ST-70 with 4 5965 driver tubes

Hi,
    first post here. I just brought an ST-70 here in Hong Kong and found it has 4 driver tubes which i have never seen before. I have added photos. Any thoughts ?

It has 2 x EL34 tubes Gold Aero Made in Great Britain and 2 x Philips 6CA7 XF2




Cheers

Peter

Now this has to one of the earliest examples of a ST70 mod that I have seen. It even retains the selenium rectifier. I have look thru all of the Audio Amateur, Elekor, Vacuum Tube Valley and Audio Express articles on ST70 mods and came up with nothing like this. Of course these were mostly from USA and Europe not in the Hong Kong region. It would be so interesting if someone could find the mod article for this. It would an interesting read. It's always possible this is from a very talented individual hobbyist. I do hope that more lite is shed on this.

Although it would be quite a job, you could trace out the circuit and draw a schematic and from that the circuit could be analyzed.

jwb474 wrote:Now this has to one of the earliest examples of a ST70 mod that I have seen. It even retains the selenium rectifier. I have look thru all of the Audio Amateur, Elekor, Vacuum Tube Valley and Audio Express articles on ST70 mods and came up with nothing like this. Of course these were mostly from USA and Europe not in the Hong Kong region. It would be so interesting if someone could find the mod article for this. It would an interesting read. It's always possible this is from a very talented individual hobbyist. I do hope that more lite is shed on this.

Although it would be quite a job, you could trace out the circuit and draw a schematic and from that the circuit could be analyzed.

That's gotten even more interesting now, thanks so much for looking into this. The dilemma now is whether to keep it as is or modify, especially the rectification.

I took all the voltages they are as follows:

EL34 LEFT CHANNEL
1 = 1.06 DC
2 = 1.56 DC
3 = 373.6 DC
4 = 374.6 DC
5 = -34.6 DC
6 = -34.6 DC
7 = 1.06 DC
8 = 1.70 DC



GZ34
2 = 386.8 DC
4 = 340.9 AC
6 = 342.2 AC
8 = 386.7 DC




EL34 RIGHT CHANNEL
1 = 0.90 DC
2 = 0.70 DC
3 = 391.1 DC
4 = 374.4 DC
5 = -27.4 DC
6 = -27.4 DC
7 = 0.80 DC
8 = 0.90 DC


These seem on the low side but i do not know the state of the tubes. The right channel i can bias down to 1.56v but the lowest the left channel will go is 2.2v

DC offset is around 50mv left channel around 100mv right channel.

Any ideas on how i should get this amp up to specs? i would like to leave it as original as possible given its unusual configuration but don't want it burning my house down.

cheers

P.

Yeh, there is something wrong with the bias circuit. Looking closer at the picture it looks like the selenium rectifier may have been replaced with a silicon diode.
What I would do is completely rebuild the bias circuit.
If you don't want to scrounge around for the parts you can get the rebuild kit from Dynakitparts.com:
https://www.dynakitparts.com/shop/st-70-bias-circuit-kit/
Although Dynakitparts sells 10K bias pots, they are kinda expensive. Antique Radio supply sell plenty of versions. I have used this one before:
https://www.tubesandmore.com/products/potentiometer-cts-audio-splined-nylon-shaft

Next would to look into replacing the multi-section can capacitors.

The easiest way to get a working reliable amp is to rebuild according to the manual. Boards are available, probably you go for a ECF80 board as these tubes are readily available, can cap and
all oter stugg is available from several sources including dynakitoarts.com

Gotta respect your courage for even turning it on.

Thanks for sharing your project. Mine is scattered across my bench right now.

The 5965 tubes are very nice linear dual triodes similar to a 12AV7.  I prefer them to other similar tubes in many ways.  They may tend to be a bit microphonic in some applications.  An O-Ring can help prevent that if needed.  

The first thing I do when working on an ST-70 is clean it up.  I find it easier to do my best with a clean amplifier.  Your milage may vary.

I'm assuming that you have some experience with vacuum tube amplifiers, so I will share what I would do for this one.  Obviously, the project is yours and so the decisions have to be yours as well.

First thing is get a manual if you don't already have one.  There are plenty of them free around the internet.  They have the original information and can be an enormous help.  I'm also grateful for the under chassis photos and diagrams I've found.  They are priceless.  

The fist thing I would do is rebuild the bias supply with a solid state diode.  The Selenium diodes can be problematic and need to go.  I would also change the capacitors in the bias circuit as a matter of course.  They could be tired and it doesn't cost much to replace them.  Check the resistors to make sure they are OK.  I'd clean the two 10K pots with DeOxit and then, once the circuit is back together, you can check the bias voltage to make sure it is smoothly adjustable and in the appropriate range.  Check it by removing the 5AR4 (that stops the B+) and watch the bias voltage with your multimeter as you rotate the pots.  Given that it adjusts OK, set it a bit higher than needed until it is time to adjust the bias for listening.  

The current power supply, while functional, is not something I would keep.  I usually replace the can cap as a matter of course.  In this case, the can caps can be removed, discarded and a new cap fitted.  There are after market cap boards designed to fit underneath the chassis, but I'd go for stock to begin with.  The manual will be helpful here.  Once the can cap is new, you may want to check on the choke.  The Dyna stock chokes were a weak spot and tended to leak goo.  If yours is fine, it is OK to clean it up and leave it.  If need be, there  is a Hammond 200 ma choke that fits the same space and bolt patten.  BTW, I'd also think about using a 5R4GA or 5R4GB in the place of the 5AR4.  The line voltage today is higher than it was when this amplifier was designed and using a 5R4GA will drop the B+ and help protect the amplifier.  Besides, the 5R4GA is a nice tall bottle that matches the EL34s.  

Now that the power supply is restored and safe, you can look at the driver section.  The first thing I would do is remove any paralleled coupling caps.  I'd keep the 0.22u Music Caps and the 0.39u Westcaps.  That makes the circuit easier to see and will probably improve the sound.  You can have too much of a good thing.  (Fine tuning can take place later.)  

Based on the tube layout and general appearance, I'd guess that the driver is similar to a Williamson design.  That is a good solid design and will probably serve you well.  I'd check the resistors and compare the layout to a Williamson PP driver.

Once the amplifier is tested and the bias is adjusted, giving it a listen is the next step.  Afterall, that's the point of all this.  :-)  I hope it provides years of pleasure. BTW, What speakers are you going to use?

I'm going to start off by saying ... I hope you are familiar with working on High Voltage tube equipment. If not, have a friend who is knowledgeable help you and always verify that the amp has no remaining high voltage on the power supply caps.

Ya can't fix "Dead"

I'll generally agree with what TubeDoc said. Definitely replace the Bias Supply caps. Then, first of all, I would also replace the two RubyCon 0.33 caps. First of all, they are electrolytics, which isn't good in the front end and secondly, they look old and I've had experience with old RubyCon caps... Not good. I would use Solens, cheap, reliable and metalized film rather than electrolytic... Once the amp is fully restored, you can always replace then with something better (if so desired).

I believe the Bias supply has already had the selenium rectifier replaced, just not removed... looks like the green wire is attached to the Transformer Bias tap lead and that goes to a diode and then to the bias caps, so I think you are good there once those caps are replaced. I personally change the bias level pots with 10 turn pots of a different value then adjust the other bias circuit resistors so I can use other tubes (KT90s, 6550, etc), but that is for later. Also verify the two 10K resistors in the bias circuit actually measure 10K, if not replace, preferably with metal film types.

Something in the HV Power supply is leaking badly, looking at the choke, I can see the waxy stuff that has flowed down from it overheating, meaning that too much current has been flowing thru that choke which was borderline (capacity wise) to begin with. Replacing the RubyCons MAY resolve that problem, after they are replaced, time to test.

At that point, I would fire up the amp again and check the voltages, if the RubyCon caps were bad, they are probably shorted and would drag down the voltages so check after adjusting the bias to the 1.56 vdc. If voltages still aren't "normal" i.e. about 425 vdc on the first cap section and  415 vdc on the second, either a new can cap ( or two ) or a replacement board using multiple caps to increase the overall capacitance are solutions.

Looking at the picture again, it appears that the original can cap on top of the chassis has several sections disconnected, you will want to verify that as well. I can't read the values on the internal can cap, but if you can post the voltage / microfarad values of the that, I can make some recommendations as well. Also, check the value of the resistors in section 3 and 4 of the high voltage supply, verify the measured value roughly matches the value printed on the resistor(+/- 10%). With the new front end section, they may be different than original values. Later, replace them with metal films which are quieter and don't drift.

And measure the two bias current resistors (white precision 15.6 ohm coming from the joined pin 8 of the output tubes to ground), assuming they are original. If they are off, that will distort the measured bias current. Verify the measured values match absolutely... If they are not the original bias resistors, please post the values on the resistor body so someone can calculate what the measured voltage should be for the roughly 50 Ma/tube Also, for longer tube life, once everything else is working correctly, I would buy the individual bias adjustment kit from DynaKit Parts, and I would bias the tubes at more like 1.40 volts instead of the 1.56 that Dynaco used, (42-44 ma) it makes it easier on the power supply and makes longer tube life as well (and sounds good).

I've rebuilt / repaired a bunch of Stereo 70s, Mark IIIs and Mark IVs. There are many, many front end options so I wouldn't change what you currently have until the amp is working properly, the test... the first two sections of the power supply should roughly match the voltages that the original manual says with the bias voltage properly set.

