Here are a couple more pics of the heatsinks:
Whole thing (1 of 2):
Diode On Heatsink:
Leach Drill Template on Heatsink:
So, I am concerned about the width of the heatsink base being able to accomidate the diodes. The heatsink is 3.2mm think at the base and the diodes are 5mm long. Just above that, the heatsink becomes 7mm thick, so a possibility would be to edit the base of some of the fins and mount diodes over the devices.
I was also thinking of using a heattransferring epoxy to mount them, which might help. Right now, i am leaning toward one diode mounted directly above each device (for thinkness). Backup is on diode beside each device with heat epoxy.
I am also concerned about my hands getting extremely tired holding the heatsinks in place. Maybe for a song or two, but it will get old quickly ;-)
I am using TO3's. The amp originally had Toshiba flat paks (4 pairs of 2sc2565/2sc1095). As you can see by the template pic, the sinks actually have plenty of room to accomidate the transistors, which will be spaced out much like the old Toshibas were. I may make a little more room (height-wise) for them with the table saw. Not sure yet. The devices just fit in right now.
Thanks-
b
Whole thing (1 of 2):
An externally hosted image should be here but it was not working when we last tested it.
Diode On Heatsink:
An externally hosted image should be here but it was not working when we last tested it.
Leach Drill Template on Heatsink:
An externally hosted image should be here but it was not working when we last tested it.
So, I am concerned about the width of the heatsink base being able to accomidate the diodes. The heatsink is 3.2mm think at the base and the diodes are 5mm long. Just above that, the heatsink becomes 7mm thick, so a possibility would be to edit the base of some of the fins and mount diodes over the devices.
I was also thinking of using a heattransferring epoxy to mount them, which might help. Right now, i am leaning toward one diode mounted directly above each device (for thinkness). Backup is on diode beside each device with heat epoxy.
I am also concerned about my hands getting extremely tired holding the heatsinks in place. Maybe for a song or two, but it will get old quickly ;-)
I am using TO3's. The amp originally had Toshiba flat paks (4 pairs of 2sc2565/2sc1095). As you can see by the template pic, the sinks actually have plenty of room to accomidate the transistors, which will be spaced out much like the old Toshibas were. I may make a little more room (height-wise) for them with the table saw. Not sure yet. The devices just fit in right now.
Thanks-
b
Hi Ace-
Just rereading one of your last post. I think thats just a machining line on the heatsink that makes it look like two pieces. It's just one.
My brother is going to do a little machining on the device mounting area at the botom of the sinks to give me a 2-3mm more clearance for the TO-3's. He'll also drill them for me (CNC machine - I am most envious).
Still trying to decide on what to do the the diodes. Do you think all four mounted as described by Prof Leach, but between only 2 devices will do it? I keep going back and forth between 4 in 1 heatsink or 2+2. This would be per side of course.
So, my method of mouting the TO3s will be modeled after how Nikko mounted the Toshibas flatpacks. They will mount on a vertical plane, with leads dropped stright down and soldered to the bottom of the output board. Leads will be about 2-2.5 cm long.
-b
Just rereading one of your last post. I think thats just a machining line on the heatsink that makes it look like two pieces. It's just one.
My brother is going to do a little machining on the device mounting area at the botom of the sinks to give me a 2-3mm more clearance for the TO-3's. He'll also drill them for me (CNC machine - I am most envious).
Still trying to decide on what to do the the diodes. Do you think all four mounted as described by Prof Leach, but between only 2 devices will do it? I keep going back and forth between 4 in 1 heatsink or 2+2. This would be per side of course.
So, my method of mouting the TO3s will be modeled after how Nikko mounted the Toshibas flatpacks. They will mount on a vertical plane, with leads dropped stright down and soldered to the bottom of the output board. Leads will be about 2-2.5 cm long.
-b
Mornin Brian,
I think the point Prof. Leach was making about mounting the diodes has to do with the temperature sharing on the heatsink. We mostly want the transistors in the middle of the heatsink area for uniform loading and sharing of the area to all devices. Your heatsink does not allow for this. I really don't know the consequences of mounting the T03's and bias diodes as you propose. Any other bright folks out there know how to help?
Prosit
I think the point Prof. Leach was making about mounting the diodes has to do with the temperature sharing on the heatsink. We mostly want the transistors in the middle of the heatsink area for uniform loading and sharing of the area to all devices. Your heatsink does not allow for this. I really don't know the consequences of mounting the T03's and bias diodes as you propose. Any other bright folks out there know how to help?
Prosit
I was thinking that, if dispersed evenly over the heatsink, the devices should all be relatively the same temp. Given that, we are just rrying to trach the average device temp with the diodes, so I think I should be okay. I'll post a drawing of what I am thinking.
-b
-b
Your transistors will be at the bottom of the heatsink, and that's where the heat is generated. It's very likely to be hotter at the bottom all the time. Perhaps the temp will even out across the heatsink and be OK. In any case, it would be better to space the transistors evenly across the bottom with the diodes in the middle of the space. But it's your amp. Try it and see.
