finally built it. big amp troubleshooting.

[birt]

a long long time ago (2010) i started drawing a conversion amp for my Super Twin Reverb.
about one year after that i started building it and stopped about halfway due to a lack of time. a headshell with half populated chassis started gathering dust until... now.

a little recap:
2 parallel preamps: Fender studio bass, Sunn 2000s and Ampeg V4B with some extra stuff. super twin/super bass poweramp.

first time i flipped the switch (no powertubes) there was a big voltage drop in the power supply. i narrowed it down to a wiring error after the first preamp node. short to ground.
when that was fixed i messed around with the resistors in the power supply to get the voltages where i wanted them.
then installed power tubes and set the bias.

there is sound! hooray!
but...issues:

-Sunn sounds quite big but it's distorted and not very loud. the bass boost switch works.

-Ampeg sounds cleaner but also too quiet. the high switch works.
the ultra low switch is a middle off 2 way switch. one way bridges the 1M to ground and the .001 cap between gainstage and volume. (this cuts volume for some reason) the other way does this as well and bridges the 47K at the cathode cap of the second gainstage. (seems to work)
the mid control on the last gainstage doesn't sound very effective but i think i need to fix the volume problem first.

-Fender... with volume and master at full there is a very faint sound when i let a chord ring. almost nothing. wiring error maybe.


Voltages:
-at the standby switch: 545Vdc
-bias: -58,4Vdc (around 30mV measured over 1 ohm cathode resistors)
-12AT7 plates: 400Vdc cathodes: 88Vdc
-12AX7 PI plates: 300Vdc cathodes: 91Vdc
-preamp node A (Ampeg): 330Vdc
-preamp node B (Sunn): 322Vdc
-preamp node C (Fender): 320Vdc

-the fuse drawn at the rectifier is not installed yet.
-the 39K 3W resistors over the 220uF caps are not needed but i kept them as they are in the original schematic. they drain the caps when standby is open.

[birt]

more inside, i know it's not the tidiest wiring and most people prefer to stick to 1 type of caps; not me 

[birt]

but the messiest part is still the original fender power amp

[Ed_Chambley]

Looks really good and I really like the idea of an Ampeg and Fender input.  Interested in hearing you thought on the SUNN after completion since SUNN amps PA section is quite different.


The only caps you find in my amps that are the same are filter caps.  Resistors are different too, except droppers are usually always ohmite.

[sluckey]

There is a 120K tail resistor under the PI tube cathodes. That seems extreme to me. I wonder if that would strangle the PI. I know the original schematic shows a 120K there. Was there indeed a 120K in your original Super Twin Reverb?

I'd want to know if the amp comes alive with a 22K tail resistor.

[birt]

i went looking for pictures of the super twin before conversion to look for the original CC resistors. The 120K was there so it has worked that way.

but i could try putting a 27K resistor over it today with alligator clips (that would make it 22K) to see what that does.

[Ed_Chambley]

Try it can't hurt,


I did not notice your PI, but it is different.

[birt]

first i got the fender preamp working. the input wire and the input of the second gainstage (12AX7 pin 2 and 7) were switched around. and on the second gainstage only the cap was connected to the cathode. a little wire bridge between cap and 2K7 and the problem was solved.

now for the volume... with all controls halfway the fender is the most quiet. ampeg and sunn are closer together with ampeg as the loudest. maybe i should change the fender master to 1Mlog as well instead of 100Klog?

the sunn preamp has a crunchy rock sound and the other 2 are clean. i'm not sure if this is normal? i thought the sunn preamp would be clean as well...

changing the PI tail resistor from 120K to 22K is hardly noticable in the sound. to give you an idea of overall volume: my B15N (with 6L6GB) on 3 is way louder than this monster on 5.

[birt]

On the fender preamp both the mid and bass control were wired backwards. I fixed that but it still sounds a lot wimpier than the others. Less full. Either there is still something wrong or i don’t like this tonestack.


The issue with the low switch for the ampeg was the 6,8uF cathode cap wired backwards.
The mid control doesn’t really do anything i can hear. That needs some tweaking as well.


I added an F 1A fuse on both HT secondaries before the rectifier.


