KT77,88,ELs,6550 etc are suited for UL and Pentode circuits, because they are pentode tubes.
Many mass manufacturers favor this tubes, its are called popular tubes for this reason, these tubes sells alot on shops.
They are easy to drive, and easy to builds a amp.
A UL circuit have the grid or screen connected on the output or plate. However Jolida made 200W power amps with the 211 big triode in UL, with just the Grid connected to plate, maybe there is a secret in this Jolida circuit.
What about the tube characteristics makes it "not suitable for triode mode? For me all these popular Pentode tubes are unsuitable for Triode, specially SET amps, because they are not Triode tubes, and had small power in a SET Class A1 amp (seems a KT88=3Watts in conservative Class A1).
I favor big triodes as 211,805,GM70 for a SET amp to 20/30Watts, not PushPull or Parrallel SE.
Last edited:
Whether or not a multi-grid power type is suited to UL mode operation depends on screen grid toughness. For example, the EL84/6П14П and the SV83/6П15П are fairly similar, but g2 in the SV83/6П15П is fragile, which means "routine" UL is off the table.
If you are willing to spend on custom O/P trafos, with separate, tertiary, g2 windings, UL mode is feasible with types whose g2 is fragile. Then, g2 B+ can be regulated at a safe value.
If you are willing to spend on custom O/P trafos, with separate, tertiary, g2 windings, UL mode is feasible with types whose g2 is fragile. Then, g2 B+ can be regulated at a safe value.
Basically, low screen voltage ratings. This would include types like the 807, 814, most TV horizontal deflection types. You may or may not get away with spec busting in pseudotriode or UL operation. Some of the RF and TV types will handle a large margin of over voltage; others poof just north of that screen voltage rating.
Whether or not you actually want to use UL, that depends. I discovered that the 807 needs the extra help of local NFB to tame some nasty higher order harmonic distortion (O. Schade recognized the problem back in 1936, and recommended feeding back 10% of the plate output into the grid. That recommend was spot-on.)
Other types didn't need any assistance beyond the usual gNFB needed to clean up the open loop response.
Bingo! Right on the spot.
UL was many times discussed here. Opinions are different including the one I share: no UL recommended tubes exist. UL was a strange fashion, artificially heated up. Approach offered by Schade himself 20 years before was better, but it did not became popular until modern days. Why? Go figure...
I converted my GU50 triode strapped SET to a SEP with a 250V G2, now it's better and louder 
ETA: I cleaned up the response with a tight feedback loop from the GU50 anode to the driver's cathode - the transformer is free of the GNFB - because in my view transformers don't belong inside a feedback loop..
ETA: I cleaned up the response with a tight feedback loop from the GU50 anode to the driver's cathode - the transformer is free of the GNFB - because in my view transformers don't belong inside a feedback loop..
Last edited:
I think you have to look at the transfer characteristics in triode connection to determine which beam power tetrodes and pentodes are suitable for triode connection. The KT88 for example works quite well in triode connection with properly chosen operating point and load - output powers of 8W or so are easily achievable with some thought. The EL34 plate curves triode connected are quite nice looking and should be capable of around 5W in triode SE..
I do have a preference for DHTs SE amps myself and like the 45, 50, 300B and GM70 for my own designs, however in some instances the expense and complexity of a DHT in a given power range may be beyond reasonable for the intended use (user) and triode connected 6BQ5, EL34, and KT66/KT77/KT88 provide a reasonable compromise and the possibility of much more power when needed in either UL or pentode mode. (See my 6V6 SE design as an example.)
I also have a preference for GM70, in SET A1, 20/30W monoblock, but so far there is no such amp at affordable prices, maybe MingDa or other chinese company could venture in the GM70 tube.
Unfortunatelly chinese factories stay away from famous Russian tubes as 6C33 and GM70, they favor only Shuguang tubes.
Seems they have become protectionists.
I have a theory about that. The 807 design I did that uses the "Schade approach" also includes active screen regulation. How many commercial designs incorporated active screen regulators? Not a one I have ever seen. The only other 807 based design that had some sort of screen regulator was designed as a theatre amp.
UL was popular because it got rid of the need for screen regulation, so cheaper designs could be built. It never made much sense to me either. Why apply a large NFB signal to a relatively insensitive grid when less may be applied to a more sensitive grid? Besides, you can tweak parallel NFB. With UL, once the OPT is wound, the die is cast.
UL Amplifiers
Hi, UL has is not a pentode conneted as a triode. It is the PP output transformer which has a 20 to 40 % tap on the primary for the screen grid of a pentode or beem tetrode tube.
It makes the tube look closer to a triode but still retaining the high sensitivity of the pentode.
Any pentode wit the same anode -screen voltage is suitable.
Eg 6L6, EL34, El84, 6GW8 etc.
I fitted 6L6's for a test in to my Leak Stereo 60 and did not find any increase in distortion.
Worry less about the brand of the output tubes, but worry about the quality of the output transformer.
KT88 where off course used in the Quad amplifiers, als in an UL function.
