Thanks Jens,
I had wondered about that. I had planned to use double-stick foam tape under them anyway to help keep everything in place.
Do these things get hot under normal use? Maybe the foam tape won't work well. How will foil help to stop ground issues? Won't it conduct electricity as well?
Thanks, Terry
I had wondered about that. I had planned to use double-stick foam tape under them anyway to help keep everything in place.
Do these things get hot under normal use? Maybe the foam tape won't work well. How will foil help to stop ground issues? Won't it conduct electricity as well?
Thanks, Terry
Terry,
During normal use, your caps should get warm (not really hot) at worst. If they get HOT, something is wrong.
Generally, things in your amp should only get warm. The warmest part wil probably be your output power transistors, but you should not burn your fingers on them (if so, your heat sink is too small).
With foil, I was thinking of insulating plastic foil.
Regarding foam tape, I think I know what you mean. I think it would be a fine supplement, especially if held in place (squeezed) by the top grounding plate.
You sound like you're already getting the hang of it, and thinking like a trained DIY'er. 😀
Jennice
During normal use, your caps should get warm (not really hot) at worst. If they get HOT, something is wrong.
Generally, things in your amp should only get warm. The warmest part wil probably be your output power transistors, but you should not burn your fingers on them (if so, your heat sink is too small).
With foil, I was thinking of insulating plastic foil.
Regarding foam tape, I think I know what you mean. I think it would be a fine supplement, especially if held in place (squeezed) by the top grounding plate.
You sound like you're already getting the hang of it, and thinking like a trained DIY'er. 😀
Jennice
Thanks Jens,
I'll stick with the tape then.
Have you had a chance to try the resistors on the filter caps yet? I'm thinking of starting out using 8 caps and arranging them like this;
except with two additional caps for the third channel.
I've go another question. On my Hafler amp there is a cap across the rectifier. What purpose does it have and should I use one on my new amp?
Thanks, Terry
I'll stick with the tape then.
Have you had a chance to try the resistors on the filter caps yet? I'm thinking of starting out using 8 caps and arranging them like this;
except with two additional caps for the third channel.
I've go another question. On my Hafler amp there is a cap across the rectifier. What purpose does it have and should I use one on my new amp?
Thanks, Terry
sounds like a snubber, you should have some form of snubber arrangment on the recitfer(s) , im still undecided what to do myself ,
this link http://www.hagtech.com/pdf/snubber.pdf has been floating round the forums for a while, its pretty full on, but describes the need for it,
it is suggests in there to use a resistor and capacitor (RC) , what im unsure of myself is wheather you use 1 of these RC filters per rectifier or for every diode
this link http://www.hagtech.com/pdf/snubber.pdf has been floating round the forums for a while, its pretty full on, but describes the need for it,
it is suggests in there to use a resistor and capacitor (RC) , what im unsure of myself is wheather you use 1 of these RC filters per rectifier or for every diode
If I understand mAJORD's reference, a capacitor alone as a snubber may make the ringing worse. It says you'll need an RC across each diode, whose value is determined by measuring inductance and stray capacitance of the transformer. IIRC, it does offer a typical range of values that should help, and won't hurt if you cannot measure the transformer.
I was told eons ago that the single cap across the ac terminals of the bridge (ala Hafler) shunts away switching noise of the diodes. It doesn't seem to make sense, since it creates an LC tank with the transformer inductance - more a resonator than a damper.
That said, both my Leach amp and A-75 DO NOT use any form of snubber and the rails show no sign of ringing on my noisy old 100 MHz scope. An amp with decent power supply rejection and good quality high frequency bypass capacitors probably does not need it. I think there is a thread discussing the reference above.
I was told eons ago that the single cap across the ac terminals of the bridge (ala Hafler) shunts away switching noise of the diodes. It doesn't seem to make sense, since it creates an LC tank with the transformer inductance - more a resonator than a damper.
That said, both my Leach amp and A-75 DO NOT use any form of snubber and the rails show no sign of ringing on my noisy old 100 MHz scope. An amp with decent power supply rejection and good quality high frequency bypass capacitors probably does not need it. I think there is a thread discussing the reference above.
Terry,
I thought you went for the 2 chaqnnel solution, considering your heat sink size!?
Anyway.. I haven't had time to really progress with my project, but I came to think of your ground thing to tie down (physically) your caps. Are your caps with threaded binding post terminals, or solder pins. if solder pins, you can't solder them to a big, solid grounding plane without serious risk of damaging the caps.
Regarding snubbers, I have never used them as my philosophy has been to keep the PSU simple and bulky. Also, I believe more in the RC filtering with caps (even if it requires a lot of them), than I believe in the progress when changing the rectifier noise waveform. The vast majority of PSU ripple doesn't come from the rectifier diode ringing, but from the peak currents and charging/discharging cycles on the caps (humming at 2x line frequency).
In reality, a capacitor bank is a complex series of ideal resistors, ideal caps and ideal inductors. I don't think there's a need to make things even more complex.
Just my 2 cents -
Jennice
I thought you went for the 2 chaqnnel solution, considering your heat sink size!?
Anyway.. I haven't had time to really progress with my project, but I came to think of your ground thing to tie down (physically) your caps. Are your caps with threaded binding post terminals, or solder pins. if solder pins, you can't solder them to a big, solid grounding plane without serious risk of damaging the caps.
Regarding snubbers, I have never used them as my philosophy has been to keep the PSU simple and bulky. Also, I believe more in the RC filtering with caps (even if it requires a lot of them), than I believe in the progress when changing the rectifier noise waveform. The vast majority of PSU ripple doesn't come from the rectifier diode ringing, but from the peak currents and charging/discharging cycles on the caps (humming at 2x line frequency).
In reality, a capacitor bank is a complex series of ideal resistors, ideal caps and ideal inductors. I don't think there's a need to make things even more complex.
Just my 2 cents -
Jennice
Hi,
I see that users are still basing their grounding scheme on the post 143 diagram.
I strongly suggest you DO NOT use multiple grounding as shown on this diagram!!!!
The peak / pulse currents flowing between the capacitors through this plate will contaminate your audio signal.
If you insist on using the cap common as your clean and dirty ground then use the Sugden trick:- put a brass ( or better still a copper) bolt through the plate and nut it up firmly. Connect all your clean and dirty grounds to the other side of the nut, then they are isolated from the pulse currents between the caps. or connect a single wire from the cap common to a remote audio star ground.
regards Andrew T.
I see that users are still basing their grounding scheme on the post 143 diagram.
I strongly suggest you DO NOT use multiple grounding as shown on this diagram!!!!
The peak / pulse currents flowing between the capacitors through this plate will contaminate your audio signal.
If you insist on using the cap common as your clean and dirty ground then use the Sugden trick:- put a brass ( or better still a copper) bolt through the plate and nut it up firmly. Connect all your clean and dirty grounds to the other side of the nut, then they are isolated from the pulse currents between the caps. or connect a single wire from the cap common to a remote audio star ground.
regards Andrew T.
Andrew,
This Bolt think is basically what Rod Elliot is doint. It is probably not without reason, that he has placed his transformer terminals on the middle of the plane, letting the currents flow as a star ground.
Jennice
This Bolt think is basically what Rod Elliot is doint. It is probably not without reason, that he has placed his transformer terminals on the middle of the plane, letting the currents flow as a star ground.
Jennice
ive used a bolt in the center of 4 caps, with all the negs from the caps, the amp ground, speaker return, chassis link, transformer centre tap, all soldered to eye terminals, then attached to the bolt as 1, with a bath of solder between them all.
using a bolt means i remove/add things to the star point easily as i experiment / get adivce from others on what i should be connnecting there .
cheers
using a bolt means i remove/add things to the star point easily as i experiment / get adivce from others on what i should be connnecting there .
cheers
So did I on my previous project. On this one, I haven't decided yet, but probably end up with something similar. I'm a little squeezed on build height, but I hope I'll think of something.
