[spaceace1014]

The SRM-717 that I just bought clicked just like it does when it turns on... then nothing... 
 
Well not exactly nothing. If I turn the volume all the way up I can hear my music faintly. 

I was listening to music when it died suddenly. There are still lights and I swapped the amp for my SRM-1/MK2 and that works just fine... 

Any Thoughts?
 
Where should I start?

 

[spaceace1014]

The plot thickens... I left it off four a couple minutes while I wrote the last post... It works now. Heat related problem?
 
EDIT: I'm going to leave it playing while I sleep tonight. We'll see if it works tomorrow morning.

 

[Michgelsen]

Yes, probably heat related. It cuts sound automatically when it overheats. Maybe there's too much dust in the amp, or the ambient temperature is simply so high that the amp can't cool itself well enough.

 

[kevin gilmore]

readjust the zero and offset.  normal behavior of detector circuit if bias is too far out of range.

 

[spritzer]

Offset is right on the detector limit so that's why it works but then doesn't.  You might have to make the measurements at the relay if the detector is engaged since the output sockets are cut off. 

 

[duncan1]

In addition to the above posts check the balanced voltage if it is too low it cuts out after a while .There is a lower limit to it I found by working on it and testing it and yes it can overheat at the wrong voltages and cut out..Same thing applies to the 727.Or its way out of balance [each channel separately ] Test at the output connections to the  output socket where they run from the mother board to the socket=red/white-/-green/yellow on a 727 and white/red-/yellow/green on a 717. Voltages must be as near equal as possible . The circuit itself does help to balance itself but not by a wide margin.

 

[kevin gilmore]

there is no thermal detector. it is strictly a limit detector on the 4 output signals.
 
but if it gets too hot, the output voltages will drift and that will trigger the detector.
 
you need to adjust the amplifier before the relay.

 

[spaceace1014]

Good call... hey maybe you guys should consider designing amps
 
The right offset was measuring in at -101.9. everything else was within a normal range.

I was about to set the offset when I decided to apply a light pressure on the right daughterboard. I had remove the board to photograph it for the recap thread. as soon as I applied pressure the offset dropped to a -8. 

I suppose the lesson here is to check the last thing you changed...

 

[spaceace1014]

Its still cutting out. all the voltages look fine. I've gone back to my SRM-1/MK2 for now Where do I go from here?

 

[duncan1]

How hot is the heat sink -can you touch it for a few seconds or as soon as you touch it it burns your finger? Are both sides of each channel -equal voltage and is that equal voltage the same as the other channel IE no large offset. Are the actual DC voltages much  lower than 30V /per side Testing on the DC range of a digital MM with 10MEG input impedance with the NEG lead to earth and the POS on each termination as I said in the above post.[do not use to test high impedance parts of the circuit unless you get meter of very high impedance-100MEG]
   Are the power supply dc input equal on both POS/NEG  fuse outputs . Usually 350V/370V [depending on the amount -value of capacitors in the power supply.] ? If okay then possible faulty BJT[transistor] /soldered connection .Get back if    okay. I have found tonal changes depending on how hot your central heating is making the heat sink even warmer -Try running without the cover on.

 

[kevin gilmore]

All output voltages within 10 volts of ground plus or minus

Plus power supply 350
Minus power supply 370

Heatsink warm. Not hot. If you can't hold on for 10 seconds something is
Very wrong.

 

[duncan1]

All output voltages within 10 volts of ground ? And that I think is your fault. I don't care what anybody says I spent months working on both models  and even if somebody says that's okay its NOT!
   AS I approached 23V and left it there after a short while it cut out -that's a down to earth practical answer. Mine is working at 50V [I know gasp!!!] BUT it works just fine at that .If you don't like that high then anything above 30V. You will find that you can adjust it to YOUR preference which might not be somebody Else's preference.for sound quality  Look at the voltage drop  across  the output BJTs[ON THE CARDS] 350V dropped to 10V big drop. Heat generation in output devices. Which are 2SC5466 They may look chunky and high current -they are NOT High voltage /800V -BUT only 50 MA.  Did I mention in my conversion of the 717/727 I fitted HEAT SINKS to the output bjts as they never fitted  any [fitted on the 727.]  You have to bend the legs of one so that it is parallel to the case to stop it hitting the case. So try raising the voltage to say 35V and see what happens. You are right about the large  heat sink  you should be able to touch it without burning your fingers for at least about 10secs.
     If there are 717/727 amps that work correctly at 10V DC then I cant comment .I can only go on what I physically found to be the case in my situation when converting the amps especially the heat given off by the output BJTs [on the cards]  .Any active [and some non active components ]when heated give off random electron noise its down to the amount in each case whether "low noise" or not Output devices generally aren't low noise  the 2SC5466 are not low noise-I am talking power amps here as the ratio of noise to signal is generally very large but if you   are going to get the last bit of fidelity out of a component that heats up then adding heat sinks to those devices will make a small difference. It is possible  because of the mods the DC voltage  range has changed .In the 717 I replaced the input Junction fets with a monolithic pair on the same piece of silicon and the same with the BJT pair. >Both have much higher gain and bandwidth. 

 

[kevin gilmore]

here is the stax protection circuit. you can figure out how it works.
 
http://gilmore.chem.northwestern.edu/staxprotect.jpg
 
for maximum voltage swing you want the outputs as close to zero as possible.
 
if you actually raise the outputs to voltages greater than zero, what you are actually
doing is increasing the power in the current sinks which causes the standing power to
go up, not down, and do so by the square of the voltage.
 
you had better take a look at the schematic again.
http://gilmore.chem.northwestern.edu/srm727.pdf

 

[duncan1]

Well going over to the heat sinks attached to the output BJTs on them they don't seem very hot . I am able to touch them for for minutes and not get burnt . Neither is the main heat sink very hot. On reducing the voltages down to the early 20s they became hotter and eventually cut off. While I appreciate what you said in practice in my case the opposite case has occurred. I have not drastically changed the circuit mainly just the feeds in the 727 by separating the 15V feeds and making them a pair for each channel  adding extra PS capacitance and extra on the 15V lines as well as applying overall feedback [direct to input devices] rather than just local and other smaller variations. On the other hand in the 717 as I said I replaced 2 sets  active devices with their dual monolithic equivalents. It leaves the question why no excess heat or distortion quite the opposite and its been working just fine for a very long time now [the 727] .?  
       Kevin I would be interested in your judgement as to why I dont have any overheating problems  as I generally like CONSTRUCTIVE criticism. 
       I notice that the circuit diagram is not the standard factory one but the amended one [privately] adding a switch with local/global feedback . mine isnt so elegant 2-1M -500V resistors direct to the input-not via the 2-250V-150K resistors in series.

 

[spaceace1014]

The problem has been solved. It was R60/61 which I've explained further in this thread:
http://www.head-fi.org/t/664534/srm-717-re-cap/15