Stumped on bias problem

[luthierwnc]

Hi Guys,
I've got a conundrum that ought to be really simple but I can't get past it.  Attached is a diagram of the bias circuit.  I am getting no voltage of any kind after the diode in both standby and full power.  VDC and VAC measurements are spot-on at the bridge.  I've checked the resistors, replaced the 1N4007, tried it with the wire removed to the master volume and reversed the leads on the DVM to be sure I was getting negative volts after the diode.  The new cap measures 51UF and doesn't leak (positive is tied to ground).  With no cap at all I still get 0 volts DC after the diode.
What am I missing?  Thanks for your indulgence, Skip

[tubeswell]

You need an AC-coupled bias supply for that type(*) of power supply. See bottom of the page on this link: http://www.valvewizard.co.uk/bias.html


* i.e. where you have a FW bridge rectifier


Adjust the size of the coupling cap to define the bias voltage range

[luthierwnc]

Thanks Tubeswell!  Worked great.  sh

[2deaf]

The present location of the standby switch will allow the bias supply to discharge when the switch is open.  It takes a long time for that bias circuit to charge once the switch is closed during which time the output tubes will be biased hot.  I recommend moving the switch as shown so that the bias supply stays charged during standby.
 

[luthierwnc]

made the changes.  thanks all

[tubeswell]

The charging time is actually quite fast as the input cap is a relatively tiny capacitance - almost instantaneous. The EC Tremolux uses a similar AC-coupled bias supply and the bias voltage doesn't switch on until you switch the standby on (the standby in that amp is on the HT centre tap).

[2deaf]

The charging time is actually quite fast as the input cap is a relatively tiny capacitance - almost instantaneous.

The JCM900/Ampeg one?  The charging time is painfully slow because the tiny input cap has to charge up and then discharge into the much larger reservoir cap every cycle.  To make matters worse, the input cap only charges for a fraction of each cycle and only discharges for another fraction of each cycle.

[tubeswell]

The charging time is actually quite fast as the input cap is a relatively tiny capacitance - almost instantaneous.

The JCM900/Ampeg one?  The charging time is painfully slow because the tiny input cap has to charge up and then discharge into the much larger reservoir cap every cycle.  To make matters worse, the input cap only charges for a fraction of each cycle and only discharges for another fraction of each cycle.

EC Tremolux

[AmberB]

The Bogan tube PA amps such as the CHB-35A use a voltage doubler for the main B+, and they use an AC coupled negative bias for the power tubes.  The AC coupled bias voltage uses a 0.22uf film cap of sufficient voltage to pass the AC through without the DC.

[PRR]

> EC Tremolux

That one as drawn here, about 1/6 seconds. Plenty fast.

[2deaf]

> EC Tremolux

That one as drawn here, about 1/6 seconds. Plenty fast.

It took the better part of a second to reach 63.2% of full bias voltage when I tested it.

[AmberB]

Hi Guys,
I've got a conundrum that ought to be really simple but I can't get past it.  Attached is a diagram of the bias circuit.  I am getting no voltage of any kind after the diode in both standby and full power.  VDC and VAC measurements are spot-on at the bridge.  I've checked the resistors, replaced the 1N4007, tried it with the wire removed to the master volume and reversed the leads on the DVM to be sure I was getting negative volts after the diode.  The new cap measures 51UF and doesn't leak (positive is tied to ground).  With no cap at all I still get 0 volts DC after the diode.
What am I missing?  Thanks for your indulgence, Skip

If you put a 0.22uf-400 volt film cap between the 100k-2watt resistor and the bridge rectifier in the spot where you have the resistor connected, that should solve the bias voltage problem.  You'll probably have to mess with resistors after the bias rectifier diode to get the voltage you want for the negative bias voltage.  Depending on what DC voltage you get to start with, you might want to change the resistance of the 2 watt resistor to adjust the basic voltage and then adjust further from there.  If you really wanted to, you could add a bisa voltage adjuster pot to fine tune your bias voltage...

[pdf64]

> EC Tremolux

That one as drawn here, about 1/6 seconds. Plenty fast.

The bias supply in post 1 has a 100k dropper and a 47uF reservoir. If all that changes to get it working is adding a largish dc blocking cap in series with the 100k dropper, wouldn’t its rise time be somewhat longer than that of the EC Trem, which has no series dropper resistor and a 10uF reservoir?

[DummyLoad]

Hi Guys,
I've got a conundrum that ought to be really simple but I can't get past it.  Attached is a diagram of the bias circuit.  I am getting no voltage of any kind after the diode in both standby and full power.  VDC and VAC measurements are spot-on at the bridge.  I've checked the resistors, replaced the 1N4007, tried it with the wire removed to the master volume and reversed the leads on the DVM to be sure I was getting negative volts after the diode.  The new cap measures 51UF and doesn't leak (positive is tied to ground).  With no cap at all I still get 0 volts DC after the diode.
What am I missing?  Thanks for your indulgence, Skip

why the FWB? why not use a FW with that PT? no cap coupled peaking ckt. needed - see attached.

--pete

[tubeswell]

> EC Tremolux

That one as drawn here, about 1/6 seconds. Plenty fast.

It took the better part of a second to reach 63.2% of full bias voltage when I tested it.

I've got one on the bench. Its pretty instantaneous

[luthierwnc]

Thanks guys. 


Dummyload; This is based on The Cesspool (renamed the Narrow Gauge).  I had no idea how he got the artificial center-tap on his schematic so I just went with a bridge rectifier.  Good to know!  I'm getting about 340VDC on the plates with a couple of old (perhaps too old) 6V6s biased around 24 ma. 


The gain was too much, especially coming off the .0047 option on the second triode split-plate resistor array.  I dropped in a 100k trimmer and a 47k leg out to the grid leak of the third triode, dropping the 150k to a sweeter-spot of 120k.  There is a 50 volt difference between the PI plates (100k each) so I may try to get those closer.  Bright is too bright so I think that will get a 56p cap rather than the 100p in there now.  The bass control is very touchy on this dude.  It has a master but I'll leave that wide open -- I should know better than to think I will use a MV at this stage of life.  So far I've only played it with a Strat in the basement shop.  I think it is a bucker amp in the wider world.


Gentlemen, thanks for all your work!  This poor chassis has been three builds and has some extra holes to prove it.  Now that it graduations to the collection, I'll need to make a cab.



Cheers and good luck with all your projects, sh 

[2deaf]

If you put a 0.22uf-400 volt film cap between the 100k-2watt resistor and the bridge rectifier in the spot where you have the resistor connected, that should solve the bias voltage problem.

The bias diode will be reverse biased at all times whether the 0.22uF capacitor is there or not.  The 0.22uF cap will never charge one way or the other because the reverse biased diode constitutes an open circuit.

[PRR]

The plans in post #1 and post #7 are two different things.

#1 plots at about 1 second for 63%.

[2deaf]

The plans in post #1 and post #7 are two different things.

#1 plots at about 1 second for 63%.

Post #1 is the original post and the plan shows a bias circuit with a diode that is never forward biased.  You're going to be waiting a really long time for that one to reach 63%.

Post #7 is reply #6.  I didn't attach any plan to that post.

Post #8 is reply #7 and it shows the EC Tremolux bias.  That one charges up to 63% in a little less than one second.

[PRR]

> Post #1 ....You're going to be waiting a really long time for that one to reach 63%.

You are correct. Sorry.