F5 Turbo Builders Thread

Hi all,
This is my first post on this forum. I've taken on Gabsta's F5T as a project with a view of building my own sometime in the Future.

As posted previously, one half of the amplifier, the right channel is completely fried. That half is now disconnected awaiting parts for a rebuild. I've just been checking the left channel as if its a new build.

I am biasing the N & P channels but I cannot get the voltage between the source resistors to match exactly. By that I mean I cannot get them closer to within 40mV of each other regardless of the DC offset. If I adjust P1 or P2 the voltages rise and fall together.

Is 40mV difference at ~300mV acceptable or is there an issue?
The power supply rails is reading ±32.35V so it doesn't seem to be an issue with the PSU. With the power off and the resistances of the source resistors measured, both reads 0.7ohms.

Any help would be greatly appreciated.

DSC_7495.jpg

have you calibrated the multimeters so you know they measure the same?
or did you try to swap the N and P ch multimeter and see how it measures?
 
when setting the output bias the load is not connected.
The only other place for bias current to escape is via the feedback resistors.

If you are measuring a different voltage drop across the source resistors then that tells me and should tell your that the Source resistors are are sufficiently matched.

Did you measure your Source resistors?
 
when setting the output bias the load is not connected.
The only other place for bias current to escape is via the feedback resistors.

If you are measuring a different voltage drop across the source resistors then that tells me and should tell your that the Source resistors are are sufficiently matched.

Did you measure your Source resistors?

Hi AndrewT,
Thanks for the response. The voltage difference I am sure is accurate. It's still approx 40 mV when measured with a single multimeter.
The output I left open circuit, the input I have shorted to make sure there is no transient input. Unfortunately I do not have a micro ohmmeter so I am unable to measure the exact resistance of the source resistors beyond one decimal place. What I measure minus the lead resistance is 0.5 ohms. The source resistors are Dale 1% 1R. But the voltage difference suggests the variation would have to be a magnitude of 16% which I should be able to pick up even at one decimal place.
As a piece of mind I may rebuild both sides with all the components measured off board as it's difficult to tell if any other part of the circuit is affecting the readings.
Regards
 
You don't need a micro ohm-meter.

String your low value resistors in series.
Pass a CCS current of ~100mA through the string.
Measure the voltage drop across each resistor.
Keep ONE resistor as a reference. Mark all your other relative to that reference.
Pick two that match. You can get better than 0.2% with care.

Bridge matching gets you even closer.
 
Thermography of F5T Circuit

I took some thermo imaging photos of the F5T amplifier I am re-building the other night as I was interested in the heat distribution of the MOSFETS. I'd thought I would share some thermographic images of the F5T circuit for discussion as there was a couple of interesting points.
The photos are taken after about 15mins with no load on the output and the input shorted. The bias voltage was low enough not to conduct the diodes as is evident by the cold diodes.

  • Looking at the MOSFETS & source resistors, the inner P-Channel is running slightly hotter than the other 3.
  • The FEts on the control board was running at about 50°C when I switched off after 30mins.
  • Its good to see the tracks not heating up at this current level.
  • Only the upper half of the heatsink is utilised in this configuration. Maybe worth installing whole circuit lower down, will decide after getting the diodes to conduct to see what they output.
  • Surprised at how warm the snubber resistors were with only a total of 85W of power input.
Entire Left Channel
IR_0686.jpg


P-Channel
IR_0692.jpg


N-Channel
IR_0694.jpg


Heatsink after 10mins
IR_0716.jpg


Heatsink after 30mins
IR_0726.jpg


Snubber resistor bundle
IR_0696.jpg


Primary Supply Thermistor
IR_0698.jpg
 
I can't identify the scale of the pics, but probably get better if dropped even more than 1", possibly as much as 2".

The hotter Pchannel can be checked for Vdrop (Vds) and for Id to determine if it is dissipating more power.
Better matching of Source resistors and Vgs will possibly eliminate this measured extra temperature.
 
the best is (as i said before) to rotate the boards 180 degree. so that the outputs points down and diodes points up.
the P ch source resistors looks as is only at ambient temp. what is the voltage drop on those resistors?

The voltage drop hasn't change from my previous post, it is still at 300mV. I have deliberately kept it low whilst I do any troubleshooting just in case whatever caused the other side to catch fire also manifests itself on this side.
See pic of the damage.

I will be looking at re populating the components to a new board anyways as the tracks have lifted off and missing in some places. Whilst I was at it I may as well make sure components are matched as close as poss.

f0358d0f-cd53-4de7-87fd-a9ac8aef12a1.jpg


Kind regards.
 
Member
Joined 2010
Paid Member
Hi to everybody,
I think that I need some advice about the F5T V2 that I just complete and biased.
2 days ago I installed the trasformer and start the bias following the suggestions found on the forum and in particular the advices from Walter and Buzzforb... slowly! And slowly I reach the 331 mv on both sides in the right channel with the best offset of -0.1.
The same I can't reach for the left channel. I have always a difference of 4.5 mV between P and N and the offset is unstabile. When I reach the 0.0 and I cover the enclosure the offset start to move to - 0.1, -0.2 up to -3.9 (now). The same for the 2 source resistors... the reading increase following the offset. This happen only in the left side while the right is stable.
I suppose that is relevant to the increasing of the temperature but why only for 1 channel? The ampli play music fantastic but I am scare for this channel... may be is normal but some advice should be greatly appreciate.
Many thanks
Enrico
Ps: Just read the meter and the offset is now -5.1. I will reopen and return back to 0.0 as I do already 3 times But I am sure that after half hour it will be the same
 
Official Court Jester
Joined 2003
Paid Member
-5.1 of what ?

MA , MV , MJ or ........ maybe mV :clown:

everything up to +/-50mV is completely OK for home gear , in temp equilibrium

btw. do not compare volatge across Rs of upper and lower part ; choose one side and measure on it ;

so - offset and current on one side is what's needed for proper setting of F5 ( and many other amps)
 
I'm a fraidy-cat, so I measured 3 different MUR3020 diodes.
I wanted to know the forward voltage drop at 53C, which is the operating temp of my own F5V2 at summer.
 

Attachments

  • MUR3020tabl.png
    MUR3020tabl.png
    56 KB · Views: 403
  • MUR3020dia.png
    MUR3020dia.png
    29.1 KB · Views: 378
  • IMG_2051.jpg
    IMG_2051.jpg
    331.4 KB · Views: 200
  • IMG_2050.jpg
    IMG_2050.jpg
    245.9 KB · Views: 241