Mini Alpeh high dc offset

[Bengali]

Hi,

Its been awhile, really rusty, decided to build another mini Aleph.

I checked the rail voltages about +/- 14VDC with a 12V 400VA antek.
I have CL60's on each primary.

No power shorts, nothing seems shorted to chassis or Gnd. Im using Brian's black boards, latest version. -Sig tied to ground.

When I power on I measure -14V at the speaker outputs for both channels, no load connected.

Im using star ground, Neg. Speaker terminals are going back to star ground.
The pos. Speaker terminal wired to pcb output.

The gnd and cgnd on the pcb are going to star ground.

I did install a 500ohm trimmer on R6 and preadjust to 250ohms.

DC offset should never be that high??? Looks like its just
Seeing the -14VDC power rail at the speaker output.

I will pull the pcb off and inspect again. I did not measure any shorts
At the irp240's.

Thanks for your help

 

[Zen Mod]

post exact schm an I'll tell what to measure

 

[Bengali]

post exact schm an I'll tell what to measure

Schematic attached. R6 has a 500ohm trimmer.
I went through older threads, Q3 might be bad?

All brand new parts and I did match the Q1&Q2.

Thanks ZM

 

[Zen Mod]

voltages:
across R8
across R18
across Z5
across R14
across R27
across R28

both inputs grounded while testing , no load at output

 

[Bengali]

R8 5.10 5.12
R18 9.08 9.08
Z5 8.85 8.86
R14 12.75 12.43
R27 0.34 0.35
R28 0.33 0.34

Q6 cold
Q7 warm

Voltage rails +\- 14.9V
Speaker outputs -14.4V

Inputs grounded No Load

Resistor values installed
R14 47.5K
R12 750
R17 100
R27 0.33
R28 0.33
R23 0.47
R24 0.47
R25,R26 DNI

 

[Zen Mod]

double check continuity of everything and possible shorts

crucial voltages are ok , but there is something wrong

I believe these pcbs are tricky - easy to make mistake when using just one pair of outputs

if nothing else , give me pics of both bottom and top of pcb

 

[dungdochi1978]

R8 5.10 5.12
R18 9.08 9.08 ---> to high that why at output -14,4V --> check circuit ,change new Q1&Q2.
Z5 8.85 8.86
R14 12.75 12.43
R27 0.34 0.35
R28 0.33 0.34

Q6 cold
Q7 warm

Voltage rails +\- 14.9V
Speaker outputs -14.4V

Inputs grounded No Load

Resistor values installed
R14 47.5K
R12 750
R17 100
R27 0.33
R28 0.33
R23 0.47
R24 0.47
R25,R26 DNI

 

[Zen Mod]

to high that why at output -14,4V --> check circuit ,change new Q1&Q2.

not necessarily to change them

from some reason there is output offset , greater then inner LTP mosfet can rectify , that's why same one is closed - entire LTP current is going through outer mosfet

however - by memory , those pcbs are made as sort of universal , and one need to triple check how to connect things for different iterations (one or more output pairs -then placed on daughter boards)

not having them and not having any pcb file , I can't check proper position of parts

in every case - lower output mosfet is wide open , resulting in minimal voltage sag across it , and there is neg voltage at output .......

so - first complete check of parts installed and possible shorts/bad solder points and - if that is OK , desoldering input mosfets and checking them in matching jig

 

[Bengali]

Thanks

Before I took measurements I reflowed all
All pcb pads

I will pull Q1-Q3. I matched them prior
To installation. Q6-Q7 were matched from
Tech diy.

ZTX450 on Q6 No Good?

 

[Zen Mod]

....
ZTX450 on Q6 No Good?

in this case - ZTX is certainly not the culprit

I think it's time for some pics

 

[Bengali]

Pics attached

I will check Q1-Q3 for shorts

 

[Bengali]

Cant seem to post more than one pic with ipad

 

[Bengali]

I do have aluminum oxide pads between irp's and heatsink, not attached directly

 

[Bengali]

R6 500ohm trim was the culprit. I removed them and offset is now 48mV and 68mV

How long do you wait for it to get stable before measuring DC offset?
1/2hr? The offset seems to drop after few minutes, this with the top cover off.

Killawatt meter showing 100W power consumption so I think its working

 

[Bengali]

I realize thats where I messed up, I installed the 500ohm trim at R6 on this pcb
When it should have gone on R8. Yikes...

back to adjusting offset, how long should I wait?

Thanks again to all

 

[Bengali]

The notes I have for the posted schematic/pcb has:

R13 Bias (47.5K)
R18 DC offset (392)
R6 TBD

Would you please clarify, changing R6 vs. R18 for offset?
R6 needs to be in 10M-30Mohms. If I want to lower DC offset by 30mV,
R6 should be 20-30Mohms or lower?

So could I use 500ohm trimmer in R18?
I just want to lower the DC offset on one of the channels

If I double R13, that should bump the bias a bit?

 

[Zen Mod]

leave it as is

when it reach temp equilibrium give us again already asked set of measurements , including output offset

 

[Bengali]

The way its mounted I cannot easily measure some of the parts.
I will have to spend some time to get all measurements.

I tried replacing Q1-Q2 matched pairs twice and still get 67mV offset after 45min
I dont understand why the offset on one channel is twice as high as the other one
I know under 100mV is ok but I want to get the offset same as other channel

2nd time I replace Q1-Q2, I also replaced Q3 and the offset is still 67mV.

I also changed R13 from 47.5k to 100k:

R27 .53v
R28 .54v

Its final bump and I will leave it. So about 1.6A bias.
Heatsink measures 43C

 

[dungdochi1978]

matching the pair Q1,2 is important ..but use need connect heat transient between them.
you need change value of R8 ( it will change offset voltage -...+) .use 01 trim pot .
this amp is very simple ..easy to build up....

 

[Zen Mod]

if offset is positive , you need sliiiiiiiiiiiightly smaller R18
for neg , sliiiiiightly increased R18

or to fiddle with R8 , if it's easier