Good luck, please post the values on the internal can cap and if you need additional input or help, PM me.

-Ed

I hope getting similar advice from Ed and I didn't confuse things too much.  

Places we agree:  

Rebuild the bias supply using a new silicon diode including checking the resistors and cleaning the pots w DeOxit.  Pull the rectifier to shut down the B+ and check to see the extremes of voltage and set the voltage a bit higher than needed until it is time to make the final adjustments.

Pull the Rubycon Oil caps in parallel with the Westcap (also oil) coupling caps.  Although they are not electolytics, they are clearly in excess of what is needed.  I have a motto--simplify and see. The same goes for the two Solen caps in parallel with smaller Musiccaps.  

We agree in principle  that the B+ power supply is not what it should be.  I would not bother to check much of it.  I'd just toss the two electrolytics, maybe the choke (depending on what you find after cleaning and testing), and return the power supply to stock configuration.  

Once you have it back into a well known and well proven state, you can better decisions about how much more you want to do.  

As Ed suggests, there area many things you might want to do.  One of my ST-70s has individual bias controls for each tube and that is nice to have since I can use EL34s that are not as well matched.  That amplifier also has increased capacitance in the power supply with a driver board from https://www.tubes4hifi.com/ST70.htm.  I can heartily recommend that.  Another of my ST-70s (that you can find on this site) has a Long Tailed Pair 6SN7 driver and a solid state power supply.  The bias supply is essentially stock.  Another ST-70 is scattered all over my bench waiting to go together after a chassis up rebuild.  

If the driver section on your current amplifier is what I think it may be, it will probably sound nice and you may not want to do much with it besides checking the resistors and checking the assembly for loose bits.  

So many Dynakits, so little time!  :-)

I'm going to reply to both TubeDoc's and Bob Latino's followup comments regarding this amp. Sorry for the long reply..... And I agree with TubeDoc, sorry if this is possibly adding confusion to the issue but there are so many options........

First of all, I blew up the image to get a better look... it helped, especially showing that 2, possibly 3 sections of the original power supply can cap were no longer in use AND that the resistor connecting the new can cap to the remaining section of the old cap appears to be burned...  Plus other stuff.

Regarding Bob's comments, I would get the amp working - mostly power supply issues, before replacing the driver section with the VTA board. I would want to see how the original front end sounded before making a big driver board change that may or may not improve the sound. Just my two cents... But Bob is correct that the older carbon comp resistors drift in value over time. If I was genuinely happy with the original overall sound, I would just replace all of them with metal film resistors. Note, the resistors I can see appear to be 1/2 watt sized which is pretty conservative in a driver stage so hopefully they didn't drift too far. And I will address the Driver Board capacitors below. Also, Bob's comment about the Triad C24-X choke is valid; although the resistance is correct, the 1.0 Hy value has less filtering than the original 1.5 Hy value of the Dynaco choke but the higher milliamp capacity and assuming the increased capacitance of the replacement power supply should negate that. And Bob is correct about the transformer, it is being pushed driving the four front end tubes. Once the amp is working to your satisfaction with the 5AR4 tube rectifier, I would replace the 5AR4 with a solid state supply and a delay relay which would free up almost 2 amps of 5 volt load on the power transformer. I would also put a CL80 NTC inline with the power cord to both slow the inrush when the unit is turned on and drop the AC voltage a bit as the ST-70 was designed to run at 117 VAC, not the 122 + volts commonly produced today. That higher AC voltage means the 425 VDC initial voltage in the ST-70 is now 443 VDC, almost 20 volts higher than it was originally designed for.

Regarding TubeDoc's follow up comments, assuming the power supply is essentially in a stock Dynaco configuration, I would agree with that, just pull the two can caps and replace with either a replacement Dynaco ST70 power supply board as TubeDoc suggests or replace the original can with an 80/40/30/20 mfd cap from Dynakit Parts( in 30/80/40/20 configuration ), get the $80 version, it has a higher voltage rating and is just a better quality than the CE version with the same capacity. With the can, replace the associated resistors with metal film types as well.

Also the coupling caps are really a mess - one set has a 0.22 mfd MusiCap, a 0 22 mfd Solen and a third smaller unidentified MusicCap all apparently paralleled. Replace that with a single 0.47 mfd Solen for the time being, likewise replace the joined 0.33 mfd Rubycon and 0.39 mfd WestCap with a 0.68 mdf Solen, total cost for all four about $10 from Michael Percy. Those changes will replace the (4)  two or three cap units with four single cap units and will clean up the interior of the amp considerably as well as possibly replacing a leaky cap(s). Be sure to insulate the leads.  Then, take a listen. If you like the sound, you can keep the Solens or replace them with something better at a later time. Or you can experiment first with smaller cap values - replace the 0.47 mfd with a 0.33 mfd or a 0.22 mfd and check out the sound, especially the bass response, likewise the 0.68 could easily be a 0.47 mfd or 0.33 mfd. Make sure they are all 600 volt rated.

