Thanks a lot!
Hi Bob,
i received OP-Amps, Boards, and parts for the power supply today professionally packed and in perfect shape. Everything looks great, a big thank you for organizing everything and everyone involved. That must have been an awful lot of logistics involved, i actually wonder how you did it, anyway thanks i will keep everyone informed how the project is proceeding.
Greetings,
Nils
Hi Bob,
i received OP-Amps, Boards, and parts for the power supply today professionally packed and in perfect shape. Everything looks great, a big thank you for organizing everything and everyone involved. That must have been an awful lot of logistics involved, i actually wonder how you did it, anyway thanks i will keep everyone informed how the project is proceeding.
Greetings,
Nils
Step Compensation
Thanks for the tip about step compensation. Until I read your board notes I had not come across step compensation before (my initial grounding dates back to a Martin Colloms book - High Performance Loudspeakers). Can you recommend any further reading?
thanks
Nigel
Thanks for the tip about step compensation. Until I read your board notes I had not come across step compensation before (my initial grounding dates back to a Martin Colloms book - High Performance Loudspeakers). Can you recommend any further reading?
thanks
Nigel
Here's a link to our own Planet10'S site. It's a good article with links to outside sources. http://www.t-linespeakers.org/tech/bafflestep/intro-bds.html
Thanks again Bob. The site is also interesting - takes me back to my first speakers I made - a coffin sized TL using the same drivers as the IMF - which I never realised was Bob Fried's creation.
Well, probably I'm at the farthest point geographically; I received the boards yesterday.
Thanks a lot, Bob.
Erhan
Thanks a lot, Bob.
Erhan
I'm glad that you got them, Erhan. I was worried that they were at the airport when the fire struck. In this case slower was better. 😉
What's everyone putting these in? Does anyone have good suggestions for something from Hammond or Par-Metal?
Par Metal's 20 series will work well if you are going the prefab route. They'll be cheaper than Hammond. Notice the date on the Par-Metal price list before you get too excited, though. I bought a 20-16125 for another project and it was ~$69 instead of the $49 listed.
If you are using the group buy power supply and heat sink remember to allow for its height. That's 2" for the sink, and maybe 1/2" for the board and standoff (3/8" standoff with a bit of clearance above.)
My metal working skills leave a bit to be desired, so I will have Front Panel Express make up my rear panel. It will likely cost more than the enclosure, but at least the jacks will be in a straight line.
If you are using the group buy power supply and heat sink remember to allow for its height. That's 2" for the sink, and maybe 1/2" for the board and standoff (3/8" standoff with a bit of clearance above.)
My metal working skills leave a bit to be desired, so I will have Front Panel Express make up my rear panel. It will likely cost more than the enclosure, but at least the jacks will be in a straight line.
Hi
A question from a newby in Opamp design.
The board uses per HP or LP filterstage 2 opamps, like e.g. in the HP section IC4A and IC4B (see the manual).
As far as I can see the second opamp (4B) has no function. Wouldn't it be better to couple the output of 4A directly to the next stage (IC5b)?
The same goes of course for 5A.
Second question: to eliminate DC offset I gather that the DC impedances at the inputs of the opamps should be equal.
Does this also go for the OPA2134 with a FET input? I understand it is less critical in that case, but can I neglect a mismatch?
If not how to deal with the output impedance of the 2134 (being practically 0)
Dick.
A question from a newby in Opamp design.
The board uses per HP or LP filterstage 2 opamps, like e.g. in the HP section IC4A and IC4B (see the manual).
As far as I can see the second opamp (4B) has no function. Wouldn't it be better to couple the output of 4A directly to the next stage (IC5b)?
The same goes of course for 5A.
Second question: to eliminate DC offset I gather that the DC impedances at the inputs of the opamps should be equal.
Does this also go for the OPA2134 with a FET input? I understand it is less critical in that case, but can I neglect a mismatch?
If not how to deal with the output impedance of the 2134 (being practically 0)
Dick.
Dick,
Jens answered your question early on in this thread. Because the filters use the equal component value Sallen-Key topology you can have quite a bit of gain in each filter stage. Note that the feed to the buffer is taken from the negative input. this means that the signal is attenuated by the amount of gain in the stage. This prevents overloading the next stage input.
The second (B) half of each opamp is a buffer that prevents interaction between the feedback loop and the following stage filter.
If you use the unity gain topology (for example R14=0 and R19=open) you could bypass the interstage buffers with a little surgery on the board. If you don't need to adjust the level of a particular filter, you can take a buffered output from JP8 pin 1 or 3 depending on filter order.
I've found that matching impedances to each input doesn't have much affect. Ideally you would, but if you keep impedances in the recommended range (a few Kohms to <100Kohms) offset is manageable. Linkwitz doesn't appear to pay too much attention to it on his boards, either.
Jens answered your question early on in this thread. Because the filters use the equal component value Sallen-Key topology you can have quite a bit of gain in each filter stage. Note that the feed to the buffer is taken from the negative input. this means that the signal is attenuated by the amount of gain in the stage. This prevents overloading the next stage input.
The second (B) half of each opamp is a buffer that prevents interaction between the feedback loop and the following stage filter.
If you use the unity gain topology (for example R14=0 and R19=open) you could bypass the interstage buffers with a little surgery on the board. If you don't need to adjust the level of a particular filter, you can take a buffered output from JP8 pin 1 or 3 depending on filter order.
