggly - I'm wondering if there's a high temp non-corrosive glue I can use to get it rock solid again. Probably not a concern but seeing as I have the thing apart anyway...

 

[Paul Birkeland]

You can use a little epoxy to secure the plastic piece of the bobbin that holds that power transformer terminal.

 

[twofires]

You can use a little epoxy to secure the plastic piece of the bobbin that holds that power transformer terminal.

Thanks for that Paul. Something like JBWeld HighHeat? It says it can withstand up to about 290C. My iron is set to 370C at the tip, but I figure the point of contact should be significantly cooler than that unless I'm dwelling far too long. Unless you recommend epoxy only after soldering? It might be a bit trickier to glue up in situ but not impossible.

 

[Paul Birkeland]

I wouldn't even worry about buying high temp products, the heat will only be there for a handful of seconds. 

 

[twofires]

And it's epoxied. That stuff (JB Weld) actually goes on really well if you have something small enough to apply it. I used one of those tiny soldering tool spatula things and iced around the base of terminal like it was a tiny cupcake. Left to cure for 24hrs and now it's rock solid. I'm happy with that.

 

[2wo]

Ah, JB WELD, let me tell you about the time my shift lever snapped off in West Virginia, or the crack in the mower or the grill handle or...

 

[twofires]

Ah, JB WELD, let me tell you about the time my shift lever snapped off in West Virginia, or the crack in the mower or the grill handle or...

It does seem like supremely useful stuff. I especially like that it doesn't stink to hell like Araldite does.

 

[twofires]

WE ARE BACK! I am an extremely happy camper.  8)

The best thing is, something I did this time around has dramatically reduced my transformer hum . Possibly that spot of JB Weld on PT terminal 18.

I added #10 stud solder tags (Keystone 4704) to the speaker terminals (between the two supplied nuts) to make soldering and speaker terminal replacement much easier. Removing the old ones when I stripped the amp taught me that is not a fun exercise when solder has clagged up the threads.

Also note that 400V .1uf Orange Drops (715P10454LD3) fit the layout just fine if you can't get the Daytons, although you need to bend the A side cap up just slightly once soldered to clear the lip of the base.

Voltages were a bit high (+6% or so), but that's likely due to Australian AC being nominal 230VAC, but usually 240VAC during non-peak times. I wired for 225-235 figuring it wouldn't hurt anything.

Now to run it in for a couple of weeks before chucking the C4S back into the mix.

 

[twofires]

Sigh.

Might have to 'UN-resolve this' - I'm still getting the pop issue.

Every component is new other than the transformers and chokes, and the tubes, but after about 40 minutes of run time I occasionally get a loud pop in one channel or other. Just one loud pop, like someone flicked your headphone cup, then nothing. Like there's a slow and silent build up of energy that releases all at once.

Yesterday it was the right channel with Focal Clear headphones (55ohm 104dB), today it was the left channel with Fostex TH-X00 (25ohm 94dB). The random channel makes me think it's not the tubes, but I have a couple of spare pairs coming in the mail next week.

I'm very confident of all the joints and the wiring, and my mains are 236VAC at the moment so I figure I'm not over or under. My only other guess is my initial one, that either I damaged the power transformer when I removed some chicken wire marks on the side by lightly hand sanding before repainting, or, more likely, when I smoked one of the 3w resistors on my initial build due to the water soluble flux shorting things.

I'll try with new tubes, but failing that I might have to source a new transformer from you, unless you have other theories.

 

[twofires]

Only other thing I can think is that there's something about my DAC it doesn't like. I'll try another source, too.

 

[Paul Birkeland]

Oddly enough, that is a noise I would expect from conductive flux arcing over periodically.  I'd leave the amp on for a very extended period (10 days or so) to see if anything more concrete develops.

 

[twofires]

Interesting. Pretty much the only place left that flux could be is inside the transformer itself, impregnated into the material around the guts of it. That could've happened when I tried to clean the initial build with isopropyl back before the rebuild. That mixture (ChipQuik and isopropyl) may have dripped in there and dried. I'm not sure that's the case, though, as I can't see any of the chalky residue that was everywhere else. That said, everything else is either new, or very, very thoroughly scrubbed, so if it's flux, the power transformer is likely where it is.

When you say something "more concrete", what kind of behaviour are we talking? I doubt I can leave it on that long as I can't keep an eye on it consistently. I should probably be close by if something dramatic could happen.
I'll see what can be done, though.

Strangely, when I bypassed my DAC (a Topping D90) and plugged the amp directly into the headphone out on my computer (so using the soundcard), I listened to music for another 2.5hrs without any pops. Hardly definitive, but, is there anything about a DAC with an internal switching supply that might cause issues like that popping? Doubtful I know, but I'm keen to rule it out.

Also hoping to rule out the thing with the sanding the sides of the power transformer - could sanding the coating off the sides and then blasting it with paint have caused some issues?

One other thing - it seems to be more common with more sensitive headphones than with speakers or high impedance headphones. That may be some other factor at play, but if it provides any clues...

Thanks for all your help.

 

[2wo]

I seriously doubt that this will end up being the transformer.

 

[twofires]

I seriously doubt that this will end up being the transformer.

That's certainly my hope, but I have to admit I'm stumped.

I realised the other component I kept was the Alps Pot I used, although I don't remember using any ChipQuik on that, and I was under the impression those are sealed anyway. I also cleaned and reflowed that breakout board pretty thoroughly.

At this point I'm holding out hope for either bad tube(s) or weird input issue, unlikely as those are.

 

[twofires]

Quick question while I'm running this amp for extended periods:

Is it normal for the amp to have a bit of a fit when you unplug headphones to switch to the attached speakers?

I was listening with headphones, and needed to step away, but wanted to keep listening for issues, so I paused the music and pulled the headphones from the socket. This resulted in 'POPOPOP POP CRACKLE POP... POP... ... ... POP... ... ... ... POPOP' for at least 10 seconds from the speakers (mainly the left). When it seemed like it wasn't going to calm down easily, I turned the amp off (another single POP), waited a couple minutes and then turned in back on. No issues after that.

Is that an expected response to me doing something silly (like changing load from 55ohms to 8 ohms by pulling a jack out of a TRS socket), or is it unusual and indicative of some other issue?

This is the socket I used: https://www.digikey.com.au/en/products/detail/cliff-electronic-components-ltd/CL1323A/13663625

Note that powering it up again with headphones connected and then tapping on the headphone jack and twisting it in the socket didn't produce any pops, and all the joints look solid.