I just looked up the DOT convention for a transformer as I have always found this marking system made my head hurt a bit.
The simplest explanation I found was this quote.
"The voltage waveform on a dotted terminal is always in phase the voltage on another dotted terminal."
This seems to confirm that Hammond reversed the phase from input to output if you look at the top terminals of the transformer in the diagram.
I wonder why they do this?
The simplest explanation I found was this quote.
"The voltage waveform on a dotted terminal is always in phase the voltage on another dotted terminal."
This seems to confirm that Hammond reversed the phase from input to output if you look at the top terminals of the transformer in the diagram.
I wonder why they do this?
I noticed that too, but it really depends on the PCB layout. Lots of opportunities to swap the polarity in a Williamson. ;-) I had a Williamson PCB designed, too, but every time I build one I have to think really hard and consult my own schematic to make sure I have the OPT hooked up right. ;-)
Hello, first thank you for your feedback 😊 It seem you found the issue !
If I start the amplifier the way its done on the last schematic, an horrible sound come out from the speakers, It sound like the audio test that speakers must pass on factory and it's very very loud even the volume completely off, that's what you call "an positive feedback" ? 😵
For the filter 1200pF + 1800K I'm gonna be honest i used your settings Grover but you are right I should start with the default 200p + 4K7
Bluesystems, for the PCB design I used EASY EDA and JLCPCB for the Production, they are not too expensive and work pretty well
So yes for now this is the wiring:
And yes you are right.... it seem to be reversed !! I don't know why they do this maybe to test newbies like me 🤣
This is the correct "way"
I'll try this modification this afternoon and fingers crossed haha
Regards
Alexis
If I start the amplifier the way its done on the last schematic, an horrible sound come out from the speakers, It sound like the audio test that speakers must pass on factory and it's very very loud even the volume completely off, that's what you call "an positive feedback" ? 😵
For the filter 1200pF + 1800K I'm gonna be honest i used your settings Grover but you are right I should start with the default 200p + 4K7
Bluesystems, for the PCB design I used EASY EDA and JLCPCB for the Production, they are not too expensive and work pretty well
So yes for now this is the wiring:
This is the correct "way"
I'll try this modification this afternoon and fingers crossed haha
Regards
Alexis
So good news, I swapped the plates wiring and it's wooooooorking !!!! 😀 Hammond are trolls
But I lost a hell lot of power, to have a decent volume I need to push the knob to the mid where previously I only needed 1/4 to get the same amount but all the power is after the mid and the end of the range, can I increase the FB resistor to get more power ?
The quality is definitely way better and it was already good but now ! 😱
I also tweaked some resistors to get closer to the original voltages:
Cheers
But I lost a hell lot of power, to have a decent volume I need to push the knob to the mid where previously I only needed 1/4 to get the same amount but all the power is after the mid and the end of the range, can I increase the FB resistor to get more power ?
The quality is definitely way better and it was already good but now ! 😱
I also tweaked some resistors to get closer to the original voltages:
Cheers
Excellent! But you didn't lose power, you lost gain! Feedback reduces the gain a lot. In fact, you'll actually have MORE power now, at least at the frequency extremes. Turning up the volume knob to 12 noon is not a problem. ;-) Do not change the feedback resistor from 10K, that is the correct value. Higher and it will reduce the feedback and decrease performance. Lower and it will add too much feedback and the amp might freak out. Don't worry, there's plenty of power there.
That is not an easy amp to build. You are to be congratulated for getting it right! I am thinking of ordering a1650NA and doing some fine tuning to stabilize that particular transformer. It would be great to have a high-powered Williamson project with a PCB and off-the-shelf parts that is guaranteed to be stable. For now you are probably okay. But later this week I'll probably order one and you can adjust your schematic accordingly. It would be a nice thing to have handy. I'd also be interested in a pair of your boards, did you post the gerbers?
You swapped the Plate leads on the output transformer. And that made it work, Good!
Now, you need to check if the Screens are connected to the correct Ultra Linear Taps.
The plate and screen of one tube need to connect to the Plate tap and UL tap that is on one side of the Center tap.
I am not saying those connections are wrong; just check them using an Ohmmeter or DMM.
Power Down the amplifier, then remove the power cord.
Caution:
I hope you have a Bleeder resistor(s) across the B+, or the result of putting your hands in the amplifier wires will be shocking!
Your schematic on Post # 124 does not show a Bleeder resistor. Put one in (but make sure the B+ capacitors are discharged first (Oh, that is what the Bleeder resistor is supposed to do). Do not use a screwdriver to discharge the capacitors, you will weld away some metal from the screwdriver tip. Use a pair of insulated alligator clip leads. Connect the bleeder resistor you should have installed already, and short from ground, through the bleeder resistor, to the B+ cap. You have 475V B+, so use three each 30k 5 Watt each resistors in series.
Safety First!
Prevent the "Surviving Spouse Syndrome"
You did not get bit by the B+ yet . . . then pick your favorite place of worship, and go on Friday, Saturday, or Sunday, and give thanks!
Example primary DCR measurement, using an Ohmmeter/DMM.
Plate to UL tap 60 Ohms
UL tap to Center tap 40 Ohms
In that case, the Plate to the Incorrect UL tap would be 60 + 40, +40 Ohms = 140 Ohms.
your numbers may differ, but the principles of measuring from one plate end to the other taps, one by one, follows the example.
Hint: Short all of the secondary leads together; there is so much inductance in the primary, which causes lots of DMMs Auto-Range feature to keep jumping between Ranges. So you will not get stable DCR readings.
But shorting the secondary leads together reduces the primary inductance, and the DMM Auto-Range goes to one Range, and stays there.
Have Fun!
Now, you need to check if the Screens are connected to the correct Ultra Linear Taps.
The plate and screen of one tube need to connect to the Plate tap and UL tap that is on one side of the Center tap.
I am not saying those connections are wrong; just check them using an Ohmmeter or DMM.
Power Down the amplifier, then remove the power cord.
Caution:
I hope you have a Bleeder resistor(s) across the B+, or the result of putting your hands in the amplifier wires will be shocking!
Your schematic on Post # 124 does not show a Bleeder resistor. Put one in (but make sure the B+ capacitors are discharged first (Oh, that is what the Bleeder resistor is supposed to do). Do not use a screwdriver to discharge the capacitors, you will weld away some metal from the screwdriver tip. Use a pair of insulated alligator clip leads. Connect the bleeder resistor you should have installed already, and short from ground, through the bleeder resistor, to the B+ cap. You have 475V B+, so use three each 30k 5 Watt each resistors in series.
Safety First!
Prevent the "Surviving Spouse Syndrome"
You did not get bit by the B+ yet . . . then pick your favorite place of worship, and go on Friday, Saturday, or Sunday, and give thanks!
Example primary DCR measurement, using an Ohmmeter/DMM.
Plate to UL tap 60 Ohms
UL tap to Center tap 40 Ohms
In that case, the Plate to the Incorrect UL tap would be 60 + 40, +40 Ohms = 140 Ohms.
your numbers may differ, but the principles of measuring from one plate end to the other taps, one by one, follows the example.
Hint: Short all of the secondary leads together; there is so much inductance in the primary, which causes lots of DMMs Auto-Range feature to keep jumping between Ranges. So you will not get stable DCR readings.
But shorting the secondary leads together reduces the primary inductance, and the DMM Auto-Range goes to one Range, and stays there.
Have Fun!
Last edited:
Thank you Grover 😊 I'm so happy for a first one !! But I got a hell lot of help here its definitely a team work
Ohhhhhhhh right, now I understand, less gain but more power that's why when I go over 12H everything go way higher than before 😆 I noticed this later in the night.
If you want I still have 3 boards left if shipping cost are not freaking high I can send you the board and also I'll post the Gerber file this afternoon, and of course with your help I'll modify the pcb accordingly with your tests.
I'll try for R11 if I find a 680Ohm resistor in my mess haha.
6A3sUMMER, thank you for your input, it's not in the schematic but there is 3 bleeding resistors, when i turn off the amplifier there is nothing left after 5 seconds.
I'll do some measurement but I swapped everything correctly, not only the plates but also the UL wires. I'm quite sure if the swap was wrong I'll have already had big trouble in the speaker no ? 😵
Ohhhhhhhh right, now I understand, less gain but more power that's why when I go over 12H everything go way higher than before 😆 I noticed this later in the night.
If you want I still have 3 boards left if shipping cost are not freaking high I can send you the board and also I'll post the Gerber file this afternoon, and of course with your help I'll modify the pcb accordingly with your tests.
I'll try for R11 if I find a 680Ohm resistor in my mess haha.
6A3sUMMER, thank you for your input, it's not in the schematic but there is 3 bleeding resistors, when i turn off the amplifier there is nothing left after 5 seconds.
I'll do some measurement but I swapped everything correctly, not only the plates but also the UL wires. I'm quite sure if the swap was wrong I'll have already had big trouble in the speaker no ? 😵
Here is the Gerber of the last production but I'm working on an optimized one with all the things that I regret a bit now 😆
This is I think a better version, there is a lot of changes but the main core is here
PCB:
For the filaments Board, traces of 7mm (before 5mm)
For the main board:
Components values changed:
If you see something you want to change do not hesitate
In the RAR file, there is the gerber file and both schematic and pcb file for Easyeda, this pcb is for 2OZ of copper and 2mm board thickness
Regards
PCB:
For the filaments Board, traces of 7mm (before 5mm)
For the main board:
- KT88 traces from 2mm to 2.5mm
- Bias traces from 0.5mm to 0.750mm
- 6SN7 power traces from 1mm to 1.3mm
- 6SN7 Filaments from 1.5mm to 2.5mm and closer to the socket
- More space between the signal caps they heat a bit
- A lot of points of measurement it's gonna be usefull
- Better positioning for the inputs, for the trimmers and for a lot of components
- The texts are bigger and more readable
- A lot of change concerning pad sizes for an easier soldering
Components values changed:
- PI 6SN7 PLATE from 39K to 42K to get 47K with the 10K trimmer
- PI Cathode Resistor from 390 to 680 Ohm
- Power supply resistor => 1K 15K and 8K8 (but it's in my case)
If you see something you want to change do not hesitate
In the RAR file, there is the gerber file and both schematic and pcb file for Easyeda, this pcb is for 2OZ of copper and 2mm board thickness
Regards
Last edited:
