Hello,
I am testing a design of mine for tube amp B+ supplies, both for input and output stages. Currently setup for the lower current input stage, running at the top of its adjustable range, 170V/75mA, with the shunt mosfet (IRF820) run in linear region (of course) dumping 40mA current. This is a stress test to check for thermal stability and worst-case scenario performance - the shunt will dump a bit less power in normal operation (25-30 mA at 150V). Looks that will make for a decent space heater if anything 😎
I am satisfied with line regulation (about 150mV pk-pk), but there is a great deal of ringing (see scope shots). First one is blurred but only there to show the main ripple voltage with the ringing around it as light green, next two pics are close-ups. Sorry with the bad shots, I have only an old scope and I'm not that good working with it. I have measured around 100mV pk-pk ringing at about 50 KHz. Is that to be expected? Note that these preliminary tests were done into a 5K dummy load, without "speed-up" capacitor C1 and output capacitor C3.
50mV/div, 0.50mS/div
20mV/div, 50uS/div
20mV/div, 20uS/div
A few more circuit details: The top device M1 is an IXYS IXCP10M45S current source; The shunt mosfet M3 has a 100 ohm stopper resistor soldered close to the gate (see board pic below). It is a dual mono board with all TO-220 devices on the same heatsink with ceramic insulating pads. In the actual builds the heatsinks will be mechanically connected to the grounded metallic chassis. This test setup doesn't use the heatsink that will be used in actual project but uses the actual adapter bracket (the aluminium flat bar the tabs screw into) and heatsinks are left floating. I will check with grounded heatsinks tomorrow.
I have measured a substantial capacitance of 50-100 nF between the TO-220 tabs (drains) and the heatsink - That can't be good, I fear HF coupling between devices and channels... I have tested only a channel at a time so far. I am not very experienced with parasitics management and would appreciate any advice on how to tame the ringing. I have often seen ferrite beads used in SPMS designs, presumably on the gate lead. Can it help here?
Thanks in advance.
- Joris
I am testing a design of mine for tube amp B+ supplies, both for input and output stages. Currently setup for the lower current input stage, running at the top of its adjustable range, 170V/75mA, with the shunt mosfet (IRF820) run in linear region (of course) dumping 40mA current. This is a stress test to check for thermal stability and worst-case scenario performance - the shunt will dump a bit less power in normal operation (25-30 mA at 150V). Looks that will make for a decent space heater if anything 😎
I am satisfied with line regulation (about 150mV pk-pk), but there is a great deal of ringing (see scope shots). First one is blurred but only there to show the main ripple voltage with the ringing around it as light green, next two pics are close-ups. Sorry with the bad shots, I have only an old scope and I'm not that good working with it. I have measured around 100mV pk-pk ringing at about 50 KHz. Is that to be expected? Note that these preliminary tests were done into a 5K dummy load, without "speed-up" capacitor C1 and output capacitor C3.
50mV/div, 0.50mS/div
20mV/div, 50uS/div
20mV/div, 20uS/div
A few more circuit details: The top device M1 is an IXYS IXCP10M45S current source; The shunt mosfet M3 has a 100 ohm stopper resistor soldered close to the gate (see board pic below). It is a dual mono board with all TO-220 devices on the same heatsink with ceramic insulating pads. In the actual builds the heatsinks will be mechanically connected to the grounded metallic chassis. This test setup doesn't use the heatsink that will be used in actual project but uses the actual adapter bracket (the aluminium flat bar the tabs screw into) and heatsinks are left floating. I will check with grounded heatsinks tomorrow.
I have measured a substantial capacitance of 50-100 nF between the TO-220 tabs (drains) and the heatsink - That can't be good, I fear HF coupling between devices and channels... I have tested only a channel at a time so far. I am not very experienced with parasitics management and would appreciate any advice on how to tame the ringing. I have often seen ferrite beads used in SPMS designs, presumably on the gate lead. Can it help here?
Thanks in advance.
- Joris
Last edited:
