Those of us who have been building tube amps forever (well it seems like forever) can often guess the power capability of power and output transformers from size and weight. The "calibration" of this ability comes from years of experience, but this "calibration" does not often hold for transformers stripped from organs. Many tube organs have tube based oscillators and dividers. The filament power for the tone generator is often huge. The large wire inside the power transformer calls for more "window" in the core than normal making the transformer bigger than it needs to be for a given VA rating. I had a Baldwin organ back in high school that used 36 6SN7's in the tone generator The power transformer was about the size of a small car battery but the amp section only ran a pair of 6V6's.
Given the 40Hz rating on the power transformer, I would guess that the organ was made before the 6L6GC was invented! That doesn't mean that you can't run a 6L6GC at 6L6 ratings, or even more conservatively. Most organ amplifiers were rather conservative designs. They wanted reliability and long tube life, not screaming power. These transformers were intended for use with 6L6, or 6L6G, or 6L6GA. There is no reason why they couldn't be used with 6L6GC set up for 15 to 25 watts.
The book says that EL84's want 8000 ohms. I choose to ignore the book and experiment. I have had good luck with EL84's at 6600 and I am getting some really low distortion numbers and great sound too. I have also tried 6AQ5's at 6600 and they do not like it. Again I wouldn't rule the EL84 / 6600 ohm combination out especially since the thread starter has them.
I like them at 8800 and about 330 on the plates. But I agree with everything you've posted right through. There's quite a bit of straight to the curve on most audio tubes operated as pentodes (or beam power). It all depends on where in AB land you want to operate. One operating point that has always puzzled me is the 3200 ohm primary that Marshall and HiWatt use for 6CA7/EL34. Only GEC (as far as I know) shows typical operation at this low impedance. Most show that 6600 at 450 or so. That's near where Dynaco runs them (I think, maybe??).
I have Heyboer make a 35 watt dual primary into 4-8-16 ohm secondary for me at ten at a time. I use 5000 and 8800 primaries. They work for a lot of tubes. I had him take apart a Hammond Organ OT and he modeled the primary after that. (It's upside down he said. ((??)))He said the secondary couldn't be deciphered because the wire kept breaking off. They sound awfully good though.
But, the OP has Hammond not Hammond Organ stuff. Completely different animal. And BTW, the way I've always calculated OT impedance is to square the turns ratio and multiply by the output impedance. I didn't quite follow what he was doing in his initial post.
There's is usually very little new under the sun as they say, and it's often true with vacuum tube amplifiers as well. Over 50 years ago, Philips of Holland endorsed and published operating parameters for the EL84 operating in class AB1 pentode mode, with fixed bias. Their stated optimum load for that condition was 6600 ohms. In that mode, it produced more power at less distortion than cathode bias operation would. Scott has run a number of their amplifiers in similar fashion, and produce wonderful sound and very good tube life.
As for the EL34, 3200 ohms is appropriate for push-pull fixed bias pentode operation with near equal screen and plate voltages at typical B+ levels for this tube. It produces max power output, and is a safe load with regards to screen dissipation as long as the screen limiters are in place. As the configuration changes to cathode bias and 40% UL operation, 6600 ohms becomes a much more optimum load under those conditions.
Dave
As for the EL34, 3200 ohms is appropriate for push-pull fixed bias pentode operation with near equal screen and plate voltages at typical B+ levels for this tube. It produces max power output, and is a safe load with regards to screen dissipation as long as the screen limiters are in place. As the configuration changes to cathode bias and 40% UL operation, 6600 ohms becomes a much more optimum load under those conditions.
Dave
If you don't know the turns ratio, and don't want to take the transformer apart, just apply a known AC voltage to one winding and measure the voltage on the other winding. Divide to get the voltage ratio. Voltage ratio = turns ratio. For most OPT's I just plug the primary into the wall outlet and measure the secondary.
I have been experimenting to see just how far I can go with the EL84. I have a bunch of surplus 6600 ohm OPT's so I connected the plate supply up to one bench power supply and the screen up to another. I kept playing with the voltage knobs to find the sweet spot for my OPT's. I have found a rather nice sounding spot that is far beyond the published maximums that works well with some old Sylvania 6BQ5's or new production JJ EL84's. Wimpy tubes will melt! I am running the plates from a 425 volt supply and the screens from 325 volts. I get about 30 watts from a pair with no red glow from anything. I am in the process of building a stereo amp with 4 tubes in each channel. I have measured 60 watts per channel with 0.7% distortion running a 3300 ohm load. I plan to back it off to about 50 WPC and see just how long it lasts.
