Ok, got it.
Red = components
Dark blue = Bottom tracks?
Green = Top tracks
Light blue = Connection betwwe top and bottom?
How am I doing
Terry
Red = components
Dark blue = Bottom tracks?
Green = Top tracks
Light blue = Connection betwwe top and bottom?
How am I doing
Terry
Terry,
The colors of the layers and silkscreen will depend on how you have the viewer configured. I posted a PIC of the PCB back in post #190. In that PIC the top layer is RED. The bottom layer is GREEN and the silkscreen (top layer) is WHITE.
The colors of the layers and silkscreen will depend on how you have the viewer configured. I posted a PIC of the PCB back in post #190. In that PIC the top layer is RED. The bottom layer is GREEN and the silkscreen (top layer) is WHITE.
Hi Carl,
I checked the PCB and could not find any errors. Some parts are not given, such as the 150nf 400v cap on the output. That is a more difficult task than it first appears.
Terry
I checked the PCB and could not find any errors. Some parts are not given, such as the 150nf 400v cap on the output. That is a more difficult task than it first appears.
Terry
Carl,
I will offer to check your board too. Rather than download the viewer, can you simply email me a 800 x 600 (or thereabouts) gif of the composite board?
Thank you,
Hugh
I will offer to check your board too. Rather than download the viewer, can you simply email me a 800 x 600 (or thereabouts) gif of the composite board?
Thank you,
Hugh
So,
Has anyone checking the PCB layout against the schematic found errors or have suggestions to improve?
Has anyone checking the PCB layout against the schematic found errors or have suggestions to improve?
Carl_Huff said:
Hi Carl,
I couldn't find any errors. How do you see us mounting to the heatsink.
Terry
Terry,
The board mounts parallel to the heatsink by way of 4 screws at each corner of the PCB making for a very space efficient solution. The output devices are sandwiched between the heatsink and the PCB. The leads of the output devices will have to be bent at 90 degrees to accomplish this. This is different from the version that appeared in the magazine article.
The board mounts parallel to the heatsink by way of 4 screws at each corner of the PCB making for a very space efficient solution. The output devices are sandwiched between the heatsink and the PCB. The leads of the output devices will have to be bent at 90 degrees to accomplish this. This is different from the version that appeared in the magazine article.
Carl,
So far I've found no problems, but give me a few more hours to check. Please excuse my delays, a lot on my plate at present,
Hugh
So far I've found no problems, but give me a few more hours to check. Please excuse my delays, a lot on my plate at present,
Hugh
HI Carl,
First, apologies for late reply. It's Sunday afternoon and I've just completed the board check.
I have found one problem only. I believe that the lag comp cap, 100pF, is terminated on the left side at the BASE of Q8, not the EMITTER of Q8 as you have it on the layout.
I have thought about D1/D21. The series orientation will protect the 220uF MUSE cap to -1.3V reverse polarisation if the output goes negative wrt ground. If it goes positive, then with 55V rail, the feedback node may rise to 0.51/12.51 * 55 (+2.2V) which is still within the 16V rating. If you connect the diodes anode/kathod and kathod/anode across the cap, then neither DC polarity from rail sticking can ever exceed 0.6V across the fb cap, which will permit using a 6.3VW cap, much smaller. This won't affect amp operation, however, and is merely a nicety. Back to back diodes will not affect AC operation, BTW.
Otherwise, it's a very nice layout and I doffs me hat..... thank you for contributing it.
Cheers,
Hugh
First, apologies for late reply. It's Sunday afternoon and I've just completed the board check.
I have found one problem only. I believe that the lag comp cap, 100pF, is terminated on the left side at the BASE of Q8, not the EMITTER of Q8 as you have it on the layout.
I have thought about D1/D21. The series orientation will protect the 220uF MUSE cap to -1.3V reverse polarisation if the output goes negative wrt ground. If it goes positive, then with 55V rail, the feedback node may rise to 0.51/12.51 * 55 (+2.2V) which is still within the 16V rating. If you connect the diodes anode/kathod and kathod/anode across the cap, then neither DC polarity from rail sticking can ever exceed 0.6V across the fb cap, which will permit using a 6.3VW cap, much smaller. This won't affect amp operation, however, and is merely a nicety. Back to back diodes will not affect AC operation, BTW.
Otherwise, it's a very nice layout and I doffs me hat..... thank you for contributing it.
Cheers,
Hugh
Hugh,
Good catch on the lag cap off of Q8. I have since fixed that on the PCB layout. Thanks for all of your help.
Good catch on the lag cap off of Q8. I have since fixed that on the PCB layout. Thanks for all of your help.
Carl et al,
You guys have been busy finalizing this amp. Do you think we are ready to implement all of the schematic changes into a pcb build? This will be my first project using the Thermal Traks. How do you see the bias on this amp. Run the outputs hot or let them have an easy life.
With six diodes in the multiplier circuit do you expect to need an adjustable pot here?
Tad
You guys have been busy finalizing this amp. Do you think we are ready to implement all of the schematic changes into a pcb build? This will be my first project using the Thermal Traks. How do you see the bias on this amp. Run the outputs hot or let them have an easy life.
With six diodes in the multiplier circuit do you expect to need an adjustable pot here?
Tad
Tad,
Now that I have included Hugh's find the PCB layout is complete. You have room to experiment with the tempco diode string. The pot is there to 'fine tune'. There is quite a lot written about that earlier in this thread and in the "Silicon Chip 200Watt LD amplifier" thread.
I just got back a quote from my favorite Chinese PCB maker. The cost for a double sided board, 2 oz copper, silkscreen and solder mask amounts to ...
100 pieces $6.00 each, 80 pieces $8.00 each, 60 pieces $10.00 each plus 'ship to' fees.
Is there sufficient interest here to do a group buy?
Now that I have included Hugh's find the PCB layout is complete. You have room to experiment with the tempco diode string. The pot is there to 'fine tune'. There is quite a lot written about that earlier in this thread and in the "Silicon Chip 200Watt LD amplifier" thread.
I just got back a quote from my favorite Chinese PCB maker. The cost for a double sided board, 2 oz copper, silkscreen and solder mask amounts to ...
100 pieces $6.00 each, 80 pieces $8.00 each, 60 pieces $10.00 each plus 'ship to' fees.
Is there sufficient interest here to do a group buy?
Carl_Huff said:
Hi Carl, I will take 20! That should start things off!
Can your supplier send out two packets, one for the USA and one for AUS. I volunteer to resend the Ausy ones out to individuals as required.
Terry
Terry said, "I volunteer to resend the Ausy ones out to individuals as required."
Great! I will hold you too it!
Great! I will hold you too it!
Carl_Huff said:
No problem Carl, I was the Ausy distributor for Morel, Audiomatica and a few others. Quite used to doing this sort of stuff.
Terry
Carl_Huff said:
Carl
Is this your latest layout?
http://www.diyaudio.com/forums/attachment.php?s=&postid=1824875&stamp=1242018735
The Transistors are very close together, how much will this affect heat dissipation?
Also them being not at the edge of the pcb, are you looking at a flanged heatsink?
Allan
- Status
- Not open for further replies.