M-125 new(er) build - copper cap rectifier

Hi!

Built two of Bob's M-125s about 6 weeks ago, have been working absolutely flawlessly from the first turn on. I originally ran them with 5AR4 rectifiers that Bob supplied, but recently tried out the "Copper Cap" rectifier kits that Bob recommended at the ebay link here: https://www.ebay.com/itm/254912521927. These resulted in slightly tighter bass with the supplied Sovtek 6550 output tubes.

Well, this morning I flipped both amps on and listened for about an hour, and then realized one channel was out. Troubleshooting further, it seemed that the slo-blo fuse on one amp had blown. Assuming a "what changed" mentality, I next took the rectifier from that amp and threw it into my Hickok 533A tube tester (which I had originally done when I built it.) Well, now the rectifier registers "zero" on the first "plate" setting. The other, working rectifier still registers good on both "plates."

Questions:

1) I assume that this is the reason my fuse blew?
2) How can I troubleshoot the rectifier further? There is still continuity between all of the pins and the components.
3) Is there a chance that one of the components (diodes, surgistor) in the rectifier failed? If so, how to isolate? And that seems...very odd that it would fail like this?

My amps are plugged into a variac providing 119V.

Thank you!

Your experience is not odd, these amps push rectifiers hard. Typically when a rectifier flashes so goes the main fuse. Helps to be rectifier savvy. These days I only run GZ37 or GZ33 in my M125's due to other options failing in unacceptable time frame (weeks), notably 5ar4. The GZ versions just sail on day in day out. Managing inlet voltage helps, but is not a panacea IME. Adds a few extra weeks maybe. Weber SS with mod to add ventilation holes I've read is reasonably stout as well, but like I said the GZ tubes are what I've settled on and serving me well.

I have a pair of m125s that I built 6+ years ago, and the only rectifier tubes that have lasted in them have been mullard gz33s and good nos 5u4gbs such as rca. New production 5ar4s usually expire in a month or two in the m125s, so I quit using them. While they work, I prefer not to use 5u4 variants because I've heard it's bad to use directly heated rectifier tubes with indirectly heated output tubes.

The mullard gz33s have been remarkable in these monoblocks though, they last for years and I've never actually had one fail. I did finally replace a set because they looked to be getting pretty used up judging by all the filament boil off at the top of the tubes. The only downside is price, a good set will cost you a $200+

Corona wrote:I have a pair of m125s that I built 6+ years ago, and the only rectifier tubes that have lasted in them have been mullard gz33s and good nos 5u4gbs such as rca.  New production 5ar4s usually expire in a month or two in the m125s, so I quit using them.  While they work, I prefer not to use 5u4 variants because I've heard it's bad to use directly heated rectifier tubes with indirectly heated output tubes.

The mullard gz33s have been remarkable in these monoblocks though, they last for years and I've never actually had one fail.  I did finally replace a set because they looked to be getting pretty used up judging by all the filament boil off at the top of the tubes.  The only downside is price, a good set will cost you a $200+

Corona.... That RCA 5u4gbs is a really beefy rectifier tube to say the least and the Mullard GZ33 is not to shabby ether.

Well, now I'm not sure what's going on but this seems worse.

I went back to basics with this amp and put in just the two tubes on the driver board, and checked operation. All fine. Then turned bias all the way down, and plugged in previously-working 5AR4, and two of four tubes. Let it come up for a minute, and checked bias. 0V! Shut everything down, and tested the 5AR4 tube. Guess what? The 5AR4 that was working just fine when I "retired" it a week ago now shows no value in the Hickok on one of the two sides!

Is it possible that something has gone wrong in my amp that is taking out rectifiers?

Now that seems strange. It sure would seem there is something else going on. I wish I had some ideas for what to look for but 2 rectifiers that fast doesn't sound right. Hope you get it sorted out soon.

I have yet to come across anyone reporting either a GZ33 or 37 failing, matching my experience. They are expensive up front, but on a per year basis start to look a lot better.

Do the "yellow sheet mod" and get a pair of GZ33s from Upscale and be done with it. I doubt Jim carries them.

At this point before I do anything more, I've contacted Bob and we're trying a few more troubleshooting steps. I'll report back here with the eventual outcome! In the meantime - back into the system goes the vintage Heathkit W5Ms (which sound nice, but not near as nice as the M-125s!)

They do sound sweet don't they.

cci1492 wrote:Do the "yellow sheet mod" and get a pair of GZ33s from Upscale and be done with it. I doubt Jim carries them.

The "yellow sheet mod"....... Exactly!!!

Ditto on the yellow sheet mod. Also on recommendation from Bob due to my crazy voltage fluctuations in this house, bought a $60 variac rated tat 5 amps and I set it daily at 117VAC to help out. I can go from 115 to 127 in this development.

I've never even needed the yellow sheet mod or a variac with gz33s, the things just don't die. I have a pair that lasted 5-6 years and the tops are all dark from filament burn off, yet they still work.

JXBJXB wrote:At this point before I do anything more, I've contacted Bob and we're trying a few more troubleshooting steps. I'll report back here with the eventual outcome! In the meantime - back into the system goes the vintage Heathkit W5Ms (which sound nice, but not near as nice as the M-125s!)

I'm running 5AR4 so decided to do the mod. Nice on the Heathkit gear. If you notice in my avatar, I restored my dad's W5M and WA P2. That's pushing a mahogany Jensen Imperial reproducer -- all 1956 vintage. Bought a vintage Eico tuner and a Garrard mono table to complete the mono set-up. Speaker still rocks crazy sound. Not needing a sub with my Fritzies, I often contemplated running the W5M and Jensen as my sub just for kicks.

cci1492 wrote:Do the "yellow sheet mod" and get a pair of GZ33s from Upscale and be done with it. I doubt Jim carries them.

IIRC, the "yellow sheet mod" is only to protect the rectifier if the amp is powered down and quickly powered back on (short cycled).
When I did the mod on my ST120, I read the entire forum thread and looked elsewhere before doing it so I knew WHY I was doing it.

It can be helpful but I don't think it's the answer in this situation. Please correct me if I'm wrong.

My 125s ate copper caps about every 9 months. Switched to GZ37 and no issues for 4 years running. Got mine from British eBay seller Langrex. Fair price and good service, tons of NOS “valves”


https://www.ebay.com/itm/GZ37-CV378-53KU-MULLARD-NOS-VALVE-TUBE-/313009944885?hash=item48e0d8b935

Just a little update as to what we think is going on. Somehow, one of the copper cap rectifiers caused an over-voltage situation that blew out at least one of the segments (40 MFD) of the quad cap. This was the immediate response from one of Bob's techs when I described the symptoms - I removed the quad-cap and indeed that appears to be the case.

I am awaiting a replacement quad cap, will swap that in, and do the diode mod, and see where I get (with a tube rectifier this time!) and then update!

-jb

Id bet that the "copper cap" kit sold on eBay has diodes spec'd for the 120 or 70 vs. the 125, or possibly the thermistor? I have been using the Weber WZ68 on my 125 and the WS1 on 120 with good success (have not done yellow sheet on any of my units). If this is case, the yellow sheet mod would double up the diodes and offer some more protection. Just a guess.... its a big amp. The good news is with the right parts you can make as many of these rectifiers for cheap, and get at least good results. You can even try FR vs 1N diodes and see if they make a difference on sound. You can use an old blown rectifier socket.

JXBJXB wrote:Just a little update as to what we think is going on. Somehow, one of the copper cap rectifiers caused an over-voltage situation that blew out at least one of the segments (40 MFD) of the quad cap. This was the immediate response from one of Bob's techs when I described the symptoms - I removed the quad-cap and indeed that appears to be the case.

I am awaiting a replacement quad cap, will swap that in, and do the diode mod, and see where I get (with a tube rectifier this time!) and then update!

-jb

Sounds kinda like the reason I've heard not to run 5u4 variant directly heated rectifiers with indirectly heated output tubes. The directly heated rectifier will start applying full voltage quickly before the output tubes are warmed up and conducting, and this can cause an overvoltage situation and damage your caps if they can't take the surge. Due to this I only run directly heated rectifiers in amps that also have directly heated output tubes such as 300bs or similar.

vtshopdog wrote:My 125s ate copper caps about every 9 months. Switched to GZ37 and no issues for 4 years running. Got mine from British eBay seller Langrex. Fair price and good service, tons of NOS “valves”


https://www.ebay.com/itm/GZ37-CV378-53KU-MULLARD-NOS-VALVE-TUBE-/313009944885?hash=item48e0d8b935

vtshopdog... That's the reason why I am using higher rated amperage diodes on my yellow sheet mod because personally I do not believe those small diode's are rated high enough for the task they are performing and those small 1A diodes are overheating shorting out and/or burning out. I can see using those small diodes in other parts of the circuit but those diode's are receiving voltage and amperage straight from the power transformer and if there is a small fluctuation in voltage or amperage those little diodes will not tolerate it.

To me higher rated diodes will keep the voltage and amperage more stable in the long run. That's just my own opinion of course.

Bob Latino wrote:I have had the same two Weber WZ68's in my two M-125's for almost 5 years now .. These amps are not played daily but they do get use every week .. These two WZ68's have never given me a bit of trouble and still work fine ..

As to why Weber is not recommending them for AUDIO tube amps, I cannot say unless someone hears from them about why ? Some thoughts as to maybe why ?

1. Weber may know now that some of their WZ68's are going into KIT BUILT audio amps. If the high voltage storage system is not wired properly and has a short or a bad connection somewhere, then the WZ68 may go up in smoke at the first turn on. They may be wary of the WZ68 not going into a factory built and pretested amp.

2. Guitar amps do not usually run as high a B+ voltage as the VTA ST-120/M-125's (490 - 500 VDC). The Weber's probably (as baddog1946 has said above) could be made to stand up better to a little bit of abuse if they used the 1N5408 diodes instead of the cheaper 1N4007 diodes. As to "abuse", I am talking about voltages much above 120 VAC or amps being short cycled either manually or by sudden ON/OFF/ON power failures.

Bob

I just now found this old post from Bob back on May 14, 2015 9:57 and it confirms what I have been saying about these small 1N4007 1A diode's

Bob is mentioning here to use the 1N5408 diodes because these are 3A diodes and they will handle a lot more abuse instead of the cheaper 1N4007 1A diodes for which is true.

When I do my yellow sheet mod I'm going to use the 10A10 10A diodes because I picked 10 of them up really cheap on eBay for about $0.64 cents per diode and I can also use these heavy duty diodes in other kits and repairs and/or use them to build a heavy duty full wave bridge rectifier for a new hand made power supply for the bench.

One thing that is very important to know about semiconductors is they are very sensitive to heat and most semiconductors have a heat sink to sink the heat away so that the semiconductor can function properly but these small diode's don't have any type of heat sink on them and even the heat from the tubes on the inside of the chassis can make them not function properly and surge from just that heat alone and because these diodes are so small they also generate a little amount of heat on their own.

Another option is to rebuild your solid-state rectifier with better components and diodes instead of buying a new one that will just fail again with time.

Have a good day.

Tom

Well, my M-125 that was blowing fuses is back in business! Bob shipped me a new quad cap, and I also did the diode mod with UF4007 diodes. Borrowed a few 5U4GB GE rectifiers from a friend, while I wait for some McShane-provided 5AR4s. Fired back up, no problems whatsoever, as shown below. Yeah, I painted the transformers in a color to most closely match the two electrolytic caps on the driver board - about four coats of Rust-Oleum satin "Espresso."

One question though! I see that the UF4007 is a perfectly reasonable diode to use for the Yellow Sheet mod, but above from Bob and Tom and in a few other places I see diodes with a higher amperage handling rating being used. Should I now stay up at night worrying about the diodes that I chose? Or just leave it be? (I'm leaning towards leave it be....)

I think I can shed some light on this - regarding which diode to use...

First off, while the 1N4007 (or the fast/soft recovery UF4007) diodes are rated for an average current of 1 amp, they are also rated for surge currents of up to 30 amps. Both diodes are rated for a PRRV (peak repetitive reverse voltage) of 1000 volts. For most "yellow sheet" applications that's adequate - keeping in mind that the voltage the diodes see is much different feeding a tube rectifier instead of a capacitive input filter circuit.

If you want to play it very safe, the 1N5408 or UF5408 is the logical choice. Using these raises the average current rating to 3.0 amps, and the surge current rating to 150 amps. Even safer would be (2) UF5408s in series. that would give you a PRRV of 2000 volts and 3 amps of average current capability.

As I mentioned above, the recommendations are based on use ahead of a vacuum tube rectifier - in a conventional full wave capacitive input solid state rectifier circuit the voltage across the diodes can be quite a bit higher. (more than double the voltage peaks across a "yellow sheet" setup!). When the AC from your power trafo goes through the "yellow sheet" diodes it is converted to pulsating DC - meaning that the tube rectifier never has to switch with the negative peak of the AC wave voltage applied on one side and the full DC voltage on the other side. Under those conditions that switching point is where a lot of tube rectifiers get in trouble and the result is a failure.