Those who have followed my antics on this forum over the years are familliar with my extreme experiments. Some work out, and some blow up, but that is the fun. I blow things up so that you don't have to.
"I am running the plates from a 425 volt supply and the screens from 325 volts. I get about 30 watts from a pair with no red glow from anything. I am in the process of building a stereo amp with 4 tubes in each channel. I have measured 60 watts per channel with 0.7% distortion running a 3300 ohm load. I plan to back it off to about 50 WPC and see just how long it lasts."
That idea is in fact the basis of one of the more esoteric commercial 6BQ5 amps produced. Ran normal screen voltages, but juiced the plate B+ up to over 500 volts as I recall. Was easily getting around 40 watts if I remember correctly, and plate dissipation was well within published limits. The quiescent current was pretty low so the tubes ran cool, and distortion was pretty low as well.
Dave
That idea is in fact the basis of one of the more esoteric commercial 6BQ5 amps produced. Ran normal screen voltages, but juiced the plate B+ up to over 500 volts as I recall. Was easily getting around 40 watts if I remember correctly, and plate dissipation was well within published limits. The quiescent current was pretty low so the tubes ran cool, and distortion was pretty low as well.
Dave
In both cases, I ask, at what voltage? According to "the book" which we and others often exceed, we work from a family group of curves and arrive at a PtoP load from the line. It's surprising how well it works. Leo Fender pushed 6V6s into 410VDC plates because that produced the tone he was looking for. Ditto Jim Marshall and all the rest. David Hafler and the folks at McIntosh, as well as Ken Fisher surprisingly, generally worked from the curves. Like I said before, it depends on where in AB land you want to be. I will say this though, Marshalls and HiWatts eat tubes alive. Gregg Levy at Hitest guitars doesn't even like to sell high dollar Mullards to HiWatt users. There are a whole variety of reasons for the short life span of course. We could go on for days about screen current, grid current and resultant conduction angle etc. A pair of Tungsram El-34s will last a lifetime in a trainwreck. 410VDC on the plates 1K screen resistors (if memory serves) and that clipper stage. Etc. Etc. This is opinion: If you operate screens 470 ohms south of the plates with no regulation you got problems with tube life; if you rock. If you country; eh, maybe not so bad. Leo designed for Western Swing.
It is interesting though that you are having such good luck with 6600 on EL-84. It's worth a try to me. Do you bias towards the high side; that is towards class A?
And yes, that's how I arrive at OT turns ratio.
Yeah, you bet. Separate screen supply. That's the ticket. I'm going, (one of these days) to go to work on a pair of 807s with regulated screens. Three hundred volts period. No matter what the plates are doing. I got a PT out in the shop with a 325-270-0-270-325 secondary looking for a job. I got a bunch of high voltage darlingtons for regulators. Someday I'll learn to wire a mosfet. Duh....
Interesting discussion guys, It's getting my brain back into valve mode after a decades long absence.
Progress so far - took covers off, cleaned the heaps of zinc oxide & cockroach dirt off, sprayed them with a hammered silver finish, they look great. Went to Dick Smith to get some heatshrink to re-dress the leads, only to find they only sell consumer electronics nowadays..ho hum... next weekend then...
Progress so far - took covers off, cleaned the heaps of zinc oxide & cockroach dirt off, sprayed them with a hammered silver finish, they look great. Went to Dick Smith to get some heatshrink to re-dress the leads, only to find they only sell consumer electronics nowadays..ho hum... next weekend then...
Re wax the innards and they'll be mo betta at their age. Thermonic Technology here's the best way, but a good long soak in hot wax will help if you don't want to hook up a vacuum. BTW, a regular shop vac will work.
Just a thought.
Chris, yeah I was going to the Penrith Jaycar next week (I'm in St Marys), but thanks for the heads up on RS -I like the sound of Free delivery, & they have the best EL84 prices I can find too.
re:'Re wax the innards ' - a good idea Danny, thanks.
