• These commercial threads are for private transactions. diyAudio.com provides these forums for the convenience of our members, but makes no warranty nor assumes any responsibility. We do not vet any members, use of this facility is at your own risk. Customers can post any issues in those threads as long as it is done in a civil manner. All diyAudio rules about conduct apply and will be enforced.

PCB order for Goldmun clone

Jump to Latest
#164
Bigpanda - No matter what the reading of the variable resistor is, you can always match it close enough with one, or two at most, fixed resistors.

To sound the best, the MOSFETS have to be run as hot as possible, but in the safe area of operation.

If they're getting too hot, simply lower the value of the BIAS resistor R23. It's a no brainer.
 
#169
Gift to Alexmm (our board designer)

smartx21 said:
Maybe a few clicks more, due to transistors pinouts and other more-than-two-legged parts.
Click to expand...

I agree with SmartX, it is much more than one click to do the mirror. Alex has done such a beautiful layout and I would not mess with it. Doing a mirror pair will add to the PCB setup cost.

Alex has done so many revisions to the board and I think that he deserve a gift from this group buy. If Alex agrees, we should send him one pair of board and we just split the cost amongst all the buyers of this group-buy.

If there are more than 50 pairs of boards in this group-buy, bigpanda can just add 2% to the price of each pair of board; that should be quite easy to do.

That is just my 2c.

- Stanley
 
#170
DanFrank - Darn, I'm wrong again regarding the mirror image. That's twice I've been wrong already! Lets not change Alex's design. Lets just get the boards already. I agree that we should buy Alex a pair of the boards for all the work that he has put in. If not everyone agrees, I volunteer to just buy a pair for him.

Regarding the MOSFET current. I really don't know... But I also don't know why this discussion is holding us down. The BIAS resistor R23 is already in the schematic and already spec'd by Goldmund to be 330 ohms. Everyone should just use a 330 ohm resistor and move on already. If your MOSFETS are getting too hot, use a bigger heatsink. If they're still too hot, lower the BIAS resistor R23 to something like 220-270 ohms. Remember one thing, Goldmund runs their MOSFETS hot, extremely hot! That's part of the reason why this amplifier sounds so good. On my Goldmund Mimesis 6, if I listen to music at close to max power for 15 minutes, I cannot even touch the heatsink for a split second without burning my hand.

Also, as per Goldmund, if you change the original Hitachi 2SK134/2SJ49 to the new Magnatec BUZ900/905, BUZ901/906, the BIAS resistor R23 will remain the SAME, 330 ohms. So there's nothing to worry about as long as you have an adequate enough heatsink.
 
#171
Never measured it, huh ?

Easy peasy wet finger from a single glance at the schematic says ~2mA driver bias, is 660mV across the bases of the output MOSFETs.
Vision impaired can finger count the bias from the posted simulation at the main GM thread
A lot of Lateral MOSFET designs from the late '80s and '90s were biased at 50mA per MOSFET.
If you take a look at the datasheets you'd notice that zero tempco occurs at different quiescent current levels for the P- and N-channel device, but a lot higher than 50mA for either one.

One does not need high bias to run a power amp hot, just pick a small heatsink.
 
#174
Heatsink of the Mimesis 9 : www.novosite.nl/editor/assets/highender/Goldmund Mimesis 9-1.jpg

Fischer SK56, the most frequent used heatsink type in these parts by both pros and sick home builders.
12" wide x 6" high in the Mim9 : 0.36 C/W
80V rails x 50mA x 12 devices makes 48W quiescent heat.
Enough to push the heatsink to +40C, without playing a single note.

The early Goldmund models had massive steel cases : high mass => high damping.
Anyone who ever made something in steel knows it's a pita to get a decent finish without visible scratches, the reason why Goldmund gave the amp cases the classic Eeew Mucus look : www.kaudio.co.kr/Gnu_Base/data/cheditor4/yongsan/1141616498.jpg
Downside of that stuff is real bad thermal conductivity.
Goldmund tops that off by hatching out the back of the chassis, and attaching the heatsink to the case poorly, in a thermal sense.
Again primarily vibration damping motivated, not thermal.

Plus, not a single hole in any side of the amp.
Great for keeping any dust out, but also prevents any heat to gently flow out : each Joule generated within the case has to invade the chassis for a chance of escape.
(the Gold Cubes were an entirely different ball game, but how many are willing to shell out a full size sedan for a pair of Rubik blocks)

THD hunters bias Laterals at over 100mA each, Golden Ear types do the same, ditto for the zero tempco minded.
Beyond 150mA is for Class A aficionados, i've never gone beyond 1/4th of max Pd dissipation for these devices, is 25W for the J50/K135.
Which makes ~300mA at 80Vdc rails the sane limit, going higher means having to spend an insane amount of cash for heatsinks which are able to keep die temperature at longevity standard level.
 
#175
right

that Goldmund heatsink looks like its fit for one channel only
and may still be on the small side
a modern stereo chassis like modu shop will at least have one on each side

it certainly does not sound very healthy if it gets hotter than classA
 
You must log in or register to reply here.
Top Bottom