They are Conrad 100mm x 150.5mm, eight per amp
http://www.conradheatsinks.com/products/flat100_350.html#MF10
http://www.conradheatsinks.com/products/flat100_350.html#MF10
The stuffed board here.
I used this board design because it allows flexibility without the need for remote mounting of other parts.
I also like to use a largish chunk of aluminium between the Diff pair as think this helps thermal stability.
The intention is to add some gold pins for the compensation caps so I can swap in and out values as required without lifting the board when I get to the testing page.
I used this board design because it allows flexibility without the need for remote mounting of other parts.
I also like to use a largish chunk of aluminium between the Diff pair as think this helps thermal stability.
The intention is to add some gold pins for the compensation caps so I can swap in and out values as required without lifting the board when I get to the testing page.
Attachments
Here is another image showing the poposed location of the final leg of the CRCRC filter capacitors.
The close proximity and local bypassing of the main 22,000 uf Pansonics with 2200 uf F series should yeild good results.
I am using 15 watt 0.45R/0.47R in parellel power resisters for the filters. According to the simulations ir takes a few hundred millseonds to charge the supply. While I will be using a soft start circuit I think the design will act as a soft start in this respect.
A sheet of aluminium will be placed under the board to minimise any likely noise from the rectifiers in the bottom of the chassis.
The close proximity and local bypassing of the main 22,000 uf Pansonics with 2200 uf F series should yeild good results.
I am using 15 watt 0.45R/0.47R in parellel power resisters for the filters. According to the simulations ir takes a few hundred millseonds to charge the supply. While I will be using a soft start circuit I think the design will act as a soft start in this respect.
A sheet of aluminium will be placed under the board to minimise any likely noise from the rectifiers in the bottom of the chassis.
Attachments
I have not had any issue with this approach and I think it helps with temp stability overall and absolute offset.
The heatsink in the commercial product is also reasonably substantial.
But it depends on the environment and how you want to look at temperature cause and effect thing.
I am going out on the limb here but its like the tail wagging the dog in many respects if you allow one side of the diff pair to immediately influenced by other or environmental influences.
Again, my own inclination is that the input diff pair move and track the temperature of the amp overall, not micro changes as such.
The biggest issue is adjusting with the lid off and running with the lid on.
As I understand it they had a special lid in production with holes in it for the tweaking and adjustment over time (hours).
The heatsink in the commercial product is also reasonably substantial.
But it depends on the environment and how you want to look at temperature cause and effect thing.
I am going out on the limb here but its like the tail wagging the dog in many respects if you allow one side of the diff pair to immediately influenced by other or environmental influences.
Again, my own inclination is that the input diff pair move and track the temperature of the amp overall, not micro changes as such.
The biggest issue is adjusting with the lid off and running with the lid on.
As I understand it they had a special lid in production with holes in it for the tweaking and adjustment over time (hours).
Zen Mod said:
Suppose the diff. pair was mounted to a giant size block of aluminium with infinitely low C/W.
It would take a lifetime to reach thermal equilibrium, but the temp difference between now and the next millenium would be a tiny fraction of 1C.
A small thin piece of tin requires only a small amount of energy to increase it's temperature by 1C if it is thermally insulated from the surrounding.
But it isn't insulated , and every doubling of dT with reference to ambient doubles the Joules per second transfer capability of the heatsink.
Third possibility: a giant size thin plate with an extremely low C/W but close to no mass.
Must be a joy to live nextdoor to the Bauxite mine, Macka.
jacco vermeulen said:
My father designed alumina plants as a matter of fact.
Thermally what is the primary influence...or what d you want the primary influence to be?
As I mentioned above its a case of keeping the air temperature stable around the diff pair and the CCS to keep the bias stable, the current share and offset(s) stable.
I think too many people try and make adjustment taking lid on and off and then have to wait for everything to settle down..take another glass of red.
I see the amp as a temperature controlled oven. The dc offset is not a problem, anything under 100 mv is quite acceptable, As I recall it was stable within -+10 mv once the amp was running for an hour.
I propose to make an adjustment lid with small holes for plastic tweakers..Muhhahaah
The absolute offset drift is more effected by other issues imho
Attachments
macka said:
My father designed alumina plants as a matter of fact.
Thermally what is the primary influence...or what d you want the primary influence to be?
As I mentioned above its a case of keeping the air temperature stable around the diff pair and the CCS to keep the bias stable, the current share and offset(s) stable.
I think too many people try and make adjustment taking lid on and off and then have to wait for everything to settle down..take another glass of red.
I see the amp as a temperature controlled oven. The dc offset is not a problem, anything under 100 mv is quite acceptable, As I recall it was stable within -+10 mv once the amp was running for an hour.
I propose to make an adjustment lid with small holes for plastic tweakers..Muhhahaah
The absolute offset drift is more effected by other issues imho
correct .
in fact- after some thinking ,I agree that approach with massive piece is at least equally good as tin bridge ; just small addendum- with big chunk - is of more importancy to have tighter matched output mosfets ; just to be sure that initial offset is in ballpark
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.