I've found that matching impedances to each input doesn't have much affect. Ideally you would, but if you keep impedances in the recommended range (a few Kohms to <100Kohms) offset is manageable. Linkwitz doesn't appear to pay too much attention to it on his boards, either.
For the enclosure I am still in doubt if i want to build something myself.I have acces to sheet metal tools like a bending press etc. but for the price it is probably wise to buy a HTPC enclosure and adapt it a little.
To get a homemade cabinet to look like this http://www.silverstonetek.com/products-lc19.htm or similar will cost a lot more in material/filler/paint etc.
Kees.
To get a homemade cabinet to look like this http://www.silverstonetek.com/products-lc19.htm or similar will cost a lot more in material/filler/paint etc.
Kees.
Alternative Case
Another alternative is ATI Research http://www.atiresearch-anodized.com/ whose cases are audio specific. The Silverstone cases seem to be all geared towards computers (at least from what I saw when browsing the site).
I was planning on using ATI myself. There have been a number of posts on this site from people that have used them.
Another alternative is ATI Research http://www.atiresearch-anodized.com/ whose cases are audio specific. The Silverstone cases seem to be all geared towards computers (at least from what I saw when browsing the site).
I was planning on using ATI myself. There have been a number of posts on this site from people that have used them.
Bob, its really easy to drill the holes in the aluminum chassis (just not square holes).
Buy a multilevel drill bit from mcmaster, it will be about $30, but its cheaper than buying a million bits, and cheaper than front panel express service. Also buy a small file.
Draw your layout from the inside, then punch each hole. Then drill a small pilot hole, and all your stuff will end up neat in a line. Then just use the multilevel bit to get to your desired hole size. Deburring will be necessary, either drill slowly and carefully with the multilevel bit on the other side, or use a file. Keys are easy to make, just use the edge of a small file.
I made some wooden blocks to set this stuff on while I drilled, and covered them with some spare felt so it didn't get scratched.
Anybody that saw a picture of my stuff, its only sloppy because I used a messy glue to keep some of the snap in connectors from jiggling. This is also a good point, its hard to get snap in connectors to fit snugly with hand drilled holes.
No replies to me on this thread, its too popular for me to subscribe.
Buy a multilevel drill bit from mcmaster, it will be about $30, but its cheaper than buying a million bits, and cheaper than front panel express service. Also buy a small file.
Draw your layout from the inside, then punch each hole. Then drill a small pilot hole, and all your stuff will end up neat in a line. Then just use the multilevel bit to get to your desired hole size. Deburring will be necessary, either drill slowly and carefully with the multilevel bit on the other side, or use a file. Keys are easy to make, just use the edge of a small file.
I made some wooden blocks to set this stuff on while I drilled, and covered them with some spare felt so it didn't get scratched.
Anybody that saw a picture of my stuff, its only sloppy because I used a messy glue to keep some of the snap in connectors from jiggling. This is also a good point, its hard to get snap in connectors to fit snugly with hand drilled holes.
No replies to me on this thread, its too popular for me to subscribe.
Re: Alternative Case
I have some ATI cases - if you are just going to buy one or two, it'll be rather expensive, since they don't take paypal or credit cards. My bank charged me $45 to wire the funds to Thailand when I got my amp cases last winter.
Nice cases, by the way, but you need to be buying more than one to justify the overhead.
Peter
nigeljking said:
I have some ATI cases - if you are just going to buy one or two, it'll be rather expensive, since they don't take paypal or credit cards. My bank charged me $45 to wire the funds to Thailand when I got my amp cases last winter.
Nice cases, by the way, but you need to be buying more than one to justify the overhead.
Peter
The reason i want to go for a HTPC chassis is the price.There is more choice and they are made in much higher numbers.This keeps the price down.I don't want to pay more for the enclosure than for its contents.
The fact that they are made for pc use doesn't bother me much,just knock out the back and replace it with a panel drilled to your own specs.
Kees.
The fact that they are made for pc use doesn't bother me much,just knock out the back and replace it with a panel drilled to your own specs.
Kees.
BobEllis said:Odd order filters were not covered in the manual. There are a few things I'd like to add. I guess I need another trip with long airport layovers to update it.
Here is a way to make an odd order filter on the AF4 board.
In case the notation is not obvious, the part numbers missing the component symbol are omitted, and the red lines indicate jumpers. Low pass is similar.
For the lowpass, would you please specify the components that you would omit and jumper? I would just like to make no mistake.
I'll second that notion. I especially like the Silverstone cases mentioned previously, but many of them come with specialized PSU's that I hate to simply throw out. I guess you could always eBay it. . . 😀kro5998 said:
For low pass the only thing that changes is the position of the remaining R and C swap positions - There is a resistor where C4 is shown and a cap where R7 is shown. The jumpers and omissions remain the same.
BobEllis said:Odd order filters were not covered in the manual. There are a few things I'd like to add. I guess I need another trip with long airport layovers to update it.
Here is a way to make an odd order filter on the AF4 board.
In case the notation is not obvious, the part numbers missing the component symbol are omitted, and the red lines indicate jumpers. Low pass is similar.
So for this high pass example, could we also bypass IC4B?
- Status
- Not open for further replies.