DIY: Pass F5 turbo v2 initiated

[omishra]

Hi Om,
I have read that Diodes doesn't even get warm.I don't know how true is this.

Sachin, thanks. I have read it about MUR diodes on Amp board not sure about rectifier board.

 

[captrajesh]

Kroyin, how hot those fairchild devices are on heat-sink?

Yeah, even I'm eager to know.

how much those MUR rectifier diodes on PS board are dissipating in heat?

I have read that Diodes doesn't even get warm.I don't know how true is this.

I have read it about MUR diodes on Amp board not sure about rectifier board.

Even I also read that the diodes on the amp do not even get kicked in during normal mode of operation and when they do during demanding tracks, they do not get too warm.

However, I guessed that those on the rectifier board may require heat sinks and bought a 4 feet long and about 2" wide heat sink which has to be cut to size to be fitted on to individual diodes.

 

[omishra]

However, I guessed that those on the rectifier board may require heat sinks and bought a 4 feet long and about 2" wide heat sink which has to be cut to size to be fitted on to individual diodes.

Yes captain, we planned 4 rectifier blocks with 16 diodes. That should conduct 0.8A in each dual rectifier, eventually 0.4A per leg of each device(standard full 0.8A biasing per device on Amp board). It's datasheet mentions 1W for 2A per leg.

So for our standard full 0.8A biasing will produce heat for that 0.2W per leg means 0.4W for every rectifier MUR diodes. Kroyin achieved 0.72A biasing per device. So we could get some idea. It should also get barely warm. As this will conduct every 1/100th sec, and 50 times/sec alternatively, can work without or with small heatsink.

 

[kroyin]

Let us just say that the rectifier diode heatsinks are lukewarm after 2 hrs of operation, so rudimentary heatsink nothing elaborate. Easiest will to bolt them to chassis using thermal paste & mica (if the chassis is aluminum)

Amp diode do not switch on under normal circumstances

Better not to bias till 400mA since somewhere above that the diode based part will kick in.

If you run out of pot turns while biasing you will have you increase the JFET drain resistors

If the plastic casing of the FQA parts are significantly hotter than your heatsink then your mounting is not good.A cheap way is to get hold of a multimeter with temp measurement and check the temp of the drain leg (not to exceed 70C)

 

[captrajesh]

@Omishra, I've noted that bibin has incorporated a heat sensor in the PCB circuit.

I feel it is good to have it for safety reasons. Since our PCB doesn't have this extra circuit, how can we retrofit it?

@Bibin, AFAIK 'Arctic Silver' paste is used for removing the air gaps between the tranny and the heat sink for achieving best contact and resultant improved conductivity. However, I see that you have used substantial quantity of the paste here. Of course, I being a noob, am talking theory here; you being an experienced electronics guy, could you kindly comment on the issue?

 

[kroyin]

Captain there are two things
---thermistor as shown in picture, used to maintain bias stability, they are there in Turbo board

---thermostatic switch (found in geysers), which is part of AC connectivity (as per diagram in F5T article,PSU circuit) which cuts off power if the case/heatsink temp reaches a rated value.

The F5 & F5T articles at firstwatt.com are good reads.

You can use artic silver,I use anabond.

Take a pea size drop, rub it in between thumb & forefinger.After that take the mica washer and spread evenly while rubbing across the thumb and forefinger

 

[bibin3210]

@Omishra, I've noted that bibin has incorporated a heat sensor in the PCB circuit.

I feel it is good to have it for safety reasons. Since our PCB doesn't have this extra circuit, how can we retrofit it?

@Bibin, AFAIK 'Arctic Silver' paste is used for removing the air gaps between the tranny and the heat sink for achieving best contact and resultant improved conductivity. However, I see that you have used substantial quantity of the paste here. Of course, I being a noob, am talking theory here; you being an experienced electronics guy, could you kindly comment on the issue?

@captrajesh, yes only a very thin layer (translucent) of paste is required for maximum efficiency. In my case i was a bit lazy and careless, i just put a drop (didn't cared about the size) and fastened the screw and make the excess squeeze out. I don't recommend this if you are not sure about the right torque to apply.

Thanks,
Bibin

 

[omishra]

@Omishra, I've noted that bibin has incorporated a heat sensor in the PCB circuit.

I feel it is good to have it for safety reasons. Since our PCB doesn't have this extra circuit, how can we retrofit it?

Captain, our PCB has it in place. Please look at the PCB near one of inner MOSFETs at each side - P,N near input above 220 ohm feedback resistors. As kroyin said, it will help correcting drift, same as bibin's F5 build.

For mains cutoff you need extra thermostat or thermostat controlled board.

@Bibin, AFAIK 'Arctic Silver' paste is used for removing the air gaps between the tranny and the heat sink for achieving best contact and resultant improved conductivity. However, I see that you have used substantial quantity of the paste here. Of course, I being a noob, am talking theory here; you being an experienced electronics guy, could you kindly comment on the issue?

Be careful, artic silver is highly poisonous (absorbed through skin also) and conductor of electricity. It can harm human as well as do short circuit by making electric flow across it.

I am going to use dry silpad in place of mica pads.
Because mica pads are no thin and its multilayered with air gaps in between. If used along with white thermal paste then please use little amount just to remove air packet in between to thermal surfaces, more is bad - less is better.

 

[quad]

Be careful, artic silver is highly poisonous (absorbed through skin also) and conductor of electricity. It can harm human as well as do short circuit by making electric flow across it.


I am pretty sure the one I used is not electrically conductive (at least to the
limit of my cheap multimeter) -

http://www.arcticsilver.com/as5.htm

 

[omishra]

Be careful, artic silver is highly poisonous (absorbed through skin also) and conductor of electricity. It can harm human as well as do short circuit by making electric flow across it.

I am pretty sure the one I used is not electrically conductive (at least to the
limit of my cheap multimeter) -

Arctic Silver Incorporated - C?ramique 2

Prasad, I used same, purchased from here.

I went by manufacturer's word. I used it on CPU core, no problem. It really got temperature of CPU down by 5C by virtue of better thermal contact. But when I used it on both sides of mica pads for MOSFETs, somehow it made contact with one lead and after 5 minutes it created spark and light with heatsink leaving black spot on that MOSFET lead. MOSFET gone. Same thing repeated twice. Then I searched over net and found that it creates some discharge over closed gaps if applied between two conducting metal gaps. Thus for amp I am back to white paste again. May be I come under wrong usecase. But not sure if I will try it again.

Here is quote, reference same link posted by you.

While much safer than electrically conductive silver and copper greases, Arctic Silver 5 should be kept away from electrical traces, pins, and leads. While it is not electrically conductive, the compound is very slightly capacitive and could potentially cause problems if it bridges two close-proximity electrical paths.

 

[Audiodoc]

I have purchased a thermal adhesive paper for this job. Will try on a gainclone before attempting to use the same on a f5 build.

 

[captrajesh]

---thermistor as shown in picture, used to maintain bias stability, they are there in Turbo board
---thermostatic switch (found in geysers), which is part of AC connectivity (as per diagram in F5T article,PSU circuit) which cuts off power if the case/heatsink temp reaches a rated value.

Captain, our PCB has it in place. Please look at the PCB near one of inner MOSFETs at each side - P,N near input above 220 ohm feedback resistors. As kroyin said, it will help correcting drift, same as bibin's F5 build.

For mains cutoff you need extra thermostat or thermostat controlled board.

OK got it. Are the sensors also part of the kit or do we have to buy them separately? I do not remember to have seen any sensors in the kit.

Take a pea size drop, rub it in between thumb & forefinger. After that take the mica washer and spread evenly while rubbing across the thumb and forefinger

yes only a very thin layer (translucent) of paste is required for maximum efficiency.

please use little amount just to remove air packet in between to thermal surfaces, more is bad - less is better.

Okie dokie. While browsing the net, I've read a very informative article about heat sinks where this is exactly what was mentioned.

I use anabond.

I correlate Anabond with their 'Cyanoacrylate' adhesive commonly known as 'Super Glue'. This piqued my curiosity and I got onto the net to check which one do you use.

Anabond Limited :: Adhesives and Sealants :: Anaerobic :: Epoxy :: RTV Silicone :: Poly Urethanes :: Rubber Based (Neoprene&Butyl) :: Acrylic :: Solvent Based Adhesive :: Speciality Coatings

I guess you use this

Is it different from Arctic silver? I feel they both would have the same chemical compound.

 

[captrajesh]

Thanks for your advices guys, they are very helpful. Will the readymade module suggested by Bininmb work out here? Will it affect the SQ in any adverse way? Sorry for the noob questions.

Repeating my question. Would the ready-made module being used by Bijinmb be useful here?

I will be using this speaker protection for my Le Monstre build:

Speaker Protection Module

 

[omishra]

OK got it. Are the sensors also part of the kit or do we have to buy them separately? I do not remember to have seen any sensors in the kit.

Yes, there are 4 green colored NTCs with the kit. You need to place them at positions TH1 and TH2 on each side of central 2 MOSFETs, on each board.

 

[kroyin]

Om..where would be the best place to an led in.
Thanks

 

[omishra]

Om..where would be the best place to an led in.
Thanks

Electronically, I suggest on secondary, with diode 1N4007 in forward bias of led and 1/4W 15K in series. 1N4007 will protect LED from reverse voltage in each cycle and 15K will be current limiter. LED Resistor value for 24V secondary will be from 3.3K (bright) to 15K (deem).

 

[kroyin]

So, it is connecting the diode -ve to led +ve via a 15k resistor??

By the way last night bhaskarcan & me had an extended listen of the F5T (not audiophile demos normal day to day music) and let us just say all F5T builders are in for a REAL TREAT :clapping:

Now just finish them uphyeah:

 

[omishra]

So, it is connecting the diode -ve to led +ve via a 15k resistor??

Yes, you are correct.

By the way last night bhaskarcan & me had an extended listen of the F5T (not audiophile demos normal day to day music) and let us just say all F5T builders are in for a REAL TREAT :clapping:

Now just finish them uphyeah:

Thanks for letting us know. Tune in to hear from us.

 

[kroyin]

Final positioning..
Brother is on the other side

 

[omishra]

Guys no machine work, only manual hand finish. Please bear with us - me and sann.

Rectifier Board -

For heatsink size, diodes are soldered alternatively at each side. Board spaced 2" from bottom. Best way solder is first PCB pins, then MUR diodes and last small caps. Carefully no heatsink shorting anywhere to PCB soldered pad to any metal body and heat sink paste used.

NOTE: captrajesh has better heatsink which is single piece strip. That way all MUR diodes could be kept one side of PCB on one heatsink with using proper electrical isolation pads. Also if space permits then all MUR diodes could be attached to bottom plate with using proper electrical isolation pads.

Power Supply filter -

In below pics you can mounting screws with mica spacers and high grade plastic separators. Please observe the closeness of adjacent PCB pins. If these screws are made of plastic then it will be good. I verified nothing is getting shorted to screws or body.

Amp PCB -

There is no preset P3 (200 ohm) used, NTCs to be glued over MOSFETs. BTW its legs are long but not shorting anywhere on PCB. You can see silpad below MOSFETs.

PCB to heatsink bolted with 5mm metal spacer.

Space utilization near back panel. See power switch with arc suppressing cap. Ground isolator/separator and mounting screw panel plate. Over that Bakelite screw plate I will mount NTC soft start and mains noise suppressing capacitor.

Some pics before wiring, almost 2" space from PS to amp boards. Transformer and PS Filters are separated by 3/4".