Hi thanks for your reply and help.
I have been looking at the link you put and reading up on the Internet about feedback loops.
It looks a bit more complicated than I first thought I was just hoping it would be a matter of taking the loop before the output transformers.
My speakers are already on the 4ohm taps.
They are monitor audio studio 12se 96db 3 way floor standers with metal dome tweeters and two 5 1/4 inch drivers.
I am new to DIY electronics and this is my first valve amp but I built my first valve dac about a year ago (coppied most of it from lampizator site).
Glad to help Dave.
As DF69 pointed out I meant that the output impedance rises when you drop the GNFB (global negative feedback). I'm not sure why he is interested in my bedtime however - he's a queer one!
Is 96dB the sensitivity of your speakers? Sounds a bit high TBH - I'd guess 85dB/watt.
Partial feedback is not a huge deal on your circuit TBH.
An externally hosted image should be here but it was not working when we last tested it.
You have what looks like a high gain SRPP input stage driving a LPT phase splitter into two cathode followers. It's a bit naughty that they use the same filament power for all lower and upper tubes but there you are.
So you could re-purpose the followers and KT88s into 4 class A partial feedback amplifiers (2 per channel), to give a dual/stereo bridged drive to the OPTs. Then you simply drive them from your LTP, perhaps with lowered gain from the input SRPP.
So you have all the bits to do the conversion with - you just need to significantly rewire the follower/driver triodes and feedback from the KT88 anodes to their cathodes. Job done. It's a very suitable amp for doing this to actually - lots of handy tubes just sitting there ready to use.
I do not think that removal of cathode followers is a good idea. Phase splitter does not like to work with variable load, which class AB power stage presents to it. And significant reduction of feedback is only possible if power tubes are biased solidly for class A operation.
Absolutely.
However the LTP will be driving a regular triode grid, and there will be more feedback - not less.
Hi you might be wright about my speakers I was just going by what he person told me.
The only specs I could find on them on the Internet said.
Monitor Audio studio 12
Gold dome tweeter
2 x13 cm aluminium and magnesium in ceramic sandwich drivers.
Frequency response 40hz-30khz +/-3db
sensivity (1w 1M) 90db
What was interesting it said 8ohm
RMS power 120/150w
Cost £1600 new
I tested my speakers with a meter and it showed 3.5 ohm perhaps I did it wrong.
It also said they was one model down from studio 20se
which was 30hz - 30khz 89db gold dome tweeter one 6 1/2 inch driver 200w rms.
Cost £1900.
Pleased when I see this because I picked mine up for £95.
They look really nice, and 90dB/W is pretty good.
An externally hosted image should be here but it was not working when we last tested it.
So if you pump in 8 watts that represents a doubling of sound pressure 3 times (1-2-4-8) so you'll get 90dB + 3 x 3dB = 99dB sound pressure - which is probably as loud as you'll ever need.
Your meter gives you the DC resistance of the bass driver(s), nothing more. Impedance is basically resistance at AC - and it usually varies quite a bit depending upon the frequency of that AC. If your drivers have a resonant frequency (they all do) then unless corrected the impedance rises dramatically at resonance because it takes so much less effort to drive something at resonance - it almost does it itself.
Sounds like an excellent deal - congrats!
Hi Globulator I think these mods are still above me at the moment.
I don't understand what LPT is.
I'm at the stage where I can copy other people's mods circuits and swap components.
I don't understand ohms law or the Maths involed.
The DAC I made was easy because the lampizator gave lots of photos with step by step explanations and mods.
It stated out as a behringer dac with tube output then I got a bit carried away.
Perhaps at a later stage I will try these mods when I understand more.
I think for now I will change the resistors as in the diagrammed you have put and feedback cap to 72mf.
Change the capacitors for better ones and put some switches so that I can switch feedback loop on or off.
I will keep watching incase someone else cleverer than me does the mod and I can copy it.
Well it's a good place to learn
. Long Tailed Pair = LTP. It's a pair of triodes with connected cathodes, so when you drive one of them the other leaps up and down in opposite phase. This therefore forms a 'phase splitter' - i.e. you put in a signal going down-up-down in, and from one anode you get a bigger down-up-down, and the other one you get a up-down-up. I.e. one side inverts the signal and one doesn't.
This allows a push-pull or 'bridged' amplifier to drive the speaker actively one way and then the other. If you imagine a very low frequency wave pushing and pulling your bass speaker cone in and out - the push comes from one side of the LTP and the pull from the other.
For a beginner the first thing is safety (read the top posts) and possibly adding a LED and 'bleed' resistor across your main power capacitors to ensure they are discharged. Seek advice to do this safely as they can hold a lethal charge for some time (weeks!).
So I'd prescribe a mild upgrade to all the main components as you suggest, as you learn more about stuff.
Ohm's law is merely a relation between voltage and current in the form of a fraction, and is R = V/I. Therefore if you wanted to connect a bleed resistor across a 400V source with a red LED (dropping 2V) that takes 5mA you'd have:
V = 400-2 = 398V. I = 5mA (5e-3A).
R = V/I = 398/0.005 = 79600ohms, closest real resistor = 81k.
Of course that resistor is now checked for voltage - most only go to 200V so you'd need 2 of them in series. Power dissipation P = V x I, so it would convert 398 x 0.005 = 1.99W into heat, which is quite a bit of heat - each resistor would need to be at least 1W rated and carefully placed in the air so they could cool.
With that in mind I'd experiment with a meter and see if I could use a bleeder resistor further down the amplifier resistor chain so you had less voltage and power to worry about. You can also work out discharge times too with capacitor equations.
As you can see the benefit of learning resistance, power and other simple equations allows you to plan out your modifications - I'm sure you'll pick it up quickly. They are also the only equations you need to learn to plot tube operating points. E.g the ECC88:
Tube operating curves show you the relation between grid/cathode voltage, cathode-anode voltage and current. For instance if you have a 200V supply ('rail') and want a 10mA current when the anode is at 100V you look up that and find the grid voltage is about -1.5V. You achieve this on the grid by attaching the grid to ground via a tiny current (big resistor like 470k - Rg) and put a resistor between the cathode and ground - Rk - to elevate the cathode.
An externally hosted image should be here but it was not working when we last tested it.
So a 1.5V cathode at 10mA = 1.5/0.010 = 150ohms on the cathode (Rk). The anode needs a resistor (Ra) to take 100V at 10mA so you need a 100/0.01 = 10k resistor there and your circuit is done (for that operating point).
Of course now you can work out the power of the resistor required etc - and choosing the best operating point is another question.
This is a page (there are lots!) about the above example I gave (with a different tube!) Designing Common-Cathode Triode Amplifiers
Also take a read of the tubelab.com articles - they are some of the best written I have seen and you'll soon learn a lot from them.
BTW which resistors were you planning to change and why?
Attachments
.gif){kind=link}
It is a good match for old (1950s - early 1960s) records of acoustical music. Small jazz bands sound is very good through it. Of cause it cannot be used as the only amplifier, but is a good compliment to high-power solid state amp (Classe, Krell etc.). Total cost was not really high - much less than if I decide to get amplifier from established high-end brand. So I can honestly call it "affordable".
Isn't this a great hobby
One thing is bugging me 'though: considering the valves you have there, the added value of the 845 is only a few extra dB of power. You get 15W now where just the 300B would give 8-10W.
The coupling cap reduces the 300B to drive the input capacity of the 845 only.
Direct or transformer coupling would allow the 300B to supply the 845 with plenty of grid current.
The 300B can be a great driver for transmitter outputs. ReinoutdV has an incredible amp where a 300B is driving a TB3-1000 (upgrade to an 833) in class A2.
Looking at tubelab's site he gets over 40W out of his 845 amp in class A2 using a SE 45 and his mosfet powerdrive. If your output trannys can handle that power, the rest of your amp can too.
One thing is bugging me 'though: considering the valves you have there, the added value of the 845 is only a few extra dB of power. You get 15W now where just the 300B would give 8-10W.
The coupling cap reduces the 300B to drive the input capacity of the 845 only.
Direct or transformer coupling would allow the 300B to supply the 845 with plenty of grid current.
The 300B can be a great driver for transmitter outputs. ReinoutdV has an incredible amp where a 300B is driving a TB3-1000 (upgrade to an 833) in class A2.
Looking at tubelab's site he gets over 40W out of his 845 amp in class A2 using a SE 45 and his mosfet powerdrive. If your output trannys can handle that power, the rest of your amp can too.
I have listening many amp in china where I am working, Yaqin in not my favourite brand as Melody or Consonance that like a lot, here in China they use a lot the 6j4p ( 6ac7) this tube have a very big impedance and normally violins and piano sound strident. I had one amp with this tube as preamp to drive the 6c33, and for me the sound was better when I change to 6sl7 paralleled but have more octal option as 6sn7 in cascade, I can no find more important things in your amp, depends of the quality of the OP, maybe need same NFB to correct the quality of the OP, you know that them make very beautiful apparent, but inside is not a Airtight.
When I buy a Chinese amp I change the POT normally is very bad, signal wires, coupled cap, but sure that th 6ac7 is gulty in 40%
Good luck
When I buy a Chinese amp I change the POT normally is very bad, signal wires, coupled cap, but sure that th 6ac7 is gulty in 40%
Good luck
As I said before 6SL7 would be a better option, but 6AC7 in triode mode is not bad either. I wouldn't use it in penthode mode in power amplifier though - it is better used when dealing with small signals.
As for limited power from 845 - if you apply the same criteria to 300B based amp, you likely get 6-8W of real power output. It is possible to get more from 845 in A2 mode, but it will require additional negative power source for driver. This is not easy to achieve without major rework of power supply circuit. Probably DC to DC converter can be used considering limited space under chassis. I might take this task, but will use different Chinese amp as base for that (Yaquin monoblocks maybe).
As for limited power from 845 - if you apply the same criteria to 300B based amp, you likely get 6-8W of real power output. It is possible to get more from 845 in A2 mode, but it will require additional negative power source for driver. This is not easy to achieve without major rework of power supply circuit. Probably DC to DC converter can be used considering limited space under chassis. I might take this task, but will use different Chinese amp as base for that (Yaquin monoblocks maybe).
Sorry I do not see the second page.
in chinese amp I have with little changes a good and cheap amp and in you case is easy , you can do a very good amp with your base:
Octal + 300b + 845 for me the question is the octal ???
think in which octal and how to implement SRPP, cascade, penthode strapped etc.
Me with the 6ac7 strapped could not drive the 6c33 at 210v
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.