Measure temperature on MOSFETS themselves too. I see gray pads in your photo. Those are not needed in this circuit when the sinks are floating in respect to ground. That would ask for isolation grommets too anyway. If there is high thermal resistance between MOSFETS and sinks, if not well pressed against, bad pads, then M1 & M2 could be boiling when the sinks don't conduct enough heat.
Only about small clarifications in the BOM was an update for some time now. Its attached in the GB's opening post since then as the only guide file.
0.351V. Both TO-220'es lukewarm as sinks. I didn't have any paste home so the pads makes good contact.
My fluke doesn't detect any AC component on vout...
Last edited:
It should have been 0.61V. You only run 750mA CCS due to marginal Vdiff (in-out) just enough to keep M2 going given your dummy. That explains the low temps. The more CCS you ask M1 for the more VGS it develops and needs space. Which is already tight for much lower targets even. Find a trafo that can hold for 12.5VDC raw DC input at your target constant current level. Yes I trust you it works stably and without ripple even now, its resilient, but it needs more juice to reach your spec.
Can parallel two 1 Ohm 2W for making 0.5 Ohm. Insulate MOSFET to external sink. Use Silpad and plastic grommet. On board sinks are floating but external sinks mounted on grounded chassis will short Vout to ground otherwise.
I have 2 questions:
1/ onboard sinks - I have insulated MOSFETS. Is there any reason I should remove insulation? (CCS 400 mA, 8v output).
2/ can anyone explain cons and pros of a FET usage instead of R6 resistor and what application FET is recommended for?
1/ onboard sinks - I have insulated MOSFETS. Is there any reason I should remove insulation? (CCS 400 mA, 8v output).
2/ can anyone explain cons and pros of a FET usage instead of R6 resistor and what application FET is recommended for?
Further tests...
Still with old psu 😀.
Picture taken after 25min.
![B]](/community/proxy.php?image=http%3A%2F%2F%5BB%5DImage+removed+by+Moderation+as+requested%5B%2FB%5D&hash=8c28cff5f24cb373de8b649712ecb17f)
The picture looks a bit weird on my iPad...
Continued for 5 more minutes watching temp climb to 66,9C. Blowed gently on the sinks and temp went down to 59ish quite quickly. Larger sinks are needed. Oh, yes. The grey pads are history - thermal paste is in place now.
I also had to feed it with 13,5V to keep voltage over R1 close to 0,61V but it fluctuated between 0,59-0,61V. Vout got down from 8,24V when cold to 8,17V where it stayed. Still not decided if this is the circuit suitable for varying current need.
Regards
Still with old psu 😀.
Picture taken after 25min.
The picture looks a bit weird on my iPad...
Continued for 5 more minutes watching temp climb to 66,9C. Blowed gently on the sinks and temp went down to 59ish quite quickly. Larger sinks are needed. Oh, yes. The grey pads are history - thermal paste is in place now.
I also had to feed it with 13,5V to keep voltage over R1 close to 0,61V but it fluctuated between 0,59-0,61V. Vout got down from 8,24V when cold to 8,17V where it stayed. Still not decided if this is the circuit suitable for varying current need.
Regards
Last edited by a moderator:
1. The extra thermal resistance of the insulation is no problem for this kind of dissipation since you already installed it. No reason to rework the mounting.
2. It brings higher internal gain to the control loop of the reg, thus it lowers its output impedance. What it takes back is it narrows the bandwidth and adds its little noise. Make no mistake, bandwidth remains very wide and noise very low by general standards. FET is recommended in systems where the whole sound could benefit to become a little tighter. Whereas R6 standard is recommended in systems that the sound already seems bit tight or somewhat aggressive. Which can be the more likely with not fully tuned systems yet. Always start with the resistor to evaluate. In other words if the sound with the REF-D built normally with R6 seems already well natural to your ears in your system, don't mod it. If to try FET remember to recheck and maybe tweak the voltage output using the dummy first because it usually runs more mA through the LEDS that can increase Vout by 0.1-0.2V extra.
Use a little 12v fan at 6-8V "shut up" mode sitting on top of the sinks I would say.
No fans in my system 😀. Bigger sinks - yes!
Will the circuit draw the designed current at all times releasing whats not needed at the load as heath thru the sinks? Should I secure functionality with IRF9530 M1 as you recomended to Bern a page or two back?
Will the circuit draw the designed current at all times releasing whats not needed at the load as heath thru the sinks? Should I secure functionality with IRF9530 M1 as you recomended to Bern a page or two back?
Last edited:
Yes, that is what CCSed shunt regs do for a living. 🙂 Although they sum the load currents at a local node and don't shake things back to the bridge and trafo for the benefit of your music, they are also notoriously inefficient. Do you know of any software that could help them hide it? 😀
😀 only fraud will help I'm afraid. Not a problem at my place as the heating is electric radiators. I have to pay for the heat in any case so I prefer to pay accompanied by good music!😀
Last edited: