Interesting - I didn't pay much attention to the FVP4 in the TPCB since Allen moved on from it, turning the voltage amplifier stage into a cascode and adding the CCS and the bootstrap devices to get to the SVP. I can see the attraction of the simple circuit, but Allen was never a fan of simplicity for its own sake. The cascode adds bandwidth to the VAS and the SLCF has higher linearity than the SRPP in the FVP4.
Alex
Alex
Having built the original FVP5 22 years ago and having happily lived with it as my main preamplifier, recently, I decided to built the SVP version (rather the FVP5A). My thought is to built 2 separate preamps: a phono preamp and a line preamp. The main reason for this decision is because I don't use my vinyl source that often and I find that it is a waste to have two more tubes working all the time the preamp is on, as they are aging without being used. The two preamps will be integrated, that means that each one will have its own PS.
So, handwound, you gave me the idea regarding the phono preamp, to use the 5842 tube instead of the 6922, as I have a bunch of them, mainly of the Raytheon brand.
So, my question to you is how you used the 5842s in you phono stage of SVP. Have you used the same passive components values as with the 6922 tubes? Could you please show or send a schematic, or at least quote the values of the resistors that you used on the two cathode coupled stages of the phono section?
Regards,
Evangelos
Trobbins
I just put 10k plate loads, and lower cathode resistors. I have about 220v input and about 90volts on the first plate, and about 110 on the second one.
I cannot remember values, but maybe 15ohms as the cathode on the input.
I left the riaa untouched kind of, but I did lower the signal cap values to work as a low pass filter. 0.015 for the first and 0.22 for the second.
I did test the original values for 6922 riaa and they were spot on. I never got to measured them on the 5842 build as the guy who has it loved it out of the box. Maybe I will see if the riaa was effected by the changes one day, but my way of measuring is archaic and time consuming
I have all signal in one box, but I can switch the phono off if required.
If you can get some Sylvania GB 5687 in the linestage it’s really worth it. Never tried the RCAs which are supposed to be the best.
I would love to see your build. As I am still working on my skills. I have young children so all my builds happen in my mind these days.
I did the same a long time ago, except I use the E88ccs (6dj8, 6922 etc family). I don' t play LPs that much so why waste the tube's
life. Besides that, if you perform design changes, I found it was easier and could still use the system while the phono stage was on the bench.
I never use solid state devices, if possible, in the signal path. I have found that even 1N914, 1N4841 diodes will bleed audiio even when
reversed biased, depending upon the impedances involved.
cheers
pos
Thank you for your reply. I will begin with the values stated and adjust accordingly.
Although I have a number of 5687 tubes, I was thinking of using the Russian 6N6P: it has slightly higher amplification factor and higher transconductance in relation to the 5687.
I might give a 5842 linestage a go, but it will have too much gain, but I like the idea of more transconductance.
The extra gain could be handy for the phono stage. Then just pad the linestage inputs with an attenuator
Too many tubes for the line stage!
By the way, searching my inventory for the 5842 tubes, I realized that I have a lot of Russian 6C45P tubes (an equivalent of the WE 437). I'm thinking of trying these instead of the 5842: a little higher μ (52 vs. 43) but much higher gm (45 vs. 25). Just a thought...
Something whithout sk170 like this << https://audioxpress.com/article/borbely-riaa-preamp-with-tubes-project-remembered-and-revisited>> for RIAA preamplifier !
Isn't over a db off a serious musical problem? Even a tenth or two of a db off is quite perceptible in a good system.
I am just wondering if you are setting your goals too low.
cheers
pos
Last edited:
I use the 6N6P in the linestage of my FVP5A. Works very well. Probably the 6H30 would be even better but not when you compare prices.
Thinking about it more thoroughly, it occurred to me that changing the tubes in the original FVP5A design could change significantly the equalization curve of the phono stage, as the gm and Rp of the upper tube of the first hybrid cascode stage affects its output impedance, which, in turn, is part of the RIAA network. So, changing to a different tube should need recalculation and readjustment of the RIAA filter. As this is beyond my knowledge, I would rather stick to the original design.
FVP: Four (or five) valve preamplifier, Allen and Joe's first single-ended preamp. This used enhancement mode MOSFETS in the SLCF output stage as bootstrap and CCS, where these were later replaced with 6922 valves.
SVP: Six valve preamplifier, basically the FVP5a (which already had six valves!).
Alex
Yes, this was a major advantage of the RTP3, where the anode resistance almost entirely determined the output impedance of the single-valve input stage that fed the RIAA network, so the frequency response of the latter hardly changed at all as the valve aged. This is probably true to some extent for the first (cascoded) stage in the SVP phono stage, but not the second one which is a simple common cathode amplifier.
Alex
On his latest edition (SVP-2), Allen offered a selection between high (18dB) and low (12 dB) gain for the inputs of the line stage, in order to cope with the high output digital sources.
I would also like to lower the gain of one of my line inputs (DAC), but without affecting my other line inputs (tuner etc). I understand that this can be done with a resistor network (voltage divider). My main question is what values of resistors should I use in order to drive the input attenuator (50 kΩ) without any problem. Any suggestions?
Evangelos
I would also like to lower the gain of one of my line inputs (DAC), but without affecting my other line inputs (tuner etc). I understand that this can be done with a resistor network (voltage divider). My main question is what values of resistors should I use in order to drive the input attenuator (50 kΩ) without any problem. Any suggestions?
Evangelos
I did this for two of the line level inputs on mine, attenuating them by 6dB (a factor of two). For some arbitrary reason I decided to keep the input impedance at 50K, so I made up a divider with 24K series and 47K shunt, but you can do this just a well by simply inserting a 47K series resistor between the input socket and the selector switch.
Alex
Hi kenev,
I agree on your too many tubes observation. And to my thinking one tube is enough for the line stage, and skip the 5842, 6C45Pi, etc high mu tubes, it then becomes a three valve preamp, skipping the SLCF altogether. With a 5687 and an output transformer like an LL1660/18mA, If that does not give enough gain for phono (sorry Allen, but line-out transformers are not against my religion) I am afraid the entire concept is flawed...if not, can use to advantage a Slagle AVC on the output, skipping the 50K Alps Black Beauty...just put the 50K (it's not really 50K based on the Black Beauties I've measured) back in the phono stage.
I've lived with a line stage in my system consisting of 5687 into LL1660/10mA for ~10 plus years and it still sounds good to my ears to this day. I know that the 5687 would be happier at 15 or 16 mA, rather than 9mA, and with LL1660/18mA, no problem.
Welcome to the TVP, that's Three Valve Preamp.
Best to all, Robert
Maybe, but if you read the TPCB, Allen explains quite clearly what the purpose of the upper and lower valves are in the SLCF. They are there to linearise the cathode follower by making it work much more closely to constant current and constant voltage conditions, without using any feedback loops.
It's a personal choice, of course, and the very low distortion of Allen's designs will make it sound different from most other circuitry out there. You may like the sound of a "naked" CF, but - to an extend dependent on your choice of valve and operating conditions - "Miss Piggy" will distort with real world loads.
Alex