F5Turbo Illustrated Build Guide

Excuse me if I missed a post about this, this is a very large thread. On the first post there is a caption "If using single-ended (RCA) inputs remember to add this link at the labeled spot." Is this jumper only for bridged operation or is there a way to configure for balanced input?
Second question, "The output boards are very straightforward, the thing to remember is to have the N channel on the right, which will make the hookups rational as well as placing the thermistor on the outside."
It seems to me that all of the connections are symetrical between the N and P boards with the driver board except for the thermistor.
Thanks for all of the great info posted.
 
A couple of other questions.
The gain stage board has 4 ground connections, I need 1 for the N channel, 1 for the P channel, 1 for the power return, 1 for the speaker return and 1 for the input return. That equals 5. There is only 4 connections on the gain stage board. What is the recommended way of connecting all of these grounds?
For the output connection, can I use the output and ground connection at the far end of the channel board that is closest to the rear panel or do I need to use the more central output from the gain stage board?
Connecting from the closest N or P channel board will make the output connection shorter but may introduce an imbalance between the P and N channels due to resistive losses in the board traces.
Is this a real or imagined concern?
 
What is the voltage across R5? Does it change when you turn P1? If turning P1 to its max results in no significant increase in voltage then P1 has likely gone bad.

What about the P2 side voltages?


nash

Nash, that was a bingo...thanks a lot...I found that why soever I have had years a go put the potis back into start position, so no current was drawn...stupid to overlook such a simple root cause...So i re-initiatilized the whole amp and rebalanced as well P3 with my new wonderful vc-80 milliohm meter...and setup was a breeze.

Only thing: cant get more than 0.366V across the 0.5ohm resistor, that is where the potis max oit, temperature at 48 degree...any hints how to raise this a bit further ? Is this a thing of the power mosfet selection or the input fets ?

By the way: I omitted the grid stopper resistors of 100ohm in front of the power mosfets and put some little ferrite breads (ft-3, tme.eu) in which have a resistance of 1k at 10mhz. So, for our NF it looks like a piece of copper wire (which it is) and for HF its an even bigger resistor. Soundwise, I used this in all of my tube gear and it was always an improvement towards more open, clear, dynamic and natural sounding...I guess as a result of avoiding that a new material mix gets into the way of the music signal (think about the construction of a resistor with its leads, end caps, film construction etc), a piece of wire always sound better than a resistor.

Power amp ran for 3hours with them, no negative temperature or offset behavior noted...sound test to be done next...
 
Last edited:
Hey, guys! So, I’ve been a member of diy audio for over a year, but I’ve never posted anything. I’m a long time fan of Nelson Pass and his creations in general. He’s a total rock star as far as I’m concerned.
I am gearing up to eventually buy the PCB’s, chassis, etc. to build an F5 Turbo V2 amp. I’ve done a lot of research and read pretty much any thread on the subject, but I’ve still got a quick question for anyone that could possibly help out.
What is the difference between an LSK170X-1 and the other LSK170 JFET’s produced by Linear Systems? Would it make sense to buy a bunch of the X-1’s to match them? Are those the highest quality of the bunch? As far as I can see, the specs are the same.
Thanks in advance to everyone at diy audio. It’s clear there’s a lot of great people here.
 
I bought 2 sets lsk170B/lsj74B and mosfets from the diystore and finished F5 turbo v2 few days ago. Everything went perfect.

4 pair fets have following Idss at 9VDC

(1)10.80/10.80 ma
(2) 10.78/10.78 ma
(3) 11.61/11.54 ma
(4) 10.88/10.65 ma

at first I used (3) and (4) for the amp, but found it is very tough to get thermal stability of Iq and Offset with (4). I ended up to use (1) to replace (4). Two channels got very good thermal stability. Power on at 20C with an offset of -80mv to near 0mv after 20min to +20mv in few hours (Iq=0.6A per device, 1.2A per channel)

what I'd like to say is that better buy them from the store, you could have good chance to make everything work at the first time.
 
wwwtttwww Thanks so much! That’s all I needed to know. I figured just buying matched sets from the store would be more advantageous than matching them myself, and if the “B’s” worked for you and most everyone else, I’m sure they’ll work for me too.

Again, thank you for your time and helping clear that up for me, and congrats on your new F5T v2!
Much appreciated!
- Nick
 
I finished upgrading my F5T monoblocks by raising PSU voltage from about 30-32VDC (depend to the time of day with mains fluctuation) to 42-43VDC. Personal thanks to 6L6 for support. The design of my speakers requires about 70W per channel to reach a full of potential of these and my 32VDC setup gave me 56W, measured. BIAS was set to 470mV and it is remained same with 42VDC PSU. I have floor standing Adagios made by Acoustic Zen and I got that info from designer-owner of that company. I didn’t measure power output yet on new setup, but volume knob of my preamp moved down by about 20% for the same loudness. Roughly, I assume I have about 85W or bit more (again, it is general feeling without measurements).

