Greetings all,
This is my first attempt to build/repair a tube amp.
I've spent a few weeks fixing previous problems on this chassis which ranged from incorrect tubes, blown cathode capacitors, Negative feedback which was hooked up as Positive feedback, Hum that just wouldn't die!, a choke that was supposed to be 15H, but was 0.8H (that one took some reading) and now I have an amp that works, but not very well.
It seems to be underpowered and frequency response seems to be from 20Hz but drops off quickly after 7.5Khz
It's an old chassis I found, it's essentially built around the Leak TL12 chassis with off the shelf components in a PP configuration, the main exception is that the power supply transformer and choke is built on a seperate chassis.
Tubes compliment is 2x6L6GC, 1xECC33 and an EF37 (was previously a 6J7GT)
B+ is 445v, 125ma using a GZ32 and a 15H choke.
Output transformer is a Beacon 48s61 ( 807 type, 10K primary, 70-20KHz, 2-16ohm output imp. Push-Pull 15watt)
I'm using the same circuit design as found here
It seems to output 8.5VRMS @1Khz @0.8% distortion before things start to head south, at 9VRMS distortion is at 2% and there is visible clipping on the bottom 1/2 of the sine wave, interesting enough if I put the scope on the anode of the EF37 there is a little bump on either peak of the cycle (approx 0.44V peak to peak), which gets progressively steeper if I push the amp harder.
Looking at the output on the scope there seems to be a definite clip at 9-10volts RMS with the output transformer configured for an 8ohm load, connected a 50w 8ohm non inductive dummy load.
Unless I have my maths wrong or made an assumption that could be totally wrong? This should work out as follows, I = V/R = 8/8 = 1amp
P=VI = 8*1 = 8WRMS output, a bit under expectation I would have thought?
So I thought I would check voltages, according to the Leak schematic...
6L6 cathode voltages are ~ 36-37v which seems right, so are the Anode voltages at 440v
ECC33 Anode voltages are way too high at ~310v each, they should be 250v and 260v (PS. I tried a 82k Resistor at R12 to try and lower the voltage, which didn't really happen), cathode voltages are 60v thich is correct.
EF37 voltages are 104v at the anode which is too low, 180v is correct, 50v at pin4 (Grid2?) which should be 35v and 1.63v at the cathode which should be 1.3v.
I cant find any oscillation on the scope, have tried 3-4 different tubes, replaced every single capacitor and resistor and still these voltages persist.
Can anyone suggest some tips? (apart from throwing it off a tall building)
Cheers.
This is my first attempt to build/repair a tube amp.
I've spent a few weeks fixing previous problems on this chassis which ranged from incorrect tubes, blown cathode capacitors, Negative feedback which was hooked up as Positive feedback, Hum that just wouldn't die!, a choke that was supposed to be 15H, but was 0.8H (that one took some reading) and now I have an amp that works, but not very well.
It seems to be underpowered and frequency response seems to be from 20Hz but drops off quickly after 7.5Khz
It's an old chassis I found, it's essentially built around the Leak TL12 chassis with off the shelf components in a PP configuration, the main exception is that the power supply transformer and choke is built on a seperate chassis.
Tubes compliment is 2x6L6GC, 1xECC33 and an EF37 (was previously a 6J7GT)
B+ is 445v, 125ma using a GZ32 and a 15H choke.
Output transformer is a Beacon 48s61 ( 807 type, 10K primary, 70-20KHz, 2-16ohm output imp. Push-Pull 15watt)
I'm using the same circuit design as found here
It seems to output 8.5VRMS @1Khz @0.8% distortion before things start to head south, at 9VRMS distortion is at 2% and there is visible clipping on the bottom 1/2 of the sine wave, interesting enough if I put the scope on the anode of the EF37 there is a little bump on either peak of the cycle (approx 0.44V peak to peak), which gets progressively steeper if I push the amp harder.
Looking at the output on the scope there seems to be a definite clip at 9-10volts RMS with the output transformer configured for an 8ohm load, connected a 50w 8ohm non inductive dummy load.
Unless I have my maths wrong or made an assumption that could be totally wrong? This should work out as follows, I = V/R = 8/8 = 1amp
P=VI = 8*1 = 8WRMS output, a bit under expectation I would have thought?
So I thought I would check voltages, according to the Leak schematic...
6L6 cathode voltages are ~ 36-37v which seems right, so are the Anode voltages at 440v
ECC33 Anode voltages are way too high at ~310v each, they should be 250v and 260v (PS. I tried a 82k Resistor at R12 to try and lower the voltage, which didn't really happen), cathode voltages are 60v thich is correct.
EF37 voltages are 104v at the anode which is too low, 180v is correct, 50v at pin4 (Grid2?) which should be 35v and 1.63v at the cathode which should be 1.3v.
I cant find any oscillation on the scope, have tried 3-4 different tubes, replaced every single capacitor and resistor and still these voltages persist.
Can anyone suggest some tips? (apart from throwing it off a tall building)
Cheers.
You did measure it at 50/60Hz and not with a DMM (that often uses 1KHz), correct?
Cheers!
I built a derivative of a similar circuit using EF37s, 6SN7 and 807s and it sounded amazing- there is definitely something wrong.
Things I noticed:
1) EF37s were set up in this amp to give ridiculously high gain, which is not necessary today. On the original TL12 circuit which is MUCH easier to read, (link below), it has the anode voltage of the EF36/EF37 as 80V rather than 180V so yours sounds about right, but 470K is a massive plate resistor- try 270K instead.
http://www.amplimos.it/images/Leak TL12.bmp
2) I wouldn't worry too much about the ECC33 anode voltages, they're pretty hardy. You are using a vintage ECC33, not a new chinese one right? Don't change the anode load resistors, they are pretty crucial.
However you could try adding an R-C stage in the power supply, right where it says turneraudio.com.au in the schematic, just add about a 5K resistor in series with around 20uF 450V cap to ground after it. This removes any potential feedback thru the power supply and gives the driver stage it's own reservoir cap.
3) At a glance, I'd try changing R21 and 22 for 1K or 5K resistors to the output stage- however this probably won't make much difference.
4) How does it sound with the feedback loop disconnected?
Things I noticed:
1) EF37s were set up in this amp to give ridiculously high gain, which is not necessary today. On the original TL12 circuit which is MUCH easier to read, (link below), it has the anode voltage of the EF36/EF37 as 80V rather than 180V so yours sounds about right, but 470K is a massive plate resistor- try 270K instead.
http://www.amplimos.it/images/Leak TL12.bmp
2) I wouldn't worry too much about the ECC33 anode voltages, they're pretty hardy. You are using a vintage ECC33, not a new chinese one right? Don't change the anode load resistors, they are pretty crucial.
However you could try adding an R-C stage in the power supply, right where it says turneraudio.com.au in the schematic, just add about a 5K resistor in series with around 20uF 450V cap to ground after it. This removes any potential feedback thru the power supply and gives the driver stage it's own reservoir cap.
3) At a glance, I'd try changing R21 and 22 for 1K or 5K resistors to the output stage- however this probably won't make much difference.
4) How does it sound with the feedback loop disconnected?
Last edited:
I measured it with a DMM which measures inductance, not sure of the frequency it uses but measurements I've made so far on other chokes seem to be true to label, Hadn't really considered that to be honest, thanks for the heads up.
Have tried the 270k at R4, voltage went to 205v is I'll have to try something in the middle.
Took the feedback off and put a 10k resistor on the input, distortion went up to 5-6%
It sounds a bit grittier with the feedback off, gain is so high I only need to move the volume control 2-3 mm from zero to get full output.
Do you think I should consider a lower gain front end tube?
Took the feedback off and put a 10k resistor on the input, distortion went up to 5-6%
It sounds a bit grittier with the feedback off, gain is so high I only need to move the volume control 2-3 mm from zero to get full output.
Do you think I should consider a lower gain front end tube?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.