This is the beginning of what is the second phase of a greater project going back several years, that being the initial salvaging of this amplifier after being found stored in a Florida attic for over 40 years. For this effort a complete video series exists where the amplifier was brought to a working order but left in the original patina also retaining the use of less that reliable parts, which have caused unpredictable issues over the years.
As a primary platform for this restoration, I've opted for the stainless steel chassis from dynakitparts.com along with new pots, switches connectors, screws and so-forth. Further work will be the cosmetic restoration of the transformers as well. An unboxing goes into each part and kit specifically. A smaller secondary unboxing comes later in this episode.
The project itself begins with the removal of the output transformers which i've opted to recolour on the ends of the cables on both sides due to fading. At that point it was safe to cut out of circuit. This process includes cable to the rectifier and the power transformer as well as the selenium rectifier, some of which have two colours. Once all cables are clipped, starting with the power transformer, each are removed one by one. Emphasis for the power transformer starts with the connections on the c-clamps. Output transformers are removed in a similar fashion, noting that the red center tap wire goes all the way to the electrolytic capacitor and needs to be pulled back on one side.
With all transformers removed the covers are unscrewed and separated so that they can be sandblasted and painted. All of the inner transformers appear ok. I take the covers to Jason's bug ranch and put the covers into the sand blaster to clean then which also textures the raw metal for good adhesion. The first painting attempt as a high gloss, didn't go well. too many imperfections in the metal, and dust and wind in the shop. A second attempt after a stripping and re-degreasing came out perfect by allowing for a light texture to come through, almost like a leather or knockdown.
Each of the transformers are masked and sanded with a Dremel for rust removal and painted as well, cleaned at intervals with spirits. During this time, custom threaded rod with acorn nuts are made for the restored transformers to replace the old screws. With all of these pieces, the reassembly of the transformers commences. Paying attention to stamp placement, the output transformers are done first, doing the measurements of the studs for the final cut. With that the covers are tightened down.
The power transformer is assembled last, also ensuring that the cover is on in the correct direction. The rod length must account for the new dampeners so no adjustments will be made at this time. Having completed the transformers, this concludes the first video in this project series.
Last edited by jrubin on Mon Apr 15, 2019 8:27 pm; edited 8 times in total
Now in Part 2 of the Dynakit ST70 restoration project, we will begin with the choke as the first component for installation, though I thought it interesting to compare both on the Genrad 1650A impedance bridge to see that the impedance and inductance is.
SPEC 62ohms 1.75h OLD CHOKE 63ohms 1.50h NEW CHOKE 64ohms 1.35h
All within acceptable range, the new choke is installed into the chassis, though, the old one is definitely a keeper for other projects. All work is done with the intent on maintaining a scratch free project and with extreme care. The tightening of all screws in this project has the slot of the screws pointing in the same direction.
The two switches followed the choke and interestingly, they were not threaded like the original ones. This was helpful though and allowed the screw heads to maintain a parallel appearance. Following the switches was the assembly and installation of the gold plated input jack board.
Returning to the old chassis, the selenium rectifier, power cable, fuse and stand-off board was removed for transfer. It became apparent that a custom screw for the rectifier would need to be fabricated. As these would be put into the new chassis it was also a good time to introduce the bias pots as well. Tightening in such a manner that the flats would be parallel to the font of the chassis.
The rear jack kit was assembled at this point. This kit did not include the 16 ohm option, instead an internal standoff for 16ohm serving as a tie-off for the feedback connection. These are much better than the screws in the original kit. Upon assembly, these too are mounted. along with the end stand-offs.
Returning from Lowes with a 6/32 screw the fuse socket was installed and the screw was cut as a rod in support of the selenium rectifier, using an acorn nut as the top cap. The standoff board was installed and then I'd assembled some hardware to test the rectifier install. It looked good and was done so that the flat again was parallel to the chassis.
The tube sockets were a little tricky because they are a bit smaller than the hole. So I had to shim it with two turns of paper to center the socket whilst tightening. It seemed to work good though. Mundane is having to repeat this several times. Moving to the power transformer was welcome.
Its challenging though, guiding this heavy thing into the chassis without scratching anything with all of these wires everywhere. Somehow I managed to pull it off. Dampeners and all installed. Immediately I install for regular nuts on the bottom to hold the transformer in for safety. One benefit of the transformer is the ability to flip it to the service position.
Power transformer wiring begins with the EL34 filaments, BROWN and GREEN. This included their center taps to the double capacitor to ground. Following this will be the connections to the rectifier. Then the primary black wires off of the power transformer to the fuse and switch. Following this, is the single cable to the selenium rectifier.
At this point I go back to the old chassis to remove the star ground hardware and install it in the new chassis. The first connection to it would be the ground off of the power transformer. At this point a 3 prong power cable is introduced and wired in. The ground of this cable also meets the star ground.
With the power testing is conducted with the Heathkit IM-11 VTVM using a fraction of the voltage. Switches are tested and then the ratio's determine that the windings are ok. The center-taps to heaters were measured by resistance across the coil.
