Ian:
That makes some sense. Not having ever designed any tube amp myself, I would not know this. I just look at the spec sheets and see the operating points on the graphs.
Tube life is a touchy subject for me, for I once owned an ARC amp that was very, very hard on it's 6550's. It got just too expensive to retube after only 1000 hrs. I got rid of it!
Of course the SP-P is running fewer and less costly tubes. But the red-plating I found when using E-H EL84's did not make me too happy. Soon, I'm going to experiment with lower plate voltages on my SP-P, and see how it sounds.
That makes some sense. Not having ever designed any tube amp myself, I would not know this. I just look at the spec sheets and see the operating points on the graphs.
Tube life is a touchy subject for me, for I once owned an ARC amp that was very, very hard on it's 6550's. It got just too expensive to retube after only 1000 hrs. I got rid of it!
Of course the SP-P is running fewer and less costly tubes. But the red-plating I found when using E-H EL84's did not make me too happy. Soon, I'm going to experiment with lower plate voltages on my SP-P, and see how it sounds.
And yet, many love ARC gear, despite the short tube life. How did the amp sound?
jeff
George did say it is for the sound quality and it looks like most of the vintage amps ran their output tubes hard for the same reason. Those Russian tubes rated for 7189 use might be the answer too. Tubestore sells them for $100 a quad. Might last much longer too. JJ's should be fine too. They sound nice too! I think many of the old tubes were similar to the JJ's in being able to take a bit higher voltage as one of my origional Eico HF-12's had all the origional OT tubes. They all have the origional preamp tubes! One had the origional rectifier tube too. Eico HF-12's and HF-81 ran tubes a bit lower and so called for only EL84's not the 7189's you see on old HH Scott's etc. I use an old 10A variac and run my old vintage amps @ just over 110V . Maybe that would be worth a try for EH EL84 and other delicate ones. Then try adjust bias R's to suit or better use Ian's LED bias.
In other words set up to your OT tubes. Maybe you can get them just below that red plating point in the dark and they will still sound great. I remember reading of a fellow running his EL34 tubes in his modded but origional circuit Dynaco ST-70 @ precisely 18 mA +/- 1 mA instead of the normal 40 mA now recommended (used to be 50 mA) That was his sweet spot. He found 35 mA was another good spot that was more forgiving but not as transparent. He didn't change the B+. Maybe just keep those delicate tubes as spares or for testing anyway.
In other words set up to your OT tubes. Maybe you can get them just below that red plating point in the dark and they will still sound great. I remember reading of a fellow running his EL34 tubes in his modded but origional circuit Dynaco ST-70 @ precisely 18 mA +/- 1 mA instead of the normal 40 mA now recommended (used to be 50 mA) That was his sweet spot. He found 35 mA was another good spot that was more forgiving but not as transparent. He didn't change the B+. Maybe just keep those delicate tubes as spares or for testing anyway.
I should have said to try adjust cathode R's as George mentioned earlier to suit instead of bias R's
Those JJ's seem like really high quality EL84's and even feel a little heavier than NOS. I think they sound really nice and aren't highly priced either. They should last a long time as they aren't redpating. These are about as low in cost as you will find for power output and great sound. I think worthwhile having a few extras around!
Those JJ's seem like really high quality EL84's and even feel a little heavier than NOS. I think they sound really nice and aren't highly priced either. They should last a long time as they aren't redpating. These are about as low in cost as you will find for power output and great sound. I think worthwhile having a few extras around!
Last edited:
When you say "pushing" the tubes, there is more than one way to do this. In hifi, running the tubes at higher plate currents (bias) tends to reduce overall distortion of the stage. This comes at the expense of some output power and, of course, tube life. You can run higher plate currents at "normal" plate voltages (assuming your power supply can handle it).
