F5X -- the EUVL Approach - The Build Thread

Jump to Latest
#621
> But with 1,5A my amp is hotter than hell.
> (6 seconds by hand test, ambient temperature 32 Cº)

Well, it was designed to heat sink temperature of 65°C for 22°C ambient.
I guess where you live and with ambient of 32°C, you should retain bias at 1.5A to keep the temperature reasonable.

The other solution would be a slow blowing fan. This will reduce temperature drastically.

> (The top panel is much hotter that the heatsinks).

That IS the case for the CRC, as Nic already reported.
This will improve when you have the regulator version later.
Another alternative would be to go back to the wire-mesh top plate, as this allows better ventilation.


Patrick
 
#625
No such thing as dumb questions. 🙂

The GB2 people have a revised protection board based on revised Gerber from Alexis, but I have not tested that myself.
If you promise to test for us, we'll send you one at nominal cost (postage).

It is AFAIK also 1.6mm. Why would you want thicker ?
If there is demand for 2mm, we can run an extra batch as a GB.

Contact Gogowatch by email if you want a test board.


Patrick
 
#629
A board of 1.6mm FR4 mounted on 120x90mm grid with 4x M3 screws.
And the force to actuate a tiny winny push button switch ........
How many Newtons ? Now many µm deformation ? Have you actually try ?

We'll do you one last service and make a batch 2mm for you if there are more than 30 subscribers.
Price tbd. (Not sure we can get small batch 2mm economically.)
Best to be distributed via the local distributors, so we only have a few packets to send.

More donkey work for WK again ........

🙁



Patrick
 
#630
I have the amp running. Now is without protection board. Feedbacks uncrossed, direct to ground.

I have polarized to 1.5 A and the power is a + -18.9 V after power resistors.

I put jumpers on board, so if unplug the jumpers the feedback is crossed and the amplifier go supersymetric. I can adjust bias without removing the board on heatsink.

I hope make the protection board this weekend.

An externally hosted image should be here but it was not working when we last tested it.
 
#631
😀
EUVL said:
For all those protection board subscribers :

Alexis tested the finished protection board for fit onto the case. Everything works as expected, but he felt that the PCB flexing when he pushes the push-button. The push-button force is only the 1N or so required to engage the micro-switch. So functionally there is no reliability issue. But it is true that there is no mechanical limiter to prevent you from exerting more force than required.

What I suggest is that before building the entire protection board, just solder on the micro-switch, mount the PCB onto the front panel and test it with the push button. If many of you feel that you do not have delicate enough fingers, and need a more rigid PCB, we can arrange a new batch made (at cost) to 2mm FR4 instead of the current 1.5mm This will increase the stiffness by a factor of 3.

Even though we consider this a physiological issue rather than technical, all future batches will automatically get 2mm PCBs, just to avoid any issues of the kind.

😉


Patrick
Click to expand...

Just to jolt the memory...
 
#632
WK forgot, I did not check. So our fault.
We all have too much to do for the normal job for the next years to come.
So we shall not do this again (F5X GBs), thus it will not happen again in the future.

But as said, we'll make a batch of 2mm, at cost, if there is sufficient demand.
I do not consider 1.6mm a design error.


Patrick
 
Last edited:
#633
Good to know someone is making progress.
Before you float the X, check that you can adjust output DC offset to zero (<10mV), and it is stable with time. Gnd the inputs first of course.
Then feed the input with a square wave, say 1kHz, and measure the output response.
Some load will help stability, say 4 ohm 50W power resistor.
When testing single ended, of course connect dummy load from output to star ground.

When all goes well, then change to floating X, repeat the same, and then time to play music.

🙂


Patrick
 
#634
My F5X without protection board, running at 1.5A (This weekend go to 2A), sounds great. This weekend I'll try to do some measurements (Distortion, noise....) and a critical listening against a regular F5 and a XA30.5.

By the moment I burned a R from the CRC (Acted like a fuse) (Beginner's luck).

As oneoclock I left a pins for shortcut by jumpers for adjust, but my unit is working with crossed feedback

An externally hosted image should be here but it was not working when we last tested it.



An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.
 
#635
Last edited:
#639
Diodes board tilts and bends the legs of LEDs with rigid transformer cable.

I made a new rectifier board for the LEDs fixed with 2 screws to the heatsink.

An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.
 
#640
EUVL said:
Good to know someone is making progress.
Before you float the X, check that you can adjust output DC offset to zero (<10mV), and it is stable with time. Gnd the inputs first of course.
Then feed the input with a square wave, say 1kHz, and measure the output response.
Some load will help stability, say 4 ohm 50W power resistor.
When testing single ended, of course connect dummy load from output to star ground.
When all goes well, then change to floating X, repeat the same, and then time to play music.
Patrick
Click to expand...

Bias 1.5 A and output DC ofset <1 mV stable all time.😀

It seems to have a good result for a 3 Amp. Class A amplifier.😀
I will measure distortion for this 1,5 A. bias before and after the X float this weekend.
 
You must log in or register to reply here.
Top Bottom