|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
63.246.183.223
Michael, I know you're a fan of the 6P3S-E, and I have quite a few myself. Have you ever measured distortion of these tubes, relative to other types in the same amplifier? I just subbed a pair into one channel of my HK A500 after rewiring the sockets, and the results are pretty disappointing. At output levels of a few watts, THD of the 6P3S-E pair is about 70% higher than the 7355s, and 60/7K IM distortion is about 40% higher. This is being measured with about +385V on the anodes and 19W dissipation each tube, using the original HK cathode (self) bias circuit. I did not try to adjust feedback values to optimize these results.
I'm also going to test with a pair of 6L6GBs, just to see if distortion is improved. If it is, I'll test with fixed bias to determine whether the effective increase in A-K voltage restores the output rating of the amplifier. (We've discussed this before; 6P3S-Es only make about 18W in this amplifier, which agrees with RCA specs for a 6L6 at this voltage and output Z of 3.6K). I might also try a pair of EL34s, which I believe will provide the rated 25W without fixed bias. The additional filament current required for EL34s is a major inconvenience though, so I'm hoping that won't be necessary.
Anyway, I'd like to compare notes if you've kept records of your work with these tubes.
FWIW, I'm only working with the power amp portion of the HK. Everything ahead of that has been disconnected.
--------------------------
Buy Chinese. Bury freedom.
Follow Ups:
Jack
Keep in mind that the 7355 is a lower plate resistance relative to the 6P3S-E which is the 6L6GB. Also,unless those tubes are matched,you will have readings all over the place.
"If it measures good and sounds bad, it is bad; if it measures bad and sounds good, you have measured the wrong thing."
- Daniel R. von Recklinghausen
"...the 6P3S-E which is the 6L6GB"
Well, this was a major surprise, but what you've said above (which is something I've also believed for a long time) doesn't appear to be true. I just finished an initial test with the 6L6GBs, and their performance in this amplifier is almost identical to the 7355s. They deliver *exactly* the same output power as the 7355s - 25W on the nose. In fact, the only reason I won't say they're identical is that distortion is better. The output stage I posted now does less than 0.1% THD at 4W output (I'm using 4W as a convenient standard for comparison at the moment).
So, I guess I need to ask how you arrived at the conclusion that the 6P3S-E is equivalent to the 6L6? I know this is a common belief around the Web, but have you verified it for yourself?
OK, I think at least one more test is needed at this point. I'll try the 6P3S-Es at higher voltage/dissipation by temporarily changing the amp to fixed bias. That will raise the effective anode voltage from about +350 to +385. I should be able to see if there's a trend toward better performance at that point. Maybe an adjustment to the feedback circuit is also in order.
About balance, all these pairs other than the 7355s have been graded for Gm on my Hickok and for idling current in a test fixture at +400V. The HK also allows balancing AC currents for minimum IMD, which I've done each time I tried a new pair of tubes.
--------------------------
Buy Chinese. Bury freedom.
First,not only is the output transformer not suited for the 6P3Se,you also don't have enough drive or B+ to make the 6P3Se develop its power and lower distortion numbers. Typically, if you run a 6L6gc at 425vdc loading with a 6.6k PP impedance,you still need about 90v of drive to bring that tube to full power.
You are starving the tube in that 7355 circuit and it's no different than someone sticking a KT120 in a KT88 circuit..It will work but it will not make anymore power and distortion will typically be higher because the tube isn't running correctly.
"If it measures good and sounds bad, it is bad; if it measures bad and sounds good, you have measured the wrong thing."
- Daniel R. von Recklinghausen
Edits: 03/15/15
Let me clarify what I'm seeing with the 6P3S-E. First, this is not an issue of insufficient drive. Saturation of the 6P3S-E itself is what's limiting the power, nothing else. With the amplifier converted to fixed bias, the 6P3S-E makes 19W max, whereas the 6L6GBs make about 27W and the 7355s do 30W. Increasing the bias current does not remedy this, and after several more tests, I can see that only an increase in plate voltage will allow the 6P3S-E to output more. So, I'm convinced Steve O is right - the high Rp of these tubes causes them to saturate earlier than a 6L6 or 7355 with the 3.6K output transformers in this HK. This means - as I've suggested during our previous discussions on this issue - that the 6P3S-E is not a suitable sub for the 7355 in the HK A500 or similar amps. As you say, it wants more voltage and a higher output Z. This implies the 6P3S-E is also not a good sub for EL-34s, at least not in circuits that similarly use a low-Z transformer. In those amps, as in this one, the 6P3S-E will saturate early, limiting output power. As for its performance in circuits designed with, say, +425V or more and at least 6K output Z, that's something else. I do have a couple amplifiers here that meet that requirement, but I have a larger number of 25-30W/ch amps with low Z OPTs that I want to resurrect first. It's not a surprise to me that the 6P3S-E won't do the job, but disappointing just the same. OTOH, the really good news was that 6L6GBs perform extremely well in this application. Based on manufacturer's data sheets, that was totally unexpected.
I'll test these 6P3S-Es again in a more suitable amplifier when time permits. If that doesn't produce numbers I can live with, I'll have to thin the herd. :)
--------------------------
Buy Chinese. Bury freedom.
"First, this is not an issue of insufficient drive."Actually it is a question of drive as well with the 6P3Se. You are comparing this to a 6L6gb which would be fair however,the 6P3Se tubes are somewhat lower dissipation when driven under the conditions you are experimenting with.Increasing bias current doesn't really increase power much but it should lower distortion a bit.
This implies the 6P3S-E is also not a good sub for EL-34s, at least not in circuits that similarly use a low-Z transformer.
Nobody disputes that.It is never a perfect replacement but it works provided you have at least 400v on the plate and have at least a 5k impedance with sufficient drive.Some EL34 circuits have this capability such as Bob Latino's amps and the AES six pack amps.
Now,try and match up the best pair of 6P3Se tubes that you can and run them in a circuit like the W5m and let me know what you find..The W5s have about 11k impedance but they also have enough drive to get the job done and the B+ is running around 390v plate to cathode.
Bottom line,you do lose some power with the 6P3Se but the fact that you can run them so much harder,that can be an advantage in some circuits. When I look at the data on the 7355,it takes about 22% less drive than a 6L6gc so I'm actually shocked that you got those good of numbers. Did you modify the driver circuit when you converted this amp to fixed bias?
Jack,since you have fixed bias in that HK amp,try and see you can red plate the 6P3Se by dropping the bias at G1. You will notice that it's quite difficult in comparison to a regular 6L6gc so that tells me if you have the capability to run the tubes into a higher impedance with sufficient drive,you can get pretty good numbers..Take a look at this. I did modify this Scott 280 but notice the B+ on the tubes and this is also a fixed bias amp. This simply shows the fact that you can run the hell out of these tubes and you almost have to get the absolute best numbers.
"If it measures good and sounds bad, it is bad; if it measures bad and sounds good, you have measured the wrong thing."
- Daniel R. von Recklinghausen
Edits: 03/16/15 03/16/15
Michael, when you say it "works" at higher voltages and Z, are you referring only to output power? Haven't you made distortion measurements? That's the only remaining question in my mind - how do they perform regarding distortion under those conditions? I don't doubt they'll make power at 592V. :)
--------------------------
Buy Chinese. Bury freedom.
I have made many measurements but mostly in 6L6 and KT66 amps..Most of the disputed argument I presented in past posts was to reiterate the fact that these tubes can handle much more than 350v rating claimed in the Data by the Russian military.They do dissipate lower power than a 6L6gc but I have run them in a Mac Mc240 and I was able to get 43wpc...Of course the 6L6gc black plates made 51 watts but distortion was within 10% of the black plates but at lower power.If you don't match those tubes up at full voltage,you will have readings all over the place.
In the W5ms,the distortion was also a bit higher than with the black plates but again,we are .2% for one and .33% for the 6P3Se..Would we hear that under most conditions? Probably not.
Match these tubes up the best you can and then give it a try in a 6L6 based amp and let me know what you find.
Mikey
"If it measures good and sounds bad, it is bad; if it measures bad and sounds good, you have measured the wrong thing."
