I just see in my notes that i had the same values, but when i past 200mA with 60R load value, who is the preamp intensity consummation in the positive Rail the tension start to drop to 15V...you should test it to see..
My understanding is the load from the preamp is ~150mA. Therefore I am happy with the PS behavior as load tested with 170mA. I will confirm proper output voltage after assembly, and if necessary address any issues at that time.
BK
Yes you are right for the preamp board, but you have to ad the values of the I/O Board
That's cutting it close. From 24.5 input you need: 2V headroom for the opamp, 4.1V for R5/R6, 1.4V for R9/R10, getting down to 17V. Get a dip in mains and it drops out of regulation.
That R5/R6 is too large and R9/R10 is too small, for me that is a design oversight.
Jan
With 100R for R5/R6 and 1k for R9/R10 with preamp board and I/O board connected for 1h..no overheat and stable at 17V, 200mA for the positive Rail and 150mA for the negative.
Perfect......Thanks a lot Jan
Yes, there should be no difference in dissipation. The dissipation is (Vin-Vout)*Iout and that doesn't change.
Jan
The main reason for that is that the input voltage drops with higher load, and the ripple voltage increases with higher load. Then there is noi enough headroom left for all that other stuff we discussed.
It does also mean that with a larger transformer it may work up to larger loads. Untill your neighbour switches on his sawing machine and the mains drops 10V ;-)
Jan
I replaced R5/R6 with 100ohm and R9/R10 with 1kohm.
Still have stable +/-17V with 170mA load, but now with 0.144V across R5/R6 and 1.41V across R9/R10. Temps are the same.
I had no issues, but made the change on Jan's suggestion as good practice. Thanks everyone for the good discussion!
BK
For those new to this stuff: in the circuit you see that the opamp drives the pass devices through R5. That drive shoulf flow into the base of the pass driver, which in turn drives the final pass device. Very little current is required, for instance for a 100mA output and both transistors have a gain of 70 (total 4900), the required current is 100mA/4900=20uA. The voltage wrt the output is 2 x Vbe, say 1.4V.
But someone decided to place R9 of 200 ohms in parallel with that 1.4V drive which in itself required 1.4/200=7mA, much much more than what the drive requires. In itself no problem for the opamp, but it has to send that current through R5 at 560ohms, which means you lose 560x7mA=3.9V. So the opamp output must be much higher than the output voltage just to overcome R5.
Probably R5 was included in an attempt to promote stability but anything above 50 ohms will do that. Similar story for R6, to promote a quick removal of the base charge thus helping fast reaction to a drop in load current. Here one or two k should be enough.
Jan
Last edited:
Thanks Jan, that help a lot the community to understand why this circuit did not work perfectly and particularly when we have to adapt to a situation like mine. (240V +-10%). We always should try to understand the situation and find a solution adapted...and do not let people with no answer, Thanks again Jan
Looks good - I'll print it out and compare with mine. I spent a couple hours today and took careful measurements + consulted the datasheets. Ended up with the attached. Need to carefully review one more time. Millimeters matter when it comes to CNC and board-mounted connectors!
BK
Yes this is the file posted by RickG post #876, i used the file posted by Jwilhelm post #1305 for my rear panel.
For the hole center heights i used the horizontal midline of the rear panel.
Did not received my panel yet with the cuts and engravings..Change txt with fdp
That looks good and thanks for clarifying the origins. I've seen a few designs now with the upper RCAs labelled as L. Are the upper RED RCAs not the R channel??
BK