Have you done the math on the heatsink to find out if it will sustain
.67 degrees C/W as Prof. Leach recommends?
Prosit
Have you done the math on the heatsink to find out if it will sustain
.67 degrees C/W as Prof. Leach recommends?
Prosit
Hi-
I am always writing these emails with my (almost) 2 year old running around behind me, so I am often less-than-articulate about what I am trying to say (like now). sorry.
I plan to have the diodes at the base of the heat sink. The TO3s will be spaced out evenly (about 4cm in between each one) and the diodes in between those (so about a 2cm gap):
------Q18L----D1/2/3/4----Q19L--------Q18R----D1/2/3/4----Q19R------
The second heatsink will be identical layout, except no diodes.
------Q20L--------------Q21L----------Q20R--------------Q21R------
(pretend those are evenly spaced 😉 )
The other option I was considering would "loop" the diodes across heatsinks (D1 and 4 on the first heatsink, D2 and 3 on the second. I think this is more compicated with not much benefit, but I could easily be convinces otherwise.
When I get the heatsinks back from my brother, I will do the cold water path tests on them. Must do some reading up on that. I am basing my estimate of their capability on their past life, but I realise that is not a very god assumption to make.
Also, I noticed the the amp as a whole had a rating of 620VA, so I am assuming (ahh!!) that the transformer is rated for 620VA (as all power runs through it).
Thanks for the help/feedback!
-b
I am always writing these emails with my (almost) 2 year old running around behind me, so I am often less-than-articulate about what I am trying to say (like now). sorry.
I plan to have the diodes at the base of the heat sink. The TO3s will be spaced out evenly (about 4cm in between each one) and the diodes in between those (so about a 2cm gap):
------Q18L----D1/2/3/4----Q19L--------Q18R----D1/2/3/4----Q19R------
The second heatsink will be identical layout, except no diodes.
------Q20L--------------Q21L----------Q20R--------------Q21R------
(pretend those are evenly spaced 😉 )
The other option I was considering would "loop" the diodes across heatsinks (D1 and 4 on the first heatsink, D2 and 3 on the second. I think this is more compicated with not much benefit, but I could easily be convinces otherwise.
When I get the heatsinks back from my brother, I will do the cold water path tests on them. Must do some reading up on that. I am basing my estimate of their capability on their past life, but I realise that is not a very god assumption to make.
Also, I noticed the the amp as a whole had a rating of 620VA, so I am assuming (ahh!!) that the transformer is rated for 620VA (as all power runs through it).
Thanks for the help/feedback!
-b
Mornin B,
Each heatsink should have Q18,Q19,Q20,Q21 and D1-4 mounted on it.
Is this one of those "less-than-articulate" moments?
D1-4 sets the bias voltage for the transistors and is somewhat heat sensitive. Are you sure you want to do it this way? Before you drill, might be a good idea to take a good second reading of the documentation.
My gramps favorite line, "No matter how many times you cut it off, it won't get any longer".
Prosit
Each heatsink should have Q18,Q19,Q20,Q21 and D1-4 mounted on it.
Is this one of those "less-than-articulate" moments?
D1-4 sets the bias voltage for the transistors and is somewhat heat sensitive. Are you sure you want to do it this way? Before you drill, might be a good idea to take a good second reading of the documentation.
My gramps favorite line, "No matter how many times you cut it off, it won't get any longer".
Prosit
Hey.
I could redraw my output board to put all four from each channel on a single heatsink, giving me the same basic tranny layout, but with a left heatsink and a right heatsink.
My diodes will be a bit furthur from the devices (than when using the Wakefield heatsink) if I put them all in the center. Should still give me a relative heat reading -- reasoning being, if the heatsink is cooler, the devices should also be cooler. He does have a suggested layout, but I think as long as the devices are even spread out on the heatsink, the diodes should do their job nicely.
Back to the drawing board.
Interestingly, however, the former resident of this chassis (the Nikko) did have the channels sharing the ehatsinks as I currently have it drawn, and mounted their biasing transistors on one of the heatsinks (actually sandwiched to one of the devices). I thought they were taking advantage of a heat averaging type of scheme: if one channel wasn;t generating as much heat, the other channel would get more cooling. If they were both generating heat, well, then same as if they each had their own heatsink.
Hmm. I will need to think about this. Maybe I will drill the heatsink holes a few different ways and experiment.
-b
I could redraw my output board to put all four from each channel on a single heatsink, giving me the same basic tranny layout, but with a left heatsink and a right heatsink.
My diodes will be a bit furthur from the devices (than when using the Wakefield heatsink) if I put them all in the center. Should still give me a relative heat reading -- reasoning being, if the heatsink is cooler, the devices should also be cooler. He does have a suggested layout, but I think as long as the devices are even spread out on the heatsink, the diodes should do their job nicely.
Back to the drawing board.
Interestingly, however, the former resident of this chassis (the Nikko) did have the channels sharing the ehatsinks as I currently have it drawn, and mounted their biasing transistors on one of the heatsinks (actually sandwiched to one of the devices). I thought they were taking advantage of a heat averaging type of scheme: if one channel wasn;t generating as much heat, the other channel would get more cooling. If they were both generating heat, well, then same as if they each had their own heatsink.
Hmm. I will need to think about this. Maybe I will drill the heatsink holes a few different ways and experiment.
-b
So, I started redrawing my output boards last night, and decided to bag it. I'll go with what I have, and keep a close eye out for thermal runaway during testing. I have a good feeling that all will be fine, based on the similar specs for the output devices.
I am using TO3's, but the original Tos's where essentially TO3's in a flatpack (metal back connected directly to the collector) -- I'll post some pics.
I am also going to drill the diode holes so that I will be able to change the configuration around if needed. I'll put four holes between each transistir on the first heatsink, and two holes between each channels' transistors on the second heatsink. I can then try a variety of combinations until it's right, but again, I have a good feeling about it working out the first time (gotta have hope).
-b
I am using TO3's, but the original Tos's where essentially TO3's in a flatpack (metal back connected directly to the collector) -- I'll post some pics.
I am also going to drill the diode holes so that I will be able to change the configuration around if needed. I'll put four holes between each transistir on the first heatsink, and two holes between each channels' transistors on the second heatsink. I can then try a variety of combinations until it's right, but again, I have a good feeling about it working out the first time (gotta have hope).
-b
Old Toshibas (and a couple fakes I got):
An externally hosted image should be here but it was not working when we last tested it.
Got my PS caps today. I think these should probably do it. Definitely gonna need some inrush protection.
I've decided to try to reuse the old turn-on delay as well. Maybe even the protection circuit...
An externally hosted image should be here but it was not working when we last tested it.
I've decided to try to reuse the old turn-on delay as well. Maybe even the protection circuit...
Two channel driver board complete (still not yet cleaned):
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Woo nice work
😎 You are giving the old Nikko amp a new life🙂 Too bad i trashed my nikko amp🙁 i could have used your output devices.
😎 You are giving the old Nikko amp a new life🙂 Too bad i trashed my nikko amp🙁 i could have used your output devices.
Hey WorkingAtHome,
Really nice job! Soon we'll know how it sounds. Congrats on a job well executed!
-Chris
Really nice job! Soon we'll know how it sounds. Congrats on a job well executed!
-Chris
Thanks Tekko, thanks anatech.
My brother still has my heatsinks (supposedly doing some machining on them) but he is going on vacation for a week on Tuesday, so I get the feeling the whole thing's gonna be delayed for a few weeks.
I think I will use that time to redesign my output board again.
I realized the other day that I am sharing chassis and signal grounds on my driver board. While it will still work, and might be fine, I have devised a jumper-mod plan to fix it. Luckily I have two ground wires in my power connector - just need to split them where they hit the board and add a jumper from one to the signal ground.
Here's a question: When I split the boards (driver and output), I moved the decoupling caps to the output board. My power goes from the PS board to the output board to the driver board in a daisy chain fashion. Should my driver board also have decoupling caps for VCC and VSS? The wire linking driver to output is about 3-4cm.
My brother still has my heatsinks (supposedly doing some machining on them) but he is going on vacation for a week on Tuesday, so I get the feeling the whole thing's gonna be delayed for a few weeks.
I think I will use that time to redesign my output board again.
I realized the other day that I am sharing chassis and signal grounds on my driver board. While it will still work, and might be fine, I have devised a jumper-mod plan to fix it. Luckily I have two ground wires in my power connector - just need to split them where they hit the board and add a jumper from one to the signal ground.
Here's a question: When I split the boards (driver and output), I moved the decoupling caps to the output board. My power goes from the PS board to the output board to the driver board in a daisy chain fashion. Should my driver board also have decoupling caps for VCC and VSS? The wire linking driver to output is about 3-4cm.
Are you still gonna make the boards after the original design of the chassis like it was before you scrapped it ?
Yup. I'm reusing the shielded torroid, the chassis, the heatsinks and the layout.
I may swap the torroid out later so I cna take advantage of the two differnt voltage levels (higher voltage for the driver stage) that this one provides. The Leach will be single voltage.
Here is a pic of the old (minus a few bits) and new driver boards:
Same footprint. Notice the old driver board has a mini power supply in the center. It was fed AC from the power supply board converted to +-63.1VDC. The power stage used 58.4VDC, which I am using for the Leach.
I may swap the torroid out later so I cna take advantage of the two differnt voltage levels (higher voltage for the driver stage) that this one provides. The Leach will be single voltage.
Here is a pic of the old (minus a few bits) and new driver boards:
An externally hosted image should be here but it was not working when we last tested it.
Same footprint. Notice the old driver board has a mini power supply in the center. It was fed AC from the power supply board converted to +-63.1VDC. The power stage used 58.4VDC, which I am using for the Leach.
- Status
- Not open for further replies.