Hi,
I built myself a small PX4 headphone amp but it's having a very peculiar issue I just can't figure out.
Schematic:
For some reason each channel draws 83mA over the 1K resistor, B+ is 130V due to very high load on the PSU. Resistance checks seem good, grid to ground is 220K, filaments to ground is 1K, double checked the wiring, tried different tubes etc.
Coleman supply at 4V, before applying B+. I'm expecting about 30mA and 200V P-K, instead I have 83ma and 130V P-K.
What would cause such a massive draw of current in an autobias setup? Why does it behave like it's biased at 0V?
I built myself a small PX4 headphone amp but it's having a very peculiar issue I just can't figure out.
Schematic:
For some reason each channel draws 83mA over the 1K resistor, B+ is 130V due to very high load on the PSU. Resistance checks seem good, grid to ground is 220K, filaments to ground is 1K, double checked the wiring, tried different tubes etc.
Coleman supply at 4V, before applying B+. I'm expecting about 30mA and 200V P-K, instead I have 83ma and 130V P-K.
What would cause such a massive draw of current in an autobias setup? Why does it behave like it's biased at 0V?
Last edited:
grid is tied to ground through 220K resistor
why is the output transformer tied to the cathode? Shouldn't it go to B+ or ground?
why is the output transformer tied to the cathode? Shouldn't it go to B+ or ground?
Yes, you're right and I updated the post.
The output transformer is a parafeed one and I was trying to see how the return to cathode would sound like. This shouldn't affect the DC operation of the output tube tho.
The output transformer is a parafeed one and I was trying to see how the return to cathode would sound like. This shouldn't affect the DC operation of the output tube tho.
Is the coupling cap leaky? Check for any DC across the output transformer should be 0.
I will assume that the heater supply is floating and not grounded.
I will assume that the heater supply is floating and not grounded.
No DC across the transformer so the caps are good. Heater supply is floating, only grounded through the 1K resistor, it's a LCRC into Coleman's.
You should ground a DH filament at it's negative end, i. e. reverse the heater supply polarity. Then returning the load through the cathode resistor for NFB reasons substantially increases the output impedance. I'm not too sure if that's a desirable effect.
Best regards!
Best regards!
Maybe both filament pins should be connected to the autobias resistor . Like you see in the internet schematics
It's the standard connection for the coleman regulator:
I even converted the whole thing to standard gapped SE with bypassed cathode. It still behaves like it's 0V biased. I'm suspecting the coleman boards might have something wrong with them. The were used for months on another amp without issues, may be a wiring issue as it's hard to believe both failed at the same time in the exact same way.
I even converted the whole thing to standard gapped SE with bypassed cathode. It still behaves like it's 0V biased. I'm suspecting the coleman boards might have something wrong with them. The were used for months on another amp without issues, may be a wiring issue as it's hard to believe both failed at the same time in the exact same way.
Is R.C. regulator raw supply "floating" (not connected to "ground")?
p.s. and regulator TO 220 transistors also insulated from heatsink?
p.s. and regulator TO 220 transistors also insulated from heatsink?
Yes, raw supply is floating, checked for any potential shorts to ground, none to be found.
Transistors are well insulated from heatsink or ground. The regulators were pulled from a working amp. I'm puzzled by this one. Say there was a mistake at some point and the regulator was shorted to ground, would that kill them in this way?
Soft start is working fine, both adjust just fine with just the heater on. This might be a sign to try out LCL.
Transistors are well insulated from heatsink or ground. The regulators were pulled from a working amp. I'm puzzled by this one. Say there was a mistake at some point and the regulator was shorted to ground, would that kill them in this way?
Soft start is working fine, both adjust just fine with just the heater on. This might be a sign to try out LCL.
Try to use another 4V 1A capable PSU (for example battery) to heat filament.
I think that there are any miswiring.
BTW what's the goal with this schematic?
It has -12dB "gain", so 2V RMS input generates only 1.42Vpp output on 50R. (5mW)
I think that there are any miswiring.
BTW what's the goal with this schematic?
It has -12dB "gain", so 2V RMS input generates only 1.42Vpp output on 50R. (5mW)
Last edited:
I'm a headphone junkey and looking for my ideal match. DHT tubes have proven themselves even if they can be a headache as is the current case.
My dac is 3.5VRMS on SE and 7V RMS on balanced out. With a high quality input transformer from audio note I can have only two stages even for speakers.
My dac is 3.5VRMS on SE and 7V RMS on balanced out. With a high quality input transformer from audio note I can have only two stages even for speakers.
A small part the puzzle not shown is the preamp/voltage gain stage.
On a previous prototype I did try a PX4 single stage with a 12.5:1 OTP with 300 ohm headphones. The result was less than satisfactory even if my source was able to drive the Miller capacitance. It lacked spacial information and presence even if all the numbers added up. Source has 0.5 ohm output and drives a slagle AVC for volume.
My subjective tests show two stages are ideal, PX4 for output and a really good input. This amp will be supplied from a 46 tube with monolith interstage or 01a and Ale's gyratory with rc coupling. There's always the possibility of 801A at 200V and filament bias and LC coupling with 2.2uf and a magnequest grid choke.
On a previous prototype I did try a PX4 single stage with a 12.5:1 OTP with 300 ohm headphones. The result was less than satisfactory even if my source was able to drive the Miller capacitance. It lacked spacial information and presence even if all the numbers added up. Source has 0.5 ohm output and drives a slagle AVC for volume.
My subjective tests show two stages are ideal, PX4 for output and a really good input. This amp will be supplied from a 46 tube with monolith interstage or 01a and Ale's gyratory with rc coupling. There's always the possibility of 801A at 200V and filament bias and LC coupling with 2.2uf and a magnequest grid choke.