Jennice
Jennice
Jennice said:
Hi Jens,
I am going to try to still do 3 channels. I figure if I have a problem with heat, it will be a simple thing to just remove one of the PCB's and it's Filter caps. I could also get larger heatsinks if necessary. I really don't think I will be running that hard anyway.
My caps are the snap-in type, so they have the little pins that I will have to solder. Are you worried that I will need too much heat to solder them to the plate? I had figured to pre-tin the plate. I haven't tried that yet, but hope that it will work.
I will have plenty of room to put a bolt in the center of the plate if necessary. My caps are fairly short. Much shorter than my transformer.
I wasn't sure if there would be too much resistance by using a bolt over direct soldering. If it's just as good to use the bolt, I would much prefer that. For that matter I could use a center bolt for the hold-down and just extend it up past the plate and use that for the star ground.
Hi BobEllis,
You said "An amp with decent power supply rejection and good quality high frequency bypass capacitors probably does not need it."
How do I know if the ESP P101 has these features? Is this something I need to do aside from what Rod Elliot has already provided in his design?
Thanks guys, Terry
hi Terry
the constant current source (Q, Q2 and associated parts) and the Q5, Q6 current mirror in the high power versionwill provide good power supply rejection. there are other ways to get it, but those work. a mirror is not as good as a CCS, hence C6 - a big electrolytic on the positve rail but not the negative.
I presume that C7 and C8 are specified as film types. If not, bypassing them with a 100nf or so film cap of suitable voltage rating will provide extra HF bypass. (electrolytic caps impedance rises with frequency, small film caps are better for hf.)
I had trouble heating up a 2oz pcb with 1.25" wide traces to solder my caps in place with a 100 watt gun. Enough that I ended up burning out the gun by having it on too long. If you use a gun, I'd let it cool 15 minutes or longer after each cap or use an iron rated for continuous use. Copper makes a good heat sink. With your strap you may need to go to a propane torch powered iron. Practice soldering wires onto a scrap piece. so that you can solder your caps without overheating them.
Bob
the constant current source (Q, Q2 and associated parts) and the Q5, Q6 current mirror in the high power versionwill provide good power supply rejection. there are other ways to get it, but those work. a mirror is not as good as a CCS, hence C6 - a big electrolytic on the positve rail but not the negative.
I presume that C7 and C8 are specified as film types. If not, bypassing them with a 100nf or so film cap of suitable voltage rating will provide extra HF bypass. (electrolytic caps impedance rises with frequency, small film caps are better for hf.)
I had trouble heating up a 2oz pcb with 1.25" wide traces to solder my caps in place with a 100 watt gun. Enough that I ended up burning out the gun by having it on too long. If you use a gun, I'd let it cool 15 minutes or longer after each cap or use an iron rated for continuous use. Copper makes a good heat sink. With your strap you may need to go to a propane torch powered iron. Practice soldering wires onto a scrap piece. so that you can solder your caps without overheating them.
Bob
Terry,
Yes, I am seriously concerned for the life of your capacitors.
My caps have solter pins, too. I have planned to hard-wire them together, and then let the middle (half way) be connected to a bolt, or just let it be a solder-on star ground point. I have also considered a strip of PCB, but (as BobEllis mentions) copper is a very good heat sink, and I fear that even if pre-soldered, it will still need too much heat energy for your caps to like it.
Jennice
Yes, I am seriously concerned for the life of your capacitors.
My caps have solter pins, too. I have planned to hard-wire them together, and then let the middle (half way) be connected to a bolt, or just let it be a solder-on star ground point. I have also considered a strip of PCB, but (as BobEllis mentions) copper is a very good heat sink, and I fear that even if pre-soldered, it will still need too much heat energy for your caps to like it.
Jennice
OK, I will give it a try with some wire or something before I try to solder the caps in place. I'm just thinking that if I pre-tin the plate so that the solder is already flowed onto it, that it will take very little heat to flow the solder onto the cap pin. I may be wrong about this so I will try with something else first.
Aside from the heat, I think once I have them bolted down they should be more secure than they would be if they were just soldered to a PCB which is how I see most of them being used.
At to the heatsink, I have purchased another one just like the first so I can make them as big as I need. (Well up to 11" tall anyway)
see here
Do you know any way I can calculate how big I need?
Thanks again for all your help.
Blessings, Terry
Aside from the heat, I think once I have them bolted down they should be more secure than they would be if they were just soldered to a PCB which is how I see most of them being used.
At to the heatsink, I have purchased another one just like the first so I can make them as big as I need. (Well up to 11" tall anyway)
see here
Do you know any way I can calculate how big I need?
Thanks again for all your help.
Blessings, Terry
See page 4 of http://www.passdiy.com/pdf/a75p2.pdf for an explanation of how to calculate your heat sink requirement.
Of course you'll need to figure out how much heat you'll have to dissipate. That's mentioned several places here, but IIRC worst case dissipation occurs at roughly 2/3 max power for a class AB amp like the p101. Multiply that current by the rail voltage and you've got an approximation of how much heat you need to get rid of. Divide that number by your accepatble temperature rise and you'll determine the required rating of your sinks if you run continuously at worst case power.
Find a similar section at AAVID, Wakefield, M&M, Conrad etc. and estimate the rating of your sink. Unfortunately, doubling the height if the sink does not double the heat rejection capability (halve the K/W rating). Nelson Pass mentioned in one of the threads that it's roughly 1.4/ doubling of length so your 11 inch sinks would be about double the performance of the 3" section used for the rating in many catalogs. AAVID has a calculator on their website that allows you to plug in the length and determine the perfomance.
You'll see that you probably want to use one of your heat sinks per channel following this method of calculating the requirement. Cutting them to half height and making the amp deep will improve performance. Now you're probably thinking, "Hey wait a minute, Bob. You told me one heat sink would be plenty for three channels. What gives?"
You will be running music through the amp and you won't be anywhere near worst case dissipation if you want to retain your hearing. That's why so many of the chip amp guys get away with small heat sinks or calling a big block of aluminum without fins their heat sink. So you can get away with smaller heat sinks than the number crunching suggests for home use. (Unless you are running class A, then you need every bit of heat sink you calculate then another 30% fudge factor since worst case is idle.)
So for a class AB amp, the heat sink requirement boils down to an educated guess, balancing your budget, sense of aesthetics and taste for ear bleeding volume to get something that will stay cool enough to be reliable. If you provide as much heat sink as a typical commercial amp of similar power rating you ought to be fairly safe. the commercial guys don't want warranty claims, so they'll lean towards spending more on heat sinks.
Sorry for being so long winded, but I hope this helps.
Of course you'll need to figure out how much heat you'll have to dissipate. That's mentioned several places here, but IIRC worst case dissipation occurs at roughly 2/3 max power for a class AB amp like the p101. Multiply that current by the rail voltage and you've got an approximation of how much heat you need to get rid of. Divide that number by your accepatble temperature rise and you'll determine the required rating of your sinks if you run continuously at worst case power.
Find a similar section at AAVID, Wakefield, M&M, Conrad etc. and estimate the rating of your sink. Unfortunately, doubling the height if the sink does not double the heat rejection capability (halve the K/W rating). Nelson Pass mentioned in one of the threads that it's roughly 1.4/ doubling of length so your 11 inch sinks would be about double the performance of the 3" section used for the rating in many catalogs. AAVID has a calculator on their website that allows you to plug in the length and determine the perfomance.
