EL-34's won't bias.

[Amled87]

Are you sure the o/p valve cathodes go to ground. and it's the same ground as the bias pot is connected to?

The op valves do go to ground, however the bias pot doesn't go to the same ground, I can move it though.

 

[johnR]

Even though each secondary is rated separately for more current then the amp can pull? Each secondary is rated for 400ma? Care to explain why the transformer would go if it's underloaded?

Sorry, I missed your post describing the transformer. That's a heavy duty transformer! It shouldn't burn out, but I'd still be inclined to revert to the original Marshall schematic as it will share the current load between both secondaries.

A rarely mentioned fact about 2204s is that Marshall changed the power transformers in the early '80s to increase the HT voltage. The older 2204s only gave around 35 watts output despite a 50W rating, and the newer ones gave closer to 60W. (I was a test engineer at Marshall in the mid '80s and have worked on many examples of both types).

Edit: Correction of my original post due to faulty memory. Both early and late JCM800 2204s used the bias scheme shown in the schematic posted by the OP. The 100W version (2203) used a bridge rectifier and a separate bias winding.

 

[abbey road d enfer]

That was just a rough drawing, I do have a standby switch implemented. The schematic I posted was just to get a working bias supply. I can't wrap my head around why the transformer when even at half current capacity (200mA or half of max rated 400mA) would go though. Because even at half current, it is in spec for the amp (it's actually closer to the spec called for as it only calls for a 150mA (213mA max) HV secondary power transformer.

So the original xfmr is rated at 213mA, for full-wave rectifying.
What you do is use half the xfmr, so 200mA for the same draw. It should work for 150mA but it's not the large headroom you seem to expect..

 

[Matador]

I would wire it up exactly as Marshall did it, with the dual secondaries with a common center tap, and full-wave rectifier. Then add the series diode and resistor to the capacitor, unconnected to anything, and make sure that you can derive a basic negative supply. Then add aspects of the circuit afterwards.

You must be using an Antek transformer, where they say not to use the two secondaries as separate windings (e.g. they say to use in series or parallel, but not isolated). I've actually used them this way many times, but it was mustly when pulling equal currents from each secondary.

The Marshall scheme should work for you, and make the most use of the copper in the transformer, however you can use a bridge rectifier, however you need to cap-couple the bias supply to get it to work. Details are here:

http://www.valvewizard.co.uk/bias.html

 

[abbey road d enfer]

however you can use a bridge rectifier, however you need to cap-couple the bias supply to get it to work. Details are here:

http://www.valvewizard.co.uk/bias.html

That's the 3rd possibility I mentioned in post #31. Making it work satisfactorily requires serious work.

 

[Gus]

There is something you need to keep in mind when using a cap coupled or cap divider.
Self healing metalized film caps can fail slowly, this will cause the value to decrease.
Work out what happens with lower cap values

 

[Tubetec]

Ive a handfull of the 50W Marshalls from the JMP era , I never saw one with a bridge rectifier in there , there is often two diodes in series in each wing of the full wave though and as you can see bias is tapped after the standby , which is problematic , it causes a high current draw at turn on which stresses everything unnessarily and causes a noise . The solution is easy , tap the bias off the hot side of the switch , now when you turn standby on bias is already where it needs to be .

I was caught out by the center tap once alright , I had replaced the mains TX with one from another amp which used a single winding ,bridge rectifier , with the standard full wave rectifier without the centre tap , I had no HT voltage . Again the fix was easy just needed to flip 2 of the 4 original diodes backwards and make the 0V connection .

 

[Johnny1234]

It does not matter what the bias supply is putting out voltage wise. The only thing to tube "sees" is the bias voltage from the control grid to the cathode. That is what is important and the meaningful voltage to measure.

 

[Amled87]

It does not matter what the bias supply is putting out voltage wise. The only thing to tube "sees" is the bias voltage from the control grid to the cathode. That is what is important and the meaningful voltage to measure.

And the issue that was present was there was nothing to measure at all as there was no negative bias voltage at all. It has since been corrected by using the second secondary to produce voltage for the bias tap and then using dropper resistors to get the correct voltage.

 

[Robinsonics]

In my experience with Marshall’s, this always one of two things (if the bias circuit is wired correctly). Either a dc blocking cap has failed, or the pcb has broken down and is now leaking DC into the bias circuit. Either way, the negative voltage gets negated by the time it gets to the tubes. It looks as if you’ve measured all the bypass caps, so I wonder if there is either a breakdown in the pcb, or a carbon trace on one of the output tube sockets that is allowing a HT leak.

