• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

Dynaco 70 restore

The interesting with these amps is, unlike some others, they got it mostly right with a very economical design, and the few faults are those of omission rather than commission
 
Pio1980 said:
They are 100 ohm 1/4 watt.
Disconnect the 1k 1/4 watt resistors between tabs 5 and 6 from tabs 6 and connect the now free ends directly to the board with wire extensions. Move the output transformer tap leads from tabs 4 to tabs 6 and connect the 100 ohm 1/watt resistors between them.
Move the feedback wires from the board from tabs 4 to tabs 6.
View attachment 2621420
View attachment 2621422
Click to expand...
So because pin 6 is a empty nothing on the socket , this mod just adds a 100 ohm to that transformer lead ? a lead that all ready must have lots of resistance from the transformer windings , im still considering it , but just trying to understand it
 
6DZ7 said:
I don't think you want to change those 1.5M. They feed the screen voltage to the pentode voltage amp.
Click to expand...
It moves the average splitter triode bias to a more favorable voltage distribution center, 25%-50%-25%.
Being direct coupled driver-splitter, the screen voltage affects plate voltage, which determines the splitter grid bias and voltage distribution.
 
6DZ7 said:
They are "stopper" resistors like grid stoppers. They can provide a filtering effect to inhibit the formation of parasitic oscillations on the screen. Don't know that Dynaco ever issued a recommendation for the those for the ST70. Other output tubes have recommendations for them on their sheets. ....Never had a problem in my amp without them. Perhaps there is an issue I'm not aware of. Would like to see a quote for it.
Click to expand...
At the worst, it's cheap enough insurance for the A-470 OPTs. At the best, it's cheap insurance for the A-470 OPTs. Dave G has discussed it in the thread and elsewhere. Marantz used them.
 
6DZ7 said:
Read the manual for test point voltages on the 7199. The voltages on pins 1,2,3,8,9 are non specified and as the manual says can be all over the place for any subject tube in place. Simply changing the 1.5M resistor won't guarantee any result in improvement. I believe the ST70 addressed the issue of inverter balance by sending some FB back in from the screen tap of the non-inverted side of the OPT.
Click to expand...
Also discussed in the base line thread. Variability of individual Tubes & subs will move it around, which is why I said "average". It's not absolutely necessary, but it does reduce stress on the splitter cathode/heater by reducing the voltage differential between them. A main cause of driver tubes becoming noisy in these amps is heater/cathode leakage.
 
When I did my 70 restore as a covid quarantine project, I encorporated as much from the thread and other credible folks to make it as idealized as practical without getting silly.
That doesn't make me an expert or the expert, I'm just an interested hobbyist with a sentimental regard for these things, relying on the experts.
 
6DZ7 said:
Pio, I'm lookin' hard for the material I have where Hafler explains how the inverter imbalance is compensated for. He explains the reason, cathode degenerative FB that exists because of the non-bypassed nature of the that cathode, and the solution. It's good info that can be employed in other circuits. ...digging..
Click to expand...
It's @dcgillespi that delves into the revised ideal splitter ratio distribution later in the thread, had earlier used a different approach. He also discusses the input DC coupling vs stability with the FB earlier re adding a blocking cap.
 
David G recommends adjusting the the splitter voltage for a 25 50 25 distribution to better optimize the output . Dyanco chose a safe for any 7199 tube value that works. Dave found that the 25 50 25 distribution provided lower THD than the 33 33 33 often sited. One means of obtaining it is to adjust the 1.5M screen resistance sufficient that the cathode resister of the triode is dropping 25% of the total B+. The means for the benefit is by feeding the output tubes this now wider differential between plate and cathode as an input signal, and not to correct the balance as you mention. I disagree with PIO only in his recommending another random fixed value, because it is rather random. I know based upon his posts here that he himself uses a variable resister, and sets the 25% based upon measurements. Optimally with the proper equipment one would use a distortion analyzer and set the voltage based on best THD. Dave's theory on providing the output tubes with the most clean signal is sound, and I also uses it. I incorporated a 1M variable resister in line with a fixed 3K0R to give me sufficient enough range to adjust for various tubes.
 
6DZ7 said:
The output tubes have no reference to the DCV on the plates or cathodes because they are isolated from that voltage by the coupling caps. The idea of having a load balance is to keep the signal (+)/(-) from clipping or going into cutoff out of sync. If the drive signal is smaller that the voltage range then there won't be any problem even if the original 33% division kept. If you are going to drive the outputs with a larger signal than the amp was designed for then yes you have to expand the range, but is that the case, that the drivers are being asked or made to push the inverter harder, passing a larger signal than can fit the 33% voltage range and so get clipped? The problem is in using the voltages read on the tube that may or may not be expressing any practical diagnostic value. And just expanding the plate/cathode differential without need isn't going to give the output grids matched inputs from the inverter.
Click to expand...
The blocking caps (The only caps in each phase of the signal chain of the oem '70) effectively isolate the output tube G1 bias from the splitter voltage distribution, so that's unrelated to the linearity of the splitter operation and the output bias.
 
The improvement of THD measurements provides evidence that the clipping you are concerned about is not occurring. Also the symmetrical traces at full output, so the 50% across the tube is not only within usable range, but it measured better. The change as presented does not undermine anything put in place prior to prevent an imbalance at 33% across the tube. David G is well known for his excellent white paper write ups, with test data backing up his findings. Hopefully you have read it.
 
I too am very cautious with regard to who and what I consider as reliable enough for me to follow. I also have to be careful when defending someone else's work, as to be honest I am no where close to being qualified enough to do so. I do know Dave is aware of Hafler's work, and explains in detail about what he did, and how he proceeded to design this change.

His assertion is, in order to allow for tube variation and simplicity, Dynaco made choices that work very well, but do leave additional performance on the plate. His suggested mod provides for adjustment for individual tubes, and a target value that he has determined improves overall performance while operating stably. Which is pretty much what he does.

Bill
 
6DZ7 said:
Unfortunately I'm just one voice fighting the ignorant tide of lemmings, with my philosophy that iconic pieces of history should be revered and kept whole. If someone wants to clone a design and mod an amp for the fun and sake of curiosity, I have no problem with that. But history in hand, take a piece of history and rebuild it as is with modern high quality parts to extend its life for the future, but don't create a different animal that breaks the chain of knowledge and respect of those who worked to create a better life for those of their time. Lemmings don't know they are lemmings until they are in freefall.
Click to expand...
I have two Mullard builds with A-470 clone OPTs I'm very happy with, the oem Dyna 70 refurb was an excersize to address compromises to get the most from the Laurent/ Tucker topology with the Dyna A-470s, of which it is capable, with modern passive parts and tubes.
 
Last edited:
Think ill stay away from any change to the 1.5 megs for now , but am doing the screen resistor safety mod for now ---------Does anyone have any thoughts on carbon resistors vs modern metal film type , ( in the signal path only) should we stay with the old carbon type ? -- and which resistors on the PCB are in the signal sound path ?
 
CC resistors should be fine for apps where little to no current is drawn thru them, such as g1 stoppers and g1 grounding ref. Series signal path no-current also should also be OK. Not ok include cathode & plate series resistors, screen resistors in gain stages, ect, where generated noise can appear on the signal path.
 
Also, fusible metal film flameproof resistors are preferable for power stage g2 screen stoppers and power stage cathode resistors.
 
Ok thanks , So have i miss any resistors on this old PCB that should be switched out even if they are still in spec . And is it ok to use the small tube adapters , to get away from the 7199 tubes to 6gh8a , any thing i should beware of here ?
 

Attachments

  • 100_4853.JPG
    100_4853.JPG
    145.9 KB · Views: 39
  • 100_4854.JPG
    100_4854.JPG
    107.7 KB · Views: 37
You must log in or register to reply here.
Back
Top Bottom