You'll see that you probably want to use one of your heat sinks per channel following this method of calculating the requirement. Cutting them to half height and making the amp deep will improve performance. Now you're probably thinking, "Hey wait a minute, Bob. You told me one heat sink would be plenty for three channels. What gives?"
You will be running music through the amp and you won't be anywhere near worst case dissipation if you want to retain your hearing. That's why so many of the chip amp guys get away with small heat sinks or calling a big block of aluminum without fins their heat sink. So you can get away with smaller heat sinks than the number crunching suggests for home use. (Unless you are running class A, then you need every bit of heat sink you calculate then another 30% fudge factor since worst case is idle.)
So for a class AB amp, the heat sink requirement boils down to an educated guess, balancing your budget, sense of aesthetics and taste for ear bleeding volume to get something that will stay cool enough to be reliable. If you provide as much heat sink as a typical commercial amp of similar power rating you ought to be fairly safe. the commercial guys don't want warranty claims, so they'll lean towards spending more on heat sinks.
Sorry for being so long winded, but I hope this helps.
Terry,
You are right that many home applications use caps with solder pins, seemingly without any other means of fastening them.
Yes, some actually do nothing, while others use hot melt at the base, or double sided adhesive tapes between the base and PCB.
However, adhesives are no good over time for holding against vibrations, which is really what kills caps. There's normally no problem taking the weight in two solder pins, but once you start vibrating it (mostly in industrial equipment, and during transport/shipping of all electronic goods), the damage happens.
Jennice
p.s.: I'm starting to wonder how you find time to actually build anything with all the posts you spend time replying to. 😀
You are right that many home applications use caps with solder pins, seemingly without any other means of fastening them.
Yes, some actually do nothing, while others use hot melt at the base, or double sided adhesive tapes between the base and PCB.
However, adhesives are no good over time for holding against vibrations, which is really what kills caps. There's normally no problem taking the weight in two solder pins, but once you start vibrating it (mostly in industrial equipment, and during transport/shipping of all electronic goods), the damage happens.
Jennice
p.s.: I'm starting to wonder how you find time to actually build anything with all the posts you spend time replying to. 😀
well I just fired mine up on full power 😀 , having some trouble setting the bias voltage on 1 channel, well not so much trouble but with the trimpot on 0ohm im getting about 0.5ish volts across the 100ohm resistors, so i adjusted it to 2v greater than this as stated, and while it sounds ok, im not happy that its 2v different the the other channel (it was 2v on minimum as expected)
it sounds great in general 🙂 , more tweaking to do yet tho of course,
i can't get over how much better it sounds with the final PSU, (2x 300va 40-0-40 + 8 x 4000uf ) , as oposed to the single 18-0-18 tranny+ 4x 4000uf) i was using for initial board testing, incredible. even at moderate volumes. So terry you can be safe in the knowledge your beefcake power supply will be worth it.
once I:
get a scope!, finish my proac 2.5 clones, BG caps + OPA2134s for preamp arive , and get a good source happening I can expermient with with a few things and get some meaningfull/helpfull results to post.
it sounds great in general 🙂 , more tweaking to do yet tho of course,
i can't get over how much better it sounds with the final PSU, (2x 300va 40-0-40 + 8 x 4000uf ) , as oposed to the single 18-0-18 tranny+ 4x 4000uf) i was using for initial board testing, incredible. even at moderate volumes. So terry you can be safe in the knowledge your beefcake power supply will be worth it.
once I:
get a scope!, finish my proac 2.5 clones, BG caps + OPA2134s for preamp arive , and get a good source happening I can expermient with with a few things and get some meaningfull/helpfull results to post.
Hi BobEllis,
Thanks for the detailed reply. I will check out the sites you mentioned. I may decide to make the amp deeper. Funny thing about this heatsink thing. I have a 200 watt Fender Bass amp that has a heatsink about 2" X 12" and it barely get warm when driving it pretty hard. My Hafler P230s however get very warm just playing at a moderate level. Do MOSFETs run hotter than FETs?
