I have just finished building my 3 watt mullard amp with 6v6 tubes instead of el84.
It fires up ok.
One channel did not produce sound at all, but i have found that there is a issue in the wiring from the rca input to the volume pot or from the volume pot to the ef86 tube. So thats gonna be fixed. So i pulled the tubes on that channel for now while i fix the channel that have sound first.
The other channel is playing. Really really low and distorts with just a little turn on the volume pot or if i turn up the power on the input signal. If i turn it really high it just crackles and nousy.i must have wired something wrong or some of the components i have used have some issues of some sort.good tips on where to check first?
It fires up ok.
One channel did not produce sound at all, but i have found that there is a issue in the wiring from the rca input to the volume pot or from the volume pot to the ef86 tube. So thats gonna be fixed. So i pulled the tubes on that channel for now while i fix the channel that have sound first.
The other channel is playing. Really really low and distorts with just a little turn on the volume pot or if i turn up the power on the input signal. If i turn it really high it just crackles and nousy.i must have wired something wrong or some of the components i have used have some issues of some sort.good tips on where to check first?
Schematic?
As you test the amplifier using normally GNFB, this shoul be disconnected at the beginning to see that all works ok. After this GNFB can be activated and adjusted to optimum.
As you test the amplifier using normally GNFB, this shoul be disconnected at the beginning to see that all works ok. After this GNFB can be activated and adjusted to optimum.
Generic experience: If there is one wiring error, there are probably more. They run in clusters.
Wiring error is far more likely than a bad part, unless you do the same build hundreds of times (factory labor), and even then wiring errors happen.
I have to check it step by step to see if i can find some more issues. And check voltages. To see if they are in the right ballpark too. Thanks for the feedback guys. Btw can a bad opt give such symptoms too or a bad tube? Nice to know if i should put that on the checklist too.
I assume you are wiring point to point. Please don't take offense if this is too obvious - however, I make a copy of the diagram and go wire by wire. I mark over each wire on the drawing with a highlighter as I check it. Whatever you can do that's pedestrian and systematic will help. Connecting to the wrong lug on a tube socket is pretty common.
I dont get offended so easy. 🙂 I will check the wiring again after a good night sleep and with a well rested mind.
The reason i suspect the opt can be a issue are that they come from a old radio . I have a lot of them. So swapping one out is no biggie.
Im really curious about what the issue might be.
Problemsolving and fixing it, are one of the parts in the learning process. So i know what to do and look for if i stumble across the same problem again. 🙂
It did not hum. so i must have done something right . Lol
The reason i suspect the opt can be a issue are that they come from a old radio . I have a lot of them. So swapping one out is no biggie.
Im really curious about what the issue might be.
Problemsolving and fixing it, are one of the parts in the learning process. So i know what to do and look for if i stumble across the same problem again. 🙂
It did not hum. so i must have done something right . Lol
Last edited:
So have you closely studied the differences in parameters between the 6V6 and the EL84. This amplifier was designed around the EL84 and is DC coupled.
The 6V6 is similar but not identical to the EL84. For example the EL84 has typically 3x the transconductance of the 6V6. (Which matters here)
I suspect beyond basic build issues there are probably some operating point issues as well. You may need to carefully adjust some component values to get it to work correctly.
When you wander away from the design intent you are potentially asking for trouble. First step is to measure all of the operating voltages and see how they compare with those listed for the EL84 - if they are close you'll probably be OK, but the overall internal voltage gain with the 6V6 will be lower which will result in higher output impedance and distortion.
The 6V6 is similar but not identical to the EL84. For example the EL84 has typically 3x the transconductance of the 6V6. (Which matters here)
I suspect beyond basic build issues there are probably some operating point issues as well. You may need to carefully adjust some component values to get it to work correctly.
When you wander away from the design intent you are potentially asking for trouble. First step is to measure all of the operating voltages and see how they compare with those listed for the EL84 - if they are close you'll probably be OK, but the overall internal voltage gain with the 6V6 will be lower which will result in higher output impedance and distortion.
No i have not. Because i was not aware until today that the 6v6 was not so similiar to the el84 as i thought in the beginning.
So i did not check the tube data so closely. But lesson learned. I think with tweaks and help from you guys it will run smoothly in the end. Im looking foward to hear the vintage brimar tubes in action . I have never heard a 6v6 tube in use before so. 🙂
So i did not check the tube data so closely. But lesson learned. I think with tweaks and help from you guys it will run smoothly in the end. Im looking foward to hear the vintage brimar tubes in action . I have never heard a 6v6 tube in use before so. 🙂
I use Mullard 6V6G in my little 6J7/6V6 design from a decade or so ago, they're quite nice. The Brimar should be very good.