- Daniel R. von Recklinghausen
Thanks, that's good info. I'll look to see which of my 6L6 amps uses the highest B+, that'll be the one to test.
--------------------------
Buy Chinese. Bury freedom.
Just to confirm the data, I installed a different pair of 6P3S-Es. With the amp still in its original configuration (cathode bias), these measured considerably better than the previous pair regarding distortion. Still not quite as good as the 7355s, but very close. Maximum output power was 19W, essentially unchanged from the first pair. Next, I grounded the wiper of the cathode balance pot and applied negative bias to the grids. Bias was adjusted for +2.5V at the cathodes, representing 50mA per tube. Maximum power increased to 23.8W, still a little shy of the factory spec, and both THD and IMD (at 4W) went up slightly. I was also able to measure distortion at 20W for comparison. At this power level, IMD was moderately worse than the 7355s and 6L6GBs, while THD fell in-between.Again, if you've taken measurements with the 6P3S-E, it would be helpful to know what you've seen. I'm confused by the fact the 6P3S-E performance seems to generally agree with manufacturers' datasheets for the 6L6, but my 6L6GBs don't. Very odd.
--------------------------
Buy Chinese. Bury freedom.
Edits: 03/14/15
Data sheet comparisons indicate that 6L6, 7355 and 6P3c-E are similar in most respects, but one outstanding difference is Ra: 6L6 ~ 22K, 7355 ~ 43K and 6P3c-E ~ 65K under similar operating conditions.
While not directly answering your questions, I've always found the 6P3c-E produces less power for a given THD in every 6L6 app I've subbed them into, usually about 20% less. As a result, I've never considered them as a 100% compatible 6L6 type. Based on specs alone, I can't see how these things could ever operate successfully in an EL34 app...and yet many report stellar sonic results with this sub. Full disclosure: never tried this sub myself. OTOH, the 6P3c-E is rugged and I kinda like the sound in the 6L6 apps I've tried them in. And they are (were?) cheap enough to be disposable test subjects for initial power up.
Yes, I remember now that you were working with these tubes. I still have the histograms you posted. Maybe the difference in sound is due to increased 2nd harmonic energy. I didn't measure the 2nd specifically during my tests, but I do know it was always the predominant harmonic.
So, I think there's a clear winner here. I've just tested the 6L6GBs with fixed bias, and harmonics at 4W are now less than half of the earlier self-biased measurement. THD is 0.04% and IMD is 0.28%. Even at 20W, THD is less than 0.5%. With the increase in effective voltage from fixed bias, max output is approaching 30W. This is all more than good enough to go on to the next step of tuning by ear.
Incidentally, this will go onto a new chassis later this year as a standalone power amp. Sensitivity is really high - 1/4V for full output - so I'll replace the 9-pins with a 'SL7 and 'SN7 during the rebuild. In addition to reducing sensitivity, that might push distortion down even further.
It's really too bad, I was hoping the 6P3S-Es could do the job. They're really inexpensive, and I have 100 or so. I guess you get what you pay for. The 6L6GBs are costly, but I have enough NOS to last until the next owner. :)
--------------------------
Buy Chinese. Bury freedom.
These will sub for a 6L6gc however they are more along the lines of a 22 watt tube.The advantage to these tubes is the fact that you can push them hard before they red plate.Notice that they seem to run relatively cool in comparison to a standard 6L6gc.Drop the G1 voltage and watch the distortion drop and then observe the plates of the tubes and tell me what you see.
"If it measures good and sounds bad, it is bad; if it measures bad and sounds good, you have measured the wrong thing."
- Daniel R. von Recklinghausen
Wanted to add that distortion levels of the 6P3S-E can be reduced to the same or less than the other two types by reducing the value of the feedback resistor. However, this substantially increases ringing seen on 1kHz and 10kHz squarewaves, and probably degrades sound quality.
--------------------------
Buy Chinese. Bury freedom.
As I have said,you are not running the tube correctly because a 6P3Se is not optimized for that 7355 circuit..The 6P3Se tubes love to be pushed hard.
"If it measures good and sounds bad, it is bad; if it measures bad and sounds good, you have measured the wrong thing."
- Daniel R. von Recklinghausen
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: