Hey Guys,
For the life of me I can not figure out what is probably a SIMPLE issue.
I have a simlpe voltage regulator circuit for my heaters. It is based on the LM317 regulator. The differemces between my circuit and the one attached are that I have 50uF as the first filter cap, 100uF at the output, and a 220 ohm and 2K pot for the voltage divider, no led and no final 0.1uF at the tubes. There is also a small CM choke before the first filter cap after the bridge.
With no load I can set the output to 12.6, no problems. But when I apply load (ie attach the circuit to the heaters in parallel) the voltage drops to 10 volts with one tube, and 7 volts with two tubes. I can follow this voltage drop all the way to the bridge, which will read 13Vdc after the bridge. I should state the my tranny is 12V/2.5A, and with no load I have about 19Vdc after rectification.
I cant seem to figure out why I am dropping so much voltage?
Thoughts?
Bryan
For the life of me I can not figure out what is probably a SIMPLE issue.
I have a simlpe voltage regulator circuit for my heaters. It is based on the LM317 regulator. The differemces between my circuit and the one attached are that I have 50uF as the first filter cap, 100uF at the output, and a 220 ohm and 2K pot for the voltage divider, no led and no final 0.1uF at the tubes. There is also a small CM choke before the first filter cap after the bridge.
With no load I can set the output to 12.6, no problems. But when I apply load (ie attach the circuit to the heaters in parallel) the voltage drops to 10 volts with one tube, and 7 volts with two tubes. I can follow this voltage drop all the way to the bridge, which will read 13Vdc after the bridge. I should state the my tranny is 12V/2.5A, and with no load I have about 19Vdc after rectification.
I cant seem to figure out why I am dropping so much voltage?
Thoughts?
Bryan
Attachments
Now how did I miss THAT statement!? Yes, boost the capacitance and all will be well. That choke right after the rectifier might be forming a little choke input filter too, which will bring the voltage down to about 90%. And if it has a DCR of even only a few ohms, 3 ohms will drop almost 2 volts at 600mA.
Here are my 2 cents.
It is designed to use a vero board and the heatsink I bought and the orientation is specifically designed for my particular layout. You may alter it for your use. By the way, I will have to enlarge it slightly to give a bit more room for heat dissipation of the components. Placing the capacitors too close to the heatsink (or the 5W wirewound resistor) may shorten their life.
I previously used LM317 and 1N5004 and it worked very well. However, my 12B4A amp ran rather hot. In order to reduce the waste of power I will be using LM1086 (you can also use LM1085) low drop regulator as well as the very cheap Shottky diodes 1N5822. The snubber resistor at the input needs to be calculated according to the leakage inductance of your transformer. I am hoping this in addition to the 0.01u caps across the diodes will eliminate all the voltage spikes and resonances. The adjustment resistor is placed as close to the Adj pin as possible to improve regulation. The 10uF capacitor needs to be low inductance type to provide improved ripple rejection. For 12B4A, I would use a 39k and a 270k as a voltage divider shunting the B+ (right before the last cap) to the ground, and the joint of the 2 resistors is to connect to the Vlift on the PCB. The two resistors will serve the purpose of lifting the filament as well as serving as a breeding resistor (@1mA) for the HV.
The 2.7R in series with the 1uF Tant is to counter the inductance of the LM1086. 2 x 220uF 100V are used to kill both the diferential and common mode noises.
I hope this gives a very quite but still highly efficient filament supply. Indeed, I am using the same circuit for my line level power supply, just with different values for a few components.
Is this an overkill for a filament supply? I don't think so. The filament will have no issue even with AC but any noise can be capacitively coupled to the HV winding in the transformer therefore the quiter the filament supply the better.
Regards,
Bill
An externally hosted image should be here but it was not working when we last tested it.
It is designed to use a vero board and the heatsink I bought and the orientation is specifically designed for my particular layout. You may alter it for your use. By the way, I will have to enlarge it slightly to give a bit more room for heat dissipation of the components. Placing the capacitors too close to the heatsink (or the 5W wirewound resistor) may shorten their life.
I previously used LM317 and 1N5004 and it worked very well. However, my 12B4A amp ran rather hot. In order to reduce the waste of power I will be using LM1086 (you can also use LM1085) low drop regulator as well as the very cheap Shottky diodes 1N5822. The snubber resistor at the input needs to be calculated according to the leakage inductance of your transformer. I am hoping this in addition to the 0.01u caps across the diodes will eliminate all the voltage spikes and resonances. The adjustment resistor is placed as close to the Adj pin as possible to improve regulation. The 10uF capacitor needs to be low inductance type to provide improved ripple rejection. For 12B4A, I would use a 39k and a 270k as a voltage divider shunting the B+ (right before the last cap) to the ground, and the joint of the 2 resistors is to connect to the Vlift on the PCB. The two resistors will serve the purpose of lifting the filament as well as serving as a breeding resistor (@1mA) for the HV.
The 2.7R in series with the 1uF Tant is to counter the inductance of the LM1086. 2 x 220uF 100V are used to kill both the diferential and common mode noises.
I hope this gives a very quite but still highly efficient filament supply. Indeed, I am using the same circuit for my line level power supply, just with different values for a few components.
Is this an overkill for a filament supply? I don't think so. The filament will have no issue even with AC but any noise can be capacitively coupled to the HV winding in the transformer therefore the quiter the filament supply the better.
Regards,
Bill
I would like to clarify that the above will give a 12.6VDC instead of the 6.3VDC as indicated, and the input is designed for 12VAC@2A with a reasonable regulation of the Tx. For a 6.3VDC, all you need is to reduce the input voltage to about 8-9VAC (You need to model this with PSUII) but you may need to use the LM1085 regulator for higher current.
Regards,
Bill
Regards,
Bill
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.