Nothing jumps out yet. It is indeed odd behavior. You said you dropped a 5k resistor in - where exactly? You have switched the EL84 in the offending channel, and also channel to channel, right? Also, where did you measure the “2.3 ma bias” for the 12AX7? I assume this is for both triodes; this seems a bit high.
If you could measure the plate, screen, grid and cathode voltages of the EL84s that might be helpful in the diagnosis.
If you could measure the plate, screen, grid and cathode voltages of the EL84s that might be helpful in the diagnosis.
Thanks for the response
I replaced the 500R potentiometer at R5 for a 5K potentiometer
I measured the voltage drop across R18 to get the 2.3mA
I did not swap out the EL84 , they bias to 34mA without any trouble at all , I can try and switch them to see if this behavior follows these tubes.
I replaced R17 with an led and used 16K8 for R16 ( as per the table )
I will measure these other voltages when I get home this evening.
with the 5K in the R5 position - it seems to hold it's voltages over time. I cannot get either channel to absolute zero, there is about 20-30mV variation
( the output is loaded with a 10R 20W resistor ) but the input is open ( not shorted to ground )
appreciate the guidance / help
..dB
I replaced the 500R potentiometer at R5 for a 5K potentiometer
I measured the voltage drop across R18 to get the 2.3mA
I did not swap out the EL84 , they bias to 34mA without any trouble at all , I can try and switch them to see if this behavior follows these tubes.
I replaced R17 with an led and used 16K8 for R16 ( as per the table )
I will measure these other voltages when I get home this evening.
with the 5K in the R5 position - it seems to hold it's voltages over time. I cannot get either channel to absolute zero, there is about 20-30mV variation
( the output is loaded with a 10R 20W resistor ) but the input is open ( not shorted to ground )
appreciate the guidance / help
..dB
I took some measurements after switching out the tubes - same behavior.
EL84:
Pin2 - control grid = -10V and -11V for each tube respectively
Pin3 - Cathode = 0.36V and 0.35V
Pin7 - Anode = 301V and 301V (B+ 316V as measured across C5)
Pin9 - Screen grid - at plate voltage 301
12AX7
Input triode section:
Pin1 - 43.6V ( hmm - going to have to check this one again )
Pin2 - mV start at about 30 and drops as I measure to 1.5mV
Pin3 - 5.8V
second triode section
Pin6 - 43.4V ( again - this is wrong - perhaps we have found the trouble )
Pin7 - mV
Pin8 - -1V
It would appear that R7 or R9 on this board is wrong - based on very low plate voltage for the 12AX7
..dB
EL84:
Pin2 - control grid = -10V and -11V for each tube respectively
Pin3 - Cathode = 0.36V and 0.35V
Pin7 - Anode = 301V and 301V (B+ 316V as measured across C5)
Pin9 - Screen grid - at plate voltage 301
12AX7
Input triode section:
Pin1 - 43.6V ( hmm - going to have to check this one again )
Pin2 - mV start at about 30 and drops as I measure to 1.5mV
Pin3 - 5.8V
second triode section
Pin6 - 43.4V ( again - this is wrong - perhaps we have found the trouble )
Pin7 - mV
Pin8 - -1V
It would appear that R7 or R9 on this board is wrong - based on very low plate voltage for the 12AX7
..dB
That's a huge disparity in bias voltages. Are you certain that the valve is OK? If so, something is not really as we think that it is.
YOS,
Chris
Hi db, there are a few issues with the PSU board. Apart from the placement of the MOSFET and the value of the electrolytic that is part of the delay startup timing function, there are also some resistors that need to be altered to set the bias supply in the right range so that it can be adjusted on the board.
Since the PSU is designed for the Toroidy transformer TSTA 250/001, which has the correct B+ for the EL34 version, and there is some duplication of PSU components on the BH EL84 boards, there is a case for just using the on board power supply components for the BH EL84 and foregoing the delayed startup.
The correct component values are probably buried in the BH EL34 thread. I'd check my board, but I am away from home. I'll continue to look, but just a heads-up.
I did intend to start a thread just for the PSU, and I thought I added some useful notes in the Wiki, but must have been a senior moment ;-/
Richard
Since the PSU is designed for the Toroidy transformer TSTA 250/001, which has the correct B+ for the EL34 version, and there is some duplication of PSU components on the BH EL84 boards, there is a case for just using the on board power supply components for the BH EL84 and foregoing the delayed startup.
The correct component values are probably buried in the BH EL34 thread. I'd check my board, but I am away from home. I'll continue to look, but just a heads-up.
