>>It's only about 10wpc so a 5-1/2"x5" square bottom plate (elevated by 1/2" tall rubber feet) should be enough (I hope).
>>
if it's thick aluminum, sure. if it's a cookie tin, maybe you should give them a little help. but if it doesn't feel hot, it's probably fine. i'd imagine what you're doing should be okay for testing. it's later on when you're cranking up floyd that you may need better sinking.
>>it appears that my volume control potentiometer is faulty - there is a 'dud spot' in the track that's breaking connection. I'll have to replace it before going much further.
>>
sounds like you're actually getting lucky. if you'd had all this bad luck back with valves then you might have hurt more than your feelings.
>>the US reputedly has the filthiest AC mains in the civilised world.
heh. i'm noticing a trend that the "filthiest AC mains in the civilised world" tend to be wherever an audiophile is having trouble.
>>It could be that RFI borne along the mains were knocking the chip into oscillation.
RFI can do that? i'd figure RFI would be an annoyance, but knock a stock chip circuit into oscillation? wouldn't that only be possible if the circuit was already borderline instable?
and fwiw, (different circuit and all but...) my headamp is unchoked and unshielded and lives quite happily on the nasty computer powerbar (several computers) in a building with lots of mains noise and 'brown' power. i'm just saying i'd think filtering is for getting improved sound. if you've got an instable circuit, i wonder if there's something else to be corrected first.
>>If true, it's time for me to get heavy-handed and put one or two more RFI filters in there, including a ferrite choke on the AC mains in.
>>
can't hurt. if you've got an obsolete mac in the house, try its psu for parts. the ones i've gutted were very nicely choked. (much more seldom with pc psu's)
>>I'll probably snub the diode bridge while I'm at it.
'scuse my ignorance again. i thought bridge snubbing just made the rectification a little smoother. essentially a trick to help out 1N4007's when one doesn't wish to order schottky's. does it help with brown power too?
>>GOD I wish I'd had the $$ to build this one as a battery powered job....
oh fiddle. how often does one get to revisit the frustrations of childhood? it's a hobby. treat it like repairing english cars: a sport.
>>
if it's thick aluminum, sure. if it's a cookie tin, maybe you should give them a little help. but if it doesn't feel hot, it's probably fine. i'd imagine what you're doing should be okay for testing. it's later on when you're cranking up floyd that you may need better sinking.
>>it appears that my volume control potentiometer is faulty - there is a 'dud spot' in the track that's breaking connection. I'll have to replace it before going much further.
>>
sounds like you're actually getting lucky. if you'd had all this bad luck back with valves then you might have hurt more than your feelings.
>>the US reputedly has the filthiest AC mains in the civilised world.
heh. i'm noticing a trend that the "filthiest AC mains in the civilised world" tend to be wherever an audiophile is having trouble.
>>It could be that RFI borne along the mains were knocking the chip into oscillation.
RFI can do that? i'd figure RFI would be an annoyance, but knock a stock chip circuit into oscillation? wouldn't that only be possible if the circuit was already borderline instable?
and fwiw, (different circuit and all but...) my headamp is unchoked and unshielded and lives quite happily on the nasty computer powerbar (several computers) in a building with lots of mains noise and 'brown' power. i'm just saying i'd think filtering is for getting improved sound. if you've got an instable circuit, i wonder if there's something else to be corrected first.
>>If true, it's time for me to get heavy-handed and put one or two more RFI filters in there, including a ferrite choke on the AC mains in.
>>
can't hurt. if you've got an obsolete mac in the house, try its psu for parts. the ones i've gutted were very nicely choked. (much more seldom with pc psu's)
>>I'll probably snub the diode bridge while I'm at it.
'scuse my ignorance again. i thought bridge snubbing just made the rectification a little smoother. essentially a trick to help out 1N4007's when one doesn't wish to order schottky's. does it help with brown power too?
>>GOD I wish I'd had the $$ to build this one as a battery powered job....
oh fiddle. how often does one get to revisit the frustrations of childhood? it's a hobby. treat it like repairing english cars: a sport.
Hi OFB;
Sorry if I've sounded a bit irritable! 🙂 This is sort of a stream-of-consciousness thread of troubleshooting....
From power testing yesterday the "chassis as a heatsink" will work fine. It barely warmed to the touch, even directly under the chip. When the second channel is up and running, it should not be too warm by a long shot!
>>>...if you'd had all this bad luck back with valves then you might have hurt more than your feelings...<<<
*LOL!* 🙂 I've had some near misses over the last couple of decades - but the worst were at work rather than here.
In all seriousness though, valve equipment is really pretty forgiving. Stay away from Futtermans or parallel output configurations, and you'll likely never encounter oscillation (save in some RF valves like the 6DJ8/6922). Most of the time you just throw a 100 to 1000 ohm resistor in as a grid stopper and you're fine.
