more bosoz problem. when will it end!

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
so, im once again trying to finally finish building my bosoz and im running into more problems.

i eventually settled on an easy fix for the problem of music playing while it was "off". i just used a 10-watt 800ohm resistor, and it works great. it gets slightly warm, but stops the music in a second or two.

but anyways, here is the problem. one of my channels (lets call it the right one) is significantly quieter than the other. its pretty noticable. its not HUGE, but if the system is setup with any distance between the two speakers, it sounds like only one speaker is on.

as far as troubleshooting goes, ive done quite a bit. first off, i checked to make sure the source was outputting an equal signal. i tried multiple sources, and even used CD's with just test tones on them. just to make sure my ears werent playing tricks on me, i used the test-tone cd and metered the outputs. i checked cables, swapped left and right.

initially, i suspected the volume control. i took the whole "volume control module" apart and rechecked wiring. i metered volume control, and for tht most part, its within a few percent.

next, i completely swapped left and right channels inside the chassis. the problem went from right to left channels, with everything else being the same.

i totally removed the volume control from the circuit and went straight from input, to source selection, to preamp, to ouputs. on this configuration, problem still existed. with an actual cd playing (i measure to check if both channels had similar information), i was getting like 300mv peaks out of the "good" channel, and a max of around 180mv out of the "bad" channel. in all scenarios, the "bad" channel was about half the output of the other channel. also, when using the volume control, the output is always very low. there is a larger differential in output when using the volume control. and, as the volume knob goes up, the differential increases. at very low volumes on the knob, they are closer to one another.

another note of interest, the power consumption on both channels is almost identical. im getting almost the exact same draw on both channels. also, this thing sounds amazing (aside from the volume difference). i had it hooked up to B&W 805S's, 804S's, 802D's, and Sonus Fabre Cremona Auditors. it sounded very respectable on all speakers. the sonus fabre's are very dependant on the gear driving them.

i could understand if one channel was flat and dull-sounding. but they both sound the same, they are just at different volumes. do i have a bad resistor, or a wrong value somewhere? where would you start?
well, i took it all apart (after i had tied all the wires nice and neat because i thought it was working) and measured it all.

all passive components measured dead on. each channel was within a few percent of eachother.

then i applied power and tested it according to the article on passdiy. im not getting anything that even resembles whats in there! i turned to the front of the sheet to make sure it wasnt the service manual for the aleph3 or aleph2 or something...

there were places where i was supposed to get like a +30v reading, i was getting 4.6v (nowhere on the board was i getting more than a ~4v reading). everything was way way off. i was getting -3.6V on the drain of T1 and T2, and 4.6v or something on the source.

but the odd thing is that BOTH channels measured almost exactly the same??? so, i guess the logical step is to replace the transistors (they can be the only problem because the power supply is working great and the passives all measure fine).

sigh. once again, this is a board that was built a long time ago and stored improperly (and also built when i didnt use anti-static stuff).

however, it is strange how both boards measure the same, yet one works fine (sound fine), and the other is much quieter, but also works ok... strange.

i guess i should have suspected something when my heatsink wasnt getting warm.
Could be the FET's. I just had a similar problem with the Zen headphone amp. Turned out to be a bad batch of IRF610's I had.
The amp played music, but never got varm. It wasnt possible to adjust the bias either. I changed all the fets with fresh ones and everything worked again.

i think thats what it is.

i replaced the 610's with some i had laying around (extras from the boards and the power supply) and tossed a pair of them in. i am very careful about static now, and have a grounding strap and everything. i plopped them in and they measured the exact same. so, im going to put in and order with digikey for a fresh new batch. i will order 8 of them, just in case.


i seem to do ok with the casework and stuff, but my electronics always have many many problems.
i hate posting to myself. oh well.

digikey is out of IRF610's. i checked mouser and they have a "IRF610B_FP001" by fairchild. will that work? im still not accustomed to all the suffixes that go along with transistors. it appears to be the same as what i have. they also have the "IRFS610B_FP001" which appears to be an insulated version. if that would work, i would rather use that.

i ordered new 610's. i got the 610B's by fairchild and their insulated version just in case.

i tried them out and i am getting the exact same readings! :mad: :dead:

this project is getting very close