Upping the plate voltage increases the power output potential. There is some fun to be had pushing the finals beyond what is in the spec and getting power out of a set of output tubes that is not often seen. But it doesn't necessarily increase tube wear because you have to look at things like plate dissipation and so forth. You may have to run less current at higher voltages to keep the dissipation in check. This may give you more watts but also may give you more distortion. The lower current actually reduces the wear on the cathode but the higher plate temperatures will wear out the getter faster.
There is more than one way to wear out a tube!
Upping the plate voltage increases the power output potential. There is some fun to be had pushing the finals beyond what is in the spec and getting power out of a set of output tubes that is not often seen. But it doesn't necessarily increase tube wear because you have to look at things like plate dissipation and so forth. You may have to run less current at higher voltages to keep the dissipation in check. This may give you more watts but also may give you more distortion. The lower current actually reduces the wear on the cathode but the higher plate temperatures will wear out the getter faster.
There is more than one way to wear out a tube!
Good comments and info from Ian, Russ and rmyauck. Thanks for helping my understanding.
Looking at some other DIY EL84 schematics floating around the 'net, I see power xfmrs with lower HV being used. Thinking about that too. Or, playing around with different RK values. Would this help less robust tubes? And what values would you guys suggest?
Jeff:
I had a VT 60, bought new in 1995. It did sound very good. At the time, I never heard anything I liked better. Still, it ran very hot and used up 6550's quickly. It sent me off tubes into SS for awhile. For 3 years now, my reference is Primaluna 3 and 7's. Great sound with matchless reliability. The 7's have GL KT88's with almost 5000 hrs and still sound good as new.
These amps have an "intelligent" auto-bias circuit that keeps the tubes at optimum values, without undue stress.
Looking at some other DIY EL84 schematics floating around the 'net, I see power xfmrs with lower HV being used. Thinking about that too. Or, playing around with different RK values. Would this help less robust tubes? And what values would you guys suggest?
Jeff:
I had a VT 60, bought new in 1995. It did sound very good. At the time, I never heard anything I liked better. Still, it ran very hot and used up 6550's quickly. It sent me off tubes into SS for awhile. For 3 years now, my reference is Primaluna 3 and 7's. Great sound with matchless reliability. The 7's have GL KT88's with almost 5000 hrs and still sound good as new.
These amps have an "intelligent" auto-bias circuit that keeps the tubes at optimum values, without undue stress.
Well after some fiddling about last night, I've probably erred in the other direction. I'd originally run just a choke in the first B+ segment ( Hammond 158S 60R / 250mA) - now it's in series and with a higher R value (270). The resulting B+ currently measures 315 plates, 317 screens, and average of 11.2 on cathodes. Of course this also drops derived voltages for the input/driver tube - which I'd image was never near its safe operating threshold.
Anticipation and expectations always cloud the perceptions in these situations, but I'm not yet sure I prefer it to the previous overdriven levels - some more fiddling is clearing in order
.
Thanks for the tip on "sweet spot" for B+ for EL84 that'll take some more fiddling (oh, I said that already ) any suggestions on the 12AT7? (JAN 6201)
George, Ian, Russ - thanks for the fun -it's been at least 5 yrs since I've enjoyed playing with an amp project this much - y'all might have started something
Anticipation and expectations always cloud the perceptions in these situations, but I'm not yet sure I prefer it to the previous overdriven levels - some more fiddling is clearing in order
. Thanks for the tip on "sweet spot" for B+ for EL84 that'll take some more fiddling (oh, I said that already
) any suggestions on the 12AT7? (JAN 6201)George, Ian, Russ - thanks for the fun -it's been at least 5 yrs since I've enjoyed playing with an amp project this much - y'all might have started something
Post 11 is what I was looking for:
http://www.diyaudio.com/forums/tubelab/176888-favorite-outputs-simple-p-p-2.html#post2360430
Edit: I believe that was with JJ EL84.
http://www.diyaudio.com/forums/tubelab/176888-favorite-outputs-simple-p-p-2.html#post2360430
Edit: I believe that was with JJ EL84.