>>>...RFI can do that? i'd figure RFI would be an annoyance, but knock a stock chip circuit into oscillation? wouldn't that only be possible if the circuit was already borderline instable?...<<<
You betcha. Any AC signal hitting an amp through the power rails will end up impressing itself at some level on the output. That's why dirty mains give things such a "grainy grunge" of a sound. That's also why you try to put decoupling caps near to the chip. Here, I was originally allowing a couple of inches of lead and that appears to have been too much.
Any AC signal with a freq higher than the "elbow freq" of the amp (that's the freq at which degeneration turns into regeneration, or to use the newfangled lingo, when NFB turns into PFB) will toss an amp using a feedback loop into oscillation. One way to avoid that problem is to choose an amp with poor frequency response characteristics above the audio range; another is to eschew the use of NFB altogether (easier said than done in some ways - fail to put in cathode bypass caps on a common cathode amp and you've just provided it with constant current NFB). In valve amps, the idea of a gridstopper is that it "mass loads" the virtual "spring system" that is oscillating (any electrical system can be modelled as a mechanical system and vice versa) with a small value resistor inline with the signal.
>>>...and fwiw, (different circuit and all but...) my headamp is unchoked and unshielded and lives quite happily on the nasty computer powerbar (several computers) in a building with lots of mains noise and 'brown' power. i'm just saying i'd think filtering is for getting improved sound. if you've got an instable circuit, i wonder if there's something else to be corrected first...<<<
Yep, the TDA2030 is a revamp of the 2003, which was designed originally for automotive radios off battery power. My guess is that it does not incorporate any protective circuitry inside for AC mains use so I'm having to put it in myself. I've been warned in separate correspondence about it's tendencies to oscillate by a friend in the UK. Apparently at some point in it's development chain, faster transistors were used that kicked it's high freq response up into the region where it starts to regenerate.
Your headamp probably has some degree of internal protection, as well as decoupling caps at the chip's pins - which is the approach I took on my later versions. Had to go out and buy some .1uF jobs that were teeny enough to fit in there though. My usual .1uF 600V Orange Drops are actually bigger than the chip.....
>>>...i thought bridge snubbing just made the rectification a little smoother. essentially a trick to help out 1N4007's when one doesn't wish to order schottky's. does it help with brown power too?...<<<
True that the usual reason for snubbing caps is to provide an alternate pathway through the poor QF cap for the RF grunge that hard switching generates. However, any RFI that's piggybacked on the 60cps AC mains will also have a nice alternate pathway out of your DC rails too....it's a pure belt and suspenders approach.
>>>...can't hurt. if you've got an obsolete mac in the house, try its psu for parts. the ones i've gutted were very nicely choked. (much more seldom with pc psu's)...<<<
Nope, no macs, but everything that friends and family toss usually ends up going across my workbench for scrapping out and I'm pretty ruthless about scavenging parts. Another great source for chokes and "X" caps is junker CD and DVD players. Even cheapies will usually have an "X" cap and a choke in there so that they can meet Part 15 of the FCC regs (SMPS are notorious for dumping RFI into mains). 'Course some companies cheat - I've read that those cordless telephone things are some of the worst offenders that way (wouldn't know, I've still got a rotary dial telephone! 🙂 ).
>>>.... how often does one get to revisit the frustrations of childhood?...<<<
*LOL!* 🙂 Too bad it's just the frustrations of childhood I get to revisit! 🙂 Be nice if I could get rid of the gray hair and wrinkles this way! 😉
All the best,
Morse
Sorry if I've sounded a bit irritable! 🙂 This is sort of a stream-of-consciousness thread of troubleshooting....
From power testing yesterday the "chassis as a heatsink" will work fine. It barely warmed to the touch, even directly under the chip. When the second channel is up and running, it should not be too warm by a long shot!
>>>...if you'd had all this bad luck back with valves then you might have hurt more than your feelings...<<<
*LOL!* 🙂 I've had some near misses over the last couple of decades - but the worst were at work rather than here.
In all seriousness though, valve equipment is really pretty forgiving. Stay away from Futtermans or parallel output configurations, and you'll likely never encounter oscillation (save in some RF valves like the 6DJ8/6922). Most of the time you just throw a 100 to 1000 ohm resistor in as a grid stopper and you're fine.
>>>...RFI can do that? i'd figure RFI would be an annoyance, but knock a stock chip circuit into oscillation? wouldn't that only be possible if the circuit was already borderline instable?...<<<
You betcha. Any AC signal hitting an amp through the power rails will end up impressing itself at some level on the output. That's why dirty mains give things such a "grainy grunge" of a sound. That's also why you try to put decoupling caps near to the chip. Here, I was originally allowing a couple of inches of lead and that appears to have been too much.
Any AC signal with a freq higher than the "elbow freq" of the amp (that's the freq at which degeneration turns into regeneration, or to use the newfangled lingo, when NFB turns into PFB) will toss an amp using a feedback loop into oscillation. One way to avoid that problem is to choose an amp with poor frequency response characteristics above the audio range; another is to eschew the use of NFB altogether (easier said than done in some ways - fail to put in cathode bypass caps on a common cathode amp and you've just provided it with constant current NFB). In valve amps, the idea of a gridstopper is that it "mass loads" the virtual "spring system" that is oscillating (any electrical system can be modelled as a mechanical system and vice versa) with a small value resistor inline with the signal.
>>>...and fwiw, (different circuit and all but...) my headamp is unchoked and unshielded and lives quite happily on the nasty computer powerbar (several computers) in a building with lots of mains noise and 'brown' power. i'm just saying i'd think filtering is for getting improved sound. if you've got an instable circuit, i wonder if there's something else to be corrected first...<<<
Yep, the TDA2030 is a revamp of the 2003, which was designed originally for automotive radios off battery power. My guess is that it does not incorporate any protective circuitry inside for AC mains use so I'm having to put it in myself. I've been warned in separate correspondence about it's tendencies to oscillate by a friend in the UK. Apparently at some point in it's development chain, faster transistors were used that kicked it's high freq response up into the region where it starts to regenerate.
Your headamp probably has some degree of internal protection, as well as decoupling caps at the chip's pins - which is the approach I took on my later versions. Had to go out and buy some .1uF jobs that were teeny enough to fit in there though. My usual .1uF 600V Orange Drops are actually bigger than the chip.....
>>>...i thought bridge snubbing just made the rectification a little smoother. essentially a trick to help out 1N4007's when one doesn't wish to order schottky's. does it help with brown power too?...<<<
True that the usual reason for snubbing caps is to provide an alternate pathway through the poor QF cap for the RF grunge that hard switching generates. However, any RFI that's piggybacked on the 60cps AC mains will also have a nice alternate pathway out of your DC rails too....it's a pure belt and suspenders approach.
>>>...can't hurt. if you've got an obsolete mac in the house, try its psu for parts. the ones i've gutted were very nicely choked. (much more seldom with pc psu's)...<<<
Nope, no macs, but everything that friends and family toss usually ends up going across my workbench for scrapping out and I'm pretty ruthless about scavenging parts. Another great source for chokes and "X" caps is junker CD and DVD players. Even cheapies will usually have an "X" cap and a choke in there so that they can meet Part 15 of the FCC regs (SMPS are notorious for dumping RFI into mains). 'Course some companies cheat - I've read that those cordless telephone things are some of the worst offenders that way (wouldn't know, I've still got a rotary dial telephone! 🙂 ).
>>>.... how often does one get to revisit the frustrations of childhood?...<<<
*LOL!* 🙂 Too bad it's just the frustrations of childhood I get to revisit! 🙂 Be nice if I could get rid of the gray hair and wrinkles this way! 😉
All the best,
Morse
m'kay. this is all quite new to me, so hang on while i try to catch up.
>>You betcha. Any AC signal hitting an amp through the power rails will end up impressing itself at some level on the output. That's why dirty mains give things such a "grainy grunge" of a sound. That's also why you try to put decoupling caps near to the chip. Here, I was originally allowing a couple of inches of lead and that appears to have been too much.
>>
"That's also why you try to put decoupling caps near to the chip." er, what is? i'm not sure how the lead distance of the decoupling cap affects things. i only know short distances are part of gainclone voodoo in general. the Why part is still mysterious.
to be certain i'm on the same playing field, the decoupling cap is the one across the power rails just before the rails connect with the chip, right?
>>Any AC signal with a freq higher than the "elbow freq" of the amp (that's the freq at which degeneration turns into regeneration, or to use the newfangled lingo, when NFB turns into PFB)
>>
"elbow freq" ... this is the same as "corner frequency" from the high-pass filter? (part of the fun of figuring out this stuff is getting a grasp of regional and generational dialects.)
>>Yep, the TDA2030 is a revamp of the 2003,
funny. when you first started having unusual trouble i dug through the scrap bin to see if i could build a circuit that did the same, as tda2030 sounded familiar. turned out i had tda2003.
>>Your headamp probably has some degree of internal protection, as well as decoupling caps at the chip's pins
>>
actually no decoupling cap, oddly enough. that was only recommended if there were oscillation problems. there were none, so i hadn't thought about it since. stability must be result of the ad8620 and quite a bit of capacitance in the regulated supply and the amp's own power supply circuit. (there's another mystery of the gc for me: just trans, bridge, and two caps. "so, where's the rest of it?" good fun.)
>>Had to go out and buy some .1uF jobs that were teeny enough to fit in there though. My usual .1uF 600V Orange Drops are actually bigger than the chip.....
>>
that's a question i should ask. does it matter if one uses the high voltage stuff when the capacitance is the same? i wouldn't think so, but i wouldn't have thought a couple of inches of lead length would have made large differences either.
>> (wouldn't know, I've still got a rotary dial telephone! ).
me too, but thanks to broadband and cell, no more telephone line. my rotary has become a figurative conversation piece.
>>You betcha. Any AC signal hitting an amp through the power rails will end up impressing itself at some level on the output. That's why dirty mains give things such a "grainy grunge" of a sound. That's also why you try to put decoupling caps near to the chip. Here, I was originally allowing a couple of inches of lead and that appears to have been too much.
>>
"That's also why you try to put decoupling caps near to the chip." er, what is? i'm not sure how the lead distance of the decoupling cap affects things. i only know short distances are part of gainclone voodoo in general. the Why part is still mysterious.
to be certain i'm on the same playing field, the decoupling cap is the one across the power rails just before the rails connect with the chip, right?
>>Any AC signal with a freq higher than the "elbow freq" of the amp (that's the freq at which degeneration turns into regeneration, or to use the newfangled lingo, when NFB turns into PFB)
>>
"elbow freq" ... this is the same as "corner frequency" from the high-pass filter? (part of the fun of figuring out this stuff is getting a grasp of regional and generational dialects.)
>>Yep, the TDA2030 is a revamp of the 2003,
funny. when you first started having unusual trouble i dug through the scrap bin to see if i could build a circuit that did the same, as tda2030 sounded familiar. turned out i had tda2003.
>>Your headamp probably has some degree of internal protection, as well as decoupling caps at the chip's pins
>>
actually no decoupling cap, oddly enough. that was only recommended if there were oscillation problems. there were none, so i hadn't thought about it since. stability must be result of the ad8620 and quite a bit of capacitance in the regulated supply and the amp's own power supply circuit. (there's another mystery of the gc for me: just trans, bridge, and two caps. "so, where's the rest of it?" good fun.)
>>Had to go out and buy some .1uF jobs that were teeny enough to fit in there though. My usual .1uF 600V Orange Drops are actually bigger than the chip.....
>>
that's a question i should ask. does it matter if one uses the high voltage stuff when the capacitance is the same? i wouldn't think so, but i wouldn't have thought a couple of inches of lead length would have made large differences either.
>> (wouldn't know, I've still got a rotary dial telephone! ).
me too, but thanks to broadband and cell, no more telephone line. my rotary has become a figurative conversation piece.
Hi OFB;
Sorry "decoupling" was a typo - I meant to write "bypass". Technically a "decoupling" cap is used to isolate one side of a line from another's DC level. Imagine that you want to tie part of a circuit that HAS to be at 250VDC to a line that HAS to be at 50VDC - you put a "coupling" or "decoupling" cap there to keep the circuit on one side of the cap at 250VDC and the other at 50VDC. That happens more than you might think with valve equipment.
Yes, a "bypass" cap is used to shunt AC off from the line. And yes, that last cap near the chip is the bypass cap. Sometimes confusion arises over what to call a small value cap in parallel with a bypass cap - usually it's best called a "shunt" cap. The idea there is that little caps are more transparent to high freqs than are large value caps, all else being equal, so it's like having a woofer and a tweeter in a loudspeaker.
Yep, "elbow freq", "corner freq", "knee freq" - take your pick. The gimmick with degeneration turning into regeneration is all tied into the reactive capacitance in the line. There's a frequency dependant phase shift, and when the phase shift goes far enough, the signal that's supposed to be cancelling out is suddenly reinforcing on every pass through the amp. Destruction follows....
I've got a pair of TDA2003's in my parts bins too; however I sourced a bunch of 2030's at under $2 each so what the heck. 2003's are actually more expensive now, so go figure... The 2030 looked good for experimenting and learning how to handle, layout, and solder teeny weeny little components instead of my usual stuff. It's been a learning experience!
>>>...does it matter if one uses the high voltage stuff when the capacitance is the same?...<<<
There's no problem at all with using a higher voltage rated cap for a low voltage application - electrically, that is. Here, my bins of Orange Drops were pretty well worthless, because they wouldn't fit in the submini chassis. I've also used this as an excuse to start stocking up on 26ga solid core wire and 1/8 watt resistors. If you haven't tried 1/8 watters, you ought to - they're wonderfully tiny and are just perfect for running things like the feedback loop between 2 pins on an IC.
The other problem with high voltage caps is cost - small value caps like Sprague Orange Drops are about $2 each and they are a bargain. Max value on OD's is .47uF.
>>>...i wouldn't have thought a couple of inches of lead length would have made large differences either...<<<
Lead length doesn't matter a bit at audio freqs. It's those pesty radio freqs where things like that make life interesting. Worse are the secondary characteristics of other parts (resistance in caps, capacitance in resistors, etc). Modern film type resistors present some problems for RF apps that make the selective use of carbon compositions a must. Pity they're getting scarce and pricey. Makes me wish I would have laid in a few more thousand of the things when they were cheap.....
>>>...stability must be result of the ad8620 and quite a bit of capacitance in the regulated supply and the amp's own power supply circuit...<<<
Yep, different circuits inside the chip are more or less stable depending on a LOT of things. Don't ask me though - I'd love to learn chip designing, but for now I'll settle with learning to best use the ready rolled ones. If you're looking for a fun "gee whiz" project, look into building a simple opamp from discrete components. It can be done; one of my long list of backburnered projects is to build one out of valves as a conversation piece ("world's biggest opamp", LOL!).
>>>...my rotary has become a figurative conversation piece...<<<
If you want a real mindblower for the kids these days, get an acoustically coupled modem and sit it with your rotary phone next to the computer. And if you want to top that, get a teletypewriter and hook it up as a printer...... 😉
All the best,
Morse
Sorry "decoupling" was a typo - I meant to write "bypass". Technically a "decoupling" cap is used to isolate one side of a line from another's DC level. Imagine that you want to tie part of a circuit that HAS to be at 250VDC to a line that HAS to be at 50VDC - you put a "coupling" or "decoupling" cap there to keep the circuit on one side of the cap at 250VDC and the other at 50VDC. That happens more than you might think with valve equipment.
Yes, a "bypass" cap is used to shunt AC off from the line. And yes, that last cap near the chip is the bypass cap. Sometimes confusion arises over what to call a small value cap in parallel with a bypass cap - usually it's best called a "shunt" cap. The idea there is that little caps are more transparent to high freqs than are large value caps, all else being equal, so it's like having a woofer and a tweeter in a loudspeaker.
Yep, "elbow freq", "corner freq", "knee freq" - take your pick. The gimmick with degeneration turning into regeneration is all tied into the reactive capacitance in the line. There's a frequency dependant phase shift, and when the phase shift goes far enough, the signal that's supposed to be cancelling out is suddenly reinforcing on every pass through the amp. Destruction follows....
I've got a pair of TDA2003's in my parts bins too; however I sourced a bunch of 2030's at under $2 each so what the heck. 2003's are actually more expensive now, so go figure... The 2030 looked good for experimenting and learning how to handle, layout, and solder teeny weeny little components instead of my usual stuff. It's been a learning experience!
>>>...does it matter if one uses the high voltage stuff when the capacitance is the same?...<<<
There's no problem at all with using a higher voltage rated cap for a low voltage application - electrically, that is. Here, my bins of Orange Drops were pretty well worthless, because they wouldn't fit in the submini chassis. I've also used this as an excuse to start stocking up on 26ga solid core wire and 1/8 watt resistors. If you haven't tried 1/8 watters, you ought to - they're wonderfully tiny and are just perfect for running things like the feedback loop between 2 pins on an IC.
The other problem with high voltage caps is cost - small value caps like Sprague Orange Drops are about $2 each and they are a bargain. Max value on OD's is .47uF.
>>>...i wouldn't have thought a couple of inches of lead length would have made large differences either...<<<
Lead length doesn't matter a bit at audio freqs. It's those pesty radio freqs where things like that make life interesting. Worse are the secondary characteristics of other parts (resistance in caps, capacitance in resistors, etc). Modern film type resistors present some problems for RF apps that make the selective use of carbon compositions a must. Pity they're getting scarce and pricey. Makes me wish I would have laid in a few more thousand of the things when they were cheap.....
>>>...stability must be result of the ad8620 and quite a bit of capacitance in the regulated supply and the amp's own power supply circuit...<<<
Yep, different circuits inside the chip are more or less stable depending on a LOT of things. Don't ask me though - I'd love to learn chip designing, but for now I'll settle with learning to best use the ready rolled ones. If you're looking for a fun "gee whiz" project, look into building a simple opamp from discrete components. It can be done; one of my long list of backburnered projects is to build one out of valves as a conversation piece ("world's biggest opamp", LOL!).
>>>...my rotary has become a figurative conversation piece...<<<
If you want a real mindblower for the kids these days, get an acoustically coupled modem and sit it with your rotary phone next to the computer. And if you want to top that, get a teletypewriter and hook it up as a printer...... 😉
All the best,
Morse
>>Yes, a "bypass" cap is used to shunt AC off from the line.
hope you don't mind a severe 101 question: just what is the AC doing there anyway? i mean, after the bridge and all. to use the hydraulic metaphor that i've been loosely grasping this stuff with, does some of the back'n'forth pressure of the AC preceding the bridge essentially resonate through? it's one of the things i've been curious about.
insert general nodding for the rest of your explanation. most helpful. thank you.
>>I've got a pair of TDA2003's in my parts bins too;
they're not much good, are they? mine showed up in a blister pack kit that a friend bought as a gift when i started messing with this stuff. cruising the web seemed to indicate they were just for cheap car amps, and that nothing else in the kit was worth saving but the heatsinks. hence it was shelved.
i suppose i could wire them up per the datasheet to try, but with LM3875's on the way there didn't seem any point.
>>If you haven't tried 1/8 watters, you ought to - they're wonderfully tiny...
actually i've started from that scale. it was the kids doing headamps in mint tins that got my attention in the first place, so what's small for you looks normal to me. and it's positively huge compared to fixing bad joints in my laptop with a weller gun. that was like whittling with an axe.
>>If you're looking for a fun "gee whiz" project, look into building a simple opamp from discrete components.
you know, that's already on the list? i figured it'd be good to spend a few evenings on.
>>If you want a real mindblower for the kids these days, get an acoustically coupled modem and sit it with your rotary phone next to the computer. And if you want to top that, get a teletypewriter and hook it up as a printer......
