• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

New DynaKitParts ST-35 Build

Jump to Latest
#44
Since I can't read Russian- only guessing- seems the 6N1 has 2 medium mu triodes... Maybe like a 12AT7?
I know the12DW7 has dissimilar triodes, as does the JJ ECC832. One low and one hi mu triode- like half a 'U7 and half a 'X7.
The hi mu half (100) is used for sig amp, the low mu half (17) is used for phase splitter. Not saying using a 6N1 can't be done, but probly have to juggle some component values.
Jim
 
#45
#46
I'm a little confused about the voltages involved in the NFB part of the circuit, namely C7 and R10.

C7 is specified as 500V. Why so high? The one in the kit appears to be mica, as are C6 and C2. Mica and ceramic seem to be the only available options.

Most 1/2 watt resistors seem to be rated 250V with some rated at 350V. What about the voltage rating of R10?

Also, why is C2 on the input specified as 500V?


1686872363420.png



1686872753220.png
 
Last edited:
#47
Small caps like 39 pF aren't any smaller or cheaper at voltages lower than 500V, so 500V ones are common. I'd use a 500V mica for C2 or C7 since I have them on hand, even though it only has to handle a few volts. C6, on the other hand is barely OK at 500V, with 330VDC plus 40% of the peak audio..
 
#48
Tom Bavis said:
I'd use a 500V mica for C2 or C7 since I have them on hand, even though it only has to handle a few volts. C6, on the other hand is barely OK at 500V, with 330VDC plus 40% of the peak audio.
Click to expand...

All three that come with the kit are mica 500V.

Most 1/2 watt resistors seem to be rated 250V with some rated at 350V. What about the voltage rating of R10? The one in the kit is a 20k with an unknown voltage rating. Technically it ought to be 18.7k. The highest voltage 1/2W 18.7k film resistor on Mouser is 350V.

Note: per Kevin at DynaKit Parts, Dave Gillespie recommended 18k, while the closest match to the original circuit would be 18.7k, and the kit comes with 20k.
 
#50
Tom Bavis said:
Less than a dB gain difference between those values
Click to expand...

Something about the time constant changing though - beyond my knowledge and expertise. Not sure why Dave G. recommended going down to 18k versus the 20k in the kit. As long as I am at it, I'll stick a few 29 cent 18.7k in my Mouser shopping cart. According to the calcs posted earlier, those would be the closest to the original design, but adjusted for moving NFB to the 8 ohm tap.
 
#51
Until proven otherwise, it's perfectly appropriate to question tolerances. In a feedback amplifier tolerances stack, so good engineering demands that we question which tolerances are more, and which less, significant. And as Tom reminds us, we need always to be thinking in log (on a dB scale). Always.

The most important component in the amplifier is unfortunately the least well specified, the output transformer (OPT). It is not only NOT the same device as the device for which the circuit's time constants were originally (1960-something?) devised, but its detailed specifics are also the limiting features in a feedback amplifier. The OPT determines almost completely the needed time constants, open loop and closed loop, of the valve stages. But we don't know its specifics without actual, unfortunately complex and difficult, multidimensional, measurement.

Paul Klipsch once said that if you can't change something by 3dB, don't bother. He was exaggerating to make a point, but not by a lot. In feedback amplifiers there are a lot of unknowns (especially the OPT) so rules-of-thumb and best practice arise. This may be intellectually unsavory, but engineering cares diddly about keeping us happy or content; it just is.

All good fortune,
Chris
 
  • Like
Reactions: jhstewart9, jcalvarez, huggygood and 1 other person
#52
Chris Hornbeck said:
... the output transformer (OPT) ... It is not only NOT the same device as the device for which the circuit's time constants were originally (1960-something?) devised ... But we don't know its specifics without actual, unfortunately complex and difficult, multidimensional, measurement.
Click to expand...

Stay tuned on that issue. More to come later. It is being addressed.

Never Get Old said:
Note: per Kevin at DynaKit Parts, Dave Gillespie recommended 18k, while the closest match to the original circuit would be 18.7k, and the kit comes with 20k.
Click to expand...

Dave is pretty knowledgeable and has tested the included transformers, so anyone who is interested in fiddling can use his recommended values 18k and 39pF.

Or ... get identical replacement transformers. The originals have been unwound, diagrammed, and tested in extreme detail a lot of times over the years. There is an original pair on Fee-Pay for $300 right now.

Michael LaFevre at MagneQuest was said to have been making the only true replacement Z-565, until he died March 25th.

https://www.facebook.com/people/MagneQuest/100054351979647/
 
Last edited:
#53
Why not build the two PC boards with the components on the bottom instead of the top to better protect them from the baking heat cast off by the tubes? The Chinese amp kit that I built has the components on the bottom of the board, spaced slightly away from the board, and the tubes are mounted on the top like this:

1686954313489.png
 
#56
directdriver,

Thanks!
The original schematic of that Stereo amplifier.

With a Single Self Bias Resistor . . .
You need a Quad of Very Well Matched 6BQ5 or EL84 tubes.

I am sure both kits and assembled amplifiers, Dyna provided a Quad of Very Well Matched 6BQ5 or EL84 Tubes.
 
#57
6A3sUMMER said:
The original schematic of that Stereo amplifier.
Click to expand...

I didn't put it in post #1 because I won't be using it.

DynakitParts provides separate schematics for the amplfier and bias adjustment kit. I wanted to see everything in one place, and I wanted the ability to change values on paper, so I drew the entire schematic over the past week. I have added it to post #1. It reflects the following changes:

  • Individual tube bias adjustment kit add-on.
  • NFB moved from 16 Ohm tap to 8 Ohm tap since the new transformers are 8/4 Ohm.
  • Added CL-90.
  • Added four 100 Ohm resistors to heaters (mod comes with the individual tube bias adjustment kit add-on).
  • Increased C8A, C8B, and C8C to 120uF 450VDC 105C since I am restuffing a cap can and they should fit inside just fine. The kit comes with a can cap that has 60/40/20 450VDC 85C, like the original amplifier.
  • Includes NFB resistor values suggested by Dave Gillespie to DynakitParts. The kit comes with 20k Ohm.
  • Optional C1, which I probably will use although my current amp doesn't use a cap there and it seems ok.
If anyone sees any errors in the schematic that I have added to post #1, please let me know.
 
Last edited:
#60
Never Get Old said:
Why not build the two PC boards with the components on the bottom instead of the top to better protect them from the baking heat cast off by the tubes? The Chinese amp kit that I built has the components on the bottom of the board, spaced slightly away from the board, and the tubes are mounted on the top like this:
Click to expand...
Since no one touched this I will offer my opinion. I think it is a good idea to put the capacitors below the PCB to protect them from cooking right next to the power tubes, but you will need skinny ones like this to fit them in the available chassis space: https://www.mouser.com/ProductDetail/Cornell-Dubilier-CDE/940C6P1K-F?qs=11kF6y5Z3OguAvn7IPlTQw==

You do NOT want to put the heat-generating resistors below due to poor ventilation in the confined space in the chassis cavity of the ST35. You Chinese amp with the much larger, relatively well-ventilated cavity seems fine with everything below the PCB.
 
You must log in or register to reply here.
Top Bottom