What determines how high bias can be on F-5?

[mk57]

I don't know how the output could be shorted the wires run directly from the pdb to the speaker posts. Iam new at this so bare with me!! tks

 

[Zen Mod]

seems that you are new to 3 R's , too ..........

I asked - are inputs shorted ......

 

[mk57]

opppppppppppppppps There is nada on the INPUTS so unless the pcb is shorted I wouldnt think so.

 

[Zen Mod]

seems that you need to reread at least one of several F5 biasing procedure instructions , floating here ....

 

[mk57]

maybe Ill just go back to elementry school and learn how to read first!! I have been reading threads on setting the bias but they were in bits and pieces so I will try and read yours complete and maybe I can do this correctly. Thanks

 

[mk57]

I reread your biasing procedure but nowhere does it state that the inputs should be shorted, I guess Iam just missing something somewhere. I will give it a try this evening.

 

[Zen Mod]

naah

that was a joke , off course

I completely understand that - if you're novice to electronics , you're probably in sort of fear and frenzy what to do ;
so - general tip : take your time , don't rush and try to find necessary info before flipping the switch ....


here it is ( read first few pages of thread thoroughly ) :

http://www.diyaudio.com/forums/pass-labs/188691-illustrated-guide-building-f5.html#post2566447

fact that you need to short inputs was already stated somewhere where that procedure was originally written ........ if I remember correctly

 

[Bksabath]

Yes shorth the Imputs newbie myself but is there most definetley.

Just croccodile clips will do.

Newbie fear and frenzy that what it is
I realy tought that I was most definetley going even more bonkers while building mine.

You know I mean like:
Where is that screw driver I just put down and what the Fuz am I doing in the kitcen? did I want a cofee? nah Just had one
a raight where is the screw driver?
well newer mind it alwayse turn up I may be beter having a coffe while here?

LULA why did you put my screw driver in the ketle?

and so on

 

[mk57]

Ok shorted the inputs but same basic results, bias and offset still jumping around.After 2hr warm up i can get offset down but it jumps + or - up to 50mv. I was wondering if the earth ground is open at the wall would this cause these problems??

 

[Bksabath]

Oky dokey define Jump around

50mV not to bad

Do you have speaker protection?
Well worth it before she goes "I told you so"

J fets stuck together and case closed?
Little thermistors next to mosfets? (Glued to drain pins of mosfets on mine)

Silly PCB with double layers tracks crossing all over the place so loads of stray capacitance? And or gate resistors to far away from gates causing oscillation?

Poor filtering on supply?


Cold solder somewhere?
Solder paste residue somewhere?
Multy turn pot out of range difficoult to adjust or damaged?

Really wish I could help this is most strange as I got 2.5 mV on mine
Picture 1 DC ofset while setting up and few hours later after quite a few traks

http://www.diyaudio.com/forums/swap-meet/182480-holco-mfr-15ppm-2w-sale.html

 

[Sangram]

@mk57: This is not usually something to be worried about.

The F5 output offset is determined by the relative temperature differences between the output devices. The Thermistors are the feedback method by which the offset is kept under control.

When one transistor warms up it passes more current, warming up even more until the thermistor attached to it starts pulling the bias down. As it goes down it cools, the thermistor then starts allowing more bias, warming it up. Now go back to the beginning of this paragraph.

What you are seeing is the Yo-yo effect between the two transistors. I had this exact issue when I built mine, so I simply removed the thermistors altogether. True to Nelson's prediction the amp took a lot longer to warm up, but the offset stayed where it was.

The most unfortunate part of this arrangement is that in major variations of ambient temperature, the bias needs to be adjusted again as there is no bias feedback mechanism. I have no major problems with this, because in my country it is inconceivable to use such a hot amp during the summer. So I set it with winter bias, and leave it at that.

 

[Bksabath]

Re reading mk57 post 89
Quote> I was wondering if the earth ground is open at the wall would this cause these problems.

Well not realy till someting go wrong and possibly kill you.

If you have any doubt about it ceck it to make shure.

 

[dazed2]

After listening to my F5 for a little while I tried to bias up to 0.8V. Yes it's definetly more dynamic, and more Prat as well.

But its Hot! I'm getting about 52C on the outside of the heatsinks. 62C at the top of the Mosfet case, and about 47C internal case temp. My transformer is reading about 55C

Previously at 0.6V bias I'm getting about 45C at the heatsink and 53C at the top of mosfet case.
Oh and I'm using a thermocouple to measure temp... I think that should be more accurate than the IR ones.

Note: I'm also using caddock MK132 as the 4 feedback resistors (not the big blue panny 3W ones)

Is this going to be ok?

 

[Sangram]

Simple matter to calculate the internal temperature of the device by using the power dissipated (your bias and supply voltage determine this) and then using the device Tj-Tc specification.

IIRC there is about a 20 degree difference between what you measure at the case and the junction itself - which is fine till about 150C. So you're within the limits of the device.

However at higher temperatures the power handling of the device drops off quite rapidly. Check the temperature derating of the device to know whether you're within safe limits. Offhand, I think you are. Mine measured about 45 degrees on case surface with default bias. The output resistors got very hot even at the 0.6V setting, and almost burning to the touch (85 degrees) at 0.9V bias. Obviously, I never stayed at that bias level for long.