| Alastair E |
Sorry, My brain has definately gone today.....
Cant remember the 'proper way to calculate the O/P power in watts RMS of my amp...
Im getting 60V Peak To Peak, 30V Peak? into 8 Ohm Non inductive resistor @ 1KHz just before apparent clipping, measured on O'Scope......
Could one of you great brains tell me the correct way to calculate this....? |
|
|
| jane |
Po = (Vpeak)^2 / (2 x Rload)
or
Po = (Vrms)^2 / Rload
Vrms = Vpeak / sqrt(2) |
|
|
| Alastair E |
| quote: | Originally posted by jane
Po = (Vpeak)^2 / (2 x Rload)
or
Po = (Vrms)^2 / Rload
Vrms = Vpeak / sqrt(2) |
Err, Not sure what '^' is supposed to mean....Sorry..... |
|
|
| dnsey |
^ Means "raised to the power of", so x^2 is "x squared".
BTW, there's no such thing as RMS power :devilr: (just kidding, we all use the term!) |
|
|
| Alastair E |
Right....
So V Peak, (half the P-P value) 30 x 30 is 900.
And 8 Ohm x 2 is 16....
900 Divided by 16 is 56.......
56W RMS from a single pair of 6C33C-B in Circlotron OTL.....Sure, That can't be right......? Thats around 7.5 AMPS per 6C33c Peak Current!-
(8 ohm, 60V peak-peak, Each one must be conducting the 7.5A, OR am I missing summit........)
(The pair is being run at 75mA idle current, and 190V per circlotron supply, with diff-amp feedback drive. The grids go positive around 30V on peaks, Checked all supplies and settings again today....)
-This sort of power from just 2 bottles in the O/P stage just aint possible surely...??!
It does sound VERY loud when you turn the wick up though, but still great sound......)
Checked the readings and scope-settings again this afternoon......
Something silly going on here.....! Told you...My brain aint that good today, so please go easy on me.....
:smash: |
|
|
| tubetvr |
Hi,
| quote: | | Im getting 60V Peak To Peak, 30V Peak? into 8 Ohm |
60V Pk-Pk is 30V Pk, 30V Pk will be 30/1.4142 = 21.21V RMS, output power is VRMS^2/Rload so in this case 21.21^2/8 = 56.25W
| quote: | | 56W RMS from a single pair of 6C33C-B in Circlotron OTL.....Sure, That can't be right......? Thats around 7.5 AMPS per 6C33c Peak Current!- |
The power calculation is OK but your calculation of Peak current is not.
Pk current for one tube is Vpk/Rload so in this case it is 30/8 = 3.75A, it is still very high but not too high, I have seen peak currents of up to 3A for a single 6C33C but I have never used positive grid voltage.
It is also interesting to calculate the dissipated power of your tubes as you are quite much over the max rating of 60W.
We calculated the peak current to 3.75A, it can be showned that the average current drawn from the power supply will be Ipk/Pi so in trhis case Iavg is 3.75/3.14 = 1.2A, the power supply voltage is 2*190V so the power drawn from the power supply is 380 * 1.2 = 456W, from this we remove the output power of 56W and end up with a dissipated power of 2 tubes of 400W or 200W per tube.
I expect that the anodes will glow red if you run full power for anything shorter than a minute or so. With music it is still probably alright as average output power is 10 - 20dB lower than peak, say 15dB as average, then you will have an average output power of ~1.8W and the average current will also be lower at around 0.2A with a power dissipation of 38W per tube so well inside the ratings.
Regards Hans |
|
|
| Alastair E |
Thanks Hans, I probably am running them pretty hard, but then again on music at a moderate volume, the dissipation as you say is very much lower. The full-power tests were only done for a matter of seconds, as any form of continuous high diss test like this would probably kill the tubes after a much more....
When I was thinking of the peak current, I was for some silly rerason thinking that the full current must be passing each tube, but logical thinking being a circlotron with one tube per 'circlotron rail' then the peak current will be halved, each supplying just half the peak current
Pretty pleased with my efforts, Time to breadboard up the other channel for some serious listening, testing, tweaking etc before I commit to a 'Proper' build in a nice box......
As to reliability, I guess providing I treat it with respect and dont show it TOO much heavy-rock, then things 'should...' be fairly reliable. Generally day-to-day running when not using my prototypes for serious listening I tend to run the TV/DVD sound through them. Ive used this single-channel circlotron for around 8 months with few if any real issues, just the odd change/mod here and there mainly on the pre-amp/phase-splitter side.
Great advice on your site about 'Pre-Conditioning' the 6C33. I have done this on all but a couple of tubes. I just let them run for around 3 days just heaters then use them. Only 'failure' I had was with a tube that hadnt been 'pre-conditioned'., when I did some power-tests some months ago....Strange thing about that tube, Left it running in the amp and it slowly came back to life.....Probably not for long though!--Maybe Ill try heaters only for a week on it to see what happens
:hot: |
|
|
| tubetvr |
Hi Alastair,
Glad to hear that it works well, and it shouldn't be a any problem when playing music.
I don't think it matter at all what kind of music you play as the peak to average ratio is at least 10 to 1. What is killing OTL amplifiers is IMHO either too high idle dissipation or current runaway due to badly matched parallell connected tubes.
My own OTL is not running positive grids but I get about 25W output power with +- 150V. The dissipation at full output power is about 110W per tube but I have done a test where I let the amplifier run at full output power for 24 hours and there where no ill effects, everything got very hot of course but nothing broke and the anodes doesn't even get red at this power level so I believe that 60W is a very conservative spec, BTW I still use the same 4 output tubes as I used from the beginning, that is now more than 6 years of use without problems. The idle dissipation is ~30W per tube and at that level they have very long life time.
Regards Hans |
|
|
| Alastair E |
| quote: | Originally posted by tubetvr
Hi Alastair,
Glad to hear that it works well, and it shouldn't be a any problem when playing music.
I don't think it matter at all what kind of music you play as the peak to average ratio is at least 10 to 1. What is killing OTL amplifiers is IMHO either too high idle dissipation or current runaway due to badly matched parallell connected tubes.
My own OTL is not running positive grids but I get about 25W output power with +- 150V. The dissipation at full output power is about 110W per tube but I have done a test where I let the amplifier run at full output power for 24 hours and there where no ill effects, everything got very hot of course but nothing broke and the anodes doesn't even get red at this power level so I believe that 60W is a very conservative spec, BTW I still use the same 4 output tubes as I used from the beginning, that is now more than 6 years of use without problems. The idle dissipation is ~30W per tube and at that level they have very long life time.
Regards Hans |
I'm of the same opinion concerning the reliability of the '33 in 'home-build/audiophile amps....
The average 'home designed'/built OTL seems to run Very high idle dissipations, and Very HOT in order to reduce O/P impedances...Seems a lot of guys consider this to be the Main factor above all in their otherwise good designs The trend for very low or No neg feedback demands that they need high currents to lower O/P impedances....
(Ive seen on the 'web, a guy running at 400 mA and 120V on his 6C33 OTL.......The way I look at it, he only has 200mA 'headroom' before the max current rating is exceeded....A little ageing of the tube or thermal run-away even a slightly higher than normal line-volts and There-You Go,-- One or several Cooked Tubes--He didnt comment on reliability!....)
My personal opinion is that neg feedback is a good thing IF used where really needed and done 'sparingly'
The 'diff-amps' (6SL7GT) at the centre of my OTL allow me 100% neg feedback Just Round the O/P stages-Only where needed, making the O/P tubes just a buffer-stage, allowing me to use as low as only 50mA idle-current, although its usually set at around 75mA. Thats only 14 watts idle-dissipation-In comparison to running 200mA @ 150V which were the settings when I first constructed an OTL, the tubes practically stay 'cold'! I havent tested the O/P Impedance, something I forgot to do yesterday--Can't be too high though, Did a quick frequency sweep, get flat response from 12Hz to 300 odd KHz-Happy with that!....
The 6C33C-B has got to be one of my favorite valves too Built like a brick-outhouse, pretty good sounding very reliable if used with care and best of all, DIRT CHEAP !
The current pair in my bread-board, have been used and REALLY abused many times both in SET designs (Running VERY hot in those...) and other OTL experiments Had the pair glowing orange in the anodes a few times during experimentation when some MOSFETS failed in the drive CCT of a previous design I was convinced they would be toast, but no, they battle on just great. with little or no apparent harm:smash:--I would Love to see a 300B that would do that! :whazzat: |
|
|
| KimBOlesen |
Hi
Can you post a schematic of your OTL, or a link? Think OTL will be my next project.
Best regards
Kim Olesen |
|
|
| Alastair E |
Sorry not to have responded sooner...
I dont actually have a diagram in 'electronic' format for this one yet. Its gone through a few alterations since I first designed it, so I wanted to get it 'all sorted' first. Ive loads of paper scrap diagrams for it though!
If you can find it on the Web, Look for the 'Allan Kimmel Sweep-Tube Monoblock' OTL. It was quite easy to find but appears to only be on an obsucure French site, or was....
Allan Kimmel used EL509 beam power-tetrodes in his design and drove them partly by cathode-follower and Mosfet, which he called 'super-triode' mode. The cathode-follower drove G1 and Mosfet drove G2. The mosfet also bootstrapped the driver and half the diff-amp--I did away with all that, and replaced the PL/EL509 with the 6C33C.....
Although my amp uses a completely different front-end, and phase-splitter, the 'heart' of it, the 'diff-amp' pair before the drivers is much the same. I have deleted the mosfet bootstrap to the drivers, and drive the 6C33C with cathode-followers. I used 7044 tubes for the drivers, but 5786 tubes work just as well I used 6SL7 as the diff-amps, and 6J5 as phase-splitter, following the RCA 47K/47K concertina type design. My input stage is with a pair of Nuvistor 7586 tubes in Mu stage. The 'circlotron' rail supplies are 190V, the diff-amps and input stages run from 350V. There is also a -170V to the diff-amp and driver-stage. The driver has its +ve supply from the appropriate circlotron rail. Bias is set by altering the grid volts on the driver tubes. Drivers are directly coupled to the 6C33C and capacitively coupled to the diff-amp.
Most alterations and updates I wanted to try have been done, and Im pretty happy with the design and sound. As soon as I get the time, I will commit it to computer and post it. I would like the feedback/comments from some of the more experienced guys here...
(I know most will probably say its so hideously complicated for an OTL...)
Im currently experimenting using the 6C41C in place of the '33C, as I saw 10 of these on fleabay for a pittance, and thought it would be interesting to experiment....
The power is well down over the bigger valve, but has its merits. These things are even cheaper again than the 6C33C....
Initially, they sounded pretty 'rough/coarse' Not at all like you would expect 'half a 6C33C' would sound.....
Now, after some hours of use they seem to have mellowed, and beginning to sound good. There have been various comments along the same lines, that the 6C41C is not a good audio tube. but I have not seen much in the way of OTL types with this tube. Mostly its been used for SE amps. I believe Bas Horneman has built an SE and is pretty pleased with the results though...
Although billed as 'half a 6C33c' I wouldnt say they were, More like a third, even though the ratings appear to be half the 6C33C.... |
|
|
|
