The BIG Dyna ST70 driver board thread

[Kegger]

I hear yu, just wanted to make sure some didn't take it as "needed" on "any" 70.

 

[shelly_d]

I hear yu, just wanted to make sure some didn't take it as "needed" on "any" 70.

Agreed. One day I do some bandwidth testing on the various stages of a stock stereo 70 to see just where roll off occurs and which stage it occurs in. That is somewhere in the distant future however.

 

[Brinkman]

Read Frank Van Alstin's rebuild article on the Stereo 70. It explains why that circuit does have that "muddy" sound in the mid-range and treble...
By the way - I tried his band pass filters set to 17 Hz and 21,000 Hz and the muddiness went away totally. Tells me he knows what he is talking about.

Thank you for the link. I recall coming across that article in a thread that had really drifted off into the "input bandwidth limiting" issue, so I understand the trepidation about bringing it up. However, I think driver topology is obviously what makes these boards different or similar to one another, so I feel it's a legitimate topic for discussion, so long as everyone is reasonable.

By the way, thank you for answering my question. But I do have issues with that article.

All issues aside, I would love to try one of his Super 70 kits. People are posting here that it sounds great and that's enough warrant my interest.

What I find perplexing is this: if the stock ST-70 driver sounds muddy and limited because of two factors; one, it needs input limiting to keep it linear, and two, the coupling caps are too small in value, then how does a similar layout like the Triode board escape these ailments? Is the Van Alstine method of improving the ST-70 but one of many ways to skin a cat?

Coincidentally, I am already heading in a direction close to what the article suggests. In an earlier post I mentioned bi-amping. An active crossover will filter out the low-pass and send it to a SS amp, while the mids and highs will go on to the ST70.

If the article still holds any water, all a stock ST70 ever needed was an active crossover/bi-amping setup (and a 17KHz upper frequency limit) to remedy all it's problems.

 

[GordonW]

I'm not going to get into a debate over this either.......

But the following clipped from above:

"the amp and output trannies can't respond that low"

Is just not true. A good working front end and adequite power supply, the ST-70 does
not "need" to be bandwidth limited. But in it's original form I could see it being a benefit.

Well, NO transformer is going to work at DC. That's just inherent physics. And few will work well below 5 Hz.

So, putting a cap in there to limit stuff below 3-5 Hz or so... usually is a good idea, IMHO. Minimal phase shift/group delay in the audible range, and the amp just runs more cleanly, as a result, IME...

However, I do NOT agree with band-limiting ANY "hi-fi" tube amp in the HIGHS (i.e, adding a low-pass filter), on the input, usually. Unless it's unstable, it won't need it... and if it IS unstable, FIX IT to where it IS stable. This may require somewhat limiting the bandwidth of the CIRCUIT itself... but it makes no sense to add MORE filtering (and more phase shift) to the signal, other than what the amp does as part of its own stability compensation circuitry...

Regards,
Gordon.

 

[Brinkman]

So, putting a cap in there to limit stuff below 3-5 Hz or so... usually is a good idea, IMHO. Minimal phase shift/group delay in the audible range, and the amp just runs more cleanly, as a result, IME...

Gordon,

Don't pre-amps commonly have cap-coupled outputs and wouldn't this filter out the DC into the next component?

Best,
Brinkman

 

[GordonW]

Gordon,

Don't pre-amps commonly have cap-coupled outputs and wouldn't this filter out the DC into the next component?

Best,
Brinkman

Tube gear, usually. Solid state, not always.

For example, I've seen CD players with tens of millivolts of DC on their outputs... anything digital has the possibility of similar behavior.

Also, there's always the potential for a leaky output cap on a tube preamp...

Regards,
Gordon.

 

[MikeyN]

VTA Mod

I ran my ST70 with the original circuit board design (from dynatube) for about a year before upgrading to the VTA High Gain mod.

I had some issues with the instalation (I'm not much with a soldering iron yet) and ended having Stereo Exchange in Chicago take a look. (If you are in the Chicago area and a tube rookie like me, Stereo Exchange is a great option)

I just brought it home a couple of weeks ago and the thing sounds great.
The most noticable difference being the low end response and overall clarity. The amp seems to be quicker and clearer.

The original board design sounded very good but I would have to say that the VTA mod was worth the effort.

 

[NOSValves]

I agree, great idea for a thread. I just finished installing the K&K Audio ST70 mod board. As you can tell from the pictures on K&K's website, the board is very well made. The design is really advanced in that it employs differential drive and output stages with constant current souces (CCS). The idea behind CCS is that it provides a more ideal current source to the amplifier. This design places the output tube's operation mode in Class A as oppose to Class AB1 of the original. I think I'll let you read the rest of this at K&K's site.

A while back, I went to this gentleman's house to audition a pair of Altec Model 19 that I eventually bought. We kept switching back between his custom made 300B SET amp (with very nice custom made OPT, btw) and a pair of monoblock Class A 6550 tube amps. Both amps have their strong points and were good in different ways. I really like how that class A PP amp sound, but due to my hectic work schedule, I haven't been able to put together a design of my own.

I have been hanging onto a mint condition ST70 with the intention of adding an upgrade board to it. I have also been keeping an eye out on new ST70 developments but it seems like that's all we are going to get (most of the list above). I wasn't aware of the Blackburn and Mapletree boards. I have heard the VTA board on my friend's ST70 and it sounds very good. When I ran across the ST70 board teaser on the K&K forum, I decided to wait to see what it's all about. I ended up buying the board with the amorphous signal transformers.

Unfortunately, I ran into a ground loop issue earlier today so I couldn't listen to it. I finally had it figured out late tonight so I haven't crank it up at all. I am burning in now with low volume music. I must say that the sound at low volume is very SET like with nice details. I hope to provide an update in a day or two.

And a side note, the power supply for the whole amp is included on the board so that you don't need to add anything underneath. Biasing is accomplished via the CCS so you don't need to mess with it much nor do you need to old bias circuit.

That K & K design looks real interesting....

To date out of all the others I have tried I prefer Shannon Parks board www.diytube.com

 

[Brinkman]

[With] a good working front end and adequite power supply, the ST-70 does
not "need" to be bandwidth limited.

You're not the only one who feels this way. For the record, here's the published Dynaco output transformer specs, from Triode Electronics' site.

 

[GordonW]

You're not the only one who feels this way. For the record, here's the published Dynaco output transformer specs, from Triode Electronics' site.

The output transformers haven't ever been in question, IMHO. To my view from testing them, it's the stock power transformer that's a little lightweight.

Once that's fixed (a suitably heavier transformer installed), and suitable capacitance is added, things get a lot better. The need for bandwidth limitation pretty much goes away... with the exception that a DC-blocker cap on the input is still a good idea... IME.

 

[Brinkman]

Gordon,

You bring up an excellent point, and a rather interesting one when you consider the appreciation of collectable vintage Dynacos vs the less sought-after new-production Dynakits (including the "Dynaclones" to be had through Triode Electronics). One thing that sets the newer kits apart from the vintage ones is an improved power transformer. There's others more qualified to weigh in on this, but I understand the filament taps on the newer power transformers are rated at 5 Amps, whereas the vintage PT is only at 3.5 or 4 Amps. This means more power tube options.:thmbsp:
This may be the only improvement made however, I'm not sure. I would definitely agree that the bass response on my new-production ST-70 is stretched a little thin at times.

With that in mind, I am somewhat surprised at the appraisal of some vintage Dynacos, especially those in need of some serious TLC and/or missing their tubes. Unless you're dead-set on restoring the genuine article back to it's original glory (down to the cheap component-laden stock driver board), I can't see the point in breaking the bank on one.

 

[smbrown]

I've done a few different ST70 board replacements. First was the "Full Curcio" power supply and driver (4 x 6DJ8). I thought it too analytical and SS sounding. I also tried the Triode Electronics board w/EF86 and 12AU7. Nice, but didn't really grab my attention for some reason. Later I built the original Hafler circuit using point to point wiring and higher quality parts - like AN PIO caps. This was the best so far. Bear in mind, all this was done over a period of several years while I focused mostly on SET projects of my own design. I kept getting pulled back to play with the ST70, however. I don't know why - mass appeal or something. Anyway, I most recently tried the VTA board with the high gain 12AT7 configuration. In retrospect, I really needed less gain - a LOT less gain. I'm now padding that down by probably 10db or so and it's okay. Next up is the newer version of the Triode Electronics board with the Ecc 99 driver. I think this might just be the ticket to match gain and drive. We'll see. Overall, none are awful, some are really pretty good. It's a great amp to play with and learn on. Personally, I think the best combo might be the traditional Mullard 5-20 design (like VTA) but with the EF86 in triode not pentode, and elimiate feedback, and put the EL34's in triode. All that together, I have a hunch, would be outstanding. As for the new kid - the K&K board, maybe - I've tried phase splitting transformers at the front of a power amp and found the low Z drive requirements to be too hard to meet, given my desire to play with a variety of preamps. Nevertheless, I've got a lot of respect for what I've seen on the K&K site, and this is likey a great contender, too.

 

[Kegger]

Vta low gain mod

If you've got a VTA high gain version it is quite easy to change it to low gain. (I call it normal gain)
Here's what you do for those who may be interested in what changes are involved to accomplish it.

Change the 68K voltage amp plate load to 43K an the phase splitters from 47k an 51K to 33k each
(all of these resistors should be at least 1 watt, preferably 2 watt)

Change the 150k RC on the voltage amp an the 7.5k on the phase splitter to 4.7k on each section
(anything from say 3.3k to 6.8k on both RC sections will work here)

Change the 2.2K cathode resistor on the voltage amp to 1.2k (remove the bypass cap on the VA)
(not all versions have the bypass cap in place to start with, you don't want one there)

Change 1k cathode resistor on phase splitter to 820ohm (820 is standard resistor value, 750 is fine)

Change the 7.5k feedback resistor to 10k (also use the 8ohm tap for feedback, not the 16ohm tap)
(use this new value along with the stock VTA feedback cap on the 8ohm tap) (5-15k can be tried here)

(these are done on both channels)

-----------------------

I believe also that a 2.2k resistor is used in the supply feeding that front end, I use 1k in that spot.
(What I like to see is between 375v an 390v to the front end/VTA board, depends on supply config)
(if your supply is original type then 2.2k or higher value is what you want here to set proper voltage)

With 10ohm cathode resistors I run between .45 and .50v for the bias, so as to run 45 to 50ma current.
(you could run it hotter with KT88's but your starting to stress the single 5AR4 rectifier tube here)

-----------------------

These changes let you run 12AU7, 12BH7, 5965, E180CC, 6829, 7062, ECC99 (and my favorite) E80CC Volt amp.
(there is plenty of 12AU7 variants as well, best sounding center tubes to me are E80CC, E180CC/7062 an 6829)

Your phase splitters now become optimized for the 12BH7 and ECC99, can use 12AU7 or it's variants as well.

I use E80CC as Volt amp with ECC99 in the phase splitters as my standard/favorite combo in these VTA amps.
(12BH7 in the phase splitters are nearly as good as the ECC99) (Voltage amp is center tube, outers Splitters)

Now the amp has a LOT less gain, you can turn your pre volume up now, ALSO it sounds much much better to me.

----------------------------------------------------------

At which time you feel a bit more gain still would not hurt you, the feedback can be removed, works well when those
EL34 are wired in Triode operation. I also like JJ KT77 output tubes in the amp with the setup/configuration here now.
(I tend to leave the feedback connected on my ST-70's, mainly to keep gain/sensitivity in check for active preamps)
(if using no feedback I suggest a 12BH7 or 12AU7 as the voltage amp and ECC99's for your 2 phase splitters)

If using feedback, on this or any other ST-70 setups, I do suggest move it from the 16 ohm tap to 8 ohm tap as well.
(I also tend to run less feedback on UL based amps, just don't feel there is all that much "needed" there)

A couple more suggestions, if the coupling cap on the driver an grid to ground cap on the phase splitter (c1 - c3) are
still .1uf I suggest upping those to .22, If the bias/grid resistors are still 270K I suggest 120k or at least down to 150k.
(the bias/grid resistors really need to be done for running KT88's, plus a benefit for the EL34's an KT77's here to)

After all that is done, if you have the newer and upgraded power tranny, KT88's will work quite nicely in this amp..

If you can not get enough bias voltage then yu increase the 27k bleeder resistors (R37 - R38) to say 33K for starters.
(Increase them higher for more bias voltage, an decrease them for less bias voltage range accordingly)

As far as power supply goes, I add in a 220uf cap underneath the chassis after the choke in parallel with the can cap.
(first cap in the power supply should be no greater then say 40uf)

--------------------

If you have an VTA low gain board these are my latest tweaks an suggestions for it, I keep this post regularly updated.

