I figure C6 and C7 are on the top side of your pcb. I think you verified before that you have picofarad values on the page where you uploaded your Excel spreadsheet.
Can you or have you done a visual inspection to make sure that C6 is indeed close to 18pf and C7 is indeed close to 39pf and not accidentally reversed?
Best,
Anand.
Can you or have you done a visual inspection to make sure that C6 is indeed close to 18pf and C7 is indeed close to 39pf and not accidentally reversed?
Best,
Anand.
Yes, several times. Except that I could not get a stable meter reading on the 18pF. All four caps are clearly marked and are in the correct spots. They are Cornell Dubilier mica. I made sure that I installed the components with their values positioned so they can be read.
I have new C0G 18pF and 39pF MLCC Kemet caps sitting here. I haven't tested their values yet.
"Ceramic caps come in a handful of popular types, with specs ranging from terrible to wonderful. A C0G, aka NP0 cap, will better a silver-mica in all regards. They approach a perfect cap, but are only available in small values." @Conrad Hoffman
https://www.diyaudio.com/community/...ica-capacitors-and-ceramic-capacitors.122573/
Here is an interesting video on capacitor types if you have very good computer audio. A different application of course, but it's interesting to note that they do have significant differences.
I have new C0G 18pF and 39pF MLCC Kemet caps sitting here. I haven't tested their values yet.
"Ceramic caps come in a handful of popular types, with specs ranging from terrible to wonderful. A C0G, aka NP0 cap, will better a silver-mica in all regards. They approach a perfect cap, but are only available in small values." @Conrad Hoffman
https://www.diyaudio.com/community/...ica-capacitors-and-ceramic-capacitors.122573/
Here is an interesting video on capacitor types if you have very good computer audio. A different application of course, but it's interesting to note that they do have significant differences.
Last edited:
They should digitally record the playing, and for the test play the same recording into each cap,
instead of playing live separately for each cap. Too much variability otherwise.
instead of playing live separately for each cap. Too much variability otherwise.
What is the function of C7? Something about a time constant, correct? The cheapo Chinese amp kit that I built just uses a resistor there.
One thing at a time. They would just be hanging loose inside the chassis because there is no place to secure them. Heavy wire -> resistor -> socket pin. Besides, I am not yet convinced they are needed.
C7 is a "phase lead" capacitor, and increases the phase margin of the amplifier, which improves stability.
Its value depends critically on the exact circuit and parts used, especially the output transformer.
To a musician, I don't think it matters much, but ok.
Another fun test (pF caps) for people with very good computer sound:
You can't play the exact same way each time, and the guitar won't behave the exact same way each time.
And the player knows which capacitor is being tested before testing it.
If you think a particular capacitor does something, you are more likely to play in a way that increases
the chances of that happening, and focus your attention on that aspect of the sound.
So the test playing should be digitally recorded, and the same recording used each time, to have a proper test.
The results could go either way. You could hear more differences among capacitors, or hear fewer differences.
But the results would certainly be more reliable and consistent.
And the player knows which capacitor is being tested before testing it.
If you think a particular capacitor does something, you are more likely to play in a way that increases
the chances of that happening, and focus your attention on that aspect of the sound.
So the test playing should be digitally recorded, and the same recording used each time, to have a proper test.
The results could go either way. You could hear more differences among capacitors, or hear fewer differences.
But the results would certainly be more reliable and consistent.
The verdict is in on the four grid stopper resistors: use them! They have solved three issues. 1) The harsh edge to the sound that caused my original complaint mostly went away. 2) The instability of the output tube bias as the amplifier warms up is solved (was unstable for about two hours). 3) The range of adjustment for the output tube bias is now within specification and the potentiometers need not be maxed out.
So I listened for a while. Better. Still not where it needed to be though. The sound was "smeared." I'm not sure how else to describe it. It was like a photo out of focus, and it wasn't limited to any particular frequency range. So I removed the four mica capacitors in the negative feedback loops and replaced them with identical value 5% tolerance C0G Kemet MLCC caps from Mouser. That snapped everything into focus immediately. So, I can recommend that swap. I decided to try them after reading positive things about the C0G caps in various discussion forums, including this one.
I will now find the "sweet spot" for the output tube bias again. It does alter the sound, which is why the adjustment pots are nice to have. I'll see if it works well at the 35mA spec, and I'll try both 6P14P and 6P14P-ER output tubes.
I realized a few weeks ago that I forgot to install the film bypass caps that I intended to put across the power supply caps. I'll do that one day.
And finally, nine days after the United States Post Office claimed to have delivered my GE 12DW7 tubes, they magically appeared in the mail today. So I'll be trying those soon. They are marked March 1985. I have no idea in what country they were made in 1985. My guess is not in the USA?
So I listened for a while. Better. Still not where it needed to be though. The sound was "smeared." I'm not sure how else to describe it. It was like a photo out of focus, and it wasn't limited to any particular frequency range. So I removed the four mica capacitors in the negative feedback loops and replaced them with identical value 5% tolerance C0G Kemet MLCC caps from Mouser. That snapped everything into focus immediately. So, I can recommend that swap. I decided to try them after reading positive things about the C0G caps in various discussion forums, including this one.
I will now find the "sweet spot" for the output tube bias again. It does alter the sound, which is why the adjustment pots are nice to have. I'll see if it works well at the 35mA spec, and I'll try both 6P14P and 6P14P-ER output tubes.
I realized a few weeks ago that I forgot to install the film bypass caps that I intended to put across the power supply caps. I'll do that one day.
And finally, nine days after the United States Post Office claimed to have delivered my GE 12DW7 tubes, they magically appeared in the mail today. So I'll be trying those soon. They are marked March 1985. I have no idea in what country they were made in 1985. My guess is not in the USA?
Maybe in former Ei factory - i have the info that they manufactured some other tubes for GE - otherwise who knows who is really OEM.
When switching from the 8 ohm output to the 4 ohm output of a tube amp, is there an automatic associated dB change?
Good condition ST-35 now reduced to the bargain price of "only" $1,478.95 USD plus tax! Way too much for me! 😵
https://www.ebay.com/itm/276074076010
Good condition ST-35 now reduced to the bargain price of "only" $1,478.95 USD plus tax! Way too much for me! 😵
https://www.ebay.com/itm/276074076010
Last edited:
Evaluation of different tubes so far:
6P14P at 35 mA (spec)
6P14P at 35 mA (spec)
- A vintage sound
- Sounds best when run at the spec 35 mA
- Soft overall, with good sustain on bass notes
- Lacking in what we would consider depth or ambience under modern standards
- Overall pleasing but not highly detailed
- A more "modern" sound compared to 6P14P
- The slight increase from 35 mA to 36.5 mA is noticeable and required
- Good depth and ambience
- Sustain on bass notes not as good as 6P14P and is slightly lacking
- Overall sound is "modern" and slightly dry
- Renders everything that goes through it totally sterile, even though the source is not
- The third time I have bought JJ tubes (12AU7, 12DW7, and EL34) because they were the cheapest on the market and lived to regret it
- Larger diameter pins almost destroyed the circuit boards trying to get the tubes installed (yes, I used a pin straightener first and they were perfect)
- Natural sounding
- Good depth and ambience
- Bass sustain and bass drive are acceptable
- Well worth a few extra dollars vs. JJ
Last edited:
I thought I would give a little update on this amp. It is FINALLY giving what I consider great sound. I had to do several things.
I spared no expense on this amplifier build ($1,336 plus paint and decals last I checked) and added every available doo-dad and add-on and upgrade and mod (see thread), and I am glad it's FINALLY producing the quality of sound that I expected from it. 👍
I'm very happy with the all-black custom paint and decals. I hate silver and chrome. The front power switch was a great addition, as were the custom binding posts spaced for dual banana plugs. That black can cap is hiding three 120uF filter caps inside. The filter choke is a nice addition. The individual bias for each output tube seems to be working very well, and I run it a little above spec. Is it really a Dynaco ST-35? No, it is not really with all of the modifications. That's why I called it a "Series ii" 😀. I bought the tube cage for it in case a future owner ever wants the protection, but it runs so hot that I can't stand to put it on the amp.
The build date that I put on it says June 2023. I had no idea that it had been that long. My how time flies.
- Added four 1k grid stoppers to the four output tubes.
- Replaced the four mica NFB caps with new C0G 18pF and 39pF MLCC Kemet caps.
- Replaced a questionable output tube that just wouldn't bias like the other three were. The seller sent me one with identical test values, and replacing that one tube in the quad made a huge difference. I am thrilled with the quads of NOS Russian 6P14P-ER/6П14П-ЕР that I have bought for this amp and another. I think I want to invest in a lifetime supply. Like gold, good tubes seem to hold thier value or go up over time.
I spared no expense on this amplifier build ($1,336 plus paint and decals last I checked) and added every available doo-dad and add-on and upgrade and mod (see thread), and I am glad it's FINALLY producing the quality of sound that I expected from it. 👍
I'm very happy with the all-black custom paint and decals. I hate silver and chrome. The front power switch was a great addition, as were the custom binding posts spaced for dual banana plugs. That black can cap is hiding three 120uF filter caps inside. The filter choke is a nice addition. The individual bias for each output tube seems to be working very well, and I run it a little above spec. Is it really a Dynaco ST-35? No, it is not really with all of the modifications. That's why I called it a "Series ii" 😀. I bought the tube cage for it in case a future owner ever wants the protection, but it runs so hot that I can't stand to put it on the amp.
The build date that I put on it says June 2023. I had no idea that it had been that long. My how time flies.
Very nice ST-35!
I'm using the same power tubes albeit branded Sovtek EL84M in mine with the same bias board.
However I did prefer the GE 12DW7 drivers over the JJ's. Maybe it has more to do with the speakers I'm using with this amp? I dunno. I tend to use a lot of JJ 12AU7's and 12AX7's though. The cage does get hot on the sides due to the power tubes. For whatever reason Dynaco only had two slots near the bottom put on each side if the cages. I want to find someone with a CNC machine to add more to the side panels. The cage should have had perforated sides like the rest of the cage.
I do wish Dynakit had sold there kits with solid core hookup wire instead of the stranded. It would have made wiring this thing easier, especially around the sockets.
I thought about adding a power switch to the front, but went with a remote controled Bluetooth power socket instead. I have another three socket Bluetooth outlet I have my VTA ST-70 and M-125 mono blocks plugged into.
I'm using the same power tubes albeit branded Sovtek EL84M in mine with the same bias board.
However I did prefer the GE 12DW7 drivers over the JJ's. Maybe it has more to do with the speakers I'm using with this amp? I dunno. I tend to use a lot of JJ 12AU7's and 12AX7's though. The cage does get hot on the sides due to the power tubes. For whatever reason Dynaco only had two slots near the bottom put on each side if the cages. I want to find someone with a CNC machine to add more to the side panels. The cage should have had perforated sides like the rest of the cage.
I do wish Dynakit had sold there kits with solid core hookup wire instead of the stranded. It would have made wiring this thing easier, especially around the sockets.
I thought about adding a power switch to the front, but went with a remote controled Bluetooth power socket instead. I have another three socket Bluetooth outlet I have my VTA ST-70 and M-125 mono blocks plugged into.
Those are reported to be 6P14P-EV/6П14П-ЕB. I am using the 6P14P-ER/6П14П-ЕР.
According to my research:
6 means 6.3V
P (cyrillic П) means power pentode
14 means fourteenth type in a series of power pentodes
P (cyrillic П) means housed in a glass tube with noval socket
after the dash - these are quality indicators
nothing indicates 3000 hour life (14 watt anode dissipation)
V (cyrillic B) without an E in front of it means 1000 hour life
K means vibration resistant with 3000 hour life
E means extended life (life depends on the character following it)
V (cyrillic B) means a higher degree of mechanical quality with 5000 hour life (14 watt anode dissipation)
R (cyrillic P) means a higher degree of mechanical quality with >5000 hour life (gold used in the grid, 14 watt anode dissipation)
From what I understand, R (cyrillic P) was developed for nuclear war enhanced resistance to radiation and shock and was the highest level achieved. The often-reported 10,000 hr life for it seems to be incorrect, instead simply being rated at greater than 5,000 hours.
I have some 6P14P-K's I picked up cheap. I'm using them in a modified Magnavox 9303 console amp, and a SEP amp I built. I thought they sounded a little better than the Sovteks and old stock Mullard EL84's I have. They're ruggedized like the EV's, EB's, etc., but aren't rated to last as long. But they do sound nice. I probably won't buy more of them though. They're isn't that much of a difference in sound, and I would rather buy the longer lasting versions.
Very pretty kit I must say.
I see in the mod the power supply input capacitor went from a 60uF input capacitor to a 120uF input capacitor.
Caution is always advised anytime you increase the capacitance value of the first capacitor in any capacitor input power supply even when solid state diodes are used.
Why?
Increasing the input capacitor value will reduce the conduction angle of the rectifier diodes.
Reducing the conduction angle of the rectifier diodes increases the peak rectifier current.
Increasing the peak rectifier current increases the transformer IR losses and the rectifier power losses by approximately the square of the increase in the peak rectifier current.
So in simplified terms a 10% increase in peak rectifier current can cause a 20% increase in transformer IR losses and so increase the transformer heating.
If your power transformer is designed to take the extra heat no problem but if not it may well live a shorter life.
Adding extra capacitance to the second capacitor after the inductor is a much safe location to increase capacitor value as it does not increase the peak rectifier current except at the point of turn on. This is also a far more effective location to reduce ripple by the addition of power supply capacitance.