I'm thinking on how I can make the most improvement with the easiest method for my Eico ST-70. The schematic shows 440v to the anodes with a 1800ohm dropping resistor to the screens (no choke)that gives 390v. The amp uses a 5AR4 rectifier.
I was wondering how I could make significant improvements without changing design much. I like this amp and am planning to replace a lot of components and tubes.
I have a thought for the power supply that I would like to throw out here. Weber has a power transformer for a Fender Twin rated for 400mA. It has a 320-0-320 and also tapped for 300-0-300. Is it possible to use a SS rectification on the 320-0-320 for the plates and then use the 5AR4 rectification on the 300-0-300 for the screens and preamp section? This would give me voltage very, very close to original specs and do it with a transformer that is made for my home's voltage (which is usually right around 125AC). I could use a CLC after the tube and have the choke be more around 100 ohms (maybe less, haven't checked) instead of 1800 ohm resistor. Much better filtering and much better impedance, correct?
Then again maybe you just can't do this with a power transformer. Thanks for any help.
I was wondering how I could make significant improvements without changing design much. I like this amp and am planning to replace a lot of components and tubes.
I have a thought for the power supply that I would like to throw out here. Weber has a power transformer for a Fender Twin rated for 400mA. It has a 320-0-320 and also tapped for 300-0-300. Is it possible to use a SS rectification on the 320-0-320 for the plates and then use the 5AR4 rectification on the 300-0-300 for the screens and preamp section? This would give me voltage very, very close to original specs and do it with a transformer that is made for my home's voltage (which is usually right around 125AC). I could use a CLC after the tube and have the choke be more around 100 ohms (maybe less, haven't checked) instead of 1800 ohm resistor. Much better filtering and much better impedance, correct?
Then again maybe you just can't do this with a power transformer. Thanks for any help.
You can do this, just remember that the total current drain off of both taps must be less than the rating for that specified for one tap. I have done this, and I have also run choke input filter using a 5AR4 and cap input using SS diodes off of the same taps to get 380 volts and 480 volts from an Allied 6K7VG.
Be careful about the protection! If the anodes (plates) run from a different circuit, please take account of what happens if the fuse blows in the anode circuit (main B+).
If the screens are left powered when the plates are unpowered, huge current will flow into the screens, and destroy your endpentodes - unless precautions are taken, eg interlocks, or current limit in the screen supply, or low value fuses.
However, if you work it all out, it is a good solution. I have done this to 1960s LANEY amps, where the B+ is 650V for EL34s. No way am I applying THAT to modern production pentodes. Luckily the Laneys have well-rated Partridge chokes, so I just add another pair of diodes (FW connexion, 2kV PIV diodes), an make a choke-input section for the screens, which pops out at ~400V. Perfect! high voltage on the anodes (for best sound), and better regulation for the screens. An individual F (flink/fast) fuse for each screen is the safest bet - its value chosen to limit the screen dissipation within the ratings, using the voltage you end up with.
If the screens are left powered when the plates are unpowered, huge current will flow into the screens, and destroy your endpentodes - unless precautions are taken, eg interlocks, or current limit in the screen supply, or low value fuses.
However, if you work it all out, it is a good solution. I have done this to 1960s LANEY amps, where the B+ is 650V for EL34s. No way am I applying THAT to modern production pentodes. Luckily the Laneys have well-rated Partridge chokes, so I just add another pair of diodes (FW connexion, 2kV PIV diodes), an make a choke-input section for the screens, which pops out at ~400V. Perfect! high voltage on the anodes (for best sound), and better regulation for the screens. An individual F (flink/fast) fuse for each screen is the safest bet - its value chosen to limit the screen dissipation within the ratings, using the voltage you end up with.
The thread starter wanted to use SS diodes for plates and tube rectifier for plates, plate voltage will be first but current will be near zero until screens come up.
This is a valid concern if there is a fuse in the plate circuit which is not too common. If you want to fuse the output circuits, place the fuse in the cathode. Be sure that the cathode bypass cap (if used) does not go around the fuse, or it will explode when the fuse blows. Other valid placements are in the B+ or the CT of the HV secondary on the power transformer.
An externally hosted image should be here but it was not working when we last tested it.
If you've already decided to replace a lot of components and tubes, then the solution is a good simpler than what you propose: active screen voltage regulation. You have a 50Vdc difference between supplies, and so any solid state regulator is going to work just fine here. You won't even need to make more holes in the chassis, or find some extra heater power.
Even if you're one of those "sand-o-phobes", doanworryboudit. SS active regulation will sound better than that simple series dropping resistor. That just says: "Cheap", and is highly detrimental to sonic performance.
Hey George, I know you like fireworks, but you don't advocate building a B+ without overcurrent protection, surely. Not with SS rectifiers, anyway.
I looked at a Trace Elliott 200W Bass amp that blew a KT88 recently. This amp has separate supplies for the screen and anode, and (presumably because of the risk to the screens) the B+ fuse is rated at T2A, and does not blow, usually. This amp blew the mains fuse first, and by then, the primary winding of the OT was fried. This was a high cost repair!
Fuse in the cathode circuit might have helped here, but really you need them in each valve's cathode, or else the heater insulation sees the whole B+ on a fuse blow. The good valves get stressed along with the bad where a common fuse is used. Beware also that a cathode fuse may permit some B+ out of G1 during an internal short, too, so precautions against the grid leak resistor or driver MOSFET setting fire to the amp should be taken.
For a 50V dropper, I think Miles's suggestion is good. I've used these reliably too, on guitar amps. Fairchild FQPF3N80C etc, wired as gyrator/cap multiplier is fine, and good sounding.
This one is in a 100W 4x EL34:
Attachments
How do you get excessive screen current while the 5AR4 is warming up? I would think current would not flow much if any before the B+ comes up on the screens.
edit: Oh, I see George already addressed this. Sorry. I'm very much in the beginner's stage.
Last edited:
I usually put fuses or sacrificial resistors in the cathode circuit on the amps that I build for myself. Another fuse goes in the power transformer primary. So far failures have been blown SS diodes or shorted electrolytics which blew the primary fuse. These have all been HiFi amps which don't see the physical abuse that a guitar amp gets.
I have seen tubes short out in guitar or bass amp heads that sit on top of the speaker cabinets and get shook up daily. Who hasn't seen an SVT blow up!
Tube shorts are rare in HiFi amps that run the tubes within their ratings. In a normal amp with reasonably good tubes a blown cathode fuse will cause the cathode voltage to rise to the point that the tube cuts off. Tubes that are gassy enough to glow blue can and will pass enough current to the grid to blow mosfets. I have seen it happen in my own amps, and both times I was using unfused bench supplies. I believe that in both cases a fuse on every pin of the tube wouldn't have saved the mosfets. They blow faster than a fuse.
Each builder is certainly encouraged to include proper circuit protection into their amps and there MUST be a fuse in the primary circuit of the transformer to avoid a flaming, smoking transformer. A B+ fuse is a good idea, and additional fusing is probably warranted if the price of the components justifies the additional protection, or the operating conditions warrant it.
I have used the multiple tap trick to generate different supply voltages from the same transformer. It can make the circuit design a bit simpler. In the case of pentode output stages it is true that application of screen voltage without plate voltage will result in fried tubes. In the case mentioned here that is an unlikely situation since the plate supply is SS rectified. SS diodes almost always fail to a short unless the short circuit current is sufficient to blow the diode in half or completly open. A primary fuse should prevent this. In other situations the circuit design should be considered to not allow this condition.
I generally prefer a mosfet regulator to generate screen voltage since it allows adjustment of the screen voltage. The screen voltage is critical in some tubes for minimizing distortion especially when running excessive plate voltage. I even have a seperate regulator for each of the 8 tubes in one of my EL84 designs. Of course mosfets usually fail to a short too which would blow up the tube, put you can't protect against each and every failure mode.
No, I'm not really a sandophobe. Without getting into true high end gear, I've done a lot of listening to both SS and tube amps in my life. A decent tube amp always sounds better to me. Now what I'm not sure of is whether a SS preamp can do as well. It seems to me the power amp is where the difference is. I have put SS guitar preamp stages into a tube amp, but I have never done an A-B with a HiFi system.
Anyway, I was studying hard about 7 - 9 years ago learning about tube electronics. We had two young ones and that just got put on hold. I have to go back and re-study all the stuff I can't remember without reference. Just before I had to stop, the guys over on Ampage were helping me learn about SS regulation. I was working on my own old Traynor amp with a 400-0-400 PT and wanted to use EL34s. I found out that EL34s can take pretty high plate voltage, but want to see (ideally) 350v for their screens. All I remember is zener diodes to set the voltage with a Mosfet to handle the current (I think). I can't remember whether the zeners dropped a specific voltage or whether the zeners actually set the specific voltage. I do remember that the circuit would only regulate if the main power supply stayed above the specified, regulated voltage. I also remember that the voltage coming off this circuit was super filtered and I was going to also use it for the preamp stages after the screens of the power tubes. Some people think there is noise from the diodes, like the hash from the SS rectifiers. I don't know about that. I have no frame of reference, but I would certainly like to relearn that circuitry and give it a try. Right now it looks a lot better that buying a another PT and making a chassis to fit all the stuff.
Could anyone help explain this a bit, links, schematics, etc.?
Hey, thanks for the help!
Any help you could give to help me relearn this type of regulation would be appreciated.
Thanks!