Could the low volume have something to do with my limited understanding of UL poweramps?
Voltages at the power tubes are:
Pin 3: 519V
Pin 4: 521V
Pin 5: -57.8V
Pin 8: 31mV 29mV 28mV 26mV 32mV 29mV

[Ed_Chambley]

On the fender preamp both the mid and bass control were wired backwards. I fixed that but it still sounds a lot wimpier than the others. Less full. Either there is still something wrong or i don’t like this tonestack.


The issue with the low switch for the ampeg was the 6,8uF cathode cap wired backwards.
The mid control doesn’t really do anything i can hear. That needs some tweaking as well.


I added an F 1A fuse on both HT secondaries before the rectifier.


Could the low volume have something to do with my limited understanding of UL poweramps?
Voltages at the power tubes are:
Pin 3: 519V
Pin 4: 521V
Pin 5: -57.8V
Pin 8: 31mV 29mV 28mV 26mV 32mV 29mV

If your power stage is still as designed there is not a lot to understand except how UL taps for screens are made and at what percentage the tap. Once you learn all this it  still works the same way. My point being if the output section .worked and has not changed it should work the same

[birt]

that's what i was thinking.
the power section is also identical to that of a Studio Bass. The Fender preamp i used is copied from that same amp. (and the 300PS uses it as well) The only difference is the 220K resistor after the master. would that bring down the volume that much?

all the voltages seem OK to me.

my next test will be to try these preamps on a different power amp and maybe a different preamp on this power amp.

[Ed_Chambley]

that's what i was thinking.
the power section is also identical to that of a Studio Bass. The Fender preamp i used is copied from that same amp. (and the 300PS uses it as well) The only difference is the 220K resistor after the master. would that bring down the volume that much?

all the voltages seem OK to me.

my next test will be to try these preamps on a different power amp and maybe a different preamp on this power amp.

That will give an answer. You can eliminate the PA.
Maybe tweak one channel at a time if you find your output is good.

[birt]

the big volume loss seems to be in the passive mixer before the PI. if i bridge one of those 220K resistors that channel gets much louder.

but i can't make those values lower because the masters will start interfering with each other. the fender and sunn masters already do that a little bit. so i might have to make them 1M like the ampeg master. in the sunn there is a 1M resistor to ground right after the treble pot. what purpose does this have? filtering in combination with the .22 right after?

does anyone have an idea on how to make the channel mixer better? or how i can recover some gain after mixing? the 12AT7 is 2 cathode followers to drive the power tubes right? does it make sense to make those gain stages? or is it more current than gain that is needed there?

[SoundmasterG]

A better way to mix the channels is to use a gain stage with local NFB for each channel. You can vary the amount of local NFB to equalize the gains from each preamp. Then you have all three sum before the phase inverter. If you have any of the KOC books, he has circuits for how to do that in TUT1 and TUT2. It would require three triodes but it allows for the best flexibility. Another option would be one of those pentagrid tubes that have quite a few control grids. Those were meant for mixing duties. Either of those methods allows for less gain loss.  Failing those, then you have to use the resistors as you are to mix the channels. Or use solid state.

Greg

[VMS]

does anyone have an idea on how to make the channel mixer better? or how i can recover some gain after mixing? the 12AT7 is 2 cathode followers to drive the power tubes right? does it make sense to make those gain stages? or is it more current than gain that is needed there?

If you're only playing one channel at a time i would use a on-on-on switch or a rotation switch to choose the channel.

[VMS]

...also if you have half of 12au7 unused maybe you can make a mixer with that tube.


Maybe something that is similar to the circuit on this page (not sure if it works with 12au7):


http://aikenamps.com/index.php/designing-single-stage-inverting-feedback-amplifiers

[SoundmasterG]

does anyone have an idea on how to make the channel mixer better? or how i can recover some gain after mixing? the 12AT7 is 2 cathode followers to drive the power tubes right? does it make sense to make those gain stages? or is it more current than gain that is needed there?

If you're only playing one channel at a time i would use a on-on-on switch or a rotation switch to choose the channel.

If he was going to go that route, he should also mute the input of the unused channels to reduce noise. Solid state would work for that purpose in addition to a switch or relays, and would likely be cheaper to implement.


Greg

[sluckey]

If he was going to go that route, he should also mute the input of the unused channels to reduce noise. Solid state would work for that purpose in addition to a switch or relays, and would likely be cheaper to implement.

His Mix Input jack already grounds any unused inputs.

[birt]

the idea here is to be able to use the preamps parallel or separate as i wish.
either with the mix input jack that feeds all of them or by plugging in one of the separate input jacks. (if not all are used the others are grounded as sluckey pointed out)

VMS; i don't have half of a 12AU7 because the ampeg preamp actually uses a 6K11 for the 3 triodes ;-)

if i would need to add a tube that pentagrid suggestion got me thinking... would it be possible to use a small signal pentode in triode mode but using the screen grid as a second input? so i would mix the fender and sunn together. (i know a double triode works here, it's just a thought)
the ampeg preamp would then feed into the now unused half of the PI as it is out of phase with the other 2. not sure how this works with the NFB though. i've only used both PI inputs on an amp without NFB.

[VMS]

How does the ampeg channel sound parallel with other channels, to me it looks like it's out of phase with the other two?

[birt]

How does the ampeg channel sound parallel with other channels, to me it looks like it's out of phase with the other two?

it is. and that makes for odd/interesting sounds depending on where volume controls are set. i am tweaking the sound of that channel still, especially that last triode (it's not good as drawn on the schematic, bot cathode resistors should be tied to either side of the pot with the wiper grounded). and the 1M grid leak on that triode can go as well i guess. there is still the possibility of that triode opening up for another function..

also, i took the ultra low switch from a B25b and i'm not happy with that yet either. the upper part of that switch (on the schem, .1, 68K to ground) puzzles me. is it supposed to be grounded when ultra low is engaged or cut off ground? (1M)

[VMS]

the ampeg preamp would then feed into the now unused half of the PI as it is out of phase with the other 2. not sure how this works with the NFB though. i've only used both PI inputs on an amp without NFB.

I don't know if it works with NFB, at least i have never seen it done with it.

also, i took the ultra low switch from a B25b and i'm not happy with that yet either. the upper part of that switch (on the schem, .1, 68K to ground) puzzles me. is it supposed to be grounded when ultra low is engaged or cut off ground? (1M)

My guess is grounded when ultra low is engaged

[SoundmasterG]

If he was going to go that route, he should also mute the input of the unused channels to reduce noise. Solid state would work for that purpose in addition to a switch or relays, and would likely be cheaper to implement.

His Mix Input jack already grounds any unused inputs.

I didn't look at it that closely yet as I had to finish work first, but thank you for pointing that out. Saves me time. :)

Greg

[birt]

so my mixer resistors are larger than my masters. this means the path to ground has a lower resistance than the path to the PI.
looking at passive mixers the mix resistors always seem to be one tenth to one third of the pot value. and the pots are all the same value.

in my amp this means i would make all masters 500K log with 50-100K mix resistors. the only issue i see is with the sunn preamp because the tone stack is after the second triode. i guess i would need to scale up the treble pot to 1M and this will mess with the frequency response of the tone stack. i just don't know how much

[VMS]

just throwing this idea here:


What if you move the mid control on the ampeg channel to the second triode cathode and then master volume after the tonestack and then mix all the channels before the third triode of the ampeg channel. This would at least solve the phase issue 

[birt]

just throwing this idea here:


What if you move the mid control on the ampeg channel to the second triode cathode and then master volume after the tonestack and then mix all the channels before the third triode of the ampeg channel. This would at least solve the phase issue 

the phase issue doesn't bother me so much. i'd rather make them sound good as they are now and fix the masters. then i could look at maybe matching the phases. if i'd go for your idea i would probably move the master before the third triode and add one extra 12AU7 tube so each channel has a triode with close to unity gain after their master. (and possibly using that triode not only to reverse the phase but also to get them at the same output level)

i haven't been able to do anything anyway as i'm on tour at the moment.

[VMS]

Changing fender channel master to 1MA pot will give little more gain for that channel. Did you try that yet?

[VMS]

in general when you mix channels with resistors you get attenuation, mixing two channels with same size resistors attenuation is 0.5 and mixing three channels with same size resistor attenuation is 0.333333 because of the two to one voltage divider.

[birt]

and months later here i am again. mixer pots are 500K, mix resistors 100K. the sunn and ampeg channels are loud now. the fender channel is still much weaker and sounds quite thin so i suspect there is something else going on there.

the mid control on the ampeg channel works but is subtle. that channel sounds clean and BIG. it has an incredible amount of bass on tap.
the sunn channel also has quite a bit of bass and has got some overdrive with all controls halfway. i expected sunn to be clean but it doesn't sound bad at all..