F5 Turbo Builders Thread

Thanks. Will compare every line. The PCB is from a India forum group buy. It has been designed by fellow member Om Mishra. While the PCBs look well made, I am probably the first guy to use these boards. There could be manufacturing issues or may be not. With time at premium, this debugging hogs quite a bit of that.

Anil, it seems I need to finish my boards early. I am waiting diodes as all earlier parts are distributed. Since your one board worked nicely, only problem with particular board. Chances of manufacturing defect are far as they are tested for continuty at vendor by automated process and marked ET OK. I still feel problem around JFets, source resistors and P3. Please test them off the board. The jefts are sensitive to static damage by hand too.


Output Fets are from buzzforb and well packed in anti static bag. So they seems okay as part are matched quads.
 
Last edited:
I dont think p3 is the problem. Even though it can effect equilibrium of FE, it cannot do so to that level. If jfets are showing proper currents across their source resistors. Thfe next thing is gate voltage for output fets. If that is in proper range, then outputs should be biased properly. We know they were before, but we had muted sound on output. I have had that happen only when something was wired wrong and signal was shunting to ground.
Alniva, check all the voltages again, starting at the FE.
Let us know what they are.
Who knows what happened after the change. Perhaps he burnt up a pot desoldering it. Had that happen before. I have never lost an output fet to static damage, but it definitrly sounds like one polariity is operable, and the other is off or shorted.
 
I dont think p3 is the problem. Even though it can effect equilibrium of FE, it cannot do so to that level. If jfets are showing proper currents across their source resistors. Thfe next thing is gate voltage for output fets. If that is in proper range, then outputs should be biased properly. We know they were before, but we had muted sound on output. I have had that happen only when something was wired wrong and signal was shunting to ground.
Alniva, check all the voltages again, starting at the FE.
Let us know what they are.
Who knows what happened after the change. Perhaps he burnt up a pot desoldering it. Had that happen before. I have never lost an output fet to static damage, but it definitrly sounds like one polariity is operable, and the other is off or shorted.

Thanks. I have disconnected both the boards and started comparing. Good news is that the MOSFETs seem OK and measure similarly between both the channels in terms of DC resistances.

I have found one loose solder on one of the MOSFET source resistors and bad solder on centre leg of P3. I have now re-soldered every thing that is there on the board. Will passively measure all the lines and components and compare both for any anomalies and report back, before I connect them to Power.

Cheers.
 
Moderator
Joined 2002
Paid Member
Might I suggest checking the circuit instead of individual components, so you can eliminate the board or soldering as an issue.

Also do check the input cables. Insulation tears can cause shorts. Try connecting the inputs together at the board, from the working board all the way to the bad one.

It's probably a simple oversight of some kind, probably before the signal even gets to the JFETs.
 
Might I suggest checking the circuit instead of individual components, so you can eliminate the board or soldering as an issue.

Also do check the input cables. Insulation tears can cause shorts. Try connecting the inputs together at the board, from the working board all the way to the bad one.

It's probably a simple oversight of some kind, probably before the signal even gets to the JFETs.

I too think so. It is not the components but basically something has gone wrong during assembly. Either a dry solder or a PTH problem or a bad input cable. I am trying to eliminate one at a time. New work week has begun and hence need to squeeze time along with way and continue. I am confident of identifying the problem soon.

Cheers.
 
OK, got the board that was giving problems (left channel), wired up again after fixing few dry solders and a loose connection at middle leg of P3.


Board biased nicely. Took it up to 200mV bias and let it settle. Then connected the speaker and played music. Seems to play well and without any gain issues, previously encountered, but I could hear distortion a little bit, compared to the other channel (right channel) which is playing music brilliantly like with any Pass amp, no distortion and full dynamics.

Tried taking the problem channel (left) to bias around 350 mV. That's when I realized there is a problem. The trimmer P1 maxed out and could not push the bias anymore than 320 mA. Then twiddled around with P3 and there seems to be a big impact on output null and the bias levels. Altered a little bit of P3 and managed to set the bias to 350mA. Connected music and played - same problem as earlier case. No gain. Very feeble. Then turned the P3 the other way and connected, while monitoring the bias. Music played back normally (with little bit of distortion) and in a short while, smoke came out from 2SK170. Quickly disconnected the power and removed the board to examine. 2SK170 split open near the base of it's legs. I suspect the P channel MOSFETs are also gone. Luckily the speaker is a test one and nothing happened to it.

Any insights what could have happened? May be it's a bad P3 component. Rest of the components are checked. Seem OK.

I will park this board for a while to debug in leisure. I cannot afford blown MOSFETs. I have a spare PCB. Will stuff it up, double checking this time with presets and everything else.

Cheers.
 
Sounds like you had bad jfet. Not being able to bias output stage is result of idss and can be dealt with using bigger pot. P3 really should not have that big of an effect on output. It is sized to provide trimming of the amount of degeneration on the FE fets, not major swings. Bad jfet or short is my quess.
 
f5 turbo manual states ""It differs mostly from the original by the inclusion of P3, which
aids in tweaking the symmetry of the circuit for the lowest possible distortion""

It is only for fixing the distortion.

That is exactly how I understood as well and hence was not careful with the tweaking. Little did I realize that it can blow the jFets if adjusted improperly. I have also noticed that it has impact on output offset and bias levels of the MOSFETs. I think I prefer now to leave it out completely or set it to middle of the trimmer range before soldering.
 
Moderator
Joined 2002
Paid Member
What happens if there is a short circuit from wiper to one of the jfet source pins? Would it not affect the DC gain (and therefore the current drawn through the jfet?).

The symmetry adjustment offered by P3 is (probably) to adjust for the difference in transconductance between P- and N- channel devices, and the only way it does that is to adjust the DC parameters around the jfet source. euvl and jackinnj had alluded to using slightly different values in the two halves of the circuit a year or so ago, P3 was a way to make it adjustable without having to desolder.

I guess my PCB and build will be redesigned to have a lower value of P3 with a couple of series resistors on each leg. Lower adjustment ratios but lower risk too.
 
Guys. pot centered, you have 9 ohms of degeneration for each jfet. Swinging it fully in either direction leaves you with 8 and 9.5. Not a significant difference. It could easily be adjusted out on the output. He had problems before adjusting the pot. THere was something wrong to begin with. Unless ZM or Nelson says otherwise. I would assume it was not the pot.