sesquialtera said:
I don't know, cost? Elegance, ease for home etchers? Maybe better usage of the bottom layer as a ground plane than just 3 lines?
Here is a quick idea. I guess feasible if moved to center it again for some very little upper and left side edge clearance.
Attachments
im very happy
Class time!
I wanted to offer some pointers on board layout for better performance. Wanting to use to use an example I started moving parts around to improve the layout. I find laying out circuit boards to be a great puzzle. Much more fun than trying to put together jig saw puzzles... You can use the layout I've come up with if you like, tweak on it some more, or just use it for a learning example. The mouting holes are on the far end from the transistors. If the transistors are mounted to a heat sink you only need to support the other end of the board.
With a shunt regulators low output impedance and good high frequency response we need to be carefull of where the signal currents flow on the regulator board. The output loop needs to be as short as possible. Just like you would do star grounding in an amplifier or pre-amplifier, you do the same thing in the regulator layout.The first thing to note on the original layout is that the ground terminal is all the way across the board from the output device (Q3). All the components that are referenced to ground are between the output and ground points and on a small trace to boot. This is not a good thing.Shunt regs work by rapidly changing the current drawn by the output device to keep the output at a fixed voltage. With the long light trace you will have bad ground loops within the regulator. This will limit the performance and posssibly be unstable at high frequencies and or high currents.
The new layout has the circuit broken up into 3 main loops. The output reference loop, the output snubber loop, and the bias loop. The output snubber loop is isolated so that when it is doing it's job at high frequencies the currents will not effect the rest of the circuit. Like wise, the reference loop is as short as possible to help with frequency response and parasitics. The bias loop is all high impedance bias currents that should not have much signal current.
I took the liberties of using 1/4 watt resistors on the gate stoppers to save a bit of space. The rest of the small signal resistors could be swithed over to 1/4 watt also. Check the layout for errors- not having the schematic to match I was not able to link the board and schematic for error checking.
Look over the layout and let me know what you think. Personal preference would have larger pads for Q1 and the wire connections.
Here is the ExpressPCB simplistic.zip file for downloading. Be careful with the file as I did not rename it! Don't want to overwrite your copy...
I wanted to offer some pointers on board layout for better performance. Wanting to use to use an example I started moving parts around to improve the layout. I find laying out circuit boards to be a great puzzle. Much more fun than trying to put together jig saw puzzles... You can use the layout I've come up with if you like, tweak on it some more, or just use it for a learning example. The mouting holes are on the far end from the transistors. If the transistors are mounted to a heat sink you only need to support the other end of the board.
With a shunt regulators low output impedance and good high frequency response we need to be carefull of where the signal currents flow on the regulator board. The output loop needs to be as short as possible. Just like you would do star grounding in an amplifier or pre-amplifier, you do the same thing in the regulator layout.The first thing to note on the original layout is that the ground terminal is all the way across the board from the output device (Q3). All the components that are referenced to ground are between the output and ground points and on a small trace to boot. This is not a good thing.Shunt regs work by rapidly changing the current drawn by the output device to keep the output at a fixed voltage. With the long light trace you will have bad ground loops within the regulator. This will limit the performance and posssibly be unstable at high frequencies and or high currents.
The new layout has the circuit broken up into 3 main loops. The output reference loop, the output snubber loop, and the bias loop. The output snubber loop is isolated so that when it is doing it's job at high frequencies the currents will not effect the rest of the circuit. Like wise, the reference loop is as short as possible to help with frequency response and parasitics. The bias loop is all high impedance bias currents that should not have much signal current.
I took the liberties of using 1/4 watt resistors on the gate stoppers to save a bit of space. The rest of the small signal resistors could be swithed over to 1/4 watt also. Check the layout for errors- not having the schematic to match I was not able to link the board and schematic for error checking.
Look over the layout and let me know what you think. Personal preference would have larger pads for Q1 and the wire connections.
An externally hosted image should be here but it was not working when we last tested it.
Here is the ExpressPCB simplistic.zip file for downloading. Be careful with the file as I did not rename it! Don't want to overwrite your copy...
salas said:
salas said:
I was just doing a simulation of this circuit in SPICE to make sure I understand how it works. The IRFP9240 is biased by the constant voltage drop across the LEDs to pass, say 100 ma of current if your circuit needs 50ma steady state. The constant current source created by the BCC560s draws a few mA, and the large resistor with the parallel 1uF capacitor locks down the voltage going to the MPSA92 at about a diode drop below the supply. As the current drawn by the load increases, the voltage drop across the 2.2k resistor in the emitter of the MPSA92 will decrease, causing the IRF840 to conduct less and keeping the current through the circuit, thus the supply voltage constant. Vice versa for a load current decrease. The resistor and capacitor in series prevent oscillation due to the negative feedback. Does that sound about right?
Hi Salas
Got all the parts, and has been arrange it PTP like the Gary's Layout.
i got question about the tube rectifier EZ81. does this shunt reg retain the claimed benefits of the tube rectifier, such as warm sounding? if not, i would rather just use the sckottky diode that i have in my drawer. it rectify more spare voltage as well that i can have the spare 50v as you recommend.
should i go with tube rectifier or solid state rectifier?
thx in adv
erwin
Got all the parts, and has been arrange it PTP like the Gary's Layout.
i got question about the tube rectifier EZ81. does this shunt reg retain the claimed benefits of the tube rectifier, such as warm sounding? if not, i would rather just use the sckottky diode that i have in my drawer. it rectify more spare voltage as well that i can have the spare 50v as you recommend.
should i go with tube rectifier or solid state rectifier?
thx in adv
erwin
You can still hear the tube rectifier but to about the 20% the impact it would normally have with a passive CLC.
I would rectify both ways and see what happens in a particular system VS the particular tube rectifier. But only by keeping the 50VDC difference in all cases. If you cant have the 50V, go solid state.
I would rectify both ways and see what happens in a particular system VS the particular tube rectifier. But only by keeping the 50VDC difference in all cases. If you cant have the 50V, go solid state.
hi Salas
nice.. i got some heatsink confusion. attached the picture of several heatsink i got. can i use the smallest one on the irfp9240? the other one is cpu heatsink with fan. then the other one is for 2n3055. thx in advance
thats the board that i arrange with Gary's suggestion layout. haven't soldered them in. but should be find as i am going to prove read it again before soldering.
erwin
nice.. i got some heatsink confusion. attached the picture of several heatsink i got. can i use the smallest one on the irfp9240? the other one is cpu heatsink with fan. then the other one is for 2n3055. thx in advance
thats the board that i arrange with Gary's suggestion layout. haven't soldered them in. but should be find as i am going to prove read it again before soldering.
erwin
Attachments