 

[drewbolce]

I've done a few different ST70 board replacements. First was the "Full Curcio" power supply and driver (4 x 6DJ8). I thought it too analytical and SS sounding.

Thanks for the input. I am redoing my ST70 (again) after an EL34 blew. I have the original four tube driver board and briefly considered using it. I ended up using a new original design board instead because of the unknowns with the Curcio four tube. I never had any intentions of using the power supply board.

My questions to anyone with any first hand experience or design knowledge is: what are the differences between the three and four tube versions of the Curcio design?

Also, I only have the original Glass Audio article (1/89). I have handed down blank boards purchased around 12/99. I am sure there have been circuit or component value changes since the original article. I have made several attempts to get a fresh manual from Curcio Audio but I have yet to get a response.

I would still like to try the board someday, but without the above knowledge I'll just stick with the 7199's on a good board with updated components.

 

[smbrown]

Kegger, thanks! I'll give the suggestions a try. Seems like that board has a lot of potential that isn't fully getting out right now. Maybe I'll make it a project for this weekend.

 

[worldoftone]

Lots of interesting reading here. As someone who has heard/used a ST-70 since dad built it in '66 (basically all my life), I actually like the nature of the stock amp, quirks and all. I use mine 3 - 6 hours a day. Every day LOL! My ears are so tuned to the thing now I'd hate to change it.

But, I can see where these mods can enhance the design to make it better.

Good, informative stuff. Maybe I'll get another one experiment a bit . . .

- WOT

 

[CharleyW]

As I type I am listening to Curcio (premium) boards purchased from Glass Audio circa 1990.

Originally I assembled it according to the article (power supply and driver) using Sovtek 6922's and Siemen's EL-34's - all purchased from Triode Electronics.

I thought it was a little sterile sounding, but, it threw an incredible stereo image, and I was very happy with it, until ...

An EL-34 died.

Two years ago I removed the power supply board and re-wired the outputs to accomodate Sovtek 6B4G outputs (with a Cetron "potato masher" 5R4WGY as a rectifier), giving me around 8 glorious watts of Class A power. Recently I replaced the 6922's with Electro-Harmonix 6CG7's (no circuit changes necessary).

This is fed by a Pioneer Elite VS-39TX receiver in stereo mode - and it is fabulous (I am sure it would be better without the Pioneer receiver in the mix). For the last 10 weeks this has provided all music and TV sound for me (in the living room - my bedroom is powered by Bottlehead Paramours).

There has been only change to the driver board that I am aware of - a 390pf cap between pins 1 and 6 on V3 and V6 (search the Dynaco Doctor forum on Audio Asylum for 390pf and you'll find Joe's post). I have not implemented that change.

 

[Brinkman]

Here's a link to some technical measurements performed on the high-gain VTA driver board.

Plenty of other interesting stickies on that forum as well.

 

[dividebytube]

A few years ago I rebuilt my '70 using the Triode EF86 board. I bought the boards bare and stuffed in my own parts (lots of carbon comp and some poly film coupling caps). I also threw in the SDS power supply board. I was quite pleased with the end results - it certainly didn't sound like the amp I remembered. Deeper bass with more control. Dynamic seemed a little soggy though - the 'images' seemed attached to springs and would recede/increase depending on power demands.

When I rebuilt my Dynaco Mark IIIs I went with the Poseidon boards once available from diyzone.

 

[dividebytube]

I forgot to add one of my more interesting experiments with the '70 -

Adding a lead and a 1/4" plate cap to pin 3 of each output socket, I was able to plug in EL38s. The problem was that the bias required for the EL38 is lower than the EL34, so I could only dial in ~20mA per tube(!).

Being lazy, I went ahead and listened to the amplifier - er, wow! Big improvement in everything. It sounded like an entirely different amplifier with more dynamics, speed, etc. I guessed that the power transformer load was decreased, bumping up the available driver voltage. Or perhaps the OTs just 'run better' with less current. Dunno - I never took it any further and got busy working on some different project.

btw - if you want to run EL38s the 'right way' with more idle current you can change the bias pot from 10k to 25k or add a dropping resistor before the bias pot.