If you don't afraid...
Let's just leave it for another day... btw, the curves look vaguely like the pentode curves instead of thestandard triode curves...
Typical right-handed triode curves.
Here is the RCA classics, 6N7 (actually, Russian version 6Н7С):
Initially I built the amplifier with 811A triode.
After completing the design I also tested it with 807 and GU50 with hi-mu connection. Below is the schematic for the original 811A version.
When GU50 is used, it is better to use one half (or both) of 6N6P triode instead of 6N1P. This is because the GU50 has lower input impedance (some 350 to 400 ohms for AC) than 811A and thus requires a little bit more driving power. The required driving voltage however is only some 10 Vrms for 15 W output.
I can also complete the schematic of GU50-version if this is interesting anybody.
Here is the image of the prototype (with 811A).
After completing the design I also tested it with 807 and GU50 with hi-mu connection. Below is the schematic for the original 811A version.
An externally hosted image should be here but it was not working when we last tested it.
When GU50 is used, it is better to use one half (or both) of 6N6P triode instead of 6N1P. This is because the GU50 has lower input impedance (some 350 to 400 ohms for AC) than 811A and thus requires a little bit more driving power. The required driving voltage however is only some 10 Vrms for 15 W output.
I can also complete the schematic of GU50-version if this is interesting anybody.
Here is the image of the prototype (with 811A).
An externally hosted image should be here but it was not working when we last tested it.
I think that hi-mu connection was meant for push-pull class B2 operation where you tie the grids together and connect them to ground (by means of an IT secondary, typically). The catode is also grounded. Basically it is a 0V bias connection with low quiescent current and high efficiency.
45
45
Last edited:
This is really interesting. How would you say it sounds? Sound signature is typical SET?
What grid currents would you use for GU-50? Do you think half 6N6P is sufficient as the current for 6N6P is almost double of the 6N1P.
Quikie22 the 811A is a different device. Firstly it's a triode and one doesn't have choice for the connection. It has 10W grid power handling and quite less grid current than the GU50 whose G2 grid handling is 5W only and much less for G1. In fact the GU50 in that hi-mu triode connection has a maximum specified quiescent anode current of 30 mA only........
Secondly one can get up to 22W from a 811A at plate dissipation around 50W (600V/80-85 mA and bias around +18.5V). I don't think one can get double figures Pout from a properly biased GU50 and also (!!!
) looking for a reliable audio amplifier to listen to music......45
Last edited:
Unfortunately I am not anymore a "golden ear". My high frequency limit is around 12 kHz. But I think it sounded very good. One reason is maybe that since the output power capacity was sufficient (15W), the linearity was good and distortion low with typical listening level. In principle my evalution of all these three versions (811A, GU50 and 807) was mainly technical.
The dc-grid current (=bias) for GU50 was around 12 mA and bias voltage about + 7 volts.
One half of 6N6P is sufficient to drive and bias GU50. Then the anode current of 6N6P should be about 25 to 30 mA.
I have full test data at my "lab", but I am now spending the winter in Spain and I have only some drawings at my computer. However I can complete the circuit diagram.
It seems that I have the 811A test results in my computer, which I could publish, since those were very equal with GU50 test results.
Good, please show us. This would be a good learning experience for me. This is the first time I have heard of a "right handed triode". Would be interesting to see how it is configured.
Test results are given without negative feedback and with 16 dB NFB and are as follows:
The schematic of version with GU50 is this:
An externally hosted image should be here but it was not working when we last tested it.
The schematic of version with GU50 is this:
An externally hosted image should be here but it was not working when we last tested it.
Man you are dangerous....a 6N6P can take 4.8W per plate if you use ONLY one section and 4W per plate (8W in total) if you use both plates. In the best case at 242V/25mA (+8V bias) plate dissipation is already more than 6W.
If then we add that the GU50 is also run well above its limit (which is 30 mA max with grids tied together) and consider that you are giving this information to a person who clearly has no idea about what you are talking about, will you take your responsibility in case of....??
In the beginning I want to emphasize that I am not publishing any instructions for building an amplifier. Instead I am presenting what I have done.
Yes, you are right. The current of 6N6P must be smaller than I recalled.
Some 16 mA / half with 250 V anode voltage would be OK, if only one section is used. As I told previously, I do not have my memos available now.....
I know that the German data from the year 1941 concerning LS50 tube gives max. Ia as 30 mA. So what ? GU50 has max. Pa as 40 W and this is not exceeded.
If the tube will last only 2000 h. instead of 3000 h. You just need to put a new tube in place after 3 years of use instead of 4 years of use.
Yes, you are right. The current of 6N6P must be smaller than I recalled.
Some 16 mA / half with 250 V anode voltage would be OK, if only one section is used. As I told previously, I do not have my memos available now.....
I know that the German data from the year 1941 concerning LS50 tube gives max. Ia as 30 mA. So what ? GU50 has max. Pa as 40 W and this is not exceeded.
If the tube will last only 2000 h. instead of 3000 h. You just need to put a new tube in place after 3 years of use instead of 4 years of use.
With the grids tied together there is no max rating for Pa simply because with 1000V max anode voltage and 30 mA max anode current you are never going to exceed it while if you bias the tube positive for an anode current more than 30 mA more likely you will exceed grid power rating. The G1 of a GU50 is not like the G1 of a 811A and G2 has got 5W only. The increase in grid current is not linear and at some point it just rockets.
The practical advantage of a so called dual grid power amplifier was its double use as driver in class A and output tube in push-pull class B.
Artosolo this is just speculation. You can only prove it with good statistics in your hands which means thousands of cases. I could accept your logic if we were talking about 10% beyond a max rating, however 40-50% starts to be a bit too much to be given for granted.
As a DIYer you free to do as you like, of course, however if you do some search for threads about tube failures such problems are mostly due to improper use.
45
Last edited:
Thank you Artosalo for generous sharing of valuable information!
Your amp is great.
Actually, GU-50 is rated rather very conservatively, for linear RF amplifiers that work constantly in military transmitters. For audio they can take more, especially for reproduction of classical music with wide dynamic range. Even when anodes are red (100+ W) they continue working. However, the problem is cooling down of glass envelopes: anodes are made of Nickel alloy that sustain high temperatures, to survive detuned antenna coils in field conditions. However, SE audio amps are similar to linear RF output stages, so I would consider 40-50W as the limit for SE amps.
Your amp is great.
Actually, GU-50 is rated rather very conservatively, for linear RF amplifiers that work constantly in military transmitters. For audio they can take more, especially for reproduction of classical music with wide dynamic range. Even when anodes are red (100+ W) they continue working. However, the problem is cooling down of glass envelopes: anodes are made of Nickel alloy that sustain high temperatures, to survive detuned antenna coils in field conditions. However, SE audio amps are similar to linear RF output stages, so I would consider 40-50W as the limit for SE amps.
Last edited:
You are wellcome. It is my pleasure if somebody is interested in my plain experiments.
I think GU50 has very great potential to many applications especially compared to the price it is available from Ebay for example.
Actually GU50 (at Ug2=250V) seems to have similar characteristics to KT88 (at Ug2=300V). But the price is only a fraction of KT88's.
No significant difference, actually. So if you have a pentode or ray tetrode with suppressor grid or ray plates connected to cathode, it is fine. If you have pentodes like 6P15P or GU-50 with suppressor grid not connected, you may tie it to other grids. Effect is minimal, due to it's density and distance from cathode.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.