Acoustic 470 PA issues

[gbono]

I'm having problems with the bias on a 470 PA board. schematic link below

http://i284.photobucket.com/albums/ll33/bonosurf/470PA.jpg

This is the 170045 version of the board since it is a 470.

The issue is I can not get the 46V bias at various points on the board. The 92V rail is fine. I have replaced Q401/2 Q412/13 tested all other devices and they pass.

The signal ground and PS ground are connected.

I have tested almost every resistor on the board and replaced all the caps but C404.

Needless to say the board does not pass a signal.

Q403 has .85V on the base and .25 volts at the emitter.

I get 63V at the node of R418/R417 but only 4V on the other side of R418 - and R418 is in spec???

If I didn't know better I would say there is a ghost in this machine.

[J M Fahey]

Hi Gbono.
Glad to help you, but the .jpg ypu posted is *almost* unreadable, and .jpegs when zoomed in in become worse.
¿Can you please re-scan it, just the power amp section, and if possible save it as .gif?
Don´t worry, it can certainly be repaired, mighty fine amp it is.
J M Fahey

[gbono]

I can send you a pdf - send me your email address.

I'll try and rescan it as well.

[gbono]

http://i284.photobucket.com/albums/ll33/bonosurf/img163.jpg

better image of the 170045 board

[Enzo]

Isn't there a nicer pdf of that same drawing in your thread over at Music Electronics Forum.  I think that is where I grabbed it.   I have a scan of the real drawing, but due to size of my scanner it is two parts.  Yours was nicer.

Oh wait, it'll fit here.

Here is a pdf version.

[gbono]

not to get to sideways off topic but - how do you do an attachement on this forum/site???

I'm desperate enough now to start modeling this board on LTSpice (too bad they don't have a simple component model for a potentiometer though).

Any other ideas - comments - suggestion would be appreciated.

[teemuk]

I modeled the circuit long time ago. Unfortunately, it doesn't reveal anything that isn't written in the schematic already.

All the important voltages are shown. I believe the service manual should give more specific details about components that require those.

If the DC bias is way off my first suspicion would be checking the condition of the voltage divider that biases the base of the differential stage. Does the voltage at the base correspond to that written in the schematic?

Is the bootstrap capacitor ok?

Are drivers and output devices intact?

What also bothers me is if the biasing circuit is drawn correctly? The SPICE model indicates that the shown circuit would cause major trouble for the drivers.

Can't really point out a fault but I doubt it's anything a systematic checking couldn't locate.

[gbono]

Teemuk,

    I believe I found the problem. When I replaced Q401/2 - the BJT parts bin had a mixed stock 2N2484 and 2N4248. The diff pair does not work with NPN's  For some reason I didn't see this until I started modeling the circuit.

How do you model a potentiometer in LT SPICE (I'm assuming that's what you use?)

Thanks

george

[teemuk]

It's a too long tutorial to write here. Basically you just make a resistive divider out of two resistors. To create a finer model (in which you don't have to calculate and input the resistance values of both resistors) you will have to play around with SPICE's user defined variables (".param") and variables ({curly_braces}). That's definitely something that is worthwhile to learn, though. Take a look at the excellent help file of the software.

Finally, you can extract that potentiometer circuit's netlist to a text file, create a subcircuit out of it and use the symbol editor to create a symbol for the said subcircuit. The LTSpice's help file and studying subcircuit models as well as the symbol editor will teach how to do that stuff eventually.

The learning curve is pretty steep but that stuff is extremely useful if you want to use SPICE software at it's fullest capability.

There is also a LTSpice Yahoo Group from which you can download the SPICE model and the LTspice symbol for a potentiometer block someone has already constructed. You need to register to Yahoo first, though.

[Enzo]

SO is it working now, with the right parts in it?

[gbono]

Sorry for the delay I've been at the Intersolar convention in SF, CA.

I replaced the differential amp transistors and the amplifier did not work. After several power up cycles of the board with a VARIAC I heard a soft poping sound and smoke appeared in the region of Q403/Q412/Q413. I turned off the power (fast wrist method) and checked these components along with the associated diodes and everything checked out okay. I couldn't see any heat markings on any other components (used an eye loop).

I decided it was time to build a light bulb current limiter and I'll use this to retest the board sans Q412/13. I'll get to it this weekend. Thanks Teemuk, Enzo and especially JM Fahey for the help. I will post outcomes soon.

[gbono]

Okay, I finally start back on this amplifier. After replacing the transistors in the differential amp, I found that Q404 was damaged and this device was then replaced. The node voltages now appear to be correct but I've run into a problem when a sine wave (1 kHz 150-200mVrms) is applied to the input of the PA board at the 47K resistor.

I have no output at the negative side of C409 - chasing the signal from the input to the output I get correct waveforms until I look at the base node of Q403. Here is the wave form at that node:

http://i284.photobucket.com/albums/ll33/bonosurf/DSCF05211.jpg

I've tried to adjust the R406 ("symmetry") but I don't get any change or lose the signal all together. I haven't chased the signal past Q403. I assumed the output of the differetial amp Q401/2 should be the same as the input with some gain. Is there a feedback loop I'm not seeing - i.e. why is the waveform non-symmetric at the base of Q403?

Also noticed that the node voltages on the differential amp and other points of the board (NOT the 90V supply or the voltages at C415, etc) take almost 15 seconds to stabilize/settle to the correct value.

What's causing the the voltage ramp? 

[J M Fahey]

Hi GBono. To begin with:
¿¿"Intersolar convention in SF"?? 
That sounds *SO* Trekkie !!!  :tu: :tu: :tu: :tu: :tu:
Unless ..... ahem! ....... don´t know how to ask it ...... you (and other members at the Convention) .... you don´t wear black suits and dark glasses, do you? 8| 8|
Well, I guess I´ll soon see a blinding flash and forget everything about it. 
Meanwhile, let´s go on with your Acoustic 470.
I´ll paste here what I answered you by mail, so it maybe can be of use to others with a similar problem.
>>>>>>>>>>>>>
Hi George.
Just a quick one, I´ll answer at length tonight or tomorrow:
to begin with, it´s *great* that you have a scope and generator.
We´ll begin with getting the correct DC voltages everywhere, and then check gain, waveforms, etc.
I´ll give you all (expected) transistor voltages in a way that
will show us which ones are bad.
Consider the amp as built in 3 blocks, which independently must work properly: differential input amp, class-A driver (Q403, that´s where you spotted a problem), and output current amplifiers (positive and negative); we´ll check and correct them one by one.
Í´ll write you later.
Good luck.
J M Fahey
<<<<<<<<<<<<<<<<<<<

[gbono]

Intersolar is the 3 year old trade show/conferance for everything to do with photovoltaics. I'm tired of working in dead industries (consumer electronics, telecom, instrumentation, etc).

-----------------------------------------------------------------------------------

Back to the 470 - here are some of the node voltages. Almost every semi has been repalced on this board (except Q403/MRF431 - it does test okay with a DMM):

Device   Base   Emitter   Collector   Node Voltage   Comments
   (V)   (V)   (V)   (V)   
Q401   23.8   24.5   0.6      
Q402   23.8      0.6      
Q403   0.6   0.6         
Q404   55.1            
Q405   57.1            
Q406               
Q407   48.2   48.2         
Q408               
Q409               
Q410               
Q411               
Q412      48   56.5      
Q413   48.2   48         
               
V420/18            57.2   node voltage between R420/418
Vsupply            91   PS voltage on board


The differential pair have an issue with the collector voltage - all devices check out okay and all the resistors (R410/11) also measure good with no open contacts/solder joints. Q403 is also not biased correctly. The output from the diff pair should be a current which drives Q403 (VAS). I can only assume that the feedback signal to the base of Q402 is the issue??

[joecool85]

Intersolar is the 3 year old trade show/conferance for everything to do with photovoltaics. I'm tired of working in dead industries (consumer electronics, telecom, instrumentation, etc).

I have no idea about your amp, but I am very interested in solar electricity.  Think you could post a bit about this convention and/or your knowledge on the subject in the "Hang Out" section of the forum?