Tube DIY Asylum

Do It Yourself (DIY) paradise for tube and SET project builders.

Return to Tube DIY Asylum


Message Sort: Post Order or Asylum Reverse Threaded

RH84 test amp: multiple problems found ...

24.140.192.106

Posted on September 23, 2016 at 17:24:17



Started working on the RH test amp again this afternoon and found a number of problems.

A) Several suspect looking solder joints on driver tube socket, re-flowed all pins with Kester 44 60/40 rosin core solder. I also reflowed the RCA input jack connections.

B) Right channel red plating of EL84. Found the LM317 cathode CCS and bypass cap to be damaged. Remove the damaged parts are reconfigured both channels to run off one LM317 cathode CCS.

C) The 5.1V X 3 zener diode string connected to right channel EL84 G2 was shorted out. Removed the damaged parts and reconfigured both channels to run off one diode string for both channel G2's.

D) reports of excess noise showing up on my distortion analyzer was due to a ground loop. When I added a ground defeat plug to the AC line cord, the THD+N measurements returned to normal.

I might need to add a bit more filtering to get the AC ripple at speaker posted below 5mV, but it is OK for listening on some bookshelf speakers.

The sound from my bookshelf speakers is much improved now that these corrections have been made.

Edit: @ 1.2W power output RT Channel THD = .77% THD+N = .79%
@ 1.2W power output LT Channel THD = .78% THD+N = .80%

 

Hide full thread outline!
    ...
RE: RH84 test amp: multiple problems found ..., posted on September 24, 2016 at 11:03:24
DAK
Audiophile

Posts: 2712
Location: PACIFIC
Joined: August 8, 2010
Hi DT, i have found that the LM317 to be less than reliable in the cathode bias duties. I have changed out almost everyone that i installed. I suspect that it is the heat from the tubes that makes them fail at a high rate but it maybe some other issue. Sometimes the power tube won't even red plate when they fail which can be tricky unless you have a test point. I was putting a heat sink on them to keep them from failing but even that did not seem to help as eventually some of them failed again. I mounted some on the outer chassis of an amp that i sold but i am not sure if that pair of ccs is still working. It is probably better to go back to a bias resistor in terms of reliability.
Thanx for posting your distortion measurements of the amp as many critics have dissed the rh amps simply because the driver tube selected is a 12at7 which they claim is a "distortion producer" and thus the design can not be "hi fi". I think your measurements show that discussion to be untrue. cheers, Dak

 

RE: RH84 test amp: multiple problems found ..., posted on September 24, 2016 at 11:50:44
Tre'
Industry Professional

Posts: 17260
Location: So. Cal.
Joined: February 9, 2002



I probably shouldn't say anything but elblanko just posted a link to a great article by Corwhurst where he explains that if the grid line spacing is not even along the load line, to the left and to the right from the idle point, harmonic distortion is the result.

Read the "Load line conclusions" starting at the bottom of page 2.

With the 12at7 I can not see where a operating point and load line can be placed where the plate voltage swing will be even to the left vs. to the right of the operating point. (P.S. 450vdc at 5ma. and 350vdc at 7.5ma. look OK but the tube is only rated to 300vdc plate so we can't use those operating points. We have to stay below 300vdc plate)

No matter where you operate the tube and no matter what the angle of the load line, the plate voltage will change much more when the tube is driven along the load line to the left of the idle point vs. to the right.

A linear tube has even spacing between the grid lines and it's easy to fine a operating point and a load impedance that causes the tube's plate voltage to move in a equal and symmetric way, to the right vs. to the left.


Tre'



Have Fun and Enjoy the Music
"Still Working the Problem"

 

RE: RH84 test amp: multiple problems found ..., posted on September 24, 2016 at 12:22:48
Thanks for the LM317 info.

The issues are high driver stage distortion and the level of odd 3rd order harmonics at the output of the amp.

I'll have to post updated graph to show this.

Maybe that is why some people either love or hate the RH circuits.

My amp does not use exact values but is the same sort of circuit.

I'm also looking at the pentode and mosfet driver stages for EL84, but I will keep listening to what I have so far to determine if I want to keep it or not.

 

RE: One is not limited to 12at7, posted on September 24, 2016 at 15:15:35
DAK
Audiophile

Posts: 2712
Location: PACIFIC
Joined: August 8, 2010
In the original rh circuit one could also use tubes of similar characteristics, like 5965, or 7062, and 12ay7, but microphonics is not a problem in this circuit. I have tried all those tube types and my current favorite driver tube is the 7062. I have not check the load lines of thee tubes so i am not sure how they stack up against the 12at7.

 

RE: One is not limited to 12at7, posted on September 24, 2016 at 15:35:05
Tre'
Industry Professional

Posts: 17260
Location: So. Cal.
Joined: February 9, 2002



The 12ay7 is by far the most linear tube of the bunch.

Pick a spot. The datasheet says 250vdc plate at 3ma. and -4vdc bias.

Load it high for a near horizontal load line. Nice even spacing between grid lines and the plate voltage moves very close to the same going to the left vs. going to the right.

With a horizontal load line I see (very close to) the plate voltage moving 87.5 volts from -4 to -2 volts grid and (very close to) the plate voltage moving 87.5 volts from -4 to -6 volts grid. If true in practice, that means that there is (very close to) no harmonic distortion.

The plate resistance is high but I don't think that would be a problem driving a pentode (pentodes don't have much Miller capacitance to worry about driving).


Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"

 

RE: One is not limited to 12at7, posted on September 24, 2016 at 17:06:56
DAK
Audiophile

Posts: 2712
Location: PACIFIC
Joined: August 8, 2010
I am breaking in a RH 6l6 amp with Norelco outputs from Switzerland. They sound very nice with the 7062 and 6L6 metal tubes. very clean sound on my Altec. Although i think it is mostly the NOS 6L6 that i just started listening to today. I will plug in some 12ay7 and let you know how that fares. cheers.

 

RE: One is not limited to 12at7, posted on September 24, 2016 at 17:25:11
Tre'
Industry Professional

Posts: 17260
Location: So. Cal.
Joined: February 9, 2002
I'm not sure you can just plug the 12ay7 in a circuit designed for a 7062.

But even if you could, to get the full benefit of the 12ay7 you would need to optimize the circuit for the 12ay7.

Plate voltage, plate load resistor, cathode bias resistor.


Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"

 

RE: harmonic spectrum data ..., posted on September 25, 2016 at 16:43:18
At 1.2W power output with 1KHz sine test signal:

Left Channel : 2nd harmonic = -44.44dB, 3rd harmonic = -46.68dB,
4th harmonic = -66.54dB, 5th harmonic = -73.93dB

THD = .77% THD+N = .79%


Right Channel : 2nd harmonic = -43.71dB, 3rd harmonic = -48.21dB,
4th harmonic = -66.05dB, 5th harmonic = -76.27dB

THD = .76% THD+N = .79%

Took these measurements tonight.

2nd and 3rd harmonic seem to dominate the spectrum in this circuit.







 

Actually, posted on September 26, 2016 at 17:17:11
Russ57
Audiophile

Posts: 3754
Location: South Florida
Joined: November 16, 2001
What you are seeing is 2nd order. The 12at7 family does have different gain at different grid voltages. In a push pull amp this can actually be used to give a reduction in higher order distortion. You will end up will a little higher overall THD but a better sounding amp.

The 7026 is kinda in the 12at7 family.

 

circuit in question, posted on September 26, 2016 at 18:03:18
Tre'
Industry Professional

Posts: 17260
Location: So. Cal.
Joined: February 9, 2002

.
Have Fun and Enjoy the Music
"Still Working the Problem"

 

RE: driver stage distortion..., posted on September 26, 2016 at 19:19:44
FWIW, the 3rd harmonic in the driver stage is 12 - 15dB below the 2nd harmonic when the overall power output is measured at 1.2W, there is some variation between 12AT7 triode sections.

The feedback is probably cancelling out some of the 2nd harmonic.

The level of 3rd harmonic at the output of the circuit is only 2 -5dB below the 2nd harmonic.

Edit: with power output of the circuit @ 1.2W, the driver stage THD measured between 7.5 - 8.25% (there was a difference between 12AT7 triode sections).

THD varies a lot with signal level.


 

RE: driver stage distortion..., posted on September 26, 2016 at 19:45:51
Tre'
Industry Professional

Posts: 17260
Location: So. Cal.
Joined: February 9, 2002
I was not intending to speak to the distortion of the amplifier as a whole.

I was just trying to point out that a 12at7 is not a very linear tube.

BTW Keep up the good work and reporting. It's very interesting and informative.


Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"

 

Page processed in 0.024 seconds.