On my list of solutions, it wouldn't be on top either. I don't like the way boxed speakers sound. But it is worth trying if for any reason one has to use boxes.
The best solution is low moving mass.
The 12" Magnavox drivers (#1 in post 19) sound remarkably well WITHOUT ANY ENCLOSIRE AT ALL wih many kinds of classical music - instrumentsl, vocal, some orchestral. Even relatively small open baffle gives them authority.
Last edited:
Member
Joined 2009
Paid Member
Jeff Larson had a no feedback pentode spud amp, the schematic is hard to find because it was sold commercially for awhile, nevertheless I did track it down, see post 22 in this thread (spoiler alert: I found feedback was needed dor decent sound, by YMMV).
https://www.diyaudio.com/community/threads/yukon-gold-a-spud-amp.280167/
https://www.diyaudio.com/community/threads/yukon-gold-a-spud-amp.280167/
Importance of voice coil inductance, Le
bondini suggested in post #6 to make an experimental 6BQ5 SE amp driving a Pioneer full range speaker. I guess it is a $17 driver from Parts Express. It is 93 dB sensitive, 8 Ohm, 8" driver. One thing bothersome about it is Le of 1.3 mH. This will be reflected to 10K:8R OPT primary as 1.6 H, or about 4 K of inductive plate load at the lower midband frequency of 400 Hz. This is bad, pentodes do not like reactive loads.
Looking at something more upscale, a 8" Lowther, also 8 Ohm, has Le of 36 uH ( microHenries). This is much more suitable as pentode load.
What's going on here? Why two - sort of - similar drivers have Le difference of almost 2 orders of magnitude? Anybody have a clue?
Hint: the Lowther was carefully designed for low Le. The Pioneer was just slapped together without any knowledge or expertise.
bondini suggested in post #6 to make an experimental 6BQ5 SE amp driving a Pioneer full range speaker. I guess it is a $17 driver from Parts Express. It is 93 dB sensitive, 8 Ohm, 8" driver. One thing bothersome about it is Le of 1.3 mH. This will be reflected to 10K:8R OPT primary as 1.6 H, or about 4 K of inductive plate load at the lower midband frequency of 400 Hz. This is bad, pentodes do not like reactive loads.
Looking at something more upscale, a 8" Lowther, also 8 Ohm, has Le of 36 uH ( microHenries). This is much more suitable as pentode load.
What's going on here? Why two - sort of - similar drivers have Le difference of almost 2 orders of magnitude? Anybody have a clue?
Hint: the Lowther was carefully designed for low Le. The Pioneer was just slapped together without any knowledge or expertise.
And that is why Zobel RC networks exists for.
I would't be so dismissive. Pioneer fullrange was carefully designed for low cost and acceptable performance, which wouldn't be possible without deep knowledge and expertise. On the other hand, Lowther is outrageously expensive for its mediocre performance (typical of fuillrange drivers).
Anyway, I am still waiting for your answer on my question:
I will post a couple of graphs, need to figure how to extract them from big pdfs.
I stick to my guns that the Pioneer is crap. I will later explain why.
Let me suggest a vertical deflection amplifier pentode. One used as pentode in that service. One could probably build a decent pentode; type 12B4 underneath a MOSFET. Mind the safe operating area under linear/DC operating conditions here.
Have to second the variable capacitance of MOSFET's as being detrimental. Cascode-ing them, even for current regs is ever-so-useful.
Now to trot out the middle-aged cars of choice; those with engine code ALH, and transmission code O2J. IOW, Mk4 TDI's, with manual transmissions 🙂
cheers,
Douglas
Have to second the variable capacitance of MOSFET's as being detrimental. Cascode-ing them, even for current regs is ever-so-useful.
Now to trot out the middle-aged cars of choice; those with engine code ALH, and transmission code O2J. IOW, Mk4 TDI's, with manual transmissions 🙂
cheers,
Douglas
Of course , mine is from first 1997 series MK4, TDI /110HP with excellent ALH engine and 5 speed manual tr.box and with full internal equipment ,still going strong without of any sign of fatigue , any way
for just in case I still preserve mine from 1981 MK1/ 1,6 D which also going strong 🙂7233 will work with even less voltage between plate and cathode than a 12B4. rock on... 🙂
Douglas
Douglas
Sorry for this bad picture. It is distortion plot of Electrovoice DL12X at 10% rated power. Complete datasheet is readily available on the web.
This is how speaker distortion should be defined - not at some arbitrary frequency, but across the whole driver's range. Few manufacturers provide this information, for a reason - there is nothing to boast about. DL12 THD averaged across its range is about 3%, but this number is deceptive because at some frequencies it is much higher. Who cares if distortion is low at 120 Hz if in the 1 kHz area it is high? The quality is not defined by the average, it is defined by the worst. And the worst is 10 dB higher than the average.
I am posting from my phone, and I am complete computer klutz, so I couldn't extract graphs from a pdf file for posting. But what I want to post about pentode distortion is readily available to everyone. On the page 13 of the 807 39-page data sheet, THD, almost exclusively 2H and 3H, is below 1% at 10W output and Ua=600 V. Page 25 shows that at 600 V, the 5th harmonic only begins to show up at power output of 30 W. Extrapolating from 500 and 600 V data (pp.11, 13), THD at 10 W 800 V should be well below 0.5%.
So, pentode stage distortion could be an order of magnitude lower than speaker distortion, and without mythical "nasty high order harmonics".
So, pentode stage distortion could be an order of magnitude lower than speaker distortion, and without mythical "nasty high order harmonics".
No, sorry, I cannot. The Pioneer is likely the best commercially available full range driver at $17. But if you insist on such low budget for a speaker, you should refrain from making statements on sound quality of different amplifier topologies.
Zobel indeed provides a uniform quasi-resistive impedance seen by the amplifier output. But what speaker with Zobel network sees is second-order high frequency roll-off, compared to first order roll-off without Zobel. Thus, in the full-range speaker, Zobel would exacerbate high frequency response deterioration caused by high VC inductance. And, for any speaker claimed as full-range, high Le is unacceptable.
With this I wholeheartedly agree. By no means I would recommend a Lowther to anyone - it has conceptual design flaws. But it's low Le is a strong point of its design as a full-range driver.
Are you sure that measured distortion is coming from an amplifier built without negative feedback?
AFAIK all distortion graphs in the tube datasheets are measured with negative feedback (usually not specified how much dB feedback).
Better yet, could you please provide us with distortion measurements on a real 807 amplifier without negative feedback?
I can't find 39-page 807 datasheet anywhere, could you post a link, please?
I can only find the 43-page 807 datasheet from STC:
http://nekhbet.com/STC807.pdf
Here, on page 12 (page 13 on pdf file) presented is distortion measurement of a 807 push-pull class AB amplifier, with implicit negative feedback, with 600 V plate voltage. At 10 W output H3 = 0.8%, at 20 W output H3 = 0.8%, and at 30 W output H3 = 1%. Why this high distortion should be better than essentially zero distortion coming from an amplifier with negative feedback?
Yes, I know the current drive reduces the loudspeaker distortion, but could you, please, provide us with real-world distortion measurements of a current driven loudspeaker at SPL = 95 dB/1m, for example? Otherwise, you are just speculating.
Here is measured distortion (by erinsaudiocorner.com) of a good quality loudspeaker (Selah Audio Purezza) costing $2900 per pair back in 2020, measured with solid state amplifier with negative feedback:
Above 90 Hz, at SPL=95dB/1m distortion from H3 is below 0.2%.
Even at SPL=99 dB/1m distortion from H3 is well below 0.3%.
Another good quality loudspeaker (Revel F208) costing $5500 per pair now, measured (by audiosciencereview.com) with solid state amplifier with negative feedback, at SPL=96dB/1m:
From 200 Hz to 4.5 kHz distortion from H3 is below 0.2%.
Moral of the story: why should we accept H3=0.8% from an 807 amplifier (with negative fedback!) alone, when we can have H3=0.2% combined distortion of SS amplifier (with NF) plus loudspeaker?
Last edited:
If you can't find 8" fullrange for the same price better than Pioneer, than you shouldn't write this:
?
Zobel network do not cause any change in the loudspeaker frequency response, when driven with low output impedance amplifier.
With current driven amplifier, Zobel network is helpful with the majority of the fullrange loudspeakers, which tend to have exaggerated high frequency output below 10 kHz.
Last edited:
Yep, the 7233 can be quite useful, but I will build first with the 6CZ5 🙂 because it is an interesting device, I can dust off and modify an existing circuit board, and because I have barely enough time as it is!!
As you can see, I am fighting the temptation to start YET ANOTHER experiment, this time a 7233+MOSFET=Fauxtode especially as I am very much a solid state earner-driver (I keep crashing them/blowing them up as I career in and out of the SOA). And, despite myself, here I sit typing "Do you have a suggested topology/circuit for such a beast?"
Turns out that I can resist anything but temptation ....