Hi Jens,
It's not just home applications. My guitar and bass amps have the caps just mountd on PCBs as well as my Carvin power amp and Hafler TA-1100 power amp.
I'm planning to use double stick foam tape at the base and bolt down the strap on top of them. I can't see there being much vibration or movement with those thing done.
I know it seems like I'm here a lot but that's not really so. I mostly come here while at work.
We have a saying, "Measure twice, cut once". I'm finding that the actual fitting and soldering of the parts is the quickest part of DIY. The learning and planning seems to take the lion share the time, IMO.
I'm trying to get a really good picture in my head before I start cutting down heatsinks and building cases and such, so I don't have to redo too many things.
You guys have been extreemly helpful and I much appreciate it.
mAJORD,
It's great to hear how you're progressing. Gives me a lot of hope.
Thanks everyone, Terry
Thanks for the detailed reply. I will check out the sites you mentioned. I may decide to make the amp deeper. Funny thing about this heatsink thing. I have a 200 watt Fender Bass amp that has a heatsink about 2" X 12" and it barely get warm when driving it pretty hard. My Hafler P230s however get very warm just playing at a moderate level. Do MOSFETs run hotter than FETs?
Hi Jens,
It's not just home applications. My guitar and bass amps have the caps just mountd on PCBs as well as my Carvin power amp and Hafler TA-1100 power amp.
I'm planning to use double stick foam tape at the base and bolt down the strap on top of them. I can't see there being much vibration or movement with those thing done.
I know it seems like I'm here a lot but that's not really so. I mostly come here while at work.
We have a saying, "Measure twice, cut once". I'm finding that the actual fitting and soldering of the parts is the quickest part of DIY. The learning and planning seems to take the lion share the time, IMO.
I'm trying to get a really good picture in my head before I start cutting down heatsinks and building cases and such, so I don't have to redo too many things.
You guys have been extreemly helpful and I much appreciate it.
mAJORD,
It's great to hear how you're progressing. Gives me a lot of hope.
Thanks everyone, Terry
Has anyone tried silicone seal to hold the caps to the board? I've used it for many other things before as a kind of glue that absorbs vibration and can be removed if necessary, and also to waterproof some electrical connections. I think it would hold up and work better than double sided sticky tape. Any negative to this?
Silicone seal from a tube will be just fine - I know it's used in industrial applications, too. I think it's just that the tape is easier and cleaner to work with. Also, the tape doesn't need curing time before it can hold. As I understand it, Terry will be using the tape to hold the caps in place before clamping them down.
Terry,
Regarding the heat of your amps, you will have to consider the duty cycle and frequency range.
Duty Cycle:
Some of the time your bass amp is not playing (there are short breaks in the music signal), which is almost never the case for a more complex music signal.
Frequency range:
Your bass produces a given sound pressure level (SPL) in the lowest audio octaves. Even though a given SPL requires less energy at higher frequencies, the energy needed to produce the same SPL over the entire audio band is higher than for bass-only applications.
Besides: If your amps get warm in idle, too, it's a matter of settiong the idle current. A bass amp can do with little idle current as possible cross-over distortion from the transistors is effectively removed by the low-pass filter function of the bass woofer itself.
Jennnice
Terry,
Regarding the heat of your amps, you will have to consider the duty cycle and frequency range.
Duty Cycle:
Some of the time your bass amp is not playing (there are short breaks in the music signal), which is almost never the case for a more complex music signal.
Frequency range:
Your bass produces a given sound pressure level (SPL) in the lowest audio octaves. Even though a given SPL requires less energy at higher frequencies, the energy needed to produce the same SPL over the entire audio band is higher than for bass-only applications.
Besides: If your amps get warm in idle, too, it's a matter of settiong the idle current. A bass amp can do with little idle current as possible cross-over distortion from the transistors is effectively removed by the low-pass filter function of the bass woofer itself.
Jennnice
- Status
- Not open for further replies.