Hi Riverty,
Rectifier tubes normally last for years. I would check the actual current draw through the B+ line if the Electroharmonix arced on power up. Normally you have to have too high a capacitance to cause that. Your schematic looked fine as far as input capacitance was concerned. We still don't know what the load is.
-Chris
Both Eli and I have said this. It is true, you will not hear the solid state diodes unless you short the tube plate to cathode. Honest. The tubes characteristics will be the character you hear. Remember, the SS diodes are in series with the tube. The device with the highest losses will determine the I-V characteristics. If you add some resistance in series with each SS diode, it will extend the life of your rectifier tube. This is because your peak currents will be reduced.
Rectifier tubes normally last for years. I would check the actual current draw through the B+ line if the Electroharmonix arced on power up. Normally you have to have too high a capacitance to cause that. Your schematic looked fine as far as input capacitance was concerned. We still don't know what the load is.
-Chris
I did the diode modification back in 2002 and mentioned it in this forum back then. I had several pairs of mono block Quicksilver amplifiers. To be honest I had three pairs of the 8417's and one pair of the KT88 version.
I was going thru 2) 5AR4's in a 3 month period. After I added the diodes I haven't gone thru another set of 5AR4's. You still retain the tube rectifier sound if you call it that. The diodes take the brunt of the load when you turn the amp on and the tube gently applies the B+ to the circuit.
I have kept one pair of the modified Quicksilvers and I'm still running on the same pair of 5AR4's. No more flashing and torching. About 5 or 6 years ago I was playing with a few Dynaco 70's and I started using 5U4's in them with my diode mod. The 5U4 is an excellent sounding rectifier and nothing is lost using a few diodes in series.
I never got caught up in the "Fast diode" line of thinking as I personally never heard any difference between them and a standard 1N4007 diode but those in the know say there is a difference so in my mind its worth a few extra pennys to simply thro them in.
The mod works period and it doesn't make the sound any different.
I was going thru 2) 5AR4's in a 3 month period. After I added the diodes I haven't gone thru another set of 5AR4's. You still retain the tube rectifier sound if you call it that. The diodes take the brunt of the load when you turn the amp on and the tube gently applies the B+ to the circuit.
I have kept one pair of the modified Quicksilvers and I'm still running on the same pair of 5AR4's. No more flashing and torching. About 5 or 6 years ago I was playing with a few Dynaco 70's and I started using 5U4's in them with my diode mod. The 5U4 is an excellent sounding rectifier and nothing is lost using a few diodes in series.
I never got caught up in the "Fast diode" line of thinking as I personally never heard any difference between them and a standard 1N4007 diode but those in the know say there is a difference so in my mind its worth a few extra pennys to simply thro them in.
The mod works period and it doesn't make the sound any different.
Without the diodes fitted, there will be around 750Volts (HT + -ve transformer emf) across the rectifier. That could cause your arcing as a 5AR4 is only rated at 550V. The diodes do not take care of the brunt of the load, refer to Kirchhoff's Law, they block high reverse voltages across the rectifier valve.
A 5U4 is rated at 1550Volts PIV so the diodes will have no effect.
A 5U4 is rated at 1550Volts PIV so the diodes will have no effect.
Maybe I don't use the correct terminology. Isn't the AC rectified by the diodes before it reaches the plates of the 5U4? Try putting a filter cap after the 5U4 larger than spec and watch the sparks fly. Now put diodes in series before the 5U4 and my experience is there is no arching. So, yes it does work.
That's sort of a philosophical question! If you have two diodes in seires, does the first one do all the rectification and the other one do nothing? Or should they be considered to be 'one big diode'? I would say the latter is a fairer appraisal.
With quality NOS 5U4GBs readily available and at reasonable prices, I have never understood why anyone would bother with JJ or EH. (I can understand why a mfr would use a modern mfr for consistent supply, but I am talking about a private individual seeking the best sounding and most reliable tube for his own equipment.) My favorite 5U4GB is a GE.
I have not tried Eli's hybrid rectifier but it sounds like an interesting idea.
I have not tried Eli's hybrid rectifier but it sounds like an interesting idea.
I am going to second the notion of adding more series resistance to the plates of the 5U4GB. General Electric datasheet states 67 ohms per plate. I have encountered this issue before and have tried both series diodes and series resistance, resistance is the proper fix. Although if you do have a hollow state rectifier internally short the diodes will save your butt.
Hi Gorgomat,
The resistors act as fuses whether you want them to or not. Adding fuses is redundant.
Hi famousmockingbird,
Yes, they are recommended. You are supposed to subtract the resistance of the windings from that value and the remainder is the absolute minimum resistance to have in series with each plate. Everyone just sticks the recommended resistance in without worrying about the calculations. I find that the added resistance really helps to reduce / eliminate rectifier switching noise. With (only) solid state diodes, the resistance is critical to put in.
Hi Salectric,
I have to side squarely with Eli on this. The Electroharmonix brand is a solid, high quality line of tubes. The quality is higher than the "NOS" tubes were back in the day. Mike Mathews did get it right in every way. I just wish the prices would drop some.
I'm going to guess you run a repair shop. From one service shop owner to another, why on earth would you gamble on anything when you have a high quality source for tubes within your own country?? Doing a job over blows any profit you made and detracts from your reputation. That really makes buying from New Sensor the only rational choice you have.
Hi JonSnell,
-Chris
The resistors act as fuses whether you want them to or not. Adding fuses is redundant.
Hi famousmockingbird,
Yes, they are recommended. You are supposed to subtract the resistance of the windings from that value and the remainder is the absolute minimum resistance to have in series with each plate. Everyone just sticks the recommended resistance in without worrying about the calculations. I find that the added resistance really helps to reduce / eliminate rectifier switching noise. With (only) solid state diodes, the resistance is critical to put in.
Hi Salectric,
I have to side squarely with Eli on this. The Electroharmonix brand is a solid, high quality line of tubes. The quality is higher than the "NOS" tubes were back in the day. Mike Mathews did get it right in every way. I just wish the prices would drop some.
I'm going to guess you run a repair shop. From one service shop owner to another, why on earth would you gamble on anything when you have a high quality source for tubes within your own country?? Doing a job over blows any profit you made and detracts from your reputation. That really makes buying from New Sensor the only rational choice you have.
Hi JonSnell,
Both those figures are rated breakdown voltages. The actual breakdown voltages are higher and variable. The voltage across each element initially depends on the relative capacitance of each device. If one breaks down the other halts the current before it really even begins. These elements are in series, so for transient voltages, look at your capacitance s. Once things become more steady state (uS to mS) the transient currents have subsided and you are into the reverse blocking time period. These diodes do extend the tube life as the tube no longer suffers arc-over. 2.5 KV rectifier diodes are available, so you can get parts that exceed what the 5U4 is rated for. You can always double the diodes up in series too (for 1N4007 and similar). Sometimes you will see capacitors across the rectifiers in high voltage applications. They are there to ensure the voltages are shared more equally.
-Chris
Soulmerchant's suggestion of using PSUD2 is well worth it if you are pondering over datasheet values
Certainly if your plate supply resistance is lower than rated, or you want to use a higher filter capacitance than rated, then the diode’s transient peak plate current (per plate), and the steady-state peak plate current (per plate) shouldn’t be exceeded. These limits are best checked using PSUD2 to simulate the worst-case operating conditions when the mains power of a hot amp is turned off for a few seconds and then turned on again. In that situation, the filter cap will have discharged but the heaters are all hot and so when the amp is turned on again the diode will conduct fully to charge the capacitor and supply the load.
Simulation can show the initial transient overshoot, and the steady-state. Different manufacturers often had slight differences in specs - so be careful to cross check the simulation model.
Certainly if your plate supply resistance is lower than rated, or you want to use a higher filter capacitance than rated, then the diode’s transient peak plate current (per plate), and the steady-state peak plate current (per plate) shouldn’t be exceeded. These limits are best checked using PSUD2 to simulate the worst-case operating conditions when the mains power of a hot amp is turned off for a few seconds and then turned on again. In that situation, the filter cap will have discharged but the heaters are all hot and so when the amp is turned on again the diode will conduct fully to charge the capacitor and supply the load.
Simulation can show the initial transient overshoot, and the steady-state. Different manufacturers often had slight differences in specs - so be careful to cross check the simulation model.
I did a project with about the same IL that used a 5U4GB (had to drop ~100V to get the DC rail voltage: Si + 650VCT over volted by that much) 40uF was too much, and exceeded the Isurge= 1.0A/plate. Dropping that to 34uF (two 68uF/300V in series) got the Isurge to a little under 800mA. Used a 7H ripple choke with a 220uF filter capacitor behind it.
So far,been using the same 5U4GB (a "Realistic Lifetime" NOS specimen) since 2007 with no problems. I'd reduce that reservoir capacitor, and get a NOS 5U4GB that's not suspect. This was a commonly used type, and there are plenty still available. Not pricey either.
Its just a bad tube. It happens. But more so with 80s - current chinese and russian tubes. Your reservoir cap is within spec. vs. "golden age" tube manufacturer sheets and many people push that cap size further than 40uf.
I'd scoop up a small lot of used US ones off ebay and don't look back.
Once they arc they are predisposed to keep doing it AFAIK.
I'd scoop up a small lot of used US ones off ebay and don't look back.
Once they arc they are predisposed to keep doing it AFAIK.
When swapping tubes or connections requiring power off, always wait 2-3 minutes for cool down before switching the amp back on to avoid creating the first arc over even with good oldie rectifiers. 20
Hi riverty,
You can do that, but also consider some series resistors for each plate circuit. Doing that will help clean up your power supply B+. Adding these resistors will give you great mileage. As for capacitance, you could go as low as 20 uF. Early equipment often used an 8 uF primary filter cap. So you can see that the industry has already pushed the input capacitor to the limits.
-Chris
You can do that, but also consider some series resistors for each plate circuit. Doing that will help clean up your power supply B+. Adding these resistors will give you great mileage. As for capacitance, you could go as low as 20 uF. Early equipment often used an 8 uF primary filter cap. So you can see that the industry has already pushed the input capacitor to the limits.
-Chris
Hi 20to20,
-Chris
I can't agree enough with you. This holds true for signal and output tubes as well. I have had more than one customer destroy their product by "tube rolling". Also, keep in mind that some recommended replacement tube types differ in their specifications. If you go with a "high reliability" type, you may find the heater current is high. With many tubes you can easily cook your power transformer.
-Chris
Hi all,
Just wanted to mention that I have a Chinese BOYUU A9 SEP (EL34's) using an old-stock 5U4B. After running it for several hours, it would arc with a big BANG! After that it would arc about every half hour. I tried Eli's fix with the diodes and all works perfectly now. As a side benefit, the hum from the speakers reduced. Before, my ear could hear the hum from about a foot or so from the speaker. Now I have to put my ear right next to the speaker to hear it. Not sure why this is, but I won't complain about it.
I first became acquainted with Eli back in the eary/mid 2000's. He's helped me off and on on many projects. He's always been spot on, so I've learned to trust his judgment. Lately he's been helping to improve the BOYUU. It's a much better amp now than it was. Thanks, Eli.
Dave
Just wanted to mention that I have a Chinese BOYUU A9 SEP (EL34's) using an old-stock 5U4B. After running it for several hours, it would arc with a big BANG! After that it would arc about every half hour. I tried Eli's fix with the diodes and all works perfectly now. As a side benefit, the hum from the speakers reduced. Before, my ear could hear the hum from about a foot or so from the speaker. Now I have to put my ear right next to the speaker to hear it. Not sure why this is, but I won't complain about it.
I first became acquainted with Eli back in the eary/mid 2000's. He's helped me off and on on many projects. He's always been spot on, so I've learned to trust his judgment. Lately he's been helping to improve the BOYUU. It's a much better amp now than it was. Thanks, Eli.
Dave
Hi hurdy_gurdyman,
I think your filter capacitors might be too high in capacitance. If the tube arcs at times other than turn-on, it sounds like you are exceeding the hot switching current rating. The rectifier diodes can help with this, but you still have a problem that should be dealt with. You don't need high capacitance filter capacitors in a tube amp.
Can you post a schematic that covers the power supply section as it is now?
-Chris
I think your filter capacitors might be too high in capacitance. If the tube arcs at times other than turn-on, it sounds like you are exceeding the hot switching current rating. The rectifier diodes can help with this, but you still have a problem that should be dealt with. You don't need high capacitance filter capacitors in a tube amp.
Can you post a schematic that covers the power supply section as it is now?
-Chris
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.