 

[metalockpick]

If your CT is grounded, you could do it this way.
Drop resistor would go after diodes...

 

[Amled87]

In my experience with Marshall’s, this always one of two things (if the bias circuit is wired correctly). Either a dc blocking cap has failed, or the pcb has broken down and is now leaking DC into the bias circuit. Either way, the negative voltage gets negated by the time it gets to the tubes. It looks as if you’ve measured all the bypass caps, so I wonder if there is either a breakdown in the pcb, or a carbon trace on one of the output tube sockets that is allowing a HT leak.

As it's a turretboard and point to point construction I don't think the PCB has broken down my friend. However, these are all good tips when working on older PCB based Marshalls or any amp for that matter.

 

[Tubetec]

Throwing away half the current capacity of the transformer to feed a bias curcuit doesnt make any sense to me , even if a single HT winding is adequate for the job , the extra juice available on tap will make a difference at moments of peak power .

 

[Amled87]

Throwing away half the current capacity of the transformer to feed a bias curcuit doesnt make any sense to me , even if a single HT winding is adequate for the job , the extra juice available on tap will make a difference at moments of peak power .

The transformer is larger then needs to be to start with, so technically current is being wasted on both HT windings regardless. I could understand being concerned with "peak power" if this was a non-master volume amp and distortion and sustain was being generated by cranking the amp. But, it's not and there's almost nowhere you'd use this at "peak power" or cranked, today especially. Even in studio there's no need to crank the master above 4-6.

 

[Tubetec]

The best advice of several members who took the time to respond to your post is in black and white ,
your captain of your own ship , do what you want .

 

[Amled87]

The best advice of several members who took the time to respond to your post is in black and white ,
your captain of your own ship , do what you want .

My friend, I've stated I've got UF4007's on order and I'll be converting the amp to follow the schematic once they arrive. However, as it currently works (after being told it wouldn't repeatedly) without issues so far I believe the issue is fixed.

After some comments, some didn't even bother to look at the transformer documentation I provided, drawing or anything else. So, while I appreciate the help, only one user actually took the information I provided, what I had to work with and helped with that. Saying it won't work (it did actually) go back to original schematic (I plainly stated I planned to, but in the mean time this is what I had to work with, help with that) and etc didn't do much good. There were even some things like grounding center tap and brodge rectifier that was recommended (which would fry the power transformer) so while I've stated I appreciate the help, some of it was far from help.

Have a good day, now please let this thread die now.

 

[Tubetec]

I take umbrage at your use of the term 'friend' in that facetious manner ,
its fine I wont be responding to anymore of your posts .

 

[Amled87]

I take umbrage at your use of the term 'friend' in that facetious manner ,
its fine I wont be responding to anymore of your posts .

There's nothing facetious in what I said, there's no humor in it nor was it meant as such. If you're annoyed I call you "friend" well my friend, I'm sorry. I've said thank you, that I appreciate the help and asked to let the thread go now as it serves no purpose.

I simply stated facts, one user stated it wouldn't work others said it would. Users made "guesses" and etc based on transformers specs they made up without even looking at the data sheets provided. There is no guess work involved, that's how people die or best case scenario components fry. So, while you take umbrage at what I've said, reality is you have no reason to, but by all means get offended I didn't do it your way.

Good day

 

[plexi]

The issue was revealed in post #8 where you stated you were only using one side of the 2nd secondary winding - there was no return path for the bias power supply, so no voltage will be developed.

 

[abbey road d enfer]

I simply stated facts, one user stated it wouldn't work others said it would. Users made "guesses" and etc based on transformers specs they made up without even looking at the data sheets provided. T

You should admit that your communication was not very clear.
In your post #52, you left us hanging with "However, I assume the starting issue will be there and I'll have no negative voltage at the bias resistors again."
Disproof came only in your post #69.
Now, a bridge rectifier is perfectly capable of delivering a positive and a negative rail. This is very common in solid state electronics, where a CT xfmr in combination with a bridge rectifier actually are configured as a bipolar full-wave.
Replacing the negative side with discrete diodes is just redundant.
However, this implies using high voltage caps for the negative side, when using diodes and associated drop-down resistors allow using must lower nominal voltage caps.