There are several ways to make a solid state regulator capable of running the output tube screens and a few preamp stages too. Some would argue that an active regulator is the best and there are valid reasons why this is true. If I was making a regulator to feed the plates, it would be active, but for a screen regulator it is hard to beat the simple zener diode plus mosfet circuit. That is what I am using in my Tube Cube Simple P-P (8 X EL84).
For a link to the circuit look at Pete Millett's Engineers Amp. The G2 regulator circuit is in the center of the power supply page.
D5, D7 and D8 are the zener's. You could use one 150 volt zener (if you could find one) or any combination of zeners in series that add up to the desired voltage. You could even use gas tubes to set the voltage, as with zeners put them in series to get the desired voltage. R33 feeds current through the zeners. The current should be enough to get the zeners into a stable region but not enough to make them hot. Pete runs about 3/4 of a milliamp which is OK, but I prefer about 1 to 1.5 mA, more if 5 watt zeners are used. If gas tubes are used you need 5 to 10 mA. C2 is a bypass cap that reduces or eliminates "zener noise" it must be reduced to .01 uF if gas tubes are used or the circuit will oscillate. D6 is a protection diode to prevent blowing the gate of the mosfet. R46 is a gate stopper resistor to prevent oscillation. C21 is a bypass cap to reduce the output impedance and R42 is a bleeder resistor to discharge the caps and provide a minimum load. R21 is a dropping resistor. It reduces the voltage at the input of the mosfet allowing it to dissipate less heat so a smaller heat sink can be used. It is not needed if adequate heat sinking is present on the mosfet and may actually reduce performance if the main power supply is not considerably higher voltage than the screen regulator puts out. In this case the main power supply is about 350 volts and the screen regulator puts out 150 volts. The resistor can drop 50 to 100 volts depending on screen current.
Link:
http://www.diyaudio.com/forums/tubes-valves/151206-posted-new-p-p-power-amp-design.html
For a link to the circuit look at Pete Millett's Engineers Amp. The G2 regulator circuit is in the center of the power supply page.
D5, D7 and D8 are the zener's. You could use one 150 volt zener (if you could find one) or any combination of zeners in series that add up to the desired voltage. You could even use gas tubes to set the voltage, as with zeners put them in series to get the desired voltage. R33 feeds current through the zeners. The current should be enough to get the zeners into a stable region but not enough to make them hot. Pete runs about 3/4 of a milliamp which is OK, but I prefer about 1 to 1.5 mA, more if 5 watt zeners are used. If gas tubes are used you need 5 to 10 mA. C2 is a bypass cap that reduces or eliminates "zener noise" it must be reduced to .01 uF if gas tubes are used or the circuit will oscillate. D6 is a protection diode to prevent blowing the gate of the mosfet. R46 is a gate stopper resistor to prevent oscillation. C21 is a bypass cap to reduce the output impedance and R42 is a bleeder resistor to discharge the caps and provide a minimum load. R21 is a dropping resistor. It reduces the voltage at the input of the mosfet allowing it to dissipate less heat so a smaller heat sink can be used. It is not needed if adequate heat sinking is present on the mosfet and may actually reduce performance if the main power supply is not considerably higher voltage than the screen regulator puts out. In this case the main power supply is about 350 volts and the screen regulator puts out 150 volts. The resistor can drop 50 to 100 volts depending on screen current.
Link:
http://www.diyaudio.com/forums/tubes-valves/151206-posted-new-p-p-power-amp-design.html
Just one parametric stabilizer (resistor and Zener) shunted by capacitor is enough to power whatecer you want through a single source follower.
In my Pyramid amps I use string of Zeners, and couple of source followers for 270V (screens) and 400V (preamp/driver).
Edit: Oops... found a drawing error, +400V stabilized connected to +400V rectified by mistake.
In my Pyramid amps I use string of Zeners, and couple of source followers for 270V (screens) and 400V (preamp/driver).
Edit: Oops... found a drawing error, +400V stabilized connected to +400V rectified by mistake.
Attachments
Last edited:
I would like to see your schematic, but it's not showing with enough resolution for me to read it.
Yes, this is the circuit the guys on Ampage were trying to help me understand some years ago. This is very helpful and I have remembered a lot but still need to understand more.
Thank you so much for your help. You are very generous with your time.
Just click on the picture to open it's full resolution image. You will see in the bottom left corner of a new window a square with arrows: click it to open the image in full size.
Nothing special there: a voltage divider formed by resistor and string of Zeners. Resistor is selected for current needed for Zeners. Zeners selected for needed voltage that goes from them to the gate of source follower. In order to reduce noises generated by Zeners they are shunted by a capacitor.
Oh, thanks! I didn't see that bottom left corner section. I'll try studying that and see if I can make sense of all this. I'm going to have to study up on basic SS circuitry.