FH9 XRK mod

@UncleMud
Your Amp looks fine, ok...the wiring is ok in my opinion. If the amp works fine and the sound is very good, let it remain as it is...
Fine to hear a opinion about the XRK pcbs from a member. I only saw pictures in the past. I need space at home... 🙄

@chermann
I think you are busy too...i have read about your problems with the Apex AA14. I am afraid i cant help....good luck to finish the amp and success to make it working.

why omg? ...its a M2Competition wih a little more power as usual...even more than the CS...but does not fly...🙂

Peter
Na oida....

Omg !!
 
Hi FH9 MOD -Friends,
our parts were already delivered, now it's time to drill holes... in MS... 🙂

uli & uli
 

Attachments

  • photo_2FH9 MOD -Parts_2024-03-02_22-56-16.jpg
    photo_2FH9 MOD -Parts_2024-03-02_22-56-16.jpg
    228 KB · Views: 192
  • Like
Reactions: chermann
Skol...with Krombacher...😉
I see you have bought a 300 Watt SMPS with 2000 uF at the output rails. If we meet next in Münster, MS, Germany, i will come along with my "bigger" FH9 amp, 500 W smps, 4700 uf pro rail, build in. Unbelievable difference...night and day...👍

Greetings

Peter
 
My FH9HVX output stage is biased at 250mA, so about P = [(2*0.25)^2]*8 = 2 watts peak Class A in 8 ohms or 1 watt peak Class A in 4 ohms before transitioning to Class B.

Class A peak for a push pull/complementary output stage is calculated as: [(2*bias in amps)^2]*load impedance.

Reading post #56 by @chermann, did you have the vertical mosfet version of this design (i.e. FH9XXX) biased at 750mA? No issues with thermal runaway? 0.75A bias will give you 18 watts peak Class A/8 ohms or 9 watts peak Class A/4 ohms before transitioning to Class B. Your voltage rails are only +/-31V. How warm were your heatsinks? Your dissipation would be 62V*0.75A = 46.5 watts. I am considering increasing my bias until my heatsinks reach about 50 degrees C. If I do, I might consider adding a cap bank immediately before connecting to the voltage rails on the main board since increased bias current puts a bigger load on the power supply - although I do have a fair amount of capacitance already.

Best,
Anand.
I looked at this and asked @keantoken to help me with a careful analysis. He said that the idea is that thermal gain needs to be well below 0 (negative), and thermal gain is given by:

Vgstempco/Gm*Vds*Rja << 1 (negative)

The effect of the temp compensation of the Vbe multiplier (using the BD139) is also important.

In the case of the FH9HVX, the simulation, according to keantoken, is that the output MOSFETs (IRFP240/9240) are over-compensated and hence, it is safe to increase the bias to 500mA.

However, the safest thing to try is to measure the bias current and apply a heat gun to the heatsink to warm it up and see if the bias current increases. If it does, that means it will be unstable and might runaway and melt.
 
  • Like
Reactions: chermann
My FH9HVX output stage is biased at 250mA, so about P = [(2*0.25)^2]*8 = 2 watts peak Class A in 8 ohms or 1 watt peak Class A in 4 ohms before transitioning to Class B.

Class A peak for a push pull/complementary output stage is calculated as: [(2*bias in amps)^2]*load impedance.

Reading post #56 by @chermann, did you have the vertical mosfet version of this design (i.e. FH9XXX) biased at 750mA? No issues with thermal runaway? 0.75A bias will give you 18 watts peak Class A/8 ohms or 9 watts peak Class A/4 ohms before transitioning to Class B. Your voltage rails are only +/-31V. How warm were your heatsinks? Your dissipation would be 62V*0.75A = 46.5 watts. I am considering increasing my bias until my heatsinks reach about 50 degrees C. If I do, I might consider adding a cap bank immediately before connecting to the voltage rails on the main board since increased bias current puts a bigger load on the power supply - although I do have a fair amount of capacitance already.

Best,
Anand.
oh sorry Anand, i miss your comment.

i use fairchild FQA16N25/FQA9P25 as MOSFET...these are sounding good and are working well.
i like the sound more "semi Class A", but yes i don not try to lower the bias and check the sound. i listening 98% in lower levels...
as i like more bias at the FX8 amp i choose the bias the same in the FH9 xrk mod. this is not the HV version.

yes the temperature is on the limit and the heat sink has about 50C°...i push often enough this amp to very loud level and i cant say it is sounding strange or compressed..
"Your dissipation would be 62V*0.75A = 46.5 watts."...yes this is correct, so i get about 23WATT each MOSFET.
bigger cap bank is for me mandatory, because i tried first at the FX8 amp following:
  1. starting with a CRC just 30mF for both channels-sound is compressed and noisy..at low levels !!! aoky ...but NO...not for me..
  2. build two amps with different cap banks-muse edition CRC with 90mF per rail and the Q17 psu board with 60mF per rail
  3. because i am using "semi Class A" biasing i need more caps-the overall power is sufficient with my 300VA(2X24V) Trani
  4. bias with 400mA was okay in sound but i like more bias...so i go to 750mA
chris
 
Hello Forum Members,
I hope I am not posting this issue prematurely. ie. (Before I have sufficient information).

I recently retrofitted my FH9HVX with a soft-start board that offers front panel on-off power switching. This replaced the original SFP soft start board by XRK audio. The new soft start with front panel switch works as expected. Unfortunately, I have picked up a hum that I suspect is caused by a ground loop. Here are some basic facts about this problem.

-The original SFP board had a physical earth ground connection which the new board does not. Other than control and output connections, the new board only offers Line and Neutral connections.
-When I run the complete system with the Wiim music streamer > ACP+ preamp > FHPHVX amp I get the hum.
-When I use my FiiO portable music player > Schiit SYS passive preamp > FH9HVX, there is no hum.
-Music plays normally on both channels despite the hum, and responds smoothly to increases/decreases in volume control.
-The volume of the hum does not change when the volume of music is increased.

I have tried plugging all three components (and nothing else) into the same power strip as an experiment. This did not help.

I am thinking about installing a bridge rectifier as a ground lifter as outlined in Hifisonix- How to Wire-up an Audio Amplifier by Andrew Russell. However, this is a new problem I haven't encountered to date, and I'm not sure I understand it completely. If I need to tear down the system to provide pictures, etc. I can do that. Any thoughts, comments or advice would be welcome. Thank you in advance.
Dave M.
 
@UncleMud

When you short the inputs of the FH9HVX and connect it to a pair of test speakers, is there any audible hum (check to make sure your RCA inputs are well soldered, tight and the jacks are isolated from the chassis) ? What are the intrinsic noise and hum levels of the Wiim streamer and ACP+ preamp? How do we know that it's not coming from the ACP+ preamp?

Understand that your FH9HVX already has a ground lift circuit (R151,C151) that is part of the PCB design for each channel. Your power supply/supplies also have individual ground lift circuits (R131/C131). I'm not saying that installing a bridge rectifier as a ground lift is necessary but if you want to isolate your mains IEC ground, you are free to do so. But I'm not sure if that will solve your problem since you already have ground lift circuits in place.

If you want to post in the original FH9HVX circuit, it's here.

Best,
Anand.
 
Last edited:
  • Like
Reactions: chermann
Uncle Mud,
As far as I know, neither the SFP or SFPP have an earth ground pin. They both operate purely to switch the AC mains. The low voltage is isolated from the mains via an optoisolator. It sounds like somehow, the SFPP low voltage power on/off switch has inadvertently touched analog clean 0v ground?

Please provide closeup photos and sketch of your wiring hookup paths.

If no hum with your portable Fiio but hum with ACP+, it looks like ground loop is between preamp and amp. The SFP and SFPP should have no effect on hum since both are optically isolated. But be careful with the low voltage ground on SFPP touching chassis or analog signal ground.

Thanks
 
Thank you X for your response. I may not have been clear but the current soft start board is neither an SFP or SFPP, but from another manufacturer. Thank you for your input though. The FH9HVX is my favorite amp and I'm anxious to get it up and running. I will dig deeper.
 
Thank you Anand for your response. Your time and patience are always welcome.

Your comments on the already in place ground lift circuits highlight my one of my challenges as a newcomer. Limited knowedge. I was just starting with the obvious, hoping to get lucky. (System works with isolated input devices. Hums with mains connected input devices).
Although I proofread my draft many times before posting, I still forgot to mention that while I was working on my FH9HVX I had the same system set up with my Honey Badger amp, with again no hum.

That said, I had this amp apart numerous times during this retrifit, since nothing goes as planned. The number of things it could be is astronomical. And you are correct I don't know for sure that some new factor hasn't affected the Wimm or ACP+.

I will test for hum with the inputs shorted as you suggested and report back. Barring some big Aha! discovery, I will proceed with taking the cover off and looking for loose connections, broken wires, etc. I'm anxious to get this resolved as the FH9HVX is my favorite amp. Best, Dave M.

Add: X corrected me on the fact that the SFP board didn't have a ground pin either. I was looking at an old picture of my original build and misidentified a ground wire that was adjacent to the SFP. Oh my. Sorry.
 
Last edited: