Pass F6 best operating temp ?

[fabrice63]

Hello everyone ,

i had to change my fan for my F6 water cooled , and the new one is reving a bit more than the previous one.

fact is that the outputs are runing 10 ° celcius above room temp now , instead of 18 ° before

i "feel " the sound to be a bit " harsh" , is it me ( my ear and brain are not brand new !! ) , or the output ( mosfet ) temperature can affect sound "quality" ?

thanks for your imputs


.

 

[Hikari1]

Have you ch checked to see if bias levels of amplifier changed when running colder? That would be my first explanation.

 

[Russellc]

My F6 never felt a little more than warm. ( Deluxe 4 chassis ) Nothing like AlephJ, F5, or even M2. Much hotter. I never felt F6 was harsh, but found it similar in character to F5, a little "thinner" for lack of better word than M2, AlephJ, or BA3....not as warm I guess.

Russellc

 

[drpro]

I am with Russell, D4 chassis works well. Definitely agree on sonics, the M2 sounds better and the F4 depending on the preamp used sounds even better.
YMMV

 

[fabrice63]

My F6 is perfectly set and has no problem at all , i just want or like to understand the relation between outputs ( mosfet ) working temperature and "sound character" , is there any ?


.

 

[ColinA123]

i just want or like to understand the relation between outputs ( mosfet ) working temperature and "sound character" , is there any ?

Difficult question.
I think most builders set the current to achieve a temp, and judge the sound on that.
To answer your question exactly I think that you would need to vary temperature while keeping bias fixed.

 

[Nelson Pass]

As there is some dependence of bias on the temperature of the
transistors, you should check the bias current if the temperature has
been changed by environment or sinking.

Better cooling does give the opportunity to increase the bias, although
you don't want to overdo it.....

 

[WBS]

...although
you don't want to overdo it.....

Hmmm. I thought we were supposed to be fearless...

😀

 

[adason]

I am fearless only if there is capacitor on the output. Otherwise very cautious. Presenting expensive speaker with 24 volts DC does not classify as fearless.

 

[Hikari1]

Hmmm. I thought we were supposed to be fearless...

😀

You don't want to overdo it because the OP's amp is water cooled with a lot of dissipation potential BUT only a pair of output devices. The "chassis" can take the dissipation, but it will probably smoke the mosfets at those super high bias levels.

 

[fabrice63]

My actual bias is 1 A , and it gives me enough power for my Tannoys 15

I can increase the bias , say 0.5 A more , but other than output power , what will be the gain , "sonically" speaking ?

.

 

[6L6]

Increasing bias does not increase output power. It does increase heat.

 

[fabrice63]

Thanks for your correction 6L6 , so what will be the gain in increasing the bias ?

.

 

[6L6]

Once you are at or more than the recommended amount? Very little. Measured distortion usually goes down a small amount, but that's not necessarily an advantage in my experience.

 

[Hikari1]

Thanks for your correction 6L6 , so what will be the gain in increasing the bias ?

.

Increased bias gives you more class A operating region and in general, a more linear amplifier and better sound.

Right now, you have about 15W in Class A until the amplifier transitions to Class A/B into a 8 hm load where it will peak at 25W. The situation is worse into 4 ohms...about 8W Class A, and class A/B until 50W. In general, more bias sounds warmer and more liquid.

Bias is what keeps the amplifier operating in Class A. It has no effect on wattage or gain.

Building a water cooled amp generally means one wants to dissipate more heat than a conventional chassis would. Which means one wants a lot of bias.

If your temps are only 10c above ambient you have a lot of room for increased bias (to get more Class A operation and better sound). The problem is you don't have enough mosfets (only a pair) to "share" the load of the increased bias currents. You reach a point where two mosfets cannot survive at those currents. Typically amplifiers with over 2A bias currents have multiple pairs of output devices.

You amp sounds worse at lower temps because the bias level varies according to the temp of the mosfets. It's not that mosfets sound better when hot, it's that the bias increases due to the temp coefficient nature of the mosfets. If you measure the bias of the amp when cold compared to hot you will see a difference in bias.

 

[fabrice63]

Thank you guy's , it cant be more clear for me now 🙂

.

 

[Bare]

Water Cooled ? just like some basement Kids' "gaming" computer.
You have No heat sinks fitted perhaps?
My rendition F6 hits a whopping 45 C during the heat of summer... driving My tannoys Measured carefully 🙂
And with a 10 k stepper as pre.. Half pot rotation is All I can manage before the Missus starts yelling
Thing simply doesn't put out much heat at all.
Something is clearly amiss.

 

[fabrice63]

Bare , not a gaming computer a real amp 😉

get there and you'll see what : liquid cooled F6

i made this amp not because the heat , but because i like to diy 🙂

.

 

[fabrice63]

Today i raise the bias up to 1.3 A , temperature goes to 14 ° over room temp , and the sound "looks" warmer indeed , something like it was before 🙂

.

 

[Hikari1]

Very interesting amplifier, Fabrice.

I have some questions:

Does the water or whatever you are using for coolant conduct electricity? On a conventional amp we need to electrically isolate the mosfet tab with mica or some sort of isolator between the mosfet and sink. The tab is connected to drain. Are you using a non-conductive fluid?

The danger of water of fan cooling is the system fails what failsafes do you have? A temp cutoff switch for the mains or something? A lot of the pass amps utilize a 75c temp activated switch that cuts power to the mains.

You have a lot of room for increased bias. Iam not familiar with the F6, but I would imagine that 1.75a of bias is entirely reasonable in your setup. And as you have found out...it sounds better.