Now, I have several F5 amp parts that I don’t need any more and these might help to someone to reduce BOM. Please PM if you would be interesting to buy these.
It is about 2 years old pats. I’ll go for 35% off from original pricing plus shipping to US.
So, I have 2 transformers made and sold by AnTek AN-8425. And I have 16 MUR3020WT which I used as rectifiers. I moved to Vishay’s HEXFreds with my new setup.
 
Last edited:
Here is some images:
 

Attachments

  • 42V_1.jpg
    42V_1.jpg
    952.2 KB · Views: 411
  • 42V_2.jpg
    42V_2.jpg
    353.6 KB · Views: 397
  • 42V_3.jpg
    42V_3.jpg
    388.5 KB · Views: 390
  • 42V_4.jpg
    42V_4.jpg
    831 KB · Views: 381
And here is mentioned above parts for DYI that I don't need anymore. It might help to someone to save a bit on BOM.
 

Attachments

  • MUR_2.JPG
    MUR_2.JPG
    728.5 KB · Views: 141
  • AN_8425_.JPG
    AN_8425_.JPG
    725.8 KB · Views: 105
  • AN_8425_5.JPG
    AN_8425_5.JPG
    761.9 KB · Views: 99
  • AN_8425_4.JPG
    AN_8425_4.JPG
    693.4 KB · Views: 101
  • AN_8425_2.JPG
    AN_8425_2.JPG
    810.9 KB · Views: 137
  • AN_8425_1.jpg
    AN_8425_1.jpg
    628.6 KB · Views: 149
  • MUR_1.jpg
    MUR_1.jpg
    496.4 KB · Views: 145
May I ask howyou like the hexfreds vs. the mur? I thoight about this as well...

It is hard to make “clean” comparison since I changed toroids, snubbers, P1/P2/P3 positions for BIAS/OFFSET and distortion calibration, in addition to rectifiers. Anyway, it is too early to judge since I did play it for 3 hours after upgrade. However, and I can tell you already, I like current SQ more since it is way more punchy, faster and “crescendo” in classics behaves correctly now. I had some sort of sag effect with 32V PSU. I assume it is SPL and general drivers speed related of your speakers. My speakers are required more power than 56W on 6R nominal impedance.
Another side is cost vs. SQ improvement… IDK, but if you are BOM tighten for Gross Margin pareto, then big NO. It is way too pricey. I paid about $160 just for my rectifiers from Mouser. But, I have no issue with overkill or over engineering since I’m not in audio high-end business of sales. I made it for myself only and do not care about additional 10% $ for top-notch quality parts.
Alex
 
Last edited:
I guess you could actually even try a choke-inpit configuration with your higher voltage transformers ? Typically choke input sounds much mkre detailed and refined than clc or crc...which is due to less stress for the transformer and less hf-hammerings of the diodes as current spikes are much, much lower...I willtry this myself, but need new xformers...
 
I guess you could actually even try a choke-inpit configuration with your higher voltage transformers ? Typically choke input sounds much mkre detailed and refined than clc or crc...which is due to less stress for the transformer and less hf-hammerings of the diodes as current spikes are much, much lower...I willtry this myself, but need new xformers...


For choke-input configuration..
Do you mean to drop DC choke directly between rectifiers and first filter Cap (like LC filter)?
Yes, and it will reduce V-ripple further (can be simulated on Duncan's PSU tool). but in other hand and for something that can handle needed current, it will be relatively low inductance (otherwise, it will be heavy artillery size device) and must be also low impedance too (otherwise, it will create unnecessary and significant V-drop for DC). It also may introduce some sort of sag effect too. I know it is quite commonly seen on valves PSUs which known for sagging in high current demand too quickly…
 
I see your point, but I must go off from CIP mode (Continuous Improvement Process). Second, I simulated my circuit with MIN inductance choke (based on Jim Hagerman) + 100% for an extra inductance and I got 3mV V-ripple drop. Not that significant to spend on probably expansive potted chokes. My setup has less then 100mV voltage ripple, calculated (not measured, CLC architecture, 44k uF and then 3.3mH air-core L, and then another 44k uF C) and I'm not sure if I would go for more elegant solution called C-Multiplier (for example explained by Rod Elliott) to improve it. Dave also discussed it on his EEBlog on YouTube.
 
Last edited: