My F4 build (questions, bragging, and other thoughts)

[Zen Mod]

....
I guess it would be ok to piggyback a 47k (or others for even smaller values) just to see how it goes?

yup, 47K, 43K,39K - whatever you have in drawer, trimpot in series will cover that

 

[Zen Mod]

do not forget to set P1 trimpot to max, prior to powering On!

 

[myleftear]

P1 Trimpot to max = clockwise 25 turns = output ≈ 5k ohm?

Not sure about how to connect the trimpot (not p1) in series? In-out is easy but where do I connect pin 3 / gnd?

 

[Zen Mod]

P1 Trimpot to max = clockwise 25 turns = output ≈ 5k ohm?

.....

be inventive - find a way to check with ohmmeter
I can't remember and I don't care much about CW/CCW even on my own pcbs, always trusting only what DMM is showing

........

Not sure about how to connect the trimpot (not p1) in series? In-out is easy but where do I connect pin 3 / gnd?

sorry, I wasn't precise - I meant P1, not additional trimpot

just correct 10K position, as already told

 

[myleftear]

Thank you ZM, good advices as always. Measure output I would have anyway [emoji1]

It would/will be SO cool if baby-F4 comes to life after just this issue!

[emoji1317] david

I‘m a bit in type-before-thinking-mode, most of the advices become clear just after i go „huh?“... [emoji51]

 

[myleftear]

Please confirm the part you used for Q11 and your value for R8.

R8 is the bummer: 22.1r instead of 22.1K

now comes what I hate: ordering 2 pieces worth $1, $20 shipping. So have to invent a new project to get to the required amount for saving shipping. (Which brings an advantage back, haha)

 

[Zen Mod]

 

[Dennis Hui]

Well, at least you've found your problem.

While you're at it, order some other values, like 27.4K in the unlikely event that
22k isn't enough.

 

[myleftear]

other values, like 27.4K in the unlikely event that
22k isn't enough.

In the basket, as well as some for r9...
many thanks, Dennis & ZM!

 

[Zen Mod]

damn Greenhorns

 

[myleftear]

 

[myleftear]

damn Greenhorns

Jaa, jaa, ha-ha-ha!

 

[myleftear]

now it's closing in...

Heya!
Got a spare hour to try doing the correction.
Replaced R8 (22.1K) on first channel, switched on (first with light-bulb, then fully) and here's BIAS! instant 0.13mA, offset between 0.00 and 0.03 mA.

Next question: Why does it not get warm? The channel has been on for ~15 minutes, no real temp-rise.... (heatsink and device are both room-temp, 18°C)

 

[Dennis Hui]

Your units are funny. Please reconfirm your measurements.

Your dc offset is a voltage. So what is it in mV?

Is your bias really 0.13mA? If so it's hardly on. So please confirm your voltage across
any of the source resistors (R16-R21).

 

[myleftear]

Argh. was too nervous and didn‘t see it !

Bias is at 0.1 V, trimpot to max
Dc offset is now a very jumpy +- 0.15 mV

Need another round of resistor-swap and will report back.

 

[myleftear]

This time it looks good!
R9 is 5K and I get 200 mV bias (slowly growing, me gently keeping it below 200 mV), still jumpy offset close to 0 mV.

It‘s warming up and I am very confident!

Thank you for chiming in so quickly, dennis!
[emoji4]

 

[Dennis Hui]

Sounds good.

Just take your time, monitor and let it warm up with the lid on and trim offset and bias if needed.

 

[myleftear]

My turn to feel embarasssed [emoji51]
But it’s on it‘s way. I connected both channels and threw it on which instantly blew a 2A fuse. On a 230V line it is running btw...
Having a break until ... thursday if I‘m lucky.

 

[myleftear]

When I was in need of a little distraction, I went to the chassis and tried my mechanic skills, making a prototype of the LED-indicator:

Since I have dozens of rectangular leds, I'm using them for the front. Cutting out a hole 5 x 2 mm is a bit of a challenge though. (As one can barely see, the LED sits snugly, but the cutout got half a mm too long). I luckily made it on a spare part.
Next try: round the led, make the hole 5 x 2 mm with round ends...

edit: That yellowish stuff is the masking tape protecting the anodized aluminium.

 

[Dennis Hui]

Did you set P1 to max value before you power up the other channel. With R9=5K
(and I'm assuming R8=22.1k) the TL431 will give fairly high Vgs voltages even with P1
set to max (~5k) and you may already get a healthy amount of current
flowing through the mosfets. And if P1 was close to zero that current will be very
high.