Searching round the datasheets, I found a reference to 6K load for distributed load connection on a Mullard Data sheet. Va 300v Ik 2x 28mA, this seems to be for cathode bias though, Rk 390+47.
re:'Re wax the innards ' - a good idea Danny, thanks.
Searching round the datasheets, I found a reference to 6K load for distributed load connection on a Mullard Data sheet. Va 300v Ik 2x 28mA, this seems to be for cathode bias though, Rk 390+47.
"In both cases, I ask, at what voltage?"
The original Philips data I mentioned operates the EL84 in fixed bias, with 300 volts on the plate, 290 volts on the screen, and -12 vdc grid #1 bias, into a 6600 ohm load. In this configuration, distortion drops to about 1/3 that of the same conditions with a more traditional 8000 ohm load (still with fixed bias), and power output is of course maximized with a 6600 ohm load as well.
Regarding the EL34, no doubt that screen limiters are very necessary, and 1K usually provides safe operation with these tubes using about a 3200 ohm load. Since these tubes are a true pentode, their screen grid does not have the protection afforded from beam power construction, so the high value limiters are very necessary. On the other hand, 6L6GC tubes can get away with as little as 100 ohm screen limiters (more generally providing protection from parasitic oscillations in this case rather than addressing screen dissipation concerns), with 470 being the most common value used for them.
Fixed bias operation of EL34s into a 3200 ohm load is most commonly used in conjunction with (at least) 1K screen limiters, and 450-475 vdc unregulated B+.
Dave
The original Philips data I mentioned operates the EL84 in fixed bias, with 300 volts on the plate, 290 volts on the screen, and -12 vdc grid #1 bias, into a 6600 ohm load. In this configuration, distortion drops to about 1/3 that of the same conditions with a more traditional 8000 ohm load (still with fixed bias), and power output is of course maximized with a 6600 ohm load as well.
Regarding the EL34, no doubt that screen limiters are very necessary, and 1K usually provides safe operation with these tubes using about a 3200 ohm load. Since these tubes are a true pentode, their screen grid does not have the protection afforded from beam power construction, so the high value limiters are very necessary. On the other hand, 6L6GC tubes can get away with as little as 100 ohm screen limiters (more generally providing protection from parasitic oscillations in this case rather than addressing screen dissipation concerns), with 470 being the most common value used for them.
Fixed bias operation of EL34s into a 3200 ohm load is most commonly used in conjunction with (at least) 1K screen limiters, and 450-475 vdc unregulated B+.
Dave
Not to be mean spirited of course but smoke is the only way I can think of, unless you can open the bells and peel back some fish paper to reach the coil wire and measure it with a dial caliper. The wire size should tell you how much current the wire is rated at but that really isn't the whole current capacity story I don't think. It lives in a DC world but the ripple creates some heat eh? If you've ever had Heyboer do any work for you and have a relationship with them I'm sure Phil could tell you right off the top of his head. <edit> Oops, ah, shed, yes, Sydney. A call to Heyboer not entirely easy. I'll see if I can help out. May take awhile. I've got a bulk order of OTs going in maybe next month. I'll ask.
Do you know VacuumVoodoo (Alexander Neimand?) He's Dutch maybe, Norwegian...something Scandia. Smartest guy I know in this stuff. He hangs in a variety of places. The Amp Garage dot com is one of them. I'll bet he's got the answer if you frequent similar environs.
That's a good question and I'd like to know too. I've got one similar in the box and it's rated at 125ma CCS. I think it's 4.5h and 128 ohms.
Last edited:
If it is 4.5H it is not a high current choke. But you can test it. Got a 1A 12V DC power supply? Put just a few volts DC through so that there is say 10ma of current. Let it run for an hour. Is the choke warm? If not rase the current and wait. When it gets warm back off two steps until it runs very cool.
re:'it is not a high current choke' - going by the size & other specs of Hammond chokes, I think you're right, it's probably capable of somewhat less than 100mA. I found some formulae so I'll try & reverse engineer from those.
re:'reach the coil wire and measure it with a dial caliper' I can see the wire, but don't want to destroy the windings - I'll have to look closely to see if I can disconect one end. Fortunately, a couple of weeks age a mate gave me an electronic caliper!!!
re:'reach the coil wire and measure it with a dial caliper' I can see the wire, but don't want to destroy the windings - I'll have to look closely to see if I can disconect one end. Fortunately, a couple of weeks age a mate gave me an electronic caliper!!!
One more thing to think about. I bet you measured 4.5H with no DC current in the choke. The inductance goes down, a lot when there is DC current. A choke's inductance is rated with DC maximum current. So when you design a circuit around this part do not assume it is 4.5H It will be less.
So you get the calipers out and you figure out the size of the wire, then what? You can't use the current carrying ability of the wire in free air. Just put the choke, a milliamp meter and a power supply in series and adjust the current a very slight about every hour until it gets a little warm, then back off 30% or so. You really do need to re-measure the inductance while this much current is present in the choke.
I updated my thread:
http://www.diyaudio.com/forums/tube...-some-help-opts-set-old-6v6gt-monoblocks.html
But since the OP was discussing Hammond organ OPTs here, I'm ask a side band question:
My amp is the Type H-1 probably from D-20 tone cabinets since I have a pair of them and not a quad that would have been required for a F40, but I have two of these amps:
http://www.captain-foldback.com/Hammond_sub/schematics/FR40.jpg
The schematic confuses me a little WRT to the secondaries heading down to the speaker jacks and back and such. I understand and realize the speakers were field coils and the coils were used as filters in the PS. But my question is, are these OPTs suitable for use on a standard stereo P-P amp with the 6V6s triode strapped? Is the iron capable of delivering full bandwidth? A little roll off in the bass is ok with me, but rolled off highs isn't and I'd really love for a bandwidth out to 60KHz like tubelab saw on his pair.
My plan would be to utilize as much of the iron as I can, and either add a input/phase splitter dual triode in front of the 6SN7s, or replace them with two novals and series connect the heaters to get 12.6V for 12AZ7As or see if I have other low heater draw 6V noval options. Otherwise I have to add a second heater transformer and potentially a second small PS for the additional tubes.
Worst case, I'll keep them PPP and triode strap them and keep them balanced monoblocks. I'll take care or the signal phase splitting via a TVP in a preamp. However, I'd really like to make them dual SE stereo amps, realizing that I'll have to buy new OPT iron for the second one as I's use the two Hammond OPTs on one chassis for the first stereo amp.
Since I'm completely rewiring the amps, would the iron and PS also be suitable for the second amp to run EL34s instead of 6V6s if I use the appropriate input/phase splitter and driver tube, or just use a D3A in a differential pair to drive the EL34s?
Thanks guys, diyA just rocks! Where else could I even find valuable info like this?
http://www.diyaudio.com/forums/tube...-some-help-opts-set-old-6v6gt-monoblocks.html
But since the OP was discussing Hammond organ OPTs here, I'm ask a side band question:
My amp is the Type H-1 probably from D-20 tone cabinets since I have a pair of them and not a quad that would have been required for a F40, but I have two of these amps:
http://www.captain-foldback.com/Hammond_sub/schematics/FR40.jpg
The schematic confuses me a little WRT to the secondaries heading down to the speaker jacks and back and such. I understand and realize the speakers were field coils and the coils were used as filters in the PS. But my question is, are these OPTs suitable for use on a standard stereo P-P amp with the 6V6s triode strapped? Is the iron capable of delivering full bandwidth? A little roll off in the bass is ok with me, but rolled off highs isn't and I'd really love for a bandwidth out to 60KHz like tubelab saw on his pair.
My plan would be to utilize as much of the iron as I can, and either add a input/phase splitter dual triode in front of the 6SN7s, or replace them with two novals and series connect the heaters to get 12.6V for 12AZ7As or see if I have other low heater draw 6V noval options. Otherwise I have to add a second heater transformer and potentially a second small PS for the additional tubes.
Worst case, I'll keep them PPP and triode strap them and keep them balanced monoblocks. I'll take care or the signal phase splitting via a TVP in a preamp. However, I'd really like to make them dual SE stereo amps, realizing that I'll have to buy new OPT iron for the second one as I's use the two Hammond OPTs on one chassis for the first stereo amp.
Since I'm completely rewiring the amps, would the iron and PS also be suitable for the second amp to run EL34s instead of 6V6s if I use the appropriate input/phase splitter and driver tube, or just use a D3A in a differential pair to drive the EL34s?
Thanks guys, diyA just rocks! Where else could I even find valuable info like this?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.