Getting ready to restore my pair of MK III's - questions

Hi all, I restored a pair of ST-70s in 2010, and still have one of them.

Now I am getting ready to restore a pair of MK-IIIs (then finally will do PAS-3). This pair of MK-IIIs is brothers from different mothers - one is much earlier issue than the other - one uses 6550s, the other KT-88s. They are all original, but they look totally different - tubes, capacitors, etc.

Here is a list (and questions) for restoration:

1). Multi-section capacitors.
a). Question 1 - Is there a big difference between German Authenticap and American CE Cap? Authenticap is significantly more expensive for the same values.
b). Question 2 - values. Do I go with original 30/20/20/20? Or upgraded CE Cap 40/20/20/20? Or some people go like ST-70 with 80/40/30/20?
c). Question 3 - I also see Dynakitparts selling selling 30/20/20/20, but they don't mention who the manufacturer is. Same price is Authenticap.

2). 1N 4007 Diode to replace Selenium rectifier.
a). Question - do I need to change the resistor value between the diode and the pot? Or it will get compensated by adjusting the pot resistance?

3). 11-Ohm bias resistors - do these need to be replaced?

4). 2 x electrolytic capacitors 50uF - replace with modern 50uf @ 250V ?

5). Replace power switches (one is stuck in On position), power cords, input and output connectors.

Thank you.

1) I prefer to use modern 450V electrolytics mounted below the chassis--Panasonic have been very reliable. I use them in series pairs with large value balancing resistors. I leave the "quad caps" in place to preserve the vintage look. I use close-to-stock values as a favor to my 5AR4 rectifiers. Those that use solid state B+ rectification will probably have a different opinion.

2) You will almost certainly need to adjust the voltage in your bias divider to recenter your bias pots after replacing the Selenium stack with 1N4007.

3) Maybe. Have they drifted?

4) 50-75µF, 100V caps work well and are fairly compact--though there is plenty of room in the MkIII for physically larger caps.

5) Three-conductor IEC cords and Mk IIIs don't always work well together. This is a safety versus hum issue that you can work through.

Thank you for your response, beede. Both my amps have completely different stacked caps. I'd like to get a matching set, so they will also look the same . That's why I'm looking at 3 available options.

I haven't measured bias resistors, to be honest. If I don't have to replace them, I definitely won't...

So for 3-conductor cords, you're saying I may get a humming amp? Damn. I can order polarized 2-prong cords, I guess?

I use 3-wire IEC cords on my Mk IIIs but I float the ground wire to avoid ground loops. My system is fussy with grounding and I try to have a single ground reference to manage hum. Others here report success while using grounded 3-wire IEC cords.

I would look at Bob Latino's tubes4hifi update boards for restoration. You will not be disappointed in any way.

There are too many advantages to mention other than your music will floor you.

ViperZ wrote:Hi all, I restored a pair of ST-70s in 2010, and still have one of them.

Now I am getting ready to restore a pair of MK-IIIs (then finally will do PAS-3). This pair of MK-IIIs is brothers from different mothers - one is much earlier issue than the other - one uses 6550s, the other KT-88s. They are all original, but they look totally different - tubes, capacitors, etc.

Here is a list (and questions) for restoration:

1). Multi-section capacitors.
a). Question 1 - Is there a big difference between German Authenticap and American CE Cap? Authenticap is significantly more expensive for the same values.
b). Question 2 - values. Do I go with original 30/20/20/20? Or upgraded CE Cap 40/20/20/20? Or some people go like ST-70 with 80/40/30/20?
c). Question 3 - I also see Dynakitparts selling selling 30/20/20/20, but they don't mention who the manufacturer is. Same price is Authenticap.

2). 1N 4007 Diode to replace Selenium rectifier.
a). Question - do I need to change the resistor value between the diode and the pot? Or it will get compensated by adjusting the pot resistance?

3). 11-Ohm bias resistors - do these need to be replaced?

4). 2 x electrolytic capacitors 50uF - replace with modern 50uf @ 250V ?

5). Replace power switches (one is stuck in On position), power cords, input and output connectors.

Thank you.

Appreciate the input, but at this time I want to concentrate on the most required changes to keep them running, not at the upgrades (boards, tubes, etc.)

Since your amps are dissimilar the updated board would make them virtually identical and therefore no mismatch.
Just a thought.

ViperZ wrote:Appreciate the input, but at this time I want to concentrate on the most required changes to keep them running, not at the upgrades (boards, tubes, etc.)

tubenutr wrote:Since your amps are dissimilar the updated board would make them virtually identical and therefore no mismatch.
Just a thought.

ViperZ wrote:Appreciate the input, but at this time I want to concentrate on the most required changes to keep them running, not at the upgrades (boards, tubes, etc.)

The boards themselves are actually the only things, which are the same (since layouts never changed). I will update the caps. Like I said before, I am focusing on the specific updates that I listed in the original post, and that's what I need help with. Thank you.

Please note the interpolations.  

ViperZ wrote:Hi all, I restored a pair of ST-70s in 2010, and still have one of them.

Now I am getting ready to restore a pair of MK-IIIs (then finally will do PAS-3). This pair of MK-IIIs is brothers from different mothers - one is much earlier issue than the other - one uses 6550s, the other KT-88s. They are all original, but they look totally different - tubes, capacitors, etc.

Here is a list (and questions) for restoration:

1). Multi-section capacitors.
a). Question 1 - Is there a big difference between German Authenticap and American CE Cap? Authenticap is significantly more expensive for the same values.
b). Question 2 - values. Do I go with original 30/20/20/20? Or upgraded CE Cap 40/20/20/20? Or some people go like ST-70 with 80/40/30/20?
c). Question 3 - I also see Dynakitparts selling selling 30/20/20/20, but they don't mention who the manufacturer is. Same price is Authenticap.

Be very, very careful here. Make sure you are getting a *new* cap, not NOS, and that it operates at the required voltages. Personally, I am of the "a cap is a cap is a cap" school of thought, and within design parameters, there is no difference between a cap rolled on the thighs of Russian Virgins on Walpurges Night and one made in an ordinary factory. On values - that is a two-edged sword. Over-reaching cap values can lead to high B+, especially in these troubled times of high wall-plate voltage. Running low reduces cap life and may allow some hum. 40/20/20/20 is about right in my book.

2). 1N 4007 Diode to replace Selenium rectifier.
a). Question - do I need to change the resistor value between the diode and the pot? Or it will get compensated by adjusting the pot resistance?

Yes, and yes. 1N4007, and yes change the resistor values. Both the work of about 10 minutes, together to re-center the trim-pot.

3). 11-Ohm bias resistors - do these need to be replaced?

Unless, they have drifted, no. These are precision resistors, and while not hard to replace, some screening is required such that they match and are well-and-truly 11 ohms.

4). 2 x electrolytic capacitors 50uF - replace with modern 50uf @ 250V ?

What was supplied was 50uF @ 75VDC. 250VDC is too many for a number of esoteric and practical reassons. 105C. Caps at 100 - 150 VDC are just fine.  

5). Replace power switches (one is stuck in On position), power cords, input and output connectors.

Yes, on the power switch, making sure that what you put in place is rated at a minimum of 200 watts. Yes, on the jacks. *POLARIZED LINE CORD - TWO-WIRE ONLY! Unless you have a thing about hum loops and their management. Further, you will void (if such things are important) the UL Label if you add an IEC connector.

Thank you.

Make sure you check each component on the boards and replace (each side) as necessary - if one is bad on one unit, replace on both.
Good quality film resistors.
Aside on resistors: In some cases, they act as the poor man's fuse. So try not to raise resistor wattage unless you understand well their function in that particular application. Noting that Dynaco tended to drive such components at the bleeding edge of their ratings.
Good quality poly-caps, 600V.

And, enjoy!

Peter W, thank you very much for your answers.

1). I'm not looking for exotic caps, just wondered what the difference is. All caps I'm looking at are brand new builds.

I had good luck with making my own kits from dynakitparts before, when I rebuilt a pair of ST-70s back in 2010, and they still run great. I think they only offer 30/20/20/20, which is the original capacitance value, with slightly higher voltage tolerance on the caps.

2). dynakitparts does sell the whole "kit" with this diode, they provide replacement resistors with new values to keep the pot centered. I found out after I asked the question here, and then I realized that I used similar kit from them for my ST-70s.

Going straight to #5:
- damn, I already have a pair of nice 3-prong power cords. I guess, I'll just go with polarized ones instead.

Both amplifiers are still running fine on all original components, maybe with slight hum. I will replace caps on the board with polycaps, buying those. I also want to replace that mica cap on the each board, one of them is really really old - I had to read the color dots on it lol.

Is there really a need to replace those resistors? I understand that carbon composition resistors floated quite a bit, but if they measure ok, would it be ok to keep them? It is going into the 'lazy' domain, since there are a lot of resistors, and it's an old board.

Is there really a need to replace those resistors? I understand that carbon composition resistors floated quite a bit, but if they measure ok, would it be ok to keep them? It is going into the 'lazy' domain, since there are a lot of resistors, and it's an old board.

Short answer: If they have not drifted under operational conditions, no.

Much longer answer:

a) What you see on those boards are, for the most part, 20% resistors. 13 of them, making 26 in total. Today, 5% resistors are pennies each.
b) Consider the concept of cumulative drift - this one drifts 5% -That- way, that one goes 5% -This- way. None the full 20%, but neither does one begin to be electrically identical to the other.
c) As you are calling this a 'restoration' vs. a basic refurbish, the perhaps-an-hour-and-$5 to do all 26 seems minimal given the certainty-of-result.

Get yourself some desoldering braid and you will find that pulling the resistors is relatively quick and easy. Do one (one) board at a time, taking pictures, reserving the other as a model.

It will be worth it!

Peter W. wrote:Is there really a need to replace those resistors? I understand that carbon composition resistors floated quite a bit, but if they measure ok, would it be ok to keep them? It is going into the 'lazy' domain, since there are a lot of resistors, and it's an old board.

Short answer: If they  have not drifted under operational conditions, no.

Much longer answer:

a) What you see on those boards are, for the most part, 20% resistors. 13 of them, making 26 in total. Today, 5% resistors are pennies each.
b) Consider the concept of cumulative drift - this one drifts 5% -That- way, that one goes 5% -This- way. None the full 20%, but neither does one begin to be electrically identical to the other.
c) As you are calling this a 'restoration' vs. a basic refurbish, the perhaps-an-hour-and-$5 to do all 26 seems minimal given the certainty-of-result.

Get yourself some desoldering braid and you will find that pulling the resistors is relatively quick and easy. Do one (one) board at a time, taking pictures, reserving the other as a model.

It will be worth it!

It was more about lazy + afraid of lifting the traces. I have all the tools and desoldering braid. I'm not sure how well those boards were made. You're right, I can go on digikey and pick those resistors if the PCB itself is good to work on.

I would definitely use 3 prong cables, and the best is to install an IEC inlet.
Hum is not fought by unsafe installations, and hum can be successfully under control with
properly grounded ( and safe) equipment.

But if ground lift is needed, it's easier to just lift it at one place, at some other living ground is
needed and wanted.

peterh wrote:I would definitely use 3 prong cables, and the best is to install an IEC inlet.
Hum is not fought by unsafe installations, and hum can be successfully under control with
properly grounded ( and safe) equipment.

But if ground lift is needed, it's easier to just lift it at one place, at some other living ground is
needed and wanted.

This is an ongoing source of disagreement for various reasons, esoteric and otherwise. But, effectively, the difference between a US-based restoration and a Euro-based restoration. Both sides are correct within local codes and standards. There is an 'attitude' for lack of a better word that the US National Electric Code (NEC) is primitive by Euro standards. However, it is what it is, and given that the US had, in common use, DC mains power until 1997, such primitive artifacts are not uncommon and to be expected. Let it be, do what is safe, and do what is best based on the design of the equipment.

I'm honestly not worried about 2-prong vs 3-prong, as long as 2-prong cords are polarized. Our neutral and ground are tied in the panel. I guess, I'll stick with 3-wire power cords that I already bought, and see/hope that there will be no ground loops.

ViperZ wrote:I'm honestly not worried about 2-prong vs 3-prong, as long as 2-prong cords are polarized. Our neutral and ground are tied in the panel. I guess, I'll stick with 3-wire power cords that I already bought, and see/hope that there will be no ground loops.

Try this.

With your VOM measure from the Hot to the Ground.
Then, measure from the Hot to the Neutral.

That neutrals and grounds are tied at the panel does not mean that they are at equal potential at any given receptacle. And if they are not at equal potential to all available decimal places, use the two-lead cord.

Peter W. wrote:

ViperZ wrote:I'm honestly not worried about 2-prong vs 3-prong, as long as 2-prong cords are polarized. Our neutral and ground are tied in the panel. I guess, I'll stick with 3-wire power cords that I already bought, and see/hope that there will be no ground loops.

Try this.

With your VOM measure from the Hot to the Ground.
Then, measure from the Hot to the Neutral.

That neutrals and grounds are tied at the panel does not mean that they are at equal potential at any given receptacle. And if they are not at equal potential to all available decimal places, use the two-lead cord.

The main thing I see what needs to be done when converting from 2-wire to 3-wire is rewiring power switch and the fuse holder in series with each other (and connected to the hot wire). Otherwise, if something shorts to ground, now my green safety wire will do its job, but the fuse that's connected to the neutral will do nothing.

I'll go with 3-wire power cords, since I already bought them (last year lol), and hope for the best. Worst case, will have to isolate the 3rd prong.

Here is my list so far (since I already have the power cords). Ignore the binding posts - they don't make them any more, will have to figure something out to replace those pesky screws. Kevin (dynakitparts) suggested these instead: https://www.zoro.com/search?q=Terminal+Strip+Adapter&gclid=EAIaIQobChMIxdzwvdfu1QIVQhppCh2F1A9iEAMYASAAEgIeIPD_BwE&gclsrc=aw.ds




Next thing I will have to figure out is how to find matching tubes. One amp has a pair of SED Winged-C, the other has a pair of Chinese ValveArt KT-88. I'd like to find another pair of Winged-C. Both also have different 6AN8 tubes (I actually have 4 different 6AN8s), and I don't have a tube tester to check if pentode/triode match in those.

As regards to tubes : do yourself a favor and get a matched quad kt88 from a good vendor.
Nothing beats fresh tubes ! You might recover som of the cost by selling the old ones on ebay.
As for the 6an8A ; don't bother. If they don't have hum problems they are fine!
New 5ar4 is optional, if the old ones works - keep em.
Whan start first time : first power up without 5ar4 , measure bias voltage on the grid of one
kt88 , turn the adjustmant to maximum negative voltage then turn off and remount 5ar4.
This way you won't have tubes redplating at first powerup. I think the procedure is described in the manual.

Some tricks-of-the-hobby:

a) On Binding posts: Purchase the best screw terminals you can find.  Remove the screws.
Purchase high-quality banana (or your favorite type) of screw-post jack.
Drill-and-tap the screw terminals to the proper size for the new posts.
Install the new posts and lock them in place with either KEP nuts or thread-locking nuts.
Done. (essentially what is suggested by that terminal-strip adapter picture, only broadening the choices and having the lock-down inside the chassis)

On matching tubes: As others may (and one already has) stated, starting over with a matched quad gets you going. Noting that matching is a temporary condition, but still desirable. I have no idea where you are, but if you are anywhere near 19027, or will be anywhere near Kutztown, PA on May 11/12 - I can match tubes - or test tubes for plate and filament current (as well as GM) - should you wish.  Not a bad idea given that you have four, they could be measured and saved for future use.

On the 6AN8, as a small-signal tube, a good one is fine. There is no need to match by brand or any other parameter, as long as they are both 'good' and not microphonic.

Harping, as I am wont to do, on the 2/3 conductor thing: William of Occam stated: Eschew needless complexity. That was about 700 years ago. Good advice then, better advice now.

I just like a finished look on those binding posts. I see tubes4hifi selling a pair of 3-terminal ones for 25 bucks plus shipping. Will probably go with those. I guess I have to convince myself that I will never use 16-Ohm output windings...

Some great recommendations by knowledgeable people in the preceding posts. They have different criteria for rebuilding than I do. I have a pair of MK IIIs that I just rebuilt for the second time. My criteria was to not change the original look (as much as possible) or deviate the circuit from the original schematic. Your criteria may differ. The original 120 V wire system works just fine - no changes. I use all three 4/8/16 ohm taps - no changes.

I have replaced the selenium rectifier with 1N4007 and all capacitors and resistors. Coupling caps were paper in oil and feedback caps were polystyrene. The resistors were Kiwame 2 watt carbon film. Only the 6AN8 pentode screen grid resistor is a metal film due to its large value. The film type will provide tight tolerances and assure your circuit is as close to the design Ed Laurent originally laid out. I had metal film types in my 1990 rebuild but I like the sound of the carbon films better with this rebuild. I ordered a few extras of each value so I could test and match between channels. They were such tight tolerances when received that it was not necessary. All are within 1% of schematic values and within 1% channel to channel. Half were exact matches. You can see my rebuild in the Photo section of this forum.

What will look different are the components on the board and the selenium rectifier if you look under it. The amps sound terrific on my efficient Altec 615A speakers.

What ever criteria use is up to you. Either mine or a mixture of the recommendations in earlier posts it should turn out fine. Enjoy them!!

I started today on my first amplifier. Here is my progress so far.

Whoever painted the chassis, did a shoddy job on painting and soldering. They cut off the wires, then resoldered them back to the leftover "butts". So I spent quite a bit of time repairing some really bad ones (and found one grid wire not even soldered, hanging by the threads).

* Changed multisection capacitor
* Changed bias resistor (old one had one end too short from previous owner)
* Installed new bias circuit (diode, resistors and electrolytic capacitors)

Plan for tomorrow:

* replace 4 capacitors on the PCB
* replace input RCA connector
* replace power switch (completely failed)
* replace power cord with 3-wire power cord

* need to replace binding posts, but they are still in the mail from tubes4hifi


2 questions:

1). I have a pair of wires going from power switch to pins 6 and 7 of the preamp socket. These wires are not even in my build manual. I guess, I'll remove them when I'll be replacing the power switch.

2). It looks like to install the 3-wire power cable, I would need to rewire the power switch and fuse configuration. Right now switch is on one wire, fuse is on the other wire. I will put the power switch and the fuse in series, and connect them to the phase wire.
This way, if some high voltage shorts to (chassis) ground, I won't fry the whole amp by not having the fuse on my hot feed (if I left the fuse connected to the Neutral wire).

WLT, you make an interesting point. I have a completely rebuilt ST-70 (10+ years ago), where I used all-metal-film resistors on the PCB. I never actually got to compare them to the carbon composition resistors... I also bought Russian PIO capacitors for that amp, and just found them in the box . I will install them hopefully this year, but I need to finish these amps first, then rebuild the power supply in PAS-3 (It still has the original multicap, and finally started to hum this year).

Well, I got stuck a little for now. I finished everything, including rewiring every tube socket, since previous owner did some shady soldering job (found at least 3 wires not attached to the sockets).

One amp has cloth-covered wires on transformers, the other has much newer transformers with modern insulation on the wires. Either way, it's all done now, except for binding posts (in the mail), and the power cords.

The original 3x14AWG power cord that i bought barely fits into the available opening, and I have no fitting for it. I checked the local hardware store, they have 3x18AWG round power cord, but I still don't have a fitting for it.

So I need to figure out where to get the right fitting, and that should be it (hopefully).

It's called a "Strain-Relief", costs a few pennies, and should be available at any electrical supply house and most big-box stores.

https://www.jameco.com/z/0705SR-6R1-Strain-Relief-Bush-Round-SJ-SJT-3C-16AWG-Nylon-66-UL94V-2-Black_182369.html?CID=GOOG&gclid=CjwKCAjww6XXBRByEiwAM-ZUINtf_MSuPLfRmyE3PwV_HA3m6isOPtIz7pS2osNbiC_jB8FKvcQ4DhoCDB4QAvD_BwE If you must order on-line.

At the same time, I very strongly suggest now that you have the option to go back to a 2-line cord, polarized to the switch.