Having completed this testing part 2, has come to an end.
Last edited by jrubin on Mon May 13, 2019 8:02 am; edited 12 times in total
Continuing with part 3 of the Dynakit ST70 restoration project, we will get right into the output transformers, though given the mass, and the nice shiny chassis, I cover everything first with masking tape in case of a mishap. All of the wires need be carefully negotiated through and the screws and nuts placed in and gently tightened just enough so the the transformer doesn't shift. Once this is accomplished, there must be an accurate account taken of the location from the end so that it remains parallel before tightening, as well as allowing the other side to be symmetrical. In the desired location, it is torques down. Wash rinse, repeat. Though the second one has the burden of matching the measurements of the first.
Sadly, having caught one wire under the chassis, I got to remove and install the same transformer, twice. I was then able to solder the newly connected transformers to the output jacks and then the connections to the EL34 sockets. The only outstanding transformer connections at this time are the red center taps to the electrolytic capacitor, not yet installed.
Having established a colour code for the circuits within the amp, I had a go first a snipped wires already in the amp from swapped components, this was predominantly the bias circuit, for which yellow was chosen. Green was chosen for Ground. Black for anything after Labor day. Also the grounds off of the output jacks were installed to the star ground.
This is a great time to pull the electrolytic cap from the old chassis, which is new, by the way. and prepare it for the new chassis. This means polish it so it doesn't look so industrial. I just ripped off the sticker and stuck it to the bottom of the power transformer. This Fiasco started at 800 and worked up to 1000 and then to 2000, and the chrome polish sealed the deal. Shiny and new, the capacitor is affixed to the new chassis, taps twisted for a tight and permanent fit.
The cap is immediately grounded to the star ground. and its terminals cleaned up, and its resistors tested. Wiring to the capacitor then begins starting from the rectifier, the HT colour will be Red. This is followed by the coil C-354. The resistors are left on the capacitor, though there are further connections being added later once the circuit board is introduced.
Smoke test time, round two. This time well take several measurements, at full voltage governed by five volts at the tube rectifier filament. This allows the testing of the bias circuit, onward through the individual pots. Also, a further inspection of the AC ripple before and after filtering shows the filter to be working just fine. A final check of the AC output on the tube rectifier socket shows an unladen value of 800 volts. Incredible.
So dropping a 5AR4 in will allow some DC measurement . The unladen rectifier output is 540VDC. This seems like a good stopping point for this Part of the project....
Last edited by jrubin on Mon Jun 10, 2019 8:06 am; edited 6 times in total
Now in part four of the Dynakit ST70 restoration project, I start off by demonstrating how dangerous of a condition i've left this unit in given the fact that nothing is connected to the electrolytic capacitor to drain it off to a load over time. Were talking about well over half an hour and the chassis is still lethal. Food for thought.....
With that out of the way the heaters are brought down from the EL34s to their respective pairs, I choose some nice fabric wire for this task to match the power transformer. After this it was an appropriate time to remove the old PC-3 circuit board from the old chassis by cutting it out and unscrewing it. It really required a cleaning and polishing too, so that was done before continuing. I was careful to leave short ends of the cut wires on that board to keep track of what would be reconnected to this board, I knew with the new bias controls there would be some modifications in the future.
The board was mounted into the new chassis and tightened down. The wiring immediately begins starting from the output transformer for the capacitive feedback. I chose blue for that. Then the bias pots were added to the board. Logically, the bias extended out of the board to the four EL34's though since it goes through 1K resistors on pin5, the cables go to unused pin6 on the EL34 and a new resistor goes to each pin 6 to pin 5.
The high voltage connections come off the electrolytic capacitors in red to the board, also making working on the amp now less dangerous upon shutoff, then the blue wires continue for the capacitive feedback circuit from the board out to one EL34 from each side. After this the main PCB ground is added.
Now is a good time to finally re-add the clamps that hold all of the cables in place by the transformer, so I stop to do this, since work now moved to the front of the chassis. This work now the front input jacks connects the jacks to the mono/stereo switch as well as the circuit board. These signals on input will be White.
The Bias Balance Control Upgrade, BBCU is the next part of the installation, the special kit i've ordered for this amp. I follow the instructions provided in the provided manual to do so. And remove the rest of the cur cables remaining in the PCB at this time. The only modification to this install is that I tie off all of the grounds to the chassis star ground, and I use my own thicker wire. Again, during socket installation I had to shim with paper to center them.
With the new kit installed, the cables are laced in and this portion of the project is now complete.
Concluding our project in part five of the Dynakit ST70 restoration, I point out a terrible mistake I made that could have caused some serious damage. A short to ground from the rectifier through the audio transformers would have not been fun. Having corrected it, i'd decided it might be a good idea to test the resistance across all point of the electrolytic capacitor for good measure.
Now sure there are no shorts that will destroy the amp I use the filament voltages of the EL34 as a reference to see how high the variac will be turned to get the unladen voltage correctly. Then, I turn the bias pots to full tilt to ensure minimum current draw when the tubes go in. As the Variac is introduced the overall current of the amp is monitored and the Variac adjusted to maintain 6.3vac on the EL34 filament pins. With nothing on fire, its a good sign. Also, looking at the rectifier output and shutting the amp we see how quickly the discharge of the electrolytic cap is now.
At this point the phase inverters are introduced into the mix, again adjusting the Variac against the EL34 filaments. The oscilloscope monitors the outputs of the phase inverters to verify operation. Satisfied with this, The EL34's are now introduced into the Amp after shorting plugs are plugged into the inputs and 8 phm 100W bricks are tied off to the output. To test the accuracy of the new system i'm adding my current meter inline to the tubes as well where:
.10 = 10ma .20 = 20ma etc etc
This was conducted on all four tubes and found to be highly accurate. Knowing I can use the front jack values, biasing could begin. Biasing is the same as the old system except that the contribution on each side must be further balanced by the pot in the front.
I take this time to give the amp with the shorting plugs in a good listen to hear of there is any hum. It is almost non-existent, i'm very pleased with this. I celebrate this by adding the rubber feet to the bottom cover of the amp so that the bottom might be screwer onto the chassis. The bench is then reshuffled for THD testing. At that point I realize I should take out the Hickok 6000 and make sure the EL34 tubes are balanced in each channel, turns out they weren't so I was able to make an improvement before tests even started.
Testing would be conducted at 11.3 volts RMS which is 15W roughly at 8 phms. I'm calling that roughly half power. I do a demo of one given frequency in the amplifier. This method takes a long time, and I lost my good cables to do this, so I get some noise at around 2000hz forming in the cables, but both sides were perfectly consistent, this tells me everything is working just fine.
Seems appropriate that I end both the video and project with a listening portion. I chose some royalty free music from Bensound.com. There will be future videos with, for, and including this amp, but as for the restoration, this is where it ends.
Last edited by jrubin on Mon Aug 12, 2019 8:20 am; edited 4 times in total
J, looks like you're having fun! I have to add that it was a project similar to what you're doing now that rather got me 'addicted' to this kind of activity. Now, some 20 years later, and with countless projects like this behind me, I have to say that's it's still as much fun today as it was then.
I hope you'll allow me to make one suggestion here; that being related to the two electrolytic caps you have on your bias supply. I say this only as it's been advised to me in the past in other such situations, so I pass this on only for consideration. If I'm looking at your pictures correctly, it appears you've got 100uf / 450V caps across your rt and lf bias supply sections. While this may work, the 'issue' (if there is one and according to how I've been advised in the past) is that the 'voltage' level of these two caps is two high. The bias supply works at about 50 to 60 volts in the ST-70. I believe Dynaco used 50uf/75V caps in the original units. Most 'kits' available today will supply a 100uf cap with a 100V rating.... which is plenty for this part of the circuit. As I've been told, use of caps should always consider the 'working voltage' range, then be selected with about a 25 to 30% margin. So, as Dynaco did it, the 55VDC of bias supply (as rated) with that same margin applied, the 75V cap came out right on the money. The problem is that caps of too large of a DC voltage when used in lessor voltage supply circuits will not 'form' correctly or completely. So, when 'filtering' as they do, if the cap is too large in a circuit it simply may not work as it should, and as such they may run the risk of failing as well. This is clearly not the same as the 'reverse' side of the coin, where a cap of too small a voltage level is doomed to fail almost immediately. However, if you consider the long term affects of using too high a voltage rating for a filter cap, you do all this time-consuming and costly efforts...with great satisfaction of a 'job well done' in the end, and then for about $5 in two component parts, you start having 'issues' with your amp down the road. Anyway, I say all this simply because I see the two caps there and it just seems worth pointing out. And, it's an 'easy' change as you're doing all this other work.
Beyond that, enjoy the project; sadly it will be over way to quickly in retrospect. And, you'll find yourself longing for the next 'resurrection/restoration' project. Good luck with that too, as those are getting harder and harder to find!! Maybe time to do as I did.....and start 'scratch' building. Lots of fun in that too. Best Regards, Tom D.
Last edited by Wharfcreek on Thu Feb 28, 2019 10:23 am; edited 1 time in total (Reason for editing : Correct a typo)
Endorsing Tom's advice: The general rule-of-thumb on electrolytic caps is no less than 150% of actual operating voltage, but no more than 300% of actual operating voltage for exactly the reasons noted. In this case, splitting the difference at 100V is just fine. I do always advise using 105C caps in any tube device, however.
For non-electrolytics, it really makes no functional difference - a 600V film cap will respond just as would a 100V film cap.
Peter W. wrote:Endorsing Tom's advice: The general rule-of-thumb on electrolytic caps is no less than 150% of actual operating voltage, but no more than 300% of actual operating voltage for exactly the reasons noted. In this case, splitting the difference at 100V is just fine. I do always advise using 105C caps in any tube device, however.
For non-electrolytics, it really makes no functional difference - a 600V film cap will respond just as would a 100V film cap.
Don't know how those got in there, they would be suitable for other projects....