Hello.
I'm thinking about building a headphones amplifier around the 6n1p and 6n6p as I'm almost done with my dac (something along those lines: http://www.tubecad.com/2006/04/blog0062.htm ). It would require a B+ of near 300V @55ma (idle).
Here are the parts I've already got on hand and can use for the PS of the new amp:
- A 333V CT @65ma B+ transformer, with 6.3V CT @3.5A for the heaters ( I can scavenge it from a broken Heathkit oscilloscope).
- A pair of 3300µF/350V Rifa Peh169
- A pair of 220µF/350V Rifa Peh 169 (I'd rather like however to use those for the output caps of the amplifiers)
- A dozen 20uF/400V MKP caps.
I'm thinking of using a choke input PS, as it would give me in theory 333*0.9= 300V. I'd simply use: xformer - mur860 - choke - 3300uF. The choke could be a Hammond 8H @100mA /259R.
It leaves me with a few question though:
-1- Most hammond chokes I can get have an insulation rating of 400V. It seems borderline in this application. Can/should I use the choke on the ground line ?
-2- What will be the voltage after the choke at startup ? The Rifa are only rated for 350 *1.1= 385V surge. PSUD, with a loading jumping from 3ma at startup (bleeders...) to 56ma after 10s, shows the voltage rising to 330V before falling down to 285V. Are the 350V caps sufficient or not ?
-3- With such a big cap, will the choke sufficient to limit inrush current ? Should I add thermistors ?
-4- What about using a standby switch to ease the startup ? A two pole switch could put in a 47K/3W resistor before the choke and a big diode in the heater supply. (see: http://www.freewebs.com/valvewizard1/standby.html ). Would it invalidate the fears exposed above ?
Thank you for reading this (way too) long post and for any help on this
I'm thinking about building a headphones amplifier around the 6n1p and 6n6p as I'm almost done with my dac (something along those lines: http://www.tubecad.com/2006/04/blog0062.htm ). It would require a B+ of near 300V @55ma (idle).
Here are the parts I've already got on hand and can use for the PS of the new amp:
- A 333V CT @65ma B+ transformer, with 6.3V CT @3.5A for the heaters ( I can scavenge it from a broken Heathkit oscilloscope).
- A pair of 3300µF/350V Rifa Peh169
- A pair of 220µF/350V Rifa Peh 169 (I'd rather like however to use those for the output caps of the amplifiers)
- A dozen 20uF/400V MKP caps.
I'm thinking of using a choke input PS, as it would give me in theory 333*0.9= 300V. I'd simply use: xformer - mur860 - choke - 3300uF. The choke could be a Hammond 8H @100mA /259R.
It leaves me with a few question though:
-1- Most hammond chokes I can get have an insulation rating of 400V. It seems borderline in this application. Can/should I use the choke on the ground line ?
-2- What will be the voltage after the choke at startup ? The Rifa are only rated for 350 *1.1= 385V surge. PSUD, with a loading jumping from 3ma at startup (bleeders...) to 56ma after 10s, shows the voltage rising to 330V before falling down to 285V. Are the 350V caps sufficient or not ?
-3- With such a big cap, will the choke sufficient to limit inrush current ? Should I add thermistors ?
-4- What about using a standby switch to ease the startup ? A two pole switch could put in a 47K/3W resistor before the choke and a big diode in the heater supply. (see: http://www.freewebs.com/valvewizard1/standby.html ). Would it invalidate the fears exposed above ?
Thank you for reading this (way too) long post and for any help on this
That mur860 is verry likely to fail. The voltage rating is far to low.
Let me illistrate just with a cap input filter. Take a 10v transformer use one diode to rectify it and a cap filter you get 14.1 V less the ~1v lost in the diode or 13.1V then on the other half cycle when the cap is not charging the diode sees -14.1v from the transformer + 13.1v from the charged capacitor for a total of 27.2v across the diode. I believe in your case your
diode rated at 600v will be exposed to ~760v !
Let me illistrate just with a cap input filter. Take a 10v transformer use one diode to rectify it and a cap filter you get 14.1 V less the ~1v lost in the diode or 13.1V then on the other half cycle when the cap is not charging the diode sees -14.1v from the transformer + 13.1v from the charged capacitor for a total of 27.2v across the diode. I believe in your case your
diode rated at 600v will be exposed to ~760v !
you're right of course.
Hum 
1 to 2H ? That seems awfully low.
A/ The preamp will idle at 55ma (both channel, 5ma per input stage, 20 ma per output stage, 5ma for bleeders). It's a class A push pull (white cathode follower). If I pull hard on both channel (let's say with low impedance headphones), I could in theory get sustained peaks at 95mA.
This limits me to chokes of at least 100mA.
B/ Rule of thumb to determine the inductance of the choke is output voltage in V / minimum current in mA. Otherwise the voltage output rises. Not a good thing with my 350V caps on the output.
To be safe, let's do the math with one of the output tube failing. Idle consumption goes down to 35ma. Thus we need a choke of about 8H (300/35).
Which brought me to the hammond 157M. The "problem" is that the 157M has high resistance (259). Not a very good thing in our application.
It also drops the output voltage at idle at around 285VDC. The 159P is a better choke, at 10H, 125ma and 155R.
I think I'm gonna go for that one. Price difference is only 5€.
