Dynaco upgrades - lets talk audio and DIY

[fini]

I'm hoping this thread stays alive long enough until I have time to start working on my amps.* Robert seems to want to tweak his amps in a certain way, while I'm mostly interested in getting mine to work, at all! Of course, quiet, safe operation is high on the priority list. I WILL study the math lessons. You've gotta attend class when it's in session!

*My wife's in Minnesota at a sales meeting, and I'm a single Dad for a bit...

[NOSValves]

Fini,

while learning is great! In reality you do not need to learn one single bit of the math! You just simply need to choose who, what and which method you want to follow and then go at it. It's really that simple. It's all laid out by the kit companies that offer these products. No science, math or tube theory required for the end result your after. No need to make this any more complicated then it has to be. With Mark 4 series amps like you have or the ST70 none of them have high voltage problems that exist in the Mark III so the normal 4 section replacment can will do a wonderful job and last for many years like 20!! Your amp never exceeds 460V naturally at start up and runs at about 430V to 440V once warmed up. Plenty of safety margin for the standard twist lock 525V can. So get to work []

Craig

[fini]

OK, that solves the can vs. board decision. I like simple (with a flavoring of complexity). You didn't want me to just roll over, did you? [] Actually, with three ST70s at my disposal, I could probably end up with a Neopolitan-like variety of amps. In reality, I should sell two of them to folks that wish to join in the fun...

BTW, one of the ST70s is a factory wired amp, and the bottom cover is rivetted to the chassis (in-between the normal 2 each side screws). Have you ever seen this?

[NOSValves]

OK, that solves the can vs. board decision. I like simple (with a flavoring of complexity). You didn't want me to just roll over, did you? [] Actually, with three ST70s at my disposal, I could probably end up with a Neopolitan-like variety of amps. In reality, I sould sell two of them to folks that wish to join in the fun...

BTW, one of the ST70s is a factory wired amp, and the bottom cover is rivetted to the chassis (in-between the normal 2 each side screws). Have you ever seen this?

Fini,

You mentioned the rivet to me before and no I have never seen one with rivets holding the bottom cover on. I have had many factory wired Dynaco amps here before so........ Me thinks the rivets are not factory wired []

[anarchist]

Alrighty,

when are we going to begin the bias schematic lesson. I need some guidance in getting started... can we try translating some of that to English so I can follow along and then lets do some math.

I would still like to know if I got this right? Otherwise I need to do some reviewing.

I will merely specify the Poseidon circuit with the difference in
the tubezone circuit and if someone can tell me I am on the right track
here.

100K resistor between 330 volts and 430 volts. (47K resistor)
Hence, this resistor is dropping 100 volts and Ohms law says the current is 1 mil. ( Ohms law says current is 2 mil.)
Connected
to pin 1 with a 121k resistor which ohms law (1 mil) says voltage is
291. (150K resistor which ohms law (2 mil) says is 300 volts).
On
v2, 430 volts connects to 68k resistor then pin 1, ohms law (1 mil)
indicates 68v drop so 362 volts on plate. (430 connects to 47K resistor
then pin 1, ohms law (2mil) hence 94v drop and 336 on the plate.)

Do I have the math down?

[anarchist]

Uh, thanks Mark. I knew that and wondered why you were telling me that again. Then I read my post and can't figure out what the hell math that is or what I was doing; now I know what 2 hours sleep does to you.

Let me try this again but I will make it simpler. Substitute a 47K for the 100K. A voltage drop of 100 would indicate current was 2 mils right. Pass that through a 150K instead of a 121K. 150000 * .002 = 300volts. Subtract that from the 330 and we have a whopping 30 volts on the plate. I believe this is not possible leading me to the conclusion the tubezone schematic is wrong. Here it is (just the relevant piece)

[NOSValves]

Uh, thanks Mark. I knew that and wondered why you were telling me that again. Then I read my post and can't figure out what the hell math that is or what I was doing; now I know what 2 hours sleep does to you.

Let me try this again but I will make it simpler. Substitute a 47K for the 100K. A voltage drop of 100 would indicate current was 2 mils right. Pass that through a 150K instead of a 121K. 150000 * .002 = 300volts. Subtract that from the 330 and we have a whopping 30 volts on the plate. I believe this is not possible leading me to the conclusion the tubezone schematic is wrong. Here it is (just the relevant piece)

I'm not all clear on what you have wrong with your figures but no way in heck is 30V correct. You change one thing on the circuit and all things related have to be changes also.

Craig

[NOSValves]

Mark,

By George I figured where the confusion is coming from. He is looking at a long tailed pair and thinking they work like your example of a simple triode. Stop hold the presses were way ahead of ours selves.

Craig

[anarchist]

Here is a more complete schematic. The only thing I am looking at is the connection running from the 430 to Pin1.

Note everything here matches the Poseidon schematic except for the changing of resistors from 100K to 47K, 121K to 150K, and 68K to 47K (BH7).

Did these lessons let me identify an error in a schematic? Who da thunk it.

[NOSValves]

Anarchist,

One thing you will notice is the actual end result change for the 12AX7 is really small 150K + 47K - 197K total 121K + 100K = 221K not all that much in this context. This is just a slight difference in the voltage operating points of the 12AX7 stage. The change on the phase splitter is to allot for the new tube type being used and even with this the change is not huge but that is not surprising the 12BH7 and 12AU7 are similar.

Craig

[anarchist]

Nah, the changes aren't huge - I am able to recognize now (via observing and doing the calcs, they are merely tailoring the plate voltage) but the schematic has a problem. There is no way the voltage at J and K are correct; if so current would jump to 2 mils and the plates would be getting minimal amounts of voltage rather than increasing the voltage as was the intention. One or the other is wrong.

My belief is the intention here was to put 230 volts on V1 and 383 on the V2 as opposed to the originals 209 and 362. The schematic just wasn't adjusted appropriately. If we read this schematic as it is written, we are putting 30 volts on the V1 and 336 on V2. I don't think so.

That bias schematic is also easy stuff once understood. Cool as hell but easy.

Edit: I saw Mark jumped in here to point out the error (uh, discrepency as well). As far as catching on, it helps when learning this stuff to have resident guru's who can prod you along quickly... thanx.

[NOSValves]

Mark,

I have a question if you don't mind. What program are you using to make these drawing they are simply awesome. Man I wish I had your software skills.

Craig

[NOSValves]

Hmmm I have a copy of that software! Couldn't get you to sell me your elements I know time is money?

Craig

[anarchist]

Another great lesson. Initially, I saw that diagram with everything on it and thought "oh, no". The color coding and preceding lessons however allowed me to quickly realize it was just an assembly of the various pieces we had dissected before with a little more detail.

Thanks Mark for the walk-through of the diagram. I am sure others appreciate it and it reinforced what we had already discussed. Great stuff.

[risingjay]

Great thread!

Quick question, when soldering boards, you use rosin core right?

[3dzapper]

Great thread!

Quick question, when soldering boards, you use rosin core right?

Correct. Acid core is for plumbing joints.

Rick

[anarchist]

This was a great thread and provides plenty of opportunity for learning & I have made some progress in my quest so I thought I would provide some updates.

The Poseidon driver boards are built save for the main caps. I have Auricaps in the house but am contemplating putting Jensen coppers or Mundorf silver/oils on the driver boards so it has been sitting waiting for me to make up my mind.

There are basically two approaches to "fixing" the Dynaco PS. Use an AES can and install some thermistors (NosValve technique) or spring for a capacitor board. You can do some more extensive reengineering (see Ricks solution or VRD's which substantially increase the values) but I wanted to retain a stock appearance, fix the voltage issue, AND maximize the potential while doing so.

I bought the SDS boards and THEN decided to do a lot more reading. I reviewed posts over the course of the last 5 years by the major contributors in this thread, read posts covering 4 or 5 years on audio asylum and the dynaco doctor (saw questions & contributions there by certain folks in this thread as well), evaluated the designs and approaches by many different "experts" in the area and frequently asked for Marks opinion.

Ultimately, I decided I would use the SDS boards but decided to deviate substantially from the "instructions" and standard build.

Due to Mullards 5ar4 being designed to see a 40uF input filter max, I decided I would use 82uf caps instead of the 100's or 120's the build sheets specified. I also decided to go with the better caps and chose Chemi-Con KMH caps offering a 105c temp rating and 400volts.

Next decision was the "tank." Consensus overall was the 60uF specified on the SDS board didn't really take advantage of the transformers in the Dynaco's. Craig, during rebuilds, uses an AES can which provides 80uF. After a great deal of looking and research, I settled on better quality Panasonic caps offering 105c temp rating, 350volts, and which offer me 90uF (2 180's in series).

The last two capacitors didn't really concern me all that much. The AES can Craig uses offers 30uF and 20uF. The SDS board stock is 60uF so they go up a bit. Mark impressed upon me his feelings that higher voltages to the driver board would offer improvements. I found some old posts from Craig where he had suggested higher values for the phase splitter and driver were beneficial. In the end, I decided to go higher and used the same Panasonics I am using for the tank.

Of course, I also decided to use "better" film bypass capacitors; speaking of which, I uncovered frequently in my reading that knowledgable folks suggested if you were using a "can", you absolutely should install bypass capacitors on it to improve the sound (for those of you continuing to use the can approach) and continue to pray the can didn't fail.

In the end, I ended up with a solution that is easy to implement, absolutely fixes the problem, and offers up 41/90/90/90. The AES/Thermistor solution offered up 40/80/30/20 but the "fix" only worked in the event of using a 5ar4. In my case, I will be able to switch between 5u4's and 5ar4's as the mood or need strikes me.

Now it is a matter of pulling the trigger on the caps for the Poseidon and getting it installed. I wonder how Rick is doing with his Poseidon's and the Jensen coppers...

[NOSValves]

In my design my "tank" is 280 uF !! The final two sections are 100uF each. I persoanally don't utilize the single can approach.

Craig

[lwhaples]

Nice update,keep the info coming.Though I don't understand most of it.[:$]

Larry

[CPlaza]

Anarchist...This is a two part reply

Part 1...I have had a pair of dead Mark III's for about 20 years, and could not get parts (paricularly the underrated filter cap on the rectifier tube). After some research I discovered the Triode Electronics power supply and the twin triode driver board. Since these amps were dead in the water I thought I would upgrade them for the fun of it. I also ordered the modification to install 2 bias controls (one for each output tube), thereby eliminating the need for matched pairs of output tubes.

Result...After completing the job, I "cooked" them for a few hours to let the new tubes settle down, then readjusted the bias on the KT88 output tubes. I then hooked them up to my Forte II's and to my surprise I found that nothing in my studio sound as sweet (and I'm talking about Crown 800 watt amp, Crown 350 amp, and a Mackie 1440i). With a good recording of a jazz piano, if you close your eyes, you feel that you can touch the piano...and all my studio friends agree with my observation. The best part is that they whisper just as well as they roar. Of course, this also has to do with the speakers you will be using. As far as reliability goes, I have been using these puppies for nearly two years and they haven't hiccupped once in that time. Since I a not a fanatic, I have only retweaked the bias once about a year later. I hope this helps you decide, if you haven't already done so.

Part 2...These kits are not for the faint of heart. You have to know how to identify the various resistors and capacitors, in addition to knowing the polarities of the filter capacitors and how to install them on the boards properly. If you lack the skills to do these upgrades, I suggest that you simply get someone with the necessary experience to help you. I am not being condescending here, and please don't take it that way. I just want you to experience these amps and hope that you enjoy them as much as I do.

Carlos