I did intend to start a thread just for the PSU, and I thought I added some useful notes in the Wiki, but must have been a senior moment ;-/
Richard
Thanks Roberto - I had based the 2.5mA on the schematic in the Wiki
If I am reading this correctly, it is just a matter of changing the value of R18 , so that the voltage drop across this resistor is 1.2mA
I had other work to do last night so didn't get to work on the amp but quickly measured the "working" channel this morning and the plate voltage on the 12AX7s is also low - 60V
..dB
and forgot to respond to @OldHector
- yes , I have modified the psu board to suite the EL84 and am not using the bias supply for the EL34 but using the onboard bias for the each EL84 board. It is working well and I like the soft start option it provides ( and the use of a single choke )
If I am reading this correctly, it is just a matter of changing the value of R18 , so that the voltage drop across this resistor is 1.2mA
I had other work to do last night so didn't get to work on the amp but quickly measured the "working" channel this morning and the plate voltage on the 12AX7s is also low - 60V
..dB
and forgot to respond to @OldHector
- yes , I have modified the psu board to suite the EL84 and am not using the bias supply for the EL34 but using the onboard bias for the each EL84 board. It is working well and I like the soft start option it provides ( and the use of a single choke )
Last edited:
You are welcome. Yes, if you double R18 you'll get half the current.
Edit:
You can also try to find the sweet spot with a trimmer. I don't remember the right value I have in my amps, but let's calculate it:
Red Leds have around 2 V across them, 2N5551 have 0.7 Vbe, so we need to drop around 1.3 V across R18.
1 kOhm sets the current at 1.3 mA, 1.2 kOhm sets at 1.08 mA, two 2.2 kOhm resistors in parallel set the current at 1.18 mA.
Edit:
You can also try to find the sweet spot with a trimmer. I don't remember the right value I have in my amps, but let's calculate it:
Red Leds have around 2 V across them, 2N5551 have 0.7 Vbe, so we need to drop around 1.3 V across R18.
1 kOhm sets the current at 1.3 mA, 1.2 kOhm sets at 1.08 mA, two 2.2 kOhm resistors in parallel set the current at 1.18 mA.
Last edited:
The sweet spot for my amp turns out to be 1.050mA but what is curious is each channel is not the same. The previously working channel has a 1k resistor while the channel that wouldnt zero needed 1k5 at R18. With these values 12AX7 plate sits at 145V and EL84 is at 298V. I can bias the el84 to 34mA and I can easily zero with R5
Thanks for all the help. Now looking forward to sorting out a way to listen
Thanks for all the help. Now looking forward to sorting out a way to listen
I've noticed I've never reported my impressions on how it sounds.
First of all it's LOUD.
My Klipsch have high efficiency (98 dB/Wm), and the amp gives out almost 20 Wrms with 23% UL ratio, so they permit to listen everything really loud.
Then it's really warm and pleasant with all music genres, but gives its best where the music content is simpler (just a comparison: better with James Taylor live at the Beacon Theater rather than Metallica's Reload).
Overall it's the best PP I've heard for most of the music I like, even if I'm sure I could still improve it by setting it better with my speakers.
I use it with 15k as shunt feedback resistor and 23%UL taps, no gnfb.
First of all it's LOUD.
My Klipsch have high efficiency (98 dB/Wm), and the amp gives out almost 20 Wrms with 23% UL ratio, so they permit to listen everything really loud.
Then it's really warm and pleasant with all music genres, but gives its best where the music content is simpler (just a comparison: better with James Taylor live at the Beacon Theater rather than Metallica's Reload).
Overall it's the best PP I've heard for most of the music I like, even if I'm sure I could still improve it by setting it better with my speakers.
I use it with 15k as shunt feedback resistor and 23%UL taps, no gnfb.
dBel84,
Good to hear the problem is licked
Have you determined the sweet spot by ear?Still puzzling this unsymmetrical behavior. I assume you carefully tested both 12AX7’s.
Roberto,
Thanks for your listening impressions. Do you have the ability to trace distortion over frequency? I’m not aware of anyone who had put the BH84 through a rigorous set of performance tests and published the results.
My thanks to you all again, I am pretty sure I would never have found this as the cause without your guidance. I had convinced myself that I had misplaced one of the parts and had not been able to identify anything on my own.
The sweet spot is purely an electrical phenomenon - the point at which the 12AX7 B+ reads what I would have anticipated and where they are closely matched ( within the tube and across channels ) - my tube tester is really geared to find fault and not trace response curves - it is an old Weston , I believe the Hickock was modelled on this design . I will try swap in different tubes to see if this bias remains the same ( I am hoping or else I need to come up with a way to adjust for this when changing tubes )
Thanks for the listening impressions Roberto , I am looking forward to hearing this amplifier myself. It is probably wishful to think that it might play well with my main speakers ( Magnepan .7) but perhaps BH84 driving the F4 into the maggies
. I do have MOTH Cicadia's which I thought would be better suited to the BH84
Now to think about all the metal work to be done - my least favourite part.
..dB
The sweet spot is purely an electrical phenomenon - the point at which the 12AX7 B+ reads what I would have anticipated and where they are closely matched ( within the tube and across channels ) - my tube tester is really geared to find fault and not trace response curves - it is an old Weston , I believe the Hickock was modelled on this design . I will try swap in different tubes to see if this bias remains the same ( I am hoping or else I need to come up with a way to adjust for this when changing tubes )
Thanks for the listening impressions Roberto , I am looking forward to hearing this amplifier myself. It is probably wishful to think that it might play well with my main speakers ( Magnepan .7) but perhaps BH84 driving the F4 into the maggies
. I do have MOTH Cicadia's which I thought would be better suited to the BH84 Now to think about all the metal work to be done - my least favourite part.
..dB