All this doesn't tell me if I'm safe with 350VDC caps at the output of the PS...
1 to 2H ? That seems awfully low.
A/ The preamp will idle at 55ma (both channel, 5ma per input stage, 20 ma per output stage, 5ma for bleeders). It's a class A push pull (white cathode follower). If I pull hard on both channel (let's say with low impedance headphones), I could in theory get sustained peaks at 95mA.
This limits me to chokes of at least 100mA.
B/ Rule of thumb to determine the inductance of the choke is output voltage in V / minimum current in mA. Otherwise the voltage output rises. Not a good thing with my 350V caps on the output.
To be safe, let's do the math with one of the output tube failing. Idle consumption goes down to 35ma. Thus we need a choke of about 8H (300/35).
Which brought me to the hammond 157M. The "problem" is that the 157M has high resistance (259). Not a very good thing in our application.
It also drops the output voltage at idle at around 285VDC. The 159P is a better choke, at 10H, 125ma and 155R.I think I'm gonna go for that one. Price difference is only 5€.
All this doesn't tell me if I'm safe with 350VDC caps at the output of the PS...
After searching the web, I've finally found a trick on this forum to avoid overvoltage even with an unloaded choke at start-up.
I just need a string of three 100V 5W zeners and a 1k/1w resistor, at the output of the choke. When the output voltage rises over 300V, they'll pull it down. Once the voltage settled under 300V by the idle load of the amplifier, they'll just sit there doing nothing. They'll also offer a nice protection in the case of pulled tubes.
This settles question 2 I guess ?
I just need a string of three 100V 5W zeners and a 1k/1w resistor, at the output of the choke. When the output voltage rises over 300V, they'll pull it down. Once the voltage settled under 300V by the idle load of the amplifier, they'll just sit there doing nothing. They'll also offer a nice protection in the case of pulled tubes.
This settles question 2 I guess ?
This thread turns into a nice experience of self-learning and beginner's mistakes.
Let's go back to the questions list:
-1- Choke on the ground side: Bad, bad, bad idea. This raises the CT of the transformer heavily and the choke sees a lot of voltage due to the transients anyway.
Fix: using a Hammond 193 with sufficient isolation rating (800v).
-2- Are 350V caps sufficient: yes but... if the choke is insufficiently loaded, the voltage will jump above V*0.9.
Fix: a string of three 100V/5W zeners + 1K/1W resistor at the output of the choke. If the voltage exceeds 300V, the resistor will start to pull the output voltage down. It shouldn't exceed 330V.
-3- Thermistors: With a choke input PS there seems to be little point.
-4- Standby switch: This also seems unnecessary in this application.
Errors corrected along the way:
- MUR860 are underrated;
- Hammond 157M are unsuitable.
Let's go back to the questions list:
-1- Choke on the ground side: Bad, bad, bad idea. This raises the CT of the transformer heavily and the choke sees a lot of voltage due to the transients anyway.
Fix: using a Hammond 193 with sufficient isolation rating (800v).
-2- Are 350V caps sufficient: yes but... if the choke is insufficiently loaded, the voltage will jump above V*0.9.
Fix: a string of three 100V/5W zeners + 1K/1W resistor at the output of the choke. If the voltage exceeds 300V, the resistor will start to pull the output voltage down. It shouldn't exceed 330V.
-3- Thermistors: With a choke input PS there seems to be little point.
-4- Standby switch: This also seems unnecessary in this application.
Errors corrected along the way:
- MUR860 are underrated;
- Hammond 157M are unsuitable.
Just a quick note.
I really like the Zener string to protect the Caps.
Thank you for pointing that out.
I will use it in my next design.
I was going to use more current in my bleeder resistors, but your solution is more elegant and causes fewer issues.
As for the thermistors, I am using them anyway. My start-up current is insane without help.
Doug
I really like the Zener string to protect the Caps.
Thank you for pointing that out.
I will use it in my next design.
I was going to use more current in my bleeder resistors, but your solution is more elegant and causes fewer issues.
As for the thermistors, I am using them anyway. My start-up current is insane without help.
Doug
I do not think so.
However, I am designing an amp with choke input. The design used 30 ma into bleeder resistors order to protect the caps with no tubes inserted. 30 ma at 330V is a bunch of heat and used 10% of the chokes current rating. Seemed excessive.
A lower current bleeder (1-2 mA) and the zener diode over-voltage protection seems "better" to me.
Doug
Hi,
since the choke input supply requires a defined load to ensure it gives the correct voltage (0.9 * Vac) then consider a shunt regulator instead of the Zener string. It can be fitted between the choke and the capacitors or after the capacitors.
There are some HV shunt regulators posted on the Forum.
If the output current drops to near zero then the capacitor voltage could rise to 1.4 * Vac * {1+reg/100}.
That will be instant explosion.
Maybe a belt and braces over-voltage crow bar set to 384Vdc?
since the choke input supply requires a defined load to ensure it gives the correct voltage (0.9 * Vac) then consider a shunt regulator instead of the Zener string. It can be fitted between the choke and the capacitors or after the capacitors.
There are some HV shunt regulators posted on the Forum.
If the output current drops to near zero then the capacitor voltage could rise to 1.4 * Vac * {1+reg/100}.
That will be instant explosion.
Maybe a belt and braces over-voltage crow bar set to 384Vdc?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.