Last edited:
Thanks Ian - I'll be playing with at least one other output & rectifier type later this Sunday afternoon, but was wondering also about suggestions for operating points for the 12AT7.
I don't think I'm out of line by suggesting that George welcomes experimentation with this board, so it's time to start. The only problem I ever have is fighting inertia - once moving I get impatient with the "scientific rigor"
of changing only one thing at a time.I don't think I'm out of line by suggesting that George welcomes experimentation with this board, so it's time to start. The only problem I ever have is fighting inertia - once moving I get impatient with the "scientific rigor"
I seem to have that problem too
The operating points for the 12AT7 would be fine for higher voltages as far as I can see. There's only around 100V across both tube sections with a 320V B+. I can't see 400V B+ doing any damage here.
The DC-coupled cathodyne input stage is a delicate balance because the anode voltage of the first stage (gain stage) sets the operating point of the second stage (cathodyne phase splitter). Increasing/decreasing the anode resistor of the first stage (gain stage) will increase/decrease the gain, but also decrease/increase the DC input to the second stage, which determines the grid voltage of that stage, which directly affects the operating point of that stage. This DC voltage is highly dependent on the anode and cathode resistors of the first stage. For the phase splitter, increasing/decreasing the anode and cathode resistors will increase/decrease the available voltage swing to drive the output tubes, and also increase/decrease the output impedance, and the gain will not change. This is how I understand it, but I could be wrong, hope I'm not. Have I caused confusion? I hope not. The short story is that I have built similar front ends in other amps but none of them seemed to outperform the Simple PP one so I left it alone.
I found tube rolling in the input stage to be interesting, at first I had some NOS RFT German 12AT7, they sound very clean with a crisp top end, later on I tried some NOS Teonex 12AT7 which seem to be a bit warmer in comparison and I left those in. A bit of net searching seems to indicate that Teonex rebranded other tubes, whatever they could get at the time, so I don't know who the real manufacturer is/was. I don't own any other 12AT7's. It seems bizarre that different tubes can seem to have slightly diferent sound characteristics/freq response, when there's just heaters, wires, plates and electrons in a vacuum.
On the subject of experimentation, I would like to try converting to fixed bias, and I am wondering if it is reasonable to use a circuit like this? If anyone can comment I'm all ears as I would like to try it out.
I seem to have that problem too
The operating points for the 12AT7 would be fine for higher voltages as far as I can see. There's only around 100V across both tube sections with a 320V B+. I can't see 400V B+ doing any damage here.
The DC-coupled cathodyne input stage is a delicate balance because the anode voltage of the first stage (gain stage) sets the operating point of the second stage (cathodyne phase splitter). Increasing/decreasing the anode resistor of the first stage (gain stage) will increase/decrease the gain, but also decrease/increase the DC input to the second stage, which determines the grid voltage of that stage, which directly affects the operating point of that stage. This DC voltage is highly dependent on the anode and cathode resistors of the first stage. For the phase splitter, increasing/decreasing the anode and cathode resistors will increase/decrease the available voltage swing to drive the output tubes, and also increase/decrease the output impedance, and the gain will not change. This is how I understand it, but I could be wrong, hope I'm not. Have I caused confusion? I hope not. The short story is that I have built similar front ends in other amps but none of them seemed to outperform the Simple PP one so I left it alone.
I found tube rolling in the input stage to be interesting, at first I had some NOS RFT German 12AT7, they sound very clean with a crisp top end, later on I tried some NOS Teonex 12AT7 which seem to be a bit warmer in comparison and I left those in. A bit of net searching seems to indicate that Teonex rebranded other tubes, whatever they could get at the time, so I don't know who the real manufacturer is/was. I don't own any other 12AT7's. It seems bizarre that different tubes can seem to have slightly diferent sound characteristics/freq response, when there's just heaters, wires, plates and electrons in a vacuum.
On the subject of experimentation, I would like to try converting to fixed bias, and I am wondering if it is reasonable to use a circuit like this? If anyone can comment I'm all ears as I would like to try it out.
Ian - that's almost too funny:
I had a nice listening session this afternoon with Jeff ( VinylKid), during which we rolled a few nice numbers from a stash he brought along.
Just tubes, friends, just tubes - the days were I could safely play with swapping tubes while "participating" are long past
NOS Mullard CV4024 vs GE JAN 6201 - waddya think?
Amperex (Holland) EL84 vs 6P14P-EV - same question
put the 2 together ....
But regarding the question of fixed vs cathode bias - that's were the fun starts. I'm in the middle of another project that is having a birthday coming up soon - rebuild of Jolida 302 ( EL34 P/P). The stock pre-map/ phase splitter stage has been replaced with Gregg van der Sluys' very cool Dynamutt board. Originally intended as (yet another
) option to replace the stock driver board for Dyna ST70, I'd been very impressed by the results in a custom rebuild from donated ST70 chassis - hence the "mutt". * Anyway, part of the process "reconfiguring" the original stock Jolida chassis and boards (although my first tube amp, it's certainly not a "treasured classic", and certainly one on which I have no aversion to performing major re-constructive surgery) is the gutting of a large PCB that houses the pre-amp/phase inverter and output stage bias network, including trim pots. It's been a rather creative bit of work to house Gregg's new board (designed to be a direct physical drop-in replacement for the original Dyna ST70 board), as well as retain that portion of the Jolida board with the PCB mounted bias trim pots. At this stage of the game, I'm thinking it likely would have been easier to spend the extra bucks on closely matched output tubes (the original stock Sveltana's are coming on 10yrs old) and reconfigure for cathode bias, which has been the case with all my other amp builds, kit or DIY.
The Simple P/P was my first P/P kit - tons of fun, and well designed with lots of latitude for experimentation that I have yet to fully explore - but for something as major a departure from George's "classic" design as you're contemplating, I'd be inclined to simply PtP the whole thing.
*
Classic Valve Design - Dynaco Clone and Original Design Boards and Repair Kits - Dynaco ST-70 Modifications
Did I say already how much fun I've had over this project? If I was previously close to ennui, that's gone (for now)
cheers, all
Biasing tubes
There is a thread in diytube.com in Stereo 35 form called " Improved SCA--35/ST-35 Performance " with link to this article:
http://www.tronola.com/A_New_Look_At_An_Old_Friend_Rev0.pdf
I think this is what George was talking about earlier instead of the LED's. Looks inexpensive to try should help OT tubes last along with possible sonic improvements. I've got to try it.
Randy
There is a thread in diytube.com in Stereo 35 form called " Improved SCA--35/ST-35 Performance " with link to this article:
http://www.tronola.com/A_New_Look_At_An_Old_Friend_Rev0.pdf
I think this is what George was talking about earlier instead of the LED's. Looks inexpensive to try should help OT tubes last along with possible sonic improvements. I've got to try it.
Randy
Last edited:
Randy:
Thank you for that link. While much of the theory presented is beyond my current level of understanding, I think I get the general concept. If you or anyone else is going to adapt this mod to their SP-P,I would appreciate a drawing detailing the connections to the SP-P board. I don't want to screw up the nice little amplifier I built.
And to think I left early to watch the Grey Cup, which turned out to be kinda boring.
Not exactly a quality piece. I had a track lift on me while replacing a trim pot. The large size is handy for big coupling caps though.
You must be special. Mine came stock with Shuguangs.
jeff
Thanks! I thought you guys would be interested too! In the thread Dave mentions you could use it with other output transformers, as long as they are 8K primary. He mentions the Heathkit 10K primary wouldn't work so well as many people build clones. The Simple PP is a much the same design of Cathodyne Splitter or Concertina and cathode bias on output tubes. Most of the vintage small PP amps used this exact design with different input tubes depending on designers. I can't see why it wouldn't work on LTP or Differential splitter amps too. George already has the necessary 1K and 100 ohm resistors added in his design which is safer for the tubes. George does have a resistor/ cap for each output tube cathode bias which is the best method of that design to work it's best also.
Dave also mentions fixed bias is possible just like some big tube amps use by adding a bias PS, but that is more complicated and money, but if you don't want any SS that is another way.
I like the idea of being able to set the bias for each tube if more pots are added as I have a number of unmatched 10BQ5's I got for very low $. I can't see why it wouldn't work with the 8K primary Edcors and others. I thought I read somewhere the 8K Edcors with UL taps have the ideal 25% ratio. Of course it takes a bit of the "Simple" out. He is also developing a design for the larger amps that don't have fixed bias.
Regards,
Randy
Dave also mentions fixed bias is possible just like some big tube amps use by adding a bias PS, but that is more complicated and money, but if you don't want any SS that is another way.
I like the idea of being able to set the bias for each tube if more pots are added as I have a number of unmatched 10BQ5's I got for very low $. I can't see why it wouldn't work with the 8K primary Edcors and others. I thought I read somewhere the 8K Edcors with UL taps have the ideal 25% ratio. Of course it takes a bit of the "Simple" out. He is also developing a design for the larger amps that don't have fixed bias.
Regards,
Randy
You must be special. Mine came stock with Shuguangs.
jeff
"special"
- but I'm sure my wife would agree with a least one interpretation of that the original purchase date was invoice for the Jolida was 2000, so I might have mis-remembered if I bought the Svetlana's separately - but certainly they're at least 9yrs old
While I will admit that some of the cheaper tubes of Chinese origin that I've heard were "sure guano", by all accounts the latest Black Treasure series by Shuguang are not to be sniffed at - for $170/pr for 6CA7, I certainly hope not.
Last edited:
A few of us have explored alternatives to LED bias. Some were discussed here:
http://www.diyaudio.com/forums/tubes-valves/171183-no-light-district-p-p-amplifier.html
It should be noted that the Dynaco ran all 4 output tubes on a single common cathode resistor. This is an obvious cost saving move in more ways than one. When an amp is driven hard the tube current goes up. This strains the power transformer. By using a common cathode resistor the cathode voltage will creep up slowly (due to the common bypass cap) when the amp is driven to full power continuously, reducing the total power supply current so a smaller power transformer can be used. This is why the measured distortion goes up and the power output goes down on the test bench when both channels are driven. This is not usually an issue with music since the average power output is far below full power.
The Simple P-P uses a seperate cathode resistor for each channel with a rather large bypass cap for each. The bias should not change at all with music applied even well into clipping. It may change a few tenths of a volt under continuous sine wave clipping.
Fixed bias or its equivalent can become advantageous if more power is desired. I have succesfully extracted 30 WPC from a Simple P-P board by using a combination of quasi fixed LED bias and regulated screen voltage. The Simple part goes out the window though.
One thing not made clear in the Dynaco mod article is the need for having the cathode bias track the supply voltage. It is OK for the "fixed cathode bias" to track the total B+ as shown in that article. This is advantageous in UL or triode mode. The usual rule of thumb is that the grid (or cathode) bias must track the SCREEN grid voltage in a pentode mode amp. In extreme power experiments with the Simple P-P it has been determined that you want the cathode voltage fixed and the screen voltage fixed. Let the total B+ vary with the line voltage, it will only affect the maximum power output slightly.
I think most people get far too serious over just what the exact ratio for UL is supposed to be. It might make a little difference in full power measurements into load resistors, but real speakers are nowhere near a constant resistive load so the plate load impedance will never be exactly 8K so theory goes out the window. My thought is, if you have a transformer, try it. I am using 6600 ohm OPT's in my P-P amp because I have a bunch of them. I have a pair of 8K OPT's too, but I think the 6600's sound better.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.