OK, I've got my parts!
I am jumping in, thanks to the confidence-boosting help of many of you on the board. Muchas gracias, y'all. I am not a total newbie - I just built a CMoy headphone amp. But that is different - it is rather hard to kill oneself with a nine-volt battery, now isn't it? I suppose I could choke on it.
Anyway, I plan to ask what may sound like very dumb questions specific to my project, especially when being wrong could cause injury or deadness. I promise to search for generic stuff. But I want to be totally sure on the critical stuff, too.
The plan is to build the PSU first on a plank, then after successful testing, build an amp channel. No sense in making it easy - I wanna go P2P, with short-as-possible paths in the (LM3875TF) amp circuit. I think I have the chip-wiring puzzle worked out. (Inverted GC.) Pins 1 and 4 are pointing straight out for the caps and such. 3,7 and 8 are pointing down. That should keep me out of trouble. NFB has to cross 7 between 3 and 8, but that is about the only place I see a potential worry. NC pins are clipped.
After building the amp successfully (fingers appropriately crossed here), I plan to get a little creative on the enclosure. Heck, worst thing I can do there is lose a thumb to the table saw, right? I have two of those, anyway. My CMoy was built in a solid (butcherblock-style) chunk of red oak. That won't be feasible with the GC, but I have some ideas...
I will post pix as I go. I am a photojournalist for a metro daily paper in the Northeast US, so one would hope at the very least, the pix should be sharp.
Looking at my Digikey and Partsexpress goodie bags that just arrived in the mail, I think I have found my first problem. The 1uf cap for running across the rails (pin 1 to 4) is a 50v Panasonic cap. I am running a 25/25/330 toriodal trafo, so I think I have to get a higher voltage cap for the voltage across the rails after DC conversion. Am I right here? If so, should 63v do it, or do I need 100v?
Also, Peter, my first executive creative decision is to build the bridge on a piece of wood, similar to your latest fiberglass bridge layout. But I am heatsinking them in a cheap, Home-Depot-kinda way. I am (gently) screwing each diode into the wood using a sandwich of small-big-small washers between the mur860 and the wood. This should give me enough heat dissipation to be safe, I figure.
I can't tell you guys how excited i am about this. I probably sound like a 15-year-old walking around with a condom in my pocket, "just in case." I already love my CMoy. This should rock. I am even thinking about building a CMoy headphone amp (as a totally separate circuit) into the final enclosure. Ahh, now there's a headphone jack one can be proud of...
Thanks much,
GnD
I am jumping in, thanks to the confidence-boosting help of many of you on the board. Muchas gracias, y'all. I am not a total newbie - I just built a CMoy headphone amp. But that is different - it is rather hard to kill oneself with a nine-volt battery, now isn't it? I suppose I could choke on it.
Anyway, I plan to ask what may sound like very dumb questions specific to my project, especially when being wrong could cause injury or deadness. I promise to search for generic stuff. But I want to be totally sure on the critical stuff, too.
The plan is to build the PSU first on a plank, then after successful testing, build an amp channel. No sense in making it easy - I wanna go P2P, with short-as-possible paths in the (LM3875TF) amp circuit. I think I have the chip-wiring puzzle worked out. (Inverted GC.) Pins 1 and 4 are pointing straight out for the caps and such. 3,7 and 8 are pointing down. That should keep me out of trouble. NFB has to cross 7 between 3 and 8, but that is about the only place I see a potential worry. NC pins are clipped.
After building the amp successfully (fingers appropriately crossed here), I plan to get a little creative on the enclosure. Heck, worst thing I can do there is lose a thumb to the table saw, right? I have two of those, anyway. My CMoy was built in a solid (butcherblock-style) chunk of red oak. That won't be feasible with the GC, but I have some ideas...
I will post pix as I go. I am a photojournalist for a metro daily paper in the Northeast US, so one would hope at the very least, the pix should be sharp.
Looking at my Digikey and Partsexpress goodie bags that just arrived in the mail, I think I have found my first problem. The 1uf cap for running across the rails (pin 1 to 4) is a 50v Panasonic cap. I am running a 25/25/330 toriodal trafo, so I think I have to get a higher voltage cap for the voltage across the rails after DC conversion. Am I right here? If so, should 63v do it, or do I need 100v?
Also, Peter, my first executive creative decision is to build the bridge on a piece of wood, similar to your latest fiberglass bridge layout. But I am heatsinking them in a cheap, Home-Depot-kinda way. I am (gently) screwing each diode into the wood using a sandwich of small-big-small washers between the mur860 and the wood. This should give me enough heat dissipation to be safe, I figure.
I can't tell you guys how excited i am about this. I probably sound like a 15-year-old walking around with a condom in my pocket, "just in case." I already love my CMoy. This should rock. I am even thinking about building a CMoy headphone amp (as a totally separate circuit) into the final enclosure. Ahh, now there's a headphone jack one can be proud of...
Thanks much,
GnD
GrahamnDodder said:
You are looking at close to 80V between rails with those secondaries, so 100V cap is a safer choice. You might not need it at all, anyway. I didn't find it useful in my circuit. So what exactly is your primary, 330V and you are using it with 110V?
You indeed sound like 15 years old with a condom in a pocket
Transformer is a 25/25/330VA toroidal, I am using a single for both channels, running 115 v AC input. Two bridges, split the rails for each channel there. Normal (2-1000 uf) caps on each channel.
What, in theory, does the pin-1- to 4-cap do if it is not really needed? Reduce relative noise between the rails? When I was first, well, memorizing the circuit I just took it for granted. But now as I start to understand the amp more it, (the cap) doesn't seem to make a whole lot of sense.
GnD
What, in theory, does the pin-1- to 4-cap do if it is not really needed? Reduce relative noise between the rails? When I was first, well, memorizing the circuit I just took it for granted. But now as I start to understand the amp more it, (the cap) doesn't seem to make a whole lot of sense.
GnD
In some circuits the cap may help to reduce noise or oscillations, but if you go for a compact p2p layout, you might not needed. Try it out and see if it does anything to the sound. In my circuit, I noticed that highs didn't sound natural with a cap, so I didn't bother.
If you go for minimized GC, you actually need only 2 resistors, series and feedback, (you might use pot or coupling cap, if your source generates offset). Other than that, you need only 2 PS filtering caps and power supply itself. Pretty simple circuit and you can't go wrong with it.
If you go for minimized GC, you actually need only 2 resistors, series and feedback, (you might use pot or coupling cap, if your source generates offset). Other than that, you need only 2 PS filtering caps and power supply itself. Pretty simple circuit and you can't go wrong with it.
GC without killing
for that hint.
Why haven't I thought of that myself?
Post without reading previous posting
and certainly not the first post (who cares of the subject
).
"GK without killing"
Is this a new "Peace-thread" ???
GrahamnDodder: OK, I've got my parts!
halojoy - think it real great to meet a guy
with some fresh and positive DIY entusiasm
too much "monkey.business" elsewhere:
- I know better than you do.
- No, you don't!
- Yes, I do.
- No, by the way, you stinks!
- You stinks yourself.
Moved to Texas by moderator X, with his little cherished Cap On Top
Thanks SkinnyboySkinnyBoy said:
Well, I didn't actually bother READING the guys post, that would make my post increments too slow..
for that hint.
Why haven't I thought of that myself?
Post without reading previous posting
and certainly not the first post (who cares of the subject
)."GK without killing"
Is this a new "Peace-thread" ???
GrahamnDodder: OK, I've got my parts!
halojoy - think it real great to meet a guy
with some fresh and positive DIY entusiasm
too much "monkey.business" elsewhere:
- I know better than you do.
- No, you don't!
- Yes, I do.
- No, by the way, you stinks!
- You stinks yourself.
Moved to Texas by moderator X, with his little cherished Cap On Top
dont take my laugh - away from me
"don't take my laugh away from me
cause I don't want to live in misery"
/paraphrase from song "Breaking Up Is Hard To do"
* * * * *
I guess moderator
roddyamam
has not got same sense of humour as we have
probably a boring type
I have a feeling that GrahamnDodder himself
would have gotten a good laugh
by our comments
he seems to be such a guy
no harm done
he can read our posts in "Off Topic" Texas-thread
and
get the laugh we wanted him to get
halo - demands freedome of laughs
(see "Marketing Terms" -thread for additional info)
"don't take my laugh away from me
cause I don't want to live in misery"
/paraphrase from song "Breaking Up Is Hard To do"
* * * * *
I guess moderator
roddyamam
has not got same sense of humour as we have
probably a boring type
I have a feeling that GrahamnDodder himself
would have gotten a good laugh
by our comments
he seems to be such a guy
no harm done
he can read our posts in "Off Topic" Texas-thread
and
get the laugh we wanted him to get
halo - demands freedome of laughs
(see "Marketing Terms" -thread for additional info)
Hmm. I will try it both ways, with regard to the pin 1 to 4 cap. All things being equal, I want the amp to be as stable as can be reasonably expected.
I am not going with stepped attenuators - they seemed like too high a percentage of the cost for a first-time-out effort. I went with the same model (cheap-but-decent) panasonic pot I used on the CMoy.
Only problem was they only came up to 50k. I may be overestimating my small-but-growing understanding of this circuit, but let me see if I have this straight.
1) The 50k pot should play louder than the 100k pot at the lower volume settings, and ramp up quicker. I am going on the hope that 50k will still be enough resistance at the volume pot to be as quiet as I would want.
2) Wouldn't lowering the value of the resistor from pin 8 to signal gnd (currently 22k, if I remember right - I'm not at home) compensate if the low volume from the 50k pot is too out of whack?
BTW, Peter, those Boeing 702 sats, with the solar array degadation problem, cost me $150,000.00 US yesterday when XM Satellite Radio's insurance company said they weren't going to cover the problem. Twenty-one percent loss in one day. Ouch. But we have had a much, much bigger run so far this year, so I guess it isn't so bad.
Besides, the insurance guys said the same thing to Thuraya, and then later ponied up $270 million. So we'll see. XM just bought a fourth 702 from Boeing, so I guess they are either otherwise happy with it, or BA has 'em by the short hairs.
GnD
I am not going with stepped attenuators - they seemed like too high a percentage of the cost for a first-time-out effort. I went with the same model (cheap-but-decent) panasonic pot I used on the CMoy.
Only problem was they only came up to 50k. I may be overestimating my small-but-growing understanding of this circuit, but let me see if I have this straight.
1) The 50k pot should play louder than the 100k pot at the lower volume settings, and ramp up quicker. I am going on the hope that 50k will still be enough resistance at the volume pot to be as quiet as I would want.
2) Wouldn't lowering the value of the resistor from pin 8 to signal gnd (currently 22k, if I remember right - I'm not at home) compensate if the low volume from the 50k pot is too out of whack?
BTW, Peter, those Boeing 702 sats, with the solar array degadation problem, cost me $150,000.00 US yesterday when XM Satellite Radio's insurance company said they weren't going to cover the problem. Twenty-one percent loss in one day. Ouch. But we have had a much, much bigger run so far this year, so I guess it isn't so bad.
Besides, the insurance guys said the same thing to Thuraya, and then later ponied up $270 million. So we'll see. XM just bought a fourth 702 from Boeing, so I guess they are either otherwise happy with it, or BA has 'em by the short hairs.
GnD
It sure was - $150K US. Biggest one-day loss ever.
But I am not selling, so it was only a paper loss. The selloff was WAY overblown. In fact, I got $60k of it back today in a partial rebound. It'll be fine. I am up over $500k for the year just on this stock. In the long term, XM is basically a license to print money. Their business model is so good it should be illegal. Buy the stock if it is available in Canada on the TSE.
I'll retire on the stock in two years, I'd bet. Then I can build cool stuff all the time...
My input resistor is 10k and the NFB is 220k. With the 50k pot I should be fine.
Taking the kids to a Sesame Street-themed amusement park in PA this weekend, so I won't get to play with the GC. But I'll be thinking about it - they can't stop me from that!
GnD
But I am not selling, so it was only a paper loss. The selloff was WAY overblown. In fact, I got $60k of it back today in a partial rebound. It'll be fine. I am up over $500k for the year just on this stock. In the long term, XM is basically a license to print money. Their business model is so good it should be illegal. Buy the stock if it is available in Canada on the TSE.
I'll retire on the stock in two years, I'd bet. Then I can build cool stuff all the time...
My input resistor is 10k and the NFB is 220k. With the 50k pot I should be fine.
Taking the kids to a Sesame Street-themed amusement park in PA this weekend, so I won't get to play with the GC. But I'll be thinking about it - they can't stop me from that!
GnD
I got a great Sun Micro SCSI drive holder off Ebay today, like another had built here. But this one was two decks high, so plenty of room for the honker trafo and good sinking.
If I use this case, I'll have to alter it to make it look cooler and get a higher WAF, which is crucial to this project as it will eventually reside, visible, in The Living Room.
Anyway, for $6.00 + $3.50 I could hardly go wrong. Even if I don't use it, it came with a nice power supply with lotsa good caps for cannibalizing. Plus, I got an A/C input module with switch and jack. Other goodies, too.
Geez. I'm hooked.
GnD
If I use this case, I'll have to alter it to make it look cooler and get a higher WAF, which is crucial to this project as it will eventually reside, visible, in The Living Room.
Anyway, for $6.00 + $3.50 I could hardly go wrong. Even if I don't use it, it came with a nice power supply with lotsa good caps for cannibalizing. Plus, I got an A/C input module with switch and jack. Other goodies, too.
Geez. I'm hooked.
GnD
My surplus case.......
http://hk.geocities.com/solomon_choi/GC2.html.
US$ 3.39 only. I used like US $17x something to build this gainclone. I'm waitting for my transformer now.
I had a Cmoy too, and I'm using the pansonic's Y10 for the GC too Cheers! and I'm going to have a switch for two stereo RCA input.
P.S: The site need an hour for recovery from my massive upload.
http://hk.geocities.com/solomon_choi/GC2.html.
US$ 3.39 only. I used like US $17x something to build this gainclone. I'm waitting for my transformer now.
I had a Cmoy too, and I'm using the pansonic's Y10 for the GC too Cheers! and I'm going to have a switch for two stereo RCA input.
P.S: The site need an hour for recovery from my massive upload.
OK, feeling pretty bipolar tonight, given that I have an impasse and an inspiration both rattling around in my head.
First, the problem. I am seeing in several different places that I am supposed to connect the (DC) power ground with the case, assuming of course the case I use conducts electricity. But I also see that I should connect the AC earth to the case in several schematics.
Can these co-exist? Isn't the DC power ground merely a "virtual" ground half-way between the rails? Won't that get me in trouble with an AC earth connected to the same chassis? Or does the transformer isolate them sufficiently? Should I buy some life insurance before I try it?
I asked this question a couple of times in another thread and still did not get an straight answer. What I am seeing in the schematics is not rationalizing in my head. It is kinda important in the, you know, not-killing-myself sense.
Thanks in advance for any answers.
Now, the inspiration. I was designing and making my bridge which, for some reason, has been a difficult part for me. Not the circuit, but the physical implementation.
Anyway, I settled on screwing the MUR860's directly to a piece of wood in two groups of four, using a three-stack of little-big-little washers in between. This is very sturdy, gives me some space to work with (diodes are raised) and is a nice heatsink for each of the diodes. the little washer spaces the big one for airflow, too.
Well, here is my idea. Why not do that with the opamps? You can get thick washers, 3" across and 1/8" thick for next-to-nothing. Sandwiching them with alternating smaller washers gives you the airflow between big-washer "ribs."
Advantages:
1) Cheap. Cheaper than cheap. Knocks out the heatsink component, cost-wise, which can be a decent chunk of the budget if you are not lucky with a surplus find.
2) Variability. You can stack as much mass/surface area as you need.