Finally, keep in mind that the higher the diode's operating temperature, the lower the DC reverse ("leakage") current. The leads on the diode function as heat sinks, so the longer you can safely leave them the better off you are. And remember too that the 3 amp rated diodes use heavier leads that can pass more heat than the one amp diodes.

I hope that's some help anyway!

Jim McShane wrote:I think I can shed some light on this - regarding which diode to use...

First off, while the 1N4007 (or the fast/soft recovery UF4007) diodes are rated for an average current of 1 amp, they are also rated for surge currents of up to 30 amps. Both diodes are rated for a PRRV (peak repetitive reverse voltage) of 1000 volts. For most "yellow sheet" applications that's adequate - keeping in mind that the voltage the diodes see is much different feeding a tube rectifier instead of a capacitive input filter circuit.

If you want to play it very safe, the 1N5408 or UF5408 is the logical choice. Using these raises the average current rating to 3.0 amps, and the surge current rating to 150 amps. Even safer would be (2) UF5408s in series. that would give you a PRRV of 2000 volts and 3 amps of average current capability.

As I mentioned above, the recommendations are based on use ahead of a vacuum tube rectifier - in a conventional full wave capacitive input solid state rectifier circuit the voltage across the diodes can be quite a bit higher. (more than double the voltage peaks across a "yellow sheet" setup!). When the AC from your power trafo goes through the "yellow sheet" diodes it is converted to pulsating DC - meaning that the tube rectifier never has to switch with the negative peak of the AC wave voltage applied on one side and the full DC voltage on the other side. Under those conditions that switching point is where a lot of tube rectifiers get in trouble and the result is a failure.

Finally, keep in mind that the higher the diode's operating temperature, the lower the DC reverse ("leakage") current. The leads on the diode function as heat sinks, so the longer you can safely leave them the better off you are. And remember too that the 3 amp rated diodes use heavier leads that can pass more heat than the one amp diodes.

I hope that's some help anyway!

I have a different opinion on this subject.  

Those small diodes only handle a continuous 1 amp load...  When you say they will handle a 30 amp surge you have to keep in mind that's only 8.3ms (milliseconds) and then it's toast. That’s not enough time to even think about it.

The problem I see here is these small diodes at that low amperage is the first component the voltage and amperage goes through ahead of the transformer before any current gets to any other operating component in the circuit. If there is the smallest of surges in the line voltage those little diodes will be the first to go and will more than likely be toast because the slow blow fuse doesn't have enough time to do it’s job because the fuse is rated at 3 amps.

This is part of the reason why I'm using heavier duty diode's so that they don't get to hot and so they can handle a little more amperage. This method I am allowing the 3 amp fuse on the amplifier to do its job instead of eating up components inside the amplifier.

You are correct about the 1N5408 or UF5408 diodes because they can handle 3 amps and 150 - 200 amp surge for 8.3ms (milliseconds). Remember that’s only 8.3ms. To me at this rating of 3 amps it’s still competing with the 3 amp slow-blow fuse so it’s a tossup which one will blow first. Pluse I just looked and my slow-blow fuse on my VTA ST 120 is 5 Amps.

What I am recommending here is to trust your slow below fuse and let it do what it was designed for to handle a surge of no more than 3 amps and/or what ever the specified rating the fuse is.

This way you are not sacrificing your diode's or any other components inside your amplifier.

Here are the datasheets on these 3 diodes.

1N4007 Datasheet: https://www.mouser.com/datasheet/2/149/1N4007-888322.pdf

1N5408 Datasheet: https://www.mouser.com/datasheet/2/149/1N5408-888344.pdf

UF5408 Datasheet: http://dalincom.ru/datasheet/UF5408.pdf

Another reason why I believe these diodes need to be beefier to handle more amperage is because I live here in Texas and we have bad thunderstorms at least once a week or every other week and if I install diode's that are of a smaller amperage than my in-line fuse I will be replacing components in my amplifier every time a storm comes along. lol That's just the way it is here where I live so I have learned to trust the in-line fuse to do its job and not take out my computer, my TV, and/or my nice tube amplifier. Lessens learned the hard way :-)