xerxes
Been here a while!
Posts: 1,115
|
Post by xerxes on Nov 18, 2006 9:09:53 GMT
I've been a bit busy with the day job, so I haven't had much time and I have to admit that I feel a little deflated about the amp not working. To date, everything has worked straight off, just lucky I guess, and the amp not working has kind of rained on my parade and taken the wind out of my sails a little. I did check the amp with a continuity tester and I don't think I have any duff solder joints. Next I will try Rick's suggestion and see if I get sound from the amp before the buffers, once I figure out where to make the connections. ;D
|
|
|
Post by PinkFloyd on Nov 18, 2006 14:14:17 GMT
I know the feeling chin up though mate, it'll be something simple and maybe you're not seeing the wood for the trees... stop looking for a while, then go back to it. So you've got all the connections the right way round? regulators in the right slots (both have different numbers remember) double check everything owen.
|
|
rickcr42
Fully Modded
Rest in peace my good friend.
Posts: 4,514
|
Post by rickcr42 on Nov 18, 2006 15:12:45 GMT
don't let it get you down man. These days of mixed PM/No-PB parts,screwy looking solder joints (dull lead free rather than the previous shiny ),higher temps to flow the unleaded solder yet too much heat for some parts is a major hit on the DIYer or small commercial operator.Even Mike ended up beating the shit out of the buffers in frustration and he has extensive experience with pcb mounted parts. Damn right you have been lucky ! That was not my experience early on at all when starting out.In fact it was SO BAD in the beginning I still flinch thirty years later every time I turn on the power after building something because I blew so much shit up my nerves were totally shot Stupid things like capacitor polarity reversed,missed solder connections or a missed solder bridge shorting out the pcb,overheating transistors/diodes,backwards voltage regulators ,reversed in the socket DIP ICs.......... all mostly from rushing to complete the project which made for sloppy build technique and piss poor habits.I wanted the "thingy" done NOW DAMMIT no matter if the dash to the finish line meant in the rush missed steps or working far too long was too tired to see simple mistakes. Considering how hard it was to procure parts in small numbers then and just how damn EXPENSIVE small quanities were at the time ('70s) AND how long thew wait for replacement parts which averaged a weeks long turnaround once I actually had the funds........ In todays dollars a single transistor would cost like $20 at retail price while the minimum wage was below $2/hour which being 15 and working part time meant electronics was a very expensive hgobby so you would think my poor ass would pay better attention and get it right the first time ,before the fkn explosion The flipside is I purchased/built a damn lot of test gear over the years and my troubleshooting skills rose to a high level quickly-had to I also am a stubborn prick so I never got discouraged even though more than a few times I had to also replace the chassis after I gunned it into the wall in a fit of rage over the POS blowing up in my face I SWEAR BY clip in test probes.we humans were just not born with enough limbs so we need a certain amount of "helper tools" to make up for the design flaw in humans. For DIP pack I.C.s it is the Aries IC Test Clips that save the day www.ap-products.com/tc01.htmcombined with those plunger operated "mini-grabber" probe tips that allow one to clip in damn near anywhere without the risk if shorting out the electronics or running the risk of slipping off while probing the problem I literally have hundreds in every color plus adapters to fit all my gear,even my old Eico VTVM that has a coax I also have a simple little GO/NO GO LED continuity tester that not only do I find invaluable,that not only frees up the DMM for other tasks but was actually the very first DIY electronics project I EVER got to work the first time out of the gate so will forever have a place in my heart. It used what was then a new technology so the first commercially available LEDs (Monsanto ? HP ? )were VERY expensive which meant Rick took no chances blowing it up. from the LED Museum : "Appearing around mid 1969, this Monsanto MV2 red LED was purchased in London at a military surplus store on Tottenham Court Road. It originally cost £1, at a time the typical wage earner made £7 a week. So it was not cheap." HELL NO THEY WEREN'T CHEAP ! Only available in red at the time,glow was/is hardly bright,the viewing angle limited,delicate by comparison to modern devices but will forever be on my test bench as a reminder to slow the fk down and do it right the first time the LED,two binding posts,a slide switch for voltage/continuity testing,a diode for polarity protection and a nine volt battery all in a 1x1x2 hammertone finich bud utility box All I can say is double then triple check all connections/polarity/parts orientation first (then check again ) and if that checks out work backwards form the output jack to the input jack to find where the fault lies. My fav piece of gear here (also built way back in the dark ages of my personal DIY jurney ) is goofy little a 1khz tone generator,also with a binding post output connection,for signal injection and a just as simple/goofy/dirt cheap LM380 "amp" with a 2" mini speaker signal tracer set. Clip the mini amp/speaker to the output then starting right at the output jack of the audio project inject the tone then work backwards to the input until there is silence then check that stage for the "why" Good luck Owen.Can be a bitch but will in the long run sharpen you troubleshooting skills AND give a better understanding of how things actually work................................or in your case don't (couldn't resist )
|
|
xerxes
Been here a while!
Posts: 1,115
|
Post by xerxes on Nov 18, 2006 15:45:25 GMT
I actually used silver loaded solder which I'm fairly certain contains lead as well: www.wbtusa.com/wbtsilversolder.htm, in any case it melts easily and is nice to work with, it seems to flow and stick readily and leaves a shiny solder joint. With regard to testing, I have a tatty old pair of headphones; I can cut the jack plug off and solder the bare wires to points on the PCB before the buffers.
|
|
rickcr42
Fully Modded
Rest in peace my good friend.
Posts: 4,514
|
Post by rickcr42 on Nov 18, 2006 17:30:07 GMT
You are a very very brave man doing a troubleshoot with headphones If at all possible it is better to clip in rather than solder in since you really want to avoid too much pcb heat,a real trace killer that if the heating vaporises the circuit path will piss you off big time and why the suggestion to just lift one end of the interstage resistor. Here is where a mini-clip test probe is at its best.You can attach to a cap,resistor,single pin of the opamp or buffer and not run a risk of it slipping off or shorting to another contact point,something SMD is fking up big time because it means way more expense for the test gear and why I am very anti SMD (going blind has a bit to do with it also ;D ) Question,what are you driving the amp with ? If that device has enough "oomph" maybe your first test needs to be to clip in directly to the RCA input jack section to at least be sure you are getting a good signal input thewn bounce over to checking out the indicidual gain and driver sections. And oh yeah,both channels going wonky on you at the same time usually points to a common source of the problem which due to each channels audio circuits being pretty much dual mono even down to the chips means the only common denominator between the two is your power supply. You might want to do a bit more detective work from the line cord inlet to the regulator output to see if maybe something there is amiss.If not look to the headphone jack for the likely answer
|
|
|
Post by PinkFloyd on Nov 18, 2006 19:53:45 GMT
My money's on that headphone socket Rick. Don't ask me why, it just is Have you measured the DC offset at the headphone socket yet Owen??? If there are VOLTS there it'll give us a clue and if there's NOTHING there it'll also give us a clue. Measure from sleeve to tip and sleeve to ring www.rock-grotto.co.uk/dcoffset.htm I remember measuring something like 20 VOLTS on mine which kinda told me something wasn't quite right
|
|
|
Post by PinkFloyd on Nov 18, 2006 20:36:13 GMT
Remember little things like the chassis mount phono sockets too.... they should not be making contact with the metal enclosure, have you fitted the insulating washers?
|
|
rickcr42
Fully Modded
Rest in peace my good friend.
Posts: 4,514
|
Post by rickcr42 on Nov 18, 2006 23:06:05 GMT
good point.Obviously it has to be something that is common to both channels so if the power supply is ruled out then what is left ?
Jacks and volume pot...................
|
|
|
Post by PinkFloyd on Nov 18, 2006 23:49:03 GMT
good point.Obviously it has to be something that is common to both channels so if the power supply is ruled out then what is left ? Jacks and volume pot................... Yup.
|
|
xerxes
Been here a while!
Posts: 1,115
|
Post by xerxes on Nov 19, 2006 1:04:51 GMT
Yep, it's not that: I'm fairly certain I havent done anything "stoopid".
|
|
rickcr42
Fully Modded
Rest in peace my good friend.
Posts: 4,514
|
Post by rickcr42 on Nov 19, 2006 2:28:05 GMT
Ignorant means you just don't "get it" Stoopid means you never will I'm bettin' on "A". It will likely be something totally stupid at fault here but those are the damndest things to track down.The really big mistakes,the ones that cause fire/smoke/sparks/bolts of lightening,they are easy to figure out.Just follow the vapor trail to the bad part
|
|
|
Post by PinkFloyd on Nov 19, 2006 12:23:27 GMT
Or burn your face when you're sniffing about on the board and BANG! goes that capacitor you fitted the wrong way round.... jesus I did that "once" and never again..... I could smell "heat" so went down on the PCB with my nose to locate the source and CRACK! went a capacitor right onto my forehead.... nearly shit myself!
|
|
rickcr42
Fully Modded
Rest in peace my good friend.
Posts: 4,514
|
Post by rickcr42 on Nov 19, 2006 15:18:42 GMT
did the same thing with the second kit I ever purchased,a 25WPC SWTPC power amp. My very first kit build was this guy here : www.swtpc.com/mholley/Catalog1971a/Page06.jpgbeing the 70's I needed another stereo amp for my EV "universal matrix (also DIY ) Quad set up so tried to buy another Plastic tiger.Got the message back in the mail that it was discontinued then superceded by a better amplifier,the 215-A (a James Biongiorno of Ampzilla fame design www.swtpc.com/mholley/Catalog1970/Catalog1970.htm ? ) so I ordered one with the idea of moving the Plastic Tiger to the rear channels..... www.swtpc.com/mholley/Catalog1975b/Page5.jpgIn my case the caps were big mamma pucker 47000uf/50 volt "Big blue Sprague" chassis clamp bolt down computer grades so stickin my nose in there could have been fatal Wired them in bass ackwards,powered up the amp,the front panel power meters pegged immediately to FULL POWER BUCKWHEAT and before I could pull the plug CRACK.......before I could recover from the shock of this mini nuclear explosion event in my bedroom the the pricks actually started to whistle like bombs falling from the sky then started to actually ooze foam out of the vent in the top like a GD Volcano Being 1977 you did not just log onto the internet and buy new parts.Parts that even though called "computer grade" were impossible to find because there were no fkn computers smaller than an average parking garage using huge tape reels for data storage and keypunch cards for data entry so not exactly common outside of the military or corporate headquarters of the larger companies or small but about as useful as tits on a bull (got the "T" word in again ) "toy kit" computers in name only. Spent fkn weeks scouring my "Yes you are a cheap bastard so here are your electronics surplus catalogs" until I could find a replacement that fit the fkn cap clamps. We won't talk about how this also introduced me to the term "thermal runaway" or how expensive it was in the late seventies to replace entire channels of discrete transistors in a balanced topology from input jack to output post Damn kit cost me like $70 to purchase and another $70 plusto correct the screw up Part of the learning curve man.You learn a hell of a lot more by falling down and getting banged up than you do never when things are always clear sailing .
|
|
xerxes
Been here a while!
Posts: 1,115
|
Post by xerxes on Nov 23, 2006 1:55:07 GMT
I measured the DC offset according to your instructions. I'm getting about -12.5v between both the left and right channels and ground. So something is definately wrong. Now what? ;D
|
|
rickcr42
Fully Modded
Rest in peace my good friend.
Posts: 4,514
|
Post by rickcr42 on Nov 23, 2006 18:18:13 GMT
damn right there is I'll see if I can get Mike to weigh in since he has experience with the build. did you try the "gain stage bypass/buffer stage bypass" trick yet ? May at least narrow it down to what stage is going FUBAR on you (Fkd Up Beyond All recognition )
|
|
xerxes
Been here a while!
Posts: 1,115
|
Post by xerxes on Nov 23, 2006 20:16:42 GMT
No I didn't, but I have a hunch that I may have cooked the buffers. I soldered them one pin straight after the other and they got pretty hot. What I should have done is solder one pin, wait for them to cool, solder the next and so on. If this turn out to be the case, I'll put it down to experience and consider it a lesson learned. The voltage regulators were easier to solder and should be OK and the Op-amps are socketed, so they wouldn't have been affected by my ham-fisted soldering. Here is a diagram indicating what and where I've measured:
|
|
|
Post by PinkFloyd on Nov 23, 2006 21:01:47 GMT
Impossible to diagnose online but, let's just say, I was also measuring around 12 volt offset with the buffers in place (the ones I ended up smashing on the head) so it's looking very likely that AGAIN these f**kin' things are the root of the problem. You can do two things.... 1: buy the buf634 in the leaded dip8 package and run hook up wire from it to the pads on the PCB. 2: try again with the same buffer package but, this time, solder the legs one at a time and let the bugger cool down. I think I have two more left, I will go and look.
Seeing as you're getting 12V offset and so was I it's very likely the buffers aren't doing what they should do, take a measurement at the buffers themselves and this should confirm they are knackered.
|
|
xerxes
Been here a while!
Posts: 1,115
|
Post by xerxes on Nov 23, 2006 21:33:58 GMT
Hi Mike, How do I check the operation of the buffers, ie. what pins should I measure between and what voltage should I get? These are the pins: BW | Vin | V- | V0 | V+ Even if you have some more buffers Mike I couldn't possibly ask you to send them to me, as I said, we'll chalk this one up to experience and try to remember we're soldering and not welding next time. I know I can get hold of the buffers at Farnell, so that's not a problem.
|
|
rickcr42
Fully Modded
Rest in peace my good friend.
Posts: 4,514
|
Post by rickcr42 on Nov 23, 2006 21:56:19 GMT
even better is a tiny clip in heat sink/soldering aid.this one is a tad big but illustrates the concept just the same www.radioshack.com/product/index.jsp?productId=2103242&cp=2032058.2032230.2032272&parentPage=familyif it IS the buffersd a simple test would be to add a jumper wire from the buffer input resistor directly to that channels headphone output jack pin then give a listen.If the opamps are fine you will hear music but if not then the buffers may be fine,the opam[ps may be bad or worst case all ICs are toast
|
|
rickcr42
Fully Modded
Rest in peace my good friend.
Posts: 4,514
|
Post by rickcr42 on Nov 23, 2006 21:59:29 GMT
Impossible to diagnose online but, let's just say, I was also measuring around 12 volt offset with the buffers in place (the ones I ended up smashing on the head) so it's looking very likely that AGAIN these f**kin' things are the root of the problem. You can do two things.... 1: buy the buf634 in the leaded dip8 package and run hook up wire from it to the pads on the PCB. 2: try again with the same buffer package but, this time, solder the legs one at a time and let the bugger cool down. I think I have two more left, I will go and look. Seeing as you're getting 12V offset and so was I it's very likely the buffers aren't doing what they should do, take a measurement at the buffers themselves and this should confirm they are knackered. never have been much of a BUF634 fanboy (maybe the TO pack in a pinch ) and this new generation is even more reson for me to stay well clear. This RoHs crap is an mechanism/attempt to take the little guy out of the competition the same going to SMD was because they know full well the only way to get it right every time is by using a wave soldering machine which now also means special prsoccesses like increased heat. Machine part set,machine soldering,humans at the end working for peanuts loading the product into boxes.......all profit once the machinery is paid for luckily the DIYers of the world figure out workarounds which in this case is soldering the parts using an oven so they failed in the attempts at shutting us out entirely but they have succeeded in shutting down cottage industry small operators if they catch them.The fines alone would wreck the little guy while being no ore than a writeoff for big business and that folks is just not right ! People suck and every year they suck even more with all this "ME FIRST" shit and money grabbing while stepping on everyone around them......... sorry.Lost myself in a mini-rant there for a moment Anyway,I have no use for any lead free parts for audio use unles the part is one that never needed the lead anyway so didn't have any.All other parts that have both a PB and a NO-PB counterpart when doing a head-to-head PROVE the new parts are a step back if having good sonics is considered.Fkn sonic butchers is what
|
|
xerxes
Been here a while!
Posts: 1,115
|
Post by xerxes on Nov 29, 2006 15:55:34 GMT
I soldered in a new set of buffers today. Exactly the same result: So the problem is not the buffers themselves and I really don't know what else to try.
|
|
rickcr42
Fully Modded
Rest in peace my good friend.
Posts: 4,514
|
Post by rickcr42 on Nov 29, 2006 17:52:17 GMT
you need to jump a wire from the output of the opamps directly to the headphone jack or you will never know what the problem is.troubleshooting is a process of elimination that takes stages beginning with "do I have power and if yes is it the correct voltage ?" then works from the output jack backwards with each stage checked for a GO/NO-GO condition. Looked at as a whole any amp would be a bitch to diagnose unless it was a single stage/single device circuit so you really need to break it down into 1-power source 2-input section 3-output section if they all check out then the obvious problem is problems on the pcb itself (cold joint,solder bridge).even a single cold joint in a critical path can screw up the entire amp as will the teeny tiniest of solder bridges : Pull the amp board,hold it up to the brightest light you have abd "looking through" from the soldered side look for even a hint of shadow between any two adjacent traces. If this checks out then you need to again study the actual solder connections (again under the brightest light you have,with a magnifying glass if neccessary) to see if maybe you missed somethin-happens to us all Finally it just may be that the new PB-Free BUF634s suck ass and you need to go through a group culling out the bad (something I personally would not do having already gunned the whole damn thing out the window by now ). Try this : Tack solder new buffers on the old "piggy back" style and if you get music then the bottom buffers are crap and need to be removed.Don't continue to run the piggy buffers without load sharing resistors so either remove the bottom buffers or add the resistor Good luck Owne,wish i could be more help but i never built this amp myself,i do not use the LM6172 opamps and i have zero use for the new BUF634s so also not used by me or I maybe would have had a definitive answer rickmeister
|
|
|
Post by PinkFloyd on Nov 29, 2006 18:17:04 GMT
It's now reached "pictures of the bottom of the PCB" time Owen, get them ASAP and make them as big as you can.
Supplementary to what Rick said I would also arm myself with a stout toothbrush, some isopropyl alcohol (or whisky / vodka if you can't get isopropyl) and scrub that board for a good 5 minutes, drying with kitchen roll. You'll be AMAZED just how some amps fire up after a damned good cleaning.
Also, I want you to remove that headphone jack totally (get it off the board). Measure offset between tip and sleeve and ring and sleeve... if all's well you've found your culprit. If not ok then remove all of the opamps and perform the same measurements but also measure each opamp output.
Personally, before doing any measuring I would go across EVERY pad on the board with a hot iron and feed a small amount of solder into each joint, until it melts and becomes shiny. When you populated the board did you solder and snip or snip and then solder?? if you snip and then solder it's possible the leg of one of the components has fallen slightly into the hole and is not making contact with anything, even though the joint "looks" good on the surface. If you did what I do and bend the legs over flush onto the pads and then solder then you're pretty sure a cold joint isn't going to prove to be the problem but still worth going over again anyway.
Another thing, get a hook like tool and go round the board (on the top) trying to pull the resistors upwards, if one pops out pretty easily, there's your culprit.... do the same with the capacitors but use your fingers. Another way is to probe the leg of a component on the top side of the board and probe the pad on the bottom to check for continuity.
Keep coming back, we need pictures and updates.
Mike.
|
|
rickcr42
Fully Modded
Rest in peace my good friend.
Posts: 4,514
|
Post by rickcr42 on Nov 29, 2006 18:54:27 GMT
damn good advice mike Listen to mike Owen,he builds WAY more shit than I do so has this down to a science dude and don't worry man,I didn't fire those boards off to you just to give up at the first sign of touble ! We WILL get this fker playing music even if I have to swim across the big pond with my soldering iron in tow ;D
|
|
|
Post by PinkFloyd on Nov 29, 2006 23:20:29 GMT
No need to look your swimming trunks out Rick (or is is a one piece bathing suit?) if the resoldering and cleaning doesn't inject some life into it I'll have a look at it for Owen though it'll have to be sent via Royal Mail, no way am I swimming anywhere in this f**kin' weather.
|
|