Thanks, I am quite deep in this topic. I have also spent time measuring and plotting impedance curves in double logarifmic scale, with various shunts. Also I have good equipment for measuring dielectric lossed of capacitors (immittance meter with down to 0,01 mOhms accuracy).
One thing from measurements is valid - the lower dielectric losses angle of a cap - the better sound. A cluster of small Low-Esr caps intended for computer boards (speced at 100 kHz) measures in 2-5 kHz range and sounds better than any brand big (22000uF or 2x10000uF Rifa or BHC) electrolytic. Just try to listen.
Also when I use cluster and put 100uF polypropylene shunt I do not see any deeps and bumps on measured impedance vs frequency curve.
Only this solution allowed me to get close to 300B SE amp vitality in sound.
Last edited:
Hi VladimirK
Im impressed with your design and handwork🙂
I have power follower 99 also but not modified. For preamp im using No compromise tube OTL headphone preamplifier (both designed by Andrea Ciuffoli)
I would like to build another PF99 and tube active crossover in hope to improve the bass region.
Do you have any suggestions about tube crossover schematics?
Im impressed with your design and handwork🙂
I have power follower 99 also but not modified. For preamp im using No compromise tube OTL headphone preamplifier (both designed by Andrea Ciuffoli)
I would like to build another PF99 and tube active crossover in hope to improve the bass region.
Do you have any suggestions about tube crossover schematics?
I understand and in general like an idea of tube crossover, but so happened that I have choosen not an easy way of high currents and power.
Unfortunately, I could not make a good suggestion for the crossover.
As for bass, with my follower version, and big PMC EB1i loudspeakers, the bass is excellently tight.
I see
Another problem is that PF99 doesnt have gain, so i cant build powerfull push-pull for bass region and use it in active setup with PF99.
Does your modified PF99 have gain?
And what do you think about this schematic
Another problem is that PF99 doesnt have gain, so i cant build powerfull push-pull for bass region and use it in active setup with PF99.
Does your modified PF99 have gain?
And what do you think about this schematic
For high enough amplification (20dB) I use transistor No NFB one-stage SE preamp with 100V power supply rail and 160mA idle current (a bit "crazy", but sounds excellently). Similar solution could be easily done with tubes.
The original Andrea's schematics, similar to various Zens by Nelson Pass, looks like a trial version, only to have a guess about sound.
For real life it could be equipped with more output caps, better transistors, higher idle current and power. These measures bring the device to a real high-end category.
The tube crossover design you mentioned seems to be quite good.
Last edited:
For output caps i use Sprague 4700Mf bypassed with Wima 22mf, and current is 3 Amp per chanell. The difference in your design as I can see is that you have two output devices in parallel and gain stage, also higher voltage and current of 3.8 A.
I cant see how you achieve the better damping factor.
I cant see how you achieve the better damping factor.
As you can see in the picture (post #7, curve 1) even with my 25000uF output cap we have slight growth of output impedance at 25 Hz. With 4700uF, impedance module and phase angle growth will be very noticable.
For No compromise solution, I would use 30000uF output cap, being a cluster of small Low-Esr caps + 100uF shunt.
Thanks VladimirK
Ill try what you suggest.
I didnt knew that output caps ahave influence on damping factor.
Ill try what you suggest.
I didnt knew that output caps ahave influence on damping factor.
they do as they raise impedance with lower frequency and do this by limiting current flow...like adding a variable resistor between the amplifier and loudspeaker...
Exactly, and since loudspeakers manufacturers fine tune their products for almost constant and low impedance modulus in the bass region (almost all mass produced amplifiers do this), SE schematics with output caps 10000uf will not cause problem only in biamping, for mid and high ranges.
However, SE approach, apart of sound quality, makes the design very reliable. In my case, I had occasionally my amp working almost one hour with short circuted one channel (during burning-in period), and fortunately nothing happened to output caps and transistors.
Last edited:
What about balancing the circuit for +- operation and omitting the output caps....???
quite sure it can be done with out sacrificing the SE nature and maintaining the HF mosfets....
i have for Raidho manufactured 200 uF copper infused film capacitors...i use these for parallel caps in base filter networks....and for bypass in the midrange....caps for large signal transfer are insanely critical..and for sure the weak point of your amplifier...
quite sure it can be done with out sacrificing the SE nature and maintaining the HF mosfets....
i have for Raidho manufactured 200 uF copper infused film capacitors...i use these for parallel caps in base filter networks....and for bypass in the midrange....caps for large signal transfer are insanely critical..and for sure the weak point of your amplifier...