Thats sounds promising 😀 i have alot of 6v6 tubes, used and
nos. so this amp will be the testpilot to see if all of them are working .
But im thinking. What would you have done? Finish this project with the 6v6 tubes. Or change the tube sockets , they are easily changed. And run the amp with el84 instead? I have not found a good 6v6 SE schematic yet. That was the reason i choose to use the 6v6 in this setup. I really liked the schematic and i had all the parts for it so. 🙂
That is one "interesting" circuit- 20V on the plate of an EF86. Should it really work any better than a conventional capacitor coupled output stage?
The EF86 operates in "starvation mode". The gain is about 3 x higher than 'normal'. This extra gain makes it possible to achieve an input sensitivity of 0.1 V while applying enough GNFB to achieve 1% distortion at 3 Watt.
The design was meant for crystal phono cartridges. The input sensitivity is rather high for normal sources.
For more details: Mullard 3-3. Three Watt Amplifier
The design was meant for crystal phono cartridges. The input sensitivity is rather high for normal sources.
For more details: Mullard 3-3. Three Watt Amplifier
Thanks for the background. Doesn't really appeal to me though. Still like a dual triode and coupling cap!
Check the plate voltage of the Left channel EF86, versus the plate voltage of the Right channel EF86.
Although the EF86 circuit is highly modified from the original Mullard circuit, the grid leak bias is a variable from tube to tube.
It is quite likely that they are not the same voltage, they may vary widely from one channel to the other.
(one tube versus the other tube, no matter which socket they are in).
And, they may be quite different than 20V.
Using an EF86, and a 10Meg Ohm g1 resistor to give grid leak bias, is just what you might expect . . .
a varying total bias voltage, and with the large tube gain, as a result the plate voltage might vary a lot from channel to channel.
Please measure the EF86 plate voltages, and report back to us.
Although the EF86 circuit is highly modified from the original Mullard circuit, the grid leak bias is a variable from tube to tube.
It is quite likely that they are not the same voltage, they may vary widely from one channel to the other.
(one tube versus the other tube, no matter which socket they are in).
And, they may be quite different than 20V.
Using an EF86, and a 10Meg Ohm g1 resistor to give grid leak bias, is just what you might expect . . .
a varying total bias voltage, and with the large tube gain, as a result the plate voltage might vary a lot from channel to channel.
Please measure the EF86 plate voltages, and report back to us.
Last edited:
@ satyrian.
Oh dear, the Mullard 3-3 design has confounded many builders in the past. For every one that works 2 fail... It is far more complicated than it would appear.
As 6A3s says it is very critical on the valves used and that is following the EF86 and EL84 original. The 6V6 is not an equivalent of the EL84 as you have found out.
I think you have very little (no) chance of making it work in the Mullard 3-3 scheme...
Surprised you cannot find a simple single ended 6V6 design, there must be plenty on the web. Or with a little imagination and a few different components you could change your build into a simple uncomplicated amp like this:
It is straight off my head so others please comment / correct!
Oh dear, the Mullard 3-3 design has confounded many builders in the past. For every one that works 2 fail... It is far more complicated than it would appear.
As 6A3s says it is very critical on the valves used and that is following the EF86 and EL84 original. The 6V6 is not an equivalent of the EL84 as you have found out.
I think you have very little (no) chance of making it work in the Mullard 3-3 scheme...
Surprised you cannot find a simple single ended 6V6 design, there must be plenty on the web. Or with a little imagination and a few different components you could change your build into a simple uncomplicated amp like this:
It is straight off my head so others please comment / correct!
Attachments
Update.
First fault fixed. One wire going to the potmeter had a small tiny strain touching ground. So that was a easy fix.
Next issue i found was that i forgot to add the 10M resistors to the rca input. That would explain some of the issues i have experienced. So im gonna solder them in and test the amp again to see what kind of issues are still remaining .
First fault fixed. One wire going to the potmeter had a small tiny strain touching ground. So that was a easy fix.
Next issue i found was that i forgot to add the 10M resistors to the rca input. That would explain some of the issues i have experienced. So im gonna solder them in and test the amp again to see what kind of issues are still remaining .
If it gets too hard to get the amp working smoothly with the 6v6 tubes. The sockets are easy changeable as you can see in the photo. I have 15 el84 tubes laying around and a small bucket full of brand new tube sockets i got as payment for a repair i did on a amp. So piece of cake to change them.