A couple of comments about the bias, first, the bias supply has already been converted to a silicon diode based unit. If you follow the green wire from the transformer you will see a reversed diode by the first bias cap. Second, it appears the the Bias measuring resistors have changed which I missed initially. This would in turn change the measured bias voltage. I see a Red / Yellow/ Orange/Brown resistor on the front panel octal sockets which may be bias measurement resistors, as well as a green resistor (which I can't read the value of) in the traditional octal socket mounting spot. Not sure which is the correct resistor OR what value is being used to measure the original 100 ma plate current. You will need to verify those to get a true plate current reading. Personally, I would reduce the plate current to about 85 ma instead of the stock 100 ma per pair of tubes on each side.

As TubeDoc says: Once you have it back into a well known and well proven state, you can make better decisions about how much more you want to do.  

And...  Experiment, it is fun.

And enjoy that 70 year old Stereo 70, I love mine, I am currently rebuilding a second pair of Dynaco Mk IVs, the mono block version of the ST-70 using a custom driver stage featuring  12AY7 and ECC99 tubes, KT-90 output tubes and film caps and four chokes in a solid state power and bias supplies.

-Ed

Thanks so much for all your very informative replies and taking so much time with details much appreciated.

I have rebuilt a PAS3 and built tube DACs but first time rebuilding a tube power amp. I’ve done a few Nelson Pass DIY amps in the past so I’m not too green.

Speaker wise I run a pair of Calix horns in one set up and LS35a’s in the other. Horns have Naim 135 monos and the LS35a’s with a PASS F5 diy.

Adding photos of the can capacitor

As this is a bit of a mish mash of parts I’m probably going to go down the tried and tested modification route. The chassis is definitely in need of some TLC and binding posts and RCAs brought into the 21st century. I think it’s much easier and safer to go the VTA route and I’m lucky in Hong Kong to have 99% of parts available in bricks and mortar form, diy HiFi shopping here is awesome..


Best

Peter

Hi Peter,

Sorry if I am repeating anything previously posted here... and sorry this is so long once again.

Looking at the pictures, the can cap on the top of the ST-70 is rated at 475 volts, very borderline given the possible 443 volts while running and likely higher when first powered on (while the tubes heat up). That is obvious by looking at the lugs underneath the chassis, with either 2 or 3 of them disconnected which I assume means those sections went bad. If the can cap underneath the chassis is rated at the same voltages, I would assume that they are already leaking or will be soon. This is confirmed by the waxy buildup on the top of the choke which is upside down in normal usage, this caused by overheating, meaning too many milliamps flowing through the choke (from capacitor leakage). Note, a friend of mine who is a Georgia Tech Electrical Engineer told me that an electrolytic cap should only be used at 80-85% of the rated voltage if you want long life.... especially in a tube amp where the parts run at a higher temperature than normal. And that can cap is rated 20/20/20/20 mfd, the original was 30/20/20/20 mfd and most people say that is too small.

I would first replace the bias caps and then both the power supply caps with either a 40/80/30/20 mfd can cap in the original position OR one of the replacement power supply boards as referenced by TubeDoc. I personally get my PS supply boards from Triode Electronics in Chicago, USA but others make good ones as well. Plus add a CL-80 NTC in series on the power cord as well to slow down the initial current inrush and drop the VAC voltage a bit, especially if you replace the 5AR4 with a solid state power supply.

Then adjust the bias for the new PS supply voltage and listen for a few days. If you like the sound of the existing front end, keep it, if not replace it with the VTA or other Dynaco ST-70 front end board. As mentioned earlier, you can also experiment, replace the coupling caps with the values I mentioned in my previous post. Note: please verify the value of the bias resistors, they have been replaced and I seriously doubt they are the original 15.6 ohm value. You can look at the drawing at the end of a downloaded copy of the ST-70 manual and see where they are attached to an octal tube socket in the drawing. Please post the resistor value and then someone can tell you the voltage you should read where Dynaco labeled 1.56 volts.

Finally, if you want really long life, replace the 5AR4 rectifier tube with a solid state rectifier. Note that most of the replacement power supply boards include this as an option. Doing this reduces the load on the power transformer by 2 amps of 5 volt load. And unfortunately, your ST-70 appears to have one of the more marginal power transformers, the height of the PS stack is less than on some of the models, it varied from year to year....

Experiment, listen and have fun,

Love the sound of my ST-70.

-Ed