3) Non-conductivity: Using a separate washer-sink for each channel, screwed to a wood subframe, means that you can use the cheaper, better heat conducting LM3875T's instead of the TF's, with no electrical shorting issues.
Disadvantages:
I dunno. You tell me?
Thanks for the responses.
GnD
First, the problem. I am seeing in several different places that I am supposed to connect the (DC) power ground with the case, assuming of course the case I use conducts electricity. But I also see that I should connect the AC earth to the case in several schematics.
Can these co-exist? Isn't the DC power ground merely a "virtual" ground half-way between the rails? Won't that get me in trouble with an AC earth connected to the same chassis? Or does the transformer isolate them sufficiently? Should I buy some life insurance before I try it?
I asked this question a couple of times in another thread and still did not get an straight answer. What I am seeing in the schematics is not rationalizing in my head. It is kinda important in the, you know, not-killing-myself sense.
Thanks in advance for any answers.
Now, the inspiration. I was designing and making my bridge which, for some reason, has been a difficult part for me. Not the circuit, but the physical implementation.
Anyway, I settled on screwing the MUR860's directly to a piece of wood in two groups of four, using a three-stack of little-big-little washers in between. This is very sturdy, gives me some space to work with (diodes are raised) and is a nice heatsink for each of the diodes. the little washer spaces the big one for airflow, too.
Well, here is my idea. Why not do that with the opamps? You can get thick washers, 3" across and 1/8" thick for next-to-nothing. Sandwiching them with alternating smaller washers gives you the airflow between big-washer "ribs."
Advantages:
1) Cheap. Cheaper than cheap. Knocks out the heatsink component, cost-wise, which can be a decent chunk of the budget if you are not lucky with a surplus find.
2) Variability. You can stack as much mass/surface area as you need.
3) Non-conductivity: Using a separate washer-sink for each channel, screwed to a wood subframe, means that you can use the cheaper, better heat conducting LM3875T's instead of the TF's, with no electrical shorting issues.
Disadvantages:
I dunno. You tell me?
Thanks for the responses.
GnD
Lots of questions
First, another committed XM'er here to. Its a killer product - it has a better uptake rate than crack. Like 70% of the folks who get a free sample go on to pay for it themselves - and less than 2% can quit. (Yeah, I own the stock too).
But to your questions. The washers idea might work IF, and its a big if, you can connect the stack in such a manner that heat is easily transferred. I sorta doubt that it would work, although its a great idea. It can be tested. Build a stack and as you point out, it probably will be inexpensive. Hmm perhaps brazing them together first?
The grounds? If you have a metal case, AC Ground must be connected. In some designs, there is a need to keep the "virtual" ground isolated because it is either above or below - meaning it has potential when compared to ground. In a gainclones case, (AT least of those Ihave seen), the two should be at the same potential. So grounding makes it so. It would require some steps to create another layer of isolation for your own safety otherwise (thats IMHO - my grandfather taught me to have a health respect for electricty). And oh yeah, all of those circuit grounds need to be made at one place, ala star ground.
The one issue might be noise. If it happens, we can tackle that afterwards.
First, another committed XM'er here to. Its a killer product - it has a better uptake rate than crack. Like 70% of the folks who get a free sample go on to pay for it themselves - and less than 2% can quit. (Yeah, I own the stock too).
But to your questions. The washers idea might work IF, and its a big if, you can connect the stack in such a manner that heat is easily transferred. I sorta doubt that it would work, although its a great idea. It can be tested. Build a stack and as you point out, it probably will be inexpensive. Hmm perhaps brazing them together first?
The grounds? If you have a metal case, AC Ground must be connected. In some designs, there is a need to keep the "virtual" ground isolated because it is either above or below - meaning it has potential when compared to ground. In a gainclones case, (AT least of those Ihave seen), the two should be at the same potential. So grounding makes it so. It would require some steps to create another layer of isolation for your own safety otherwise (thats IMHO - my grandfather taught me to have a health respect for electricty). And oh yeah, all of those circuit grounds need to be made at one place, ala star ground.
The one issue might be noise. If it happens, we can tackle that afterwards.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.