>>
definitely. my osborne executive has caused some very entertaining guesses to be voiced. usually they figure it's obscure test gear. there's even been flat disbelief that it's a computer.
hope you don't mind a severe 101 question: just what is the AC doing there anyway? i mean, after the bridge and all. to use the hydraulic metaphor that i've been loosely grasping this stuff with, does some of the back'n'forth pressure of the AC preceding the bridge essentially resonate through? it's one of the things i've been curious about.
insert general nodding for the rest of your explanation. most helpful. thank you.
>>I've got a pair of TDA2003's in my parts bins too;
they're not much good, are they? mine showed up in a blister pack kit that a friend bought as a gift when i started messing with this stuff. cruising the web seemed to indicate they were just for cheap car amps, and that nothing else in the kit was worth saving but the heatsinks. hence it was shelved.
i suppose i could wire them up per the datasheet to try, but with LM3875's on the way there didn't seem any point.
>>If you haven't tried 1/8 watters, you ought to - they're wonderfully tiny...
actually i've started from that scale. it was the kids doing headamps in mint tins that got my attention in the first place, so what's small for you looks normal to me. and it's positively huge compared to fixing bad joints in my laptop with a weller gun. that was like whittling with an axe.
>>If you're looking for a fun "gee whiz" project, look into building a simple opamp from discrete components.
you know, that's already on the list? i figured it'd be good to spend a few evenings on.
>>If you want a real mindblower for the kids these days, get an acoustically coupled modem and sit it with your rotary phone next to the computer. And if you want to top that, get a teletypewriter and hook it up as a printer......
>>
definitely. my osborne executive has caused some very entertaining guesses to be voiced. usually they figure it's obscure test gear. there's even been flat disbelief that it's a computer.
Hi OFB;
>>>...hope you don't mind a severe 101 question: just what is the AC doing there anyway? i mean, after the bridge and all. to use the hydraulic metaphor that i've been loosely grasping this stuff with, does some of the back'n'forth pressure of the AC preceding the bridge essentially resonate through? it's one of the things i've been curious about...<<<
Essentially, yes. Here's how.
Let's go back and look at what a AC is and how a rectifier works to understand it better. There are a number of different rectification schemes, but the 2 most popular are the bridge and the full wave rectifier (sometimes a bridge is called a 'full wave bridge' adding to the confusion). Think of your hydraulic model as being the side view of a water wave (or sine wave, if you will). In any full wave rectification, the "up" portion of the sine wave remains unaltered, but the "down" portion of the sine wave is inverted. Thus, you get a wave pattern after full wave rectification that looks like a row of small hills.
Next comes filtering. When you tie a capacitor to ground, you're "filling in the gaps" between the "hills". However, there's a maximum rate at which a cap can "fill in the gaps" and if you have some RF signals they may be so quick the cap can't respond effectively. If you also add a large value inductor (my 2A3 uses a 10 Henry inductor that's about the same size as this chipamp I'm playing with - it's considerably heavier, too!) in series, it will "knock down the tips" of the hills. Commonly in older electronics we used "pi filters" consisting of a cap tied to ground, an inductor in series, and a second cap tied to ground. A resistor can be used in place of the inductor, but you'll lose some voltage and generate some heat under the hood.
In solid state electronics, it's become common to just throw a few thousand microfarad of capacitance in a single cap tied to ground and call it 'filtered'. This will work, though it has it's downsides. One is that it works the transformer harder, since it has to charge up the cap at poweron - and the bigger the cap, the harder the trafo has to struggle to put out surge current 'till it's reached stable operation. The other problem with this approach is that you are limited to low voltage applications. If you don't believe me, go out and try to find a 10,000 uF cap rated to 500VDC. Yes, you can parallel a hundred 100uF @ 500VDC caps - at about $9 each. And each one will be anywhere from the size of a 35mm film cannister to the size of a roll of half dollars. Easier by far to throw 2 caps at the problem and a $60 to $100 choke - which is why most valve electronics take this general approach.
Okay, so after you go through all this it's '100% filtered', right? Well, no. There's always some residual AC lurking in the shadows - caps are not 100% efficient as shunting RFI (due to it's frequency, and the maximum rate at which a cap of a given size can respond to AC fluctuations), and even inductors let some percent of things sneak past. And the tiniest bit of line noise can be enough to start an unstable device oscillating, or add some serious grunge to the sound. If you're up for an experiment, put a couple of large ferrite chokes on your TV's AC mains, as near the set as possible. If the picture cleans up significantly, congratulation, line noise is a real problem in your area. It is in mine...
>>>... it's positively huge compared to fixing bad joints in my laptop with a weller gun. that was like whittling with an axe...<<<
Yep, you definitely need a good soldering station if you're going to do anything with surface mount devices. I've done a little of that using boards and it's tricky all right. Here, the learning curve involves the fact that I'm trying to go point to point (no circuit board) with these teeny little buggers. One thing I'm discovering is that if anything the positioning of the wires is at least as critical as it is on valve equipment.
>>>...definitely. my osborne executive has caused some very entertaining guesses to be voiced. usually they figure it's obscure test gear. there's even been flat disbelief that it's a computer....<<<
Oho, the Osborne Executive System! I rember that one....couldn't afford it myself. My last computer from the 'good old days' is a Commodore C128D. Remember the Vic 20 and the C64?
All the best,
Morse
>>>...hope you don't mind a severe 101 question: just what is the AC doing there anyway? i mean, after the bridge and all. to use the hydraulic metaphor that i've been loosely grasping this stuff with, does some of the back'n'forth pressure of the AC preceding the bridge essentially resonate through? it's one of the things i've been curious about...<<<
Essentially, yes. Here's how.
Let's go back and look at what a AC is and how a rectifier works to understand it better. There are a number of different rectification schemes, but the 2 most popular are the bridge and the full wave rectifier (sometimes a bridge is called a 'full wave bridge' adding to the confusion). Think of your hydraulic model as being the side view of a water wave (or sine wave, if you will). In any full wave rectification, the "up" portion of the sine wave remains unaltered, but the "down" portion of the sine wave is inverted. Thus, you get a wave pattern after full wave rectification that looks like a row of small hills.
Next comes filtering. When you tie a capacitor to ground, you're "filling in the gaps" between the "hills". However, there's a maximum rate at which a cap can "fill in the gaps" and if you have some RF signals they may be so quick the cap can't respond effectively. If you also add a large value inductor (my 2A3 uses a 10 Henry inductor that's about the same size as this chipamp I'm playing with - it's considerably heavier, too!) in series, it will "knock down the tips" of the hills. Commonly in older electronics we used "pi filters" consisting of a cap tied to ground, an inductor in series, and a second cap tied to ground. A resistor can be used in place of the inductor, but you'll lose some voltage and generate some heat under the hood.
In solid state electronics, it's become common to just throw a few thousand microfarad of capacitance in a single cap tied to ground and call it 'filtered'. This will work, though it has it's downsides. One is that it works the transformer harder, since it has to charge up the cap at poweron - and the bigger the cap, the harder the trafo has to struggle to put out surge current 'till it's reached stable operation. The other problem with this approach is that you are limited to low voltage applications. If you don't believe me, go out and try to find a 10,000 uF cap rated to 500VDC. Yes, you can parallel a hundred 100uF @ 500VDC caps - at about $9 each. And each one will be anywhere from the size of a 35mm film cannister to the size of a roll of half dollars. Easier by far to throw 2 caps at the problem and a $60 to $100 choke - which is why most valve electronics take this general approach.
Okay, so after you go through all this it's '100% filtered', right? Well, no. There's always some residual AC lurking in the shadows - caps are not 100% efficient as shunting RFI (due to it's frequency, and the maximum rate at which a cap of a given size can respond to AC fluctuations), and even inductors let some percent of things sneak past. And the tiniest bit of line noise can be enough to start an unstable device oscillating, or add some serious grunge to the sound. If you're up for an experiment, put a couple of large ferrite chokes on your TV's AC mains, as near the set as possible. If the picture cleans up significantly, congratulation, line noise is a real problem in your area. It is in mine...
>>>... it's positively huge compared to fixing bad joints in my laptop with a weller gun. that was like whittling with an axe...<<<
Yep, you definitely need a good soldering station if you're going to do anything with surface mount devices. I've done a little of that using boards and it's tricky all right. Here, the learning curve involves the fact that I'm trying to go point to point (no circuit board) with these teeny little buggers. One thing I'm discovering is that if anything the positioning of the wires is at least as critical as it is on valve equipment.
>>>...definitely. my osborne executive has caused some very entertaining guesses to be voiced. usually they figure it's obscure test gear. there's even been flat disbelief that it's a computer....<<<
Oho, the Osborne Executive System! I rember that one....couldn't afford it myself. My last computer from the 'good old days' is a Commodore C128D. Remember the Vic 20 and the C64?
All the best,
Morse
>>Essentially, yes. Here's how.
sounds good. thanks for that.
>> My last computer from the 'good old days' is a Commodore C128D. Remember the Vic 20 and the C64?
>>
you bet. my vic 20 was a big step up from the PET, but i dreamed of the C128. it was the Unobtanium for a while.
sounds good. thanks for that.
>> My last computer from the 'good old days' is a Commodore C128D. Remember the Vic 20 and the C64?
>>
you bet. my vic 20 was a big step up from the PET, but i dreamed of the C128. it was the Unobtanium for a while.
Hi OFB;
>>>...my vic 20 was a big step up from the PET...<<<
Yep; the VIC was quite the machine in it's day - I had one of the buggers that I "raced" against a TI 99/4a, back when they were competitors. All it involved was a loop counting from 1 to some large number and displaying the value on the screen. The VIC beat the pants off the TI99/4A, since the TI didn't have on chip registers and only had an 8 bit data bus - 16 bit chip or not.
