New Marshall project - 1970 50w Lead Restoration

[sluckey]

Also, is it possible that the rectifier is original? I’m finding pictures that show both the monolithic and the diode string. Which is preferable for reliability and longevity?

Looks original to me. No need to change it. That block is actually a full wave bridge rectifier but since the negative terminal is not connected it is being used as a simple two diode conventional full wave rectifier. Should last another 50 years or more. Silicon diodes, whether single diodes or encased in a block of epoxy, do not gradually wear out or go bad. Most silicon diode rectifiers fail due to failure of another component (such as shorted filter cap). Death usually comes in the blink of an eye.

[neskor]

original for sure

[Raoul Duke]

Thanks Steve, I appreciate the info as always! Wasn’t sure if these being an “assembly” of sorts made it a weak link.

Thanks as well neskor! Your Marshall knowledge is very much appreciated!

Also, are the snubber caps to the left of the rectifier needed? Another case of “I sometimes see them, I sometimes don’t”. I’ve read they came from the factory both ways as well and my understanding is that they reduce switch “pop”?

[Roe]

All very helpful info guys, I do appreciate it!

Roe, when you say 1k8-100k-3n3 - I’m interpreting that as 1k8/3n3 as the cathode bypass and 100k as the plate load? You gave me some values for V1 when I built my JTM that really made the difference in its sound, so I just want to get it right. I think I have a few 1W/100k Iskras here somewhere that I’ll likely replace those carbon comps with.

Also, is it possible that the rectifier is original? I’m finding pictures that show both the monolithic and the diode string. Which is preferable for reliability and longevity?

Thank you!

correct but 3n3 is suggested coupling cap (with a big bypass cap)

[Raoul Duke]

Thanks Roe!

So I did some research on the rectifier block and found a couple of tests to perform. Long story short: both negative to AC read OL and both AC to positive read 1.1. I’m thinking now is the time to replace this before I get the new PT. Question is: do I replace with a similar part or build a diode string? The board has holes where turrets would go, but I’m also reading that a 55 year old paxolin board may not be amenable to staking turrets at this point in life.

Searching for modern replacement blocks just to see what’s available, but certainly open to opinions about the best way to get this straight. The excavation continues…

[sluckey]

Long story short: both negative to AC read OL and both AC to positive read 1.1.

Nothing wrong with those readings. But the test is only half completed. Now you need to swap the meter leads and repeat the same test. The readings should all flip. IOW, a previous reading of "OL" will now read "1.1" and a previous reading of "1.1" will now read "OL". This completes the test. These readings show that there is a good front to back ratio for all four diodes in the block. This means the block is good. If you don't get a front to back reading on any of the four diodes, then the block is bad.

I bet your block is good. Let me remind you again... Notice the negative terminal is not connected? This means that only two diodes are being used.

[neskor]

[quote=
Also, are the snubber caps to the left of the rectifier needed? Another case of “I sometimes see them, I sometimes don’t”. I’ve read they came from the factory both ways as well and my understanding is that they reduce switch “pop”?
[/quote]
If you are referring to the 0.22uF caps that appeared (briefly) in some Marshall amps circa 1970, wired between the power supply filter capacitors and standby switch, these are best omitted.
Have a look at where they are placed in the circuit. If one shorts, it crashes at least one side of the PT secondary. Those caps are notorious for doing exactly that. IIRC, Marshall advised authorized service centers to remove them back in the 1970s.

[Raoul Duke]

Steve, I’ll test every terminal against the others and post what I get. I appreciate your coaching as I’ve never worked with one of these before and have no idea about their intricacies. Seen pictures though, lol…

Nestor - thanks! That’s the kind of info that is very helpful. I’ve been reading up on those wherever I can find info and was getting a hunch that they might not be worth replacing so your info has saved me some searching. Appreciated!

Back to scraping electrolyte of the inner chassis - after I test the diode block.

[Raoul Duke]

Black lead on DC+
* red lead on AC1 (upper left corner) = 1.137
* red lead on DC- = OL
* red lead on AC2 = 1.166

Black lead on DC- = OL with red on all other turrets

Black lead on AC2 = OL with red lead on all other turrets

Black lead on AC1 = OL with red lead on all other turrets

Red lead on DC+ = OL with red lead on all other turrets

Red lead on DC- OL with red lead on all other turrets

Red lead on AC2 = 1.164 with black lead on DC+, OL with black lead on all other turrets

Red lead on AC1 = 1.135 with black lead on DC+, OL with black lead on all other turrets

I think that’s every combination. I’m a little confused…

[sluckey]

I think that’s every combination. I’m a little confused…

I bet you are! Rather than trying to measure everything all at once, break it down. There are four diodes in that block. Measure one diode to get a front to back ratio. Then move to another diode and repeat. Do this for all four diodes. The attached pic should make it crystal clear.

[Raoul Duke]

Ahh, now I see. The drawing does make it crystal clear. Thanks Steve!

So using the drawing as reference, I get forward voltage with the common lead on DC+, red lead on either AC with OL if I put red on DC+ and common to the ACs. However, I get no forward voltage on anything else/any combination. This leads me to believe that I only have one side working - but that’s all I need because this circuit uses full wave with no bridge, correct?

Is the higher forward voltage (1.2 avg) a sign that it’s starting to degrade? Also, the DC+ terminal looks crooked - but it might be bent vs internally melted at some point as the bottom turret flange seems flat all the way around.

Don’t mean to be a PITA, just trying to asses risk in an informed manner. I’d hate for it to fail shorted and take the new PT with it.

Anyhoo, thanks again Steve - I always appreciate the lessons!

[sluckey]

So using the drawing as reference, I get forward voltage with the common lead on DC+, red lead on either AC with OL if I put red on DC+ and common to the ACs.

That's good.

However, I get no forward voltage on anything else/any combination. This leads me to believe that I only have one side working - but that’s all I need because this circuit uses full wave with no bridge, correct?

That seems to be correct. However, your circuit does not use either of the diodes that connect to the negative terminal. Just to be sure, connect the red probe to the negative terminal and see if you have a low voltage reading when you touch the black probe to either AC terminal.

Just a thought... there may not be any diodes connected to the negative terminal(meaning this is not a four diode FWB) and that would give the readings you are getting. No way to prove that thought without destroying the block.

Is the higher forward voltage (1.2 avg) a sign that it’s starting to degrade?

No. Silicon diodes do not degrade. They are good right up to the instant that they catastrophically fail then they are dead. There is no in between. It's good or it's bad. But the 1.2v forward voltage reading makes me think that each diode in my drawing is actually two diodes connected in series and acting as a single diode. A single diode should have a forward voltage drop of about 0.6V. Two diodes in series would have a voltage drop of 1.2V.

You can buy a 1000V/3A FWB from Hoffman for only $1.17. Scroll down near the bottom of the page. Should be able to find a similar FWB at most other retailers too.
https://hoffmanamps.com/MyStore/perlsho ... =!ORDERID!

Or build your own conventional rectifier using four 1N4007 diodes (two in series on each AC leg).

[Raoul Duke]

Interesting stuff Steve. Definitely a learning experience for me.

Red probe to the negative terminal gets nothing. I took it out to clean the burst cap goo around it and it’s an International Rectifier GB256/PA dated 26 week of 70.

Torn, but I’ll check out the Hoffman page just to read up on modern alternatives.

Thanks again!

[neskor]

The problem with the diode block as used by Marshall is it only uses two of the diodes in the block (two are unused as they are for fullwave rectification).
My suggestion is do not use the block but four of the heavy duty ones ((you need at least a 1200v rating and 2A per each diode). The board does not need drilling, just turrets putting in the holes.

[Raoul Duke]

Thanks neskor!

I have the usual 1N4007 on hand (which doesn’t meet the spec mentioned) as well as some 1N5408. If neither are ideal - please suggest another. As I’ve been researching, it seems many in the same situation I now find myself in build the diode string. There’s certainly plenty of room there as well as pre-drilled holes. I’ve also found some threaded turrets at Mouser so that if done right, I won’t need to stake/hammer the board.

Thanks again!