Regards, Horst.
Hi, UL has is not a pentode conneted as a triode. It is the PP output transformer which has a 20 to 40 % tap on the primary for the screen grid of a pentode or beem tetrode tube.
It makes the tube look closer to a triode but still retaining the high sensitivity of the pentode.
Any pentode wit the same anode -screen voltage is suitable.
Eg 6L6, EL34, El84, 6GW8 etc.
I fitted 6L6's for a test in to my Leak Stereo 60 and did not find any increase in distortion.
Worry less about the brand of the output tubes, but worry about the quality of the output transformer.
KT88 where off course used in the Quad amplifiers, als in an UL function.
Regards, Horst.
So if a tube data sheet for the KT77 says that it was made specifically for Ultralinear, they are only saying that the screen grid may not hold up under triode configuration?
Not that I may be listening to more distortion (i don't have an ocilloscope)
Or that it will have reduced output/headroom,
Or.
Thanks
Paul
No. That means that screen grid is rugged and can dissipate significant power when plate voltage swings below screen grid voltage: the higher is screen grid voltage, the longer it dissipates higher power. Look at typical plate curves that show as well screen grid current, and you will see for yourself.
I sense a bit of confusion here. The OP seems to be assuming that being appropriate for UL means that it is not appropriate for Triode mode. As far as I know that is not the case. In fact having a sturdy screen grid would seem to be a very nice feature for a tube that is wired in Triode mode.
If the anode/plate swings below the screen grid then the screen grid becomes a temporary anode and may dissipate considerable power, depending on the perameters of the valve.
I have recently been playing about with triode/tetrode/UL connection with some 5B255M that I have, these are ruggedised 807 and so cannot be considered the same. I melted some old 807 screens was fun but it was expected.
There is no such thing as a UL specified valve. Usually its on screen voltage but you need to be very careful here as in push pull the anode can swing to twice B+ and so the screen can swing to whatever percentage of this the UL taps are at.
I have found that abusing the screen voltage rating in triode is ok within reason but I would not do it unless you can see or envisage some operating point past the published curves that looks good ie you dont mind trouble for the sake of experiment. For 807 it is stated that the max for triode is 400V I have had no trouble at 480V. Of course in triode the screen is connected to the anode via a resistor so cannot really get in too much trouble.
Anyway after all this I have come to realise that UL is a waste of time. At least with 807alikes. A bit of local feedback and tetrode connection sounds great. Of course as Miles mentions screen regulation is a must, preferably with a good regulator that does not change its output impedence significantly with frequency.
Cheers Matt.
I have recently been playing about with triode/tetrode/UL connection with some 5B255M that I have, these are ruggedised 807 and so cannot be considered the same. I melted some old 807 screens
was fun but it was expected.There is no such thing as a UL specified valve. Usually its on screen voltage but you need to be very careful here as in push pull the anode can swing to twice B+ and so the screen can swing to whatever percentage of this the UL taps are at.
I have found that abusing the screen voltage rating in triode is ok within reason but I would not do it unless you can see or envisage some operating point past the published curves that looks good ie you dont mind trouble for the sake of experiment. For 807 it is stated that the max for triode is 400V I have had no trouble at 480V. Of course in triode the screen is connected to the anode via a resistor so cannot really get in too much trouble.
Anyway after all this I have come to realise that UL is a waste of time. At least with 807alikes. A bit of local feedback and tetrode connection sounds great. Of course as Miles mentions screen regulation is a must, preferably with a good regulator that does not change its output impedence significantly with frequency.
Cheers Matt.
It is absolutely fine. When anode swings above B+ G2 swings less, it increases voltage on anode in respect to the voltage on G2. What is bad in UL, when anode swings below B+. In such case voltage on G2 is more positive than on anode. That increases power that G2 has to dissipate, that increases it's temperature.
Agreed, I melted the screens of those flakey 807 in UL at extreme voltages.
As I stated in the first sentence of my post:
"If the anode/plate swings below the screen grid then the screen grid becomes a temporary anode and may dissipate considerable power, depending on the perameters of the valve."
I only said that you need to be careful in UL because I melted them in UL rather than triode at the same voltages because of, I think, inductive effects in the OPT. I should have stated that. It could have also been oscillation but I did not see it and I had a fast enough scope on the screens to see it.
I should have also stated that screen dissipation rather then voltage is the thing to watch for.
Cheers Matt.
As I stated in the first sentence of my post:
"If the anode/plate swings below the screen grid then the screen grid becomes a temporary anode and may dissipate considerable power, depending on the perameters of the valve."
I only said that you need to be careful in UL because I melted them in UL rather than triode at the same voltages because of, I think, inductive effects in the OPT. I should have stated that. It could have also been oscillation but I did not see it and I had a fast enough scope on the screens to see it.
I should have also stated that screen dissipation rather then voltage is the thing to watch for.
Cheers Matt.
Inductive effects are the same no matter is it UL, or triode. What is better in triode VS UL, in triode mode anode voltage can't swing below G2 voltage.
Right.
Right.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.