>>>...but i dreamed of the C128. it was the Unobtanium for a while...<<<
Frankly it was Unobtainium for me too - 'til one of my better heeled friends decided to sell his after he went with an early generation pentium. I got the thing for a song.
If you ever decide to track one down at a garage sale or such, the only caution is to NEVER hot swap components on ANY of the Commodores - and that includes joysticks. I've seen more than one of 'em get the CIA (Complex Interface Adaptor) chip fried that way. On the bright side, all the way up through the C128, they were using through mount devices that are relatively easy to replace compared to the modern SMD stuff - so if a generic component like an electrolytic cap or a resistor fries it's easy to replace.
Sorry for the trip down memory lane - I guess it's just a sign of age that I'm doing that more and more.... 😉
All the best,
Morse
>>>...my vic 20 was a big step up from the PET...<<<
Yep; the VIC was quite the machine in it's day - I had one of the buggers that I "raced" against a TI 99/4a, back when they were competitors. All it involved was a loop counting from 1 to some large number and displaying the value on the screen. The VIC beat the pants off the TI99/4A, since the TI didn't have on chip registers and only had an 8 bit data bus - 16 bit chip or not.
>>>...but i dreamed of the C128. it was the Unobtanium for a while...<<<
Frankly it was Unobtainium for me too - 'til one of my better heeled friends decided to sell his after he went with an early generation pentium. I got the thing for a song.
If you ever decide to track one down at a garage sale or such, the only caution is to NEVER hot swap components on ANY of the Commodores - and that includes joysticks. I've seen more than one of 'em get the CIA (Complex Interface Adaptor) chip fried that way. On the bright side, all the way up through the C128, they were using through mount devices that are relatively easy to replace compared to the modern SMD stuff - so if a generic component like an electrolytic cap or a resistor fries it's easy to replace.
Sorry for the trip down memory lane - I guess it's just a sign of age that I'm doing that more and more.... 😉
All the best,
Morse
>>If you ever decide to track one down at a garage sale or such,
heh. i'm cured! had my retro binge already. nowadays i'm quite happy to watch this stuff fill other people's closets.
>>Sorry for the trip down memory lane - I guess it's just a sign of age that I'm doing that more and more....
there's also memory alleyways. check this link that one of my younger friends just inflicted on me. a whole lot of music i'd flat forgotten, and/or repressed. quite the retrospective.
http://weddingspartieseverything.blogspot.com/2003_12_01_weddingspartieseverything_archive.html#107102807472234623
nice chat though, thank you. to turn back on subject, i look forward to you next update on the reluctant tda2030.
heh. i'm cured! had my retro binge already. nowadays i'm quite happy to watch this stuff fill other people's closets.
>>Sorry for the trip down memory lane - I guess it's just a sign of age that I'm doing that more and more....
there's also memory alleyways. check this link that one of my younger friends just inflicted on me. a whole lot of music i'd flat forgotten, and/or repressed. quite the retrospective.
http://weddingspartieseverything.blogspot.com/2003_12_01_weddingspartieseverything_archive.html#107102807472234623
nice chat though, thank you. to turn back on subject, i look forward to you next update on the reluctant tda2030.
Thanks for the link, OFB;
Back OT, I did get one channel working properly before I decided to do a total rebuild. Not bad sounding (at least it was smooth without the usual ss grain), but no where near what my OPA134 headphone amps will do, and not in the same league by a long shot as my DIY valve stuff. Still, this is a 'training wheels' project.
Next up, I'm going to try carefully cutting some pretinned perfboard and opening up the holes with a small drill. With a little careful bending of the pins I ought to be able to fit each chip with a 'miniboard' that will hold the actual filter parts and allow for larger components to be placed closer (I'd like to put the 1r resistor/.22uF cap combo on the board rather than on the speaker terminals for example).
That's just a stopgap though. Time for this old dog to learn some new tricks, like etching my own boards. Seems kind of ironic, since some of my first serious DIY electronics a few decades back was fixing some cracked boards by hand wiring point to point jumpers with 30 gauge wrapping wire after epoxying the boards back together....
All the best,
Morse
Back OT, I did get one channel working properly before I decided to do a total rebuild. Not bad sounding (at least it was smooth without the usual ss grain), but no where near what my OPA134 headphone amps will do, and not in the same league by a long shot as my DIY valve stuff. Still, this is a 'training wheels' project.
Next up, I'm going to try carefully cutting some pretinned perfboard and opening up the holes with a small drill. With a little careful bending of the pins I ought to be able to fit each chip with a 'miniboard' that will hold the actual filter parts and allow for larger components to be placed closer (I'd like to put the 1r resistor/.22uF cap combo on the board rather than on the speaker terminals for example).
That's just a stopgap though. Time for this old dog to learn some new tricks, like etching my own boards. Seems kind of ironic, since some of my first serious DIY electronics a few decades back was fixing some cracked boards by hand wiring point to point jumpers with 30 gauge wrapping wire after epoxying the boards back together....
All the best,
Morse
- Status
- Not open for further replies.