Proceeding from general intuition, seems that you are right, but actually the only industrial system, that I heared sounded as good as this my solution, was MBL9007 (monoblock configuration) + Rockport Technology speakers. I spend few days in Singapore (The Adelphi) listening to dozens of high-end systems, including tube ones. Bought XRCD K2U discs for listening.
Also I have 300B SE tube integrated amp from Eraudio.
I would say, 99% of designs without output cap sounded worse than this follower. Not forget 1000 times repeated argument, absence of output cap does not mean that PS electrolytics are out of game.
Last edited:
i think the beautiful is not from the output-cap but from the simple single ended design...This is like some of the low wattage simple PASS designs...and the use of the good mossfet devices...
but making it with balanced supplies is by no means simpler.. and one can with some right argue that there still will be caps in the signal path....as the power supply caps are somewhere in the loop...
but making it with balanced supplies is by no means simpler.. and one can with some right argue that there still will be caps in the signal path....as the power supply caps are somewhere in the loop...
MiiB,
Something else: have you tried to drive your nice speaker drivers from a very high impedance source?
Absolutely right.
Absolutely right, also, don´t forget the even worse impact of the rectifying diodes.
Something else: have you tried to drive your nice speaker drivers from a very high impedance source?
Only tried Zen4 with 1,6 Ohms output impedance, for PMC EB1i rather low output impedance is mandatory. However, with 300B SE amp (with custom made output transformers, output impedance of the amp is 0,6 Ohms) bass is not bad at all.
But with present follower it is definitely more tight and better.
WuYit
my speakers Raidho Acoustics are quite suitable for current amplifiers..because of the way the filters are made..they work with current diversion..instead of impedance raising and thus current limiting filters...A quite different approach that has the clear advantage that in the crossover region the drivers share current..and thus in locked phase wise together...
I've had SET's .. 805's and 211...and they do have some magic...but so does the VLT 450's or the old Adyton opera.. amplifier...I personally think it has a lot to do with the execution of the design...more than the actual design topology...
my speakers Raidho Acoustics are quite suitable for current amplifiers..because of the way the filters are made..they work with current diversion..instead of impedance raising and thus current limiting filters...A quite different approach that has the clear advantage that in the crossover region the drivers share current..and thus in locked phase wise together...
I've had SET's .. 805's and 211...and they do have some magic...but so does the VLT 450's or the old Adyton opera.. amplifier...I personally think it has a lot to do with the execution of the design...more than the actual design topology...
I agree 100%, design topology contribute to final sound only partially. And moreover, among design topologies, it seems enough to keep only some few rules of thumb, and to concentrate on parts and device execution.
For transistor schematics, first my rule is to keep electrolytics fare from important signal paths. If one has to use them nevetheless, like the output cap in SE design, then pay special attention to their arrangement.
Another rule is try to keep current consumption from PS as constant as possible, without excessive complexity.
Last edited:
VladimirK,
the capacitances of RD100HHF1 are not specified, but cannot be small.
Why would the audio signal bypass any eligible path of travel?
How could both the input and output signal bypass C3/C4?
the capacitances of RD100HHF1 are not specified, but cannot be small.
It is possible to make pecise statements about foreseeable things.
Luckily, the definitions are readily provided by others through many lifetimes of work.
Hopefully, Andrea will find an opportunity to answer these questions:
Why would the audio signal bypass any eligible path of travel?
How could both the input and output signal bypass C3/C4?
The capacitances are well specified in datasheet's figures.
As for Andrea's answer, in his place I just would make reference to his site Audio Design Guide
He wrote very clear, what is the difference between his follower and, for instance, Pavel Macura follower. Not needed to waist storage space.
Saying about electrolytics in signal path, I clearly refer to NO SIGNAL THROUGH POWER SUPPLY electrolytics, not output ones. The last ones must be "cooked" very carefully, and everybody understands that. However, it is not the case with PS electrolytics, which most of people believe have minor effect on sound. Noway, the effect very strong.
As for your disign philosophy, I understand that you believe, that good sound can be disigned theoretically, playing with models and books. I would say, this is correct for PA amp, that must provide power and reliability. But not for High-End. There is few available in schematics theory leading to genuine high-end sound, IMHO.
- Status
- Not open for further replies.