M³ amplifier build discussions thread
Oct 10, 2005 at 3:37 AM Post #811 of 828
Recently I had added and additional rca output for pushing my newly-build 3886 gainclone (certainly, I do remember to use signal in ground as rca output ground). The result is great and make my Jmlab 706 resurrect
biggrin.gif


While here comes some problems: I found that M3's DC offset is keep on swinging between 0-80mV when playing music. At first I am not even aware of this during building of kit, but when I checking voltage before input cap of gainclone, I suddenly discover that there's such DC keep on happening on the M3.

I had followed the initial setup procedure, and get 0.1mV(OG), 1.0mV(OL) and -1.5mV(OR) vs IG when no signal input, and checked my CDP's DC level is 0.1-0.2mV during playing music. The measurement is get from using 627AP x3, @30VDC. If changed to AD8610 x3 @26VDC, it even swing between 0-160mV. I asking this because when i measure gainclone's output dc level, I got no such condition(it only stay at 0.1mV even shouting loud) And, this only happen on OL/OR vs IG, OG vs IG do not have such change, keeped on 0.1mV.

Is that normal, or something wrong with my M3?

Again trobuling question here, thanks fo advise!
 
Oct 10, 2005 at 5:16 AM Post #812 of 828
SiuGuyGuy, many multimeters can't reliably measure DC voltage while there is a lot of AC voltage riding on top of that (such as when you're playing music). If your DC offset looks reasonable when the volume is turned completely off, then everything is fine.
 
Oct 10, 2005 at 8:34 AM Post #813 of 828
Quote:

What amount of current is your power supply capable of? I'd guess you're overloading it.


Well, first time I tried the PSU the regulator blew due to the high inrush current. With only ~3000-3500uF low ESR I though it would hold, but I was wrong.
I connected a soft start in between and got it starting ok, perhaps I got the softstart wrong so it limits when I start to pull more current.
The PSU should be able to give 1.5A, but it seems it doesn't with the softstart.

Anyway, I tried with a bench supply that can supply 2A, and put in 3*AD8610 and all voltages seems fine. Finetunes all bias settings and then measuring 0.0mV, 0.0mV, 0.5mV offsets on outputs.
All looks fine, so it tried it with music as well, and first impression is that it sounds perfect.
600smile.gif
Real listening impressions will have to wait.

Thanks for the help.
Next step will be to check up on that PSU. Not very practical or nice to use the bench supply forever.
 
Oct 10, 2005 at 9:04 AM Post #814 of 828
Spent some quality time weekend trying to finish and case up my second MMM, but ran into a snag. So then I spent some quality time with my DMM and searching this thread for info. My first MMM went perfectly and sounds great, so I really don't understand this:

Setup steps 1-3, looked good. Voltages from each DIP pin 4 and 7 to IG and across DIP 4 and 7 were good on all sockets. No shorts anywhere. Installed OPA627 in OPAMPG. BIASG set to 0.500V across R5G, BIASOPG set to 80mV across R9G+. No problem. Heat sinks warmed up nicely.
cool.gif


Installed OPA637 in OPAMPR. Could NOT raise BIASR higher than about 0.1V.
Installed OPA637 in OPAMPL. Exact same thing. BIASL wouldn't go higher than 0.1V. Both maxed at 0.1V and both were steady.

Removed OPA637 and OPA627. Repeated everything with OPA604. Exact same thing: G channel working, R and L BIAS wouldn't go above about 0.1V.

I've compared the boards of both MMM and cannot see any differences. I checked the value of each R and each BIAS pot with the DMM and they are all correct (match their printed values, nothing is fried).

Am I just blowing some easy detail?
confused.gif
Thanks to all.

ottopig
 
Oct 10, 2005 at 11:17 AM Post #815 of 828
did you tried to put the OPA627 into the Left and Right channel, to see?

It seems also the OPA637 is less tolerant regarding low gain. Your gain is ok for OPA637?

For the rest I let specialists answer, that's my two cents ideas.

All the best

GregVDS
 
Oct 10, 2005 at 8:09 PM Post #816 of 828
ottopig, check the voltage across opamps' pins 4 and 7 with the OPA637 installed. Is it still what you expect?

Check that the correct MOSFETs are installed in each position.

Also, is the BIAS pot turned all the way to their counterclockwise position when you are doing this step? If so, check the voltage across the C2 capacitor and tell me what it is.
 
Oct 10, 2005 at 10:58 PM Post #817 of 828
Amb and GregDVS,

Thanks. I'm going to chalk this up to "don't do this stuff at 4 in the morning".

In preparing to take your advice and make new measurements, all I did was pull all the buffers, recheck the voltages without them, and then sequentially installed OPA in G, R and L and everything was perfectly normal. Easy to set BIASOP for 500mV and BIAS to 100mV, nothing fried.

???!

ottopig
 
Oct 12, 2005 at 6:17 PM Post #820 of 828
Quote:

Originally Posted by morsel
Moderators, please close this thread.
M³ build discussion has been moved here:

New Headwize M³ Build Discussion


that's their Headwize discussion, why we have to close this thread and move our discussion to headwize?
confused.gif
 
Oct 12, 2005 at 8:08 PM Post #823 of 828
Oct 12, 2005 at 10:49 PM Post #824 of 828
If I glean correctly, it seems that the M3 designers as well as the MINT and PIMETA deisngers -- i.e. some of the best open-source headphone amp designers around -- seem to have been driven away from this site by its overly authoritarian new group-buy rules.

If I'm misreading, please correct me, and I will edit.

If not, probably best to take your show over to Headwize.
 
Oct 13, 2005 at 1:20 AM Post #825 of 828
Quote:

Originally Posted by Dougigs
If I glean correctly, it seems that the M3 designers as well as the MINT and PIMETA deisngers -- i.e. some of the best open-source headphone amp designers around -- seem to have been driven away from this site by its overly authoritarian new group-buy rules.

If I'm misreading, please correct me, and I will edit.

If not, probably best to take your show over to Headwize.


I see your point, but I am still waiting for Jude's final decision. PMed him but no response. Depending on the policy, I might shift to Headwize, too, since I am mainly concentrating on DIY, now.
 
You must log in or register to reply here.

Users who are viewing this thread

Total: 2 (members: 0, guests: 2)
Back
Top