The 2K2 at the bottom of the SLCF, if 15V on the Gate, then you should have approx 11.5V across the 2K2 and that is what sets the current. Do 11.5/2200 = 5.2mA and if you decrease that to 1K2 and you will get 9.5mA and 1K 11.5mA etc.
Once you have set the current, adjust R4 (shown as 680K) to drop around 20V. That should always be done anyway. If in doubt and you get the circuit working, you can always post the voltages here and then we can tweak it. I assume you are around 250V HT, then the input on the Follower Grid should be roughly half that value. About 20V drop across R4 and hence also across MPSA29, assuming you have already see R7 to give you near 10mA.
Studying further the FVP6A circuit, an ignorant question came up: How the gain of the T7 gain stage be 5-6X when μ of T7 is 60?
OK, I see. It's because the cathode current sink is in parallel with the cathode resistor.
What I don't get is how this configuration works and how the actual gain of this circuit is set.
The T7 circuit gain can be adjusted where it says 2.6V and the 2K2 cathode resistor, increase value to decrease gain and decrease value to increase gain.
After that, adjust 2K7 to get you an Anode voltage that is close to half the HT rail voltage. We have 250V HT and hence 125-130V is about right. The circuit can work with as low as 200V, but in that case, adjust for around 90V. But 200V would be an absolute minimum, IMO.
Next, let us consider changes to the SLCF (unity gain follower). If you change from T7 here to a 6922 and want to increase the current, then decrease 2K2 value will increase current. If you want 10mA, then around 1K2 will achieve that.
The final remaining adjustment is the 560K and adjust this value, so that there is around 20-25V drop across MPSA29. This adjustment should always be done, regardless of which tube you choose.
Hope that helps.
Hi Joe,
Thank you for your response.
I was asking about the FVP6A version, the one with the -15V supply for the current sink. Could you please give me the relevant instructions for that circuit?
Thank you!
Thank you for your response.
I was asking about the FVP6A version, the one with the -15V supply for the current sink. Could you please give me the relevant instructions for that circuit?
Thank you!
OK.
Pretty much the same with regards to SLCF.
To change T7 gain stage, change the 6K8 cathode value. Increase gain by using a lower value and vice versa.
The problem now is setting the +130V as that will change. Hence the current source needs to change. Now adjust the current source and "???" value, or whatever controls the current in the current source you are using. Tweak to 125-130V when HT rail is 250V.\
BTW, this is similar to Allen's SVP circuit.
Pretty much the same with regards to SLCF.
To change T7 gain stage, change the 6K8 cathode value. Increase gain by using a lower value and vice versa.
The problem now is setting the +130V as that will change. Hence the current source needs to change. Now adjust the current source and "???" value, or whatever controls the current in the current source you are using. Tweak to 125-130V when HT rail is 250V.\
BTW, this is similar to Allen's SVP circuit.
Could someone else comment on this? I'd like to build a full function pre with FVP-5 phono and FVP-6B line.
I'm reading up on inverse RIAA and trying to figure out how the circuit would be measured and tweaked to minimize deviation on the response.
Are there updated values per the original schematic which highlights these resistors with a marked box and clearly says they may need to be adjusted.
The circuit is widely promoted as "state of the art" for its time. Sarunas in what way does it fail?
A square wave test failed. Some pictures you can see here:
https://www.diyaudio.com/community/threads/vacuum-state-rtp3c.151733/page-46
I got the same results with LTSpice simulation and with practical circuit. Unfortunately nobody commented this fact 🙁 There is a lot of builders of FVP5 but strange that nobody tested it.
https://www.diyaudio.com/community/threads/vacuum-state-rtp3c.151733/page-46
I got the same results with LTSpice simulation and with practical circuit. Unfortunately nobody commented this fact 🙁 There is a lot of builders of FVP5 but strange that nobody tested it.
I have a copy of Secrets of Phono in which Allen details the various points about the FVP5 phono design. It suggests the R3, 1K4, in series with 3n3 is the resistor that controls the high frequency boost. The article suggests matching all RIAA caps to better than 0.1% and that the whole response will almost certainly have to be tweaked using the indicated resistors. I wonder what the simulated result would look like increasing or decreasing this R3 value. Or even omitting it all together as I've also come across comments saying that the claim of this cutter head protection is incorrect.
To quote from section 4 of the arcticle. All credit to Allen Wright "Secrets of the Phono Stage"
And if I understand correctly an overshoot in square wave response can be related to excessive and uncontrolled high frequency response.
To quote from section 4 of the arcticle. All credit to Allen Wright "Secrets of the Phono Stage"
And if I understand correctly an overshoot in square wave response can be related to excessive and uncontrolled high frequency response.
I missed the final part of that quote.
"c/ Fit it in series with the cap—and tweak for sonic satisfaction and exact upper octave ch to ch balance. N.B. It’s phase accuracy across all four bands that gives you that real life image the bottom feeders say don’t exist!"
"c/ Fit it in series with the cap—and tweak for sonic satisfaction and exact upper octave ch to ch balance. N.B. It’s phase accuracy across all four bands that gives you that real life image the bottom feeders say don’t exist!"
Probably this overshooting is because of this missing time constant. Inverse RIAA doesn’t have this time constant and probably it is why impossible to reproduce clear square wave. I am trying to understand how Allen came to these RIAA values and if problem is with this fourth time constant then you should get square wave without it. Unfortunately Allen can not comment it but maybe Joe can. I believe that nothing is wrong with FVP5 circuit. Allen and all other team members worked hard and with passion. And just a lack of knowledge rises such questions.
The quote I posted seems to have all the necessary info. I am not that familiar with LTspice. Could you run the sim again without the 1k4 in play or alter its value and see if a better shape square could be obtained?
Why do you say "failed"? How do you define a "correct" response to a square wave?
I am interested to know how you applied the square wave signal. Did you apply it directly to the phono input, or is it equalised with an inverse RIAA function? I assume the latter is the case: the fact that the standard three-pole equalisation is not the same as Allen's four-pole function will guarantee you will never see a perfect square wave on your scope.
Alex
Sure with inverse RIAA, just I used slightly different circuit as Allen used. Just different R and C values but it does the same job. And it "failed" for me. I said that with circuit probably everything is okay just lack of knowledge rises such questions. And I would like to find an answer. Allen also said "Input a square wave via precise reverse RIAA network and then tweak the relevant resistors to get the best possible square wave out..." So, I decided that in any way you should get more or less correct square wave. 1.4k resistor definitely plays a role for 50kHz zero line. And this overshooting is probably just interpretation of data. I understand that RIAA filter is split int two parts 50Hz+500Hz (3180uS+318uS) and 2122Hz+50kHz (75uS+3.18uS). Also I want to understand if this overshooting is okay or needs some more precise tuning.
And maybe for others will be interesting the article about square wave testing and data interpretation:
https://sound-au.com/articles/squarewave.htm
And maybe for others will be interesting the article about square wave testing and data interpretation:
https://sound-au.com/articles/squarewave.htm
Did some more tests with inverse RIAA. I have three circuits for it, first one is original suggested by Reg Williamson (I built this one), second one is suggested by P.Lipshitz and used by Allen and last one is published by Merlin Blencowe with zero pole at 100kHz but not at 50Khz.
And all are almost the same, just some difference in output signal level and inverse RIAA with 100kHz zero pole gets flat from this point as stated. But others also are more or less flat from that point.
Square wave test results with these filters are slightly different. With first two they are identical but with Merlin Blencowe filter spikes are a little bit lower:
So, square wave form depends from inverse RIAA filter. If 1.4k resistor is simply shorted there no more overshooting but wave also not perfectly square.
And all are almost the same, just some difference in output signal level and inverse RIAA with 100kHz zero pole gets flat from this point as stated. But others also are more or less flat from that point.
Square wave test results with these filters are slightly different. With first two they are identical but with Merlin Blencowe filter spikes are a little bit lower:
So, square wave form depends from inverse RIAA filter. If 1.4k resistor is simply shorted there no more overshooting but wave also not perfectly square.
Probably question should be asked in another way. I just found JLTi test with a 100kHz square wave:
https://www.customanalogue.com/jlti_phono_stage.htm
Well it is solid state but in the picture perfect square wave. So, question would be how to perform square wave test correctly?
https://www.customanalogue.com/jlti_phono_stage.htm
Well it is solid state but in the picture perfect square wave. So, question would be how to perform square wave test correctly?
If you want to test the circuit without the high-frequency pole, you should test with the 1.4k resistor disconnected from the 3n3 capacitor, rather than shorting the resistor.
Alex
