It was more for a comparison, it would have been fun and definitive way to see which caps I wanted for my music. As Mr. Pass mentioned in an article, the types of caps may not be very important in this design as again suggested in the preceding post.
With the "just one more Mouser order Honey..." the mere $18 added on for 16 caps made me stick with what I already have on hand, the Nichicon Fine Gold. That amount is a drop in the bucket but the bucket is filled and don't want one more drop to break the molecular cohesion and create a problem. The Silk will have to wait until I go some amount of time without placing another order with Mouser if such a time exists... but by that time the amps will not be as identical with some parts more burned in. Of course if there is someone in Austin who wants to listen and has those few dollars go ahead and order them and I'll build them in for a comparison should that person have a listening space with everything but the F4 amps. I wouldn't trust my subjective description, I have a difficult enough time describing whiskey using language.
Personal I drink single barrel Bourbon from two countys in Ky or single malt . Smooth is all the words that I need for whiskey and prefer it cold . Regards
Thank you, 6L6 for the build guide. I've switched out my F5 for an F4, and while my BA-3 is easily a month away from completion, the meager volume I'm able to get from my 5687 Aikido through the F4 sounds incredible. I am in love with this thing and will be keeping it as my main amp.
It seems to provide more depth and 'bounce' to the music than the F5 was able to.
On question, though - Bias settled in at .2, after an hour, and offset on either channel settled to between 0 and .01. But the right heat sink is noticeably warmer than the left. What might cause that at the same bias level, and same perceived volume from each channel? I set the MOSFETs on either side, to as close to the same torque as I could by hand, and I don't have any loose ones. I'll measure the actual temp diff this weekend when I can open it back up.
It seems to provide more depth and 'bounce' to the music than the F5 was able to.
On question, though - Bias settled in at .2, after an hour, and offset on either channel settled to between 0 and .01. But the right heat sink is noticeably warmer than the left. What might cause that at the same bias level, and same perceived volume from each channel? I set the MOSFETs on either side, to as close to the same torque as I could by hand, and I don't have any loose ones. I'll measure the actual temp diff this weekend when I can open it back up.
The channel that was hotter evened out and is now very close to the 'good' side.
On the previously hot side: warmed up after an hour, and at zero offset, R20 above Q7 is at .28v, while R21 and R19 are at .07v. R16 and R18 are both at .13v, with R17 at .16v. Using a 4.75k in R9.
From reading previous posts, would using 10R source resistors help in this case? In all source spots, or just above the Q7 and Q4 current hogs?
On the previously hot side: warmed up after an hour, and at zero offset, R20 above Q7 is at .28v, while R21 and R19 are at .07v. R16 and R18 are both at .13v, with R17 at .16v. Using a 4.75k in R9.
From reading previous posts, would using 10R source resistors help in this case? In all source spots, or just above the Q7 and Q4 current hogs?
Thanks for hanging in there with me, really appreciate the help to get this operating the way it needs to.
* Source resistors = 3W .047R KOA Speer Metal Oxide 5%
* MOSFETs from alweit on ebay, pieced together from a couple of different orders but all fell between 4.15 and 3.89 vgs (I don't know the parameters of his testing rig, but his parts worked well in my F5).
The 'good' channel is also using a less-than-tightly-matched set of output devices from the same collection...but biased up nicely. Do I have any options short of replacing the fets that are hogging current?
* Source resistors = 3W .047R KOA Speer Metal Oxide 5%
* MOSFETs from alweit on ebay, pieced together from a couple of different orders but all fell between 4.15 and 3.89 vgs (I don't know the parameters of his testing rig, but his parts worked well in my F5).
The 'good' channel is also using a less-than-tightly-matched set of output devices from the same collection...but biased up nicely. Do I have any options short of replacing the fets that are hogging current?
I purchased from the same source asking for matched for 4x F4's and hoping for a closer match. Didn't go through eBay but contacted him through the board. I'm away from home right now so can't check - did he print / write those values on the packaging?
His order was received fast and he was polite, no complaints there.
On eBay he claims they are matched to 1% but claims 2% to be conservative.
While I'm not going build a testing rig, I am now considering paying someone local to test for me. I'm probably worrying too much.
This is how closely mine were matched.
In the F4 Manual (Found here: http://www.firstwatt.com/pdf/prod_f4_man.pdf) it says "All devices are matched for Vgs to .02V."
In the F4 Manual (Found here: http://www.firstwatt.com/pdf/prod_f4_man.pdf) it says "All devices are matched for Vgs to .02V."
Jeffs don't worry, you will receive a very close match within a single order from him. This channel on my F4 was put together with leftovers from a few different orders from his store.
As this is clearly not a circuit I should be putting leftovers into, I will replace them with a tighter group. (does this qualify for a dodo award, from Zen?)
Until the order arrives I think I'll try some 1R 3W source resistors. Thank you 6L6.
As this is clearly not a circuit I should be putting leftovers into, I will replace them with a tighter group. (does this qualify for a dodo award, from Zen?)
Until the order arrives I think I'll try some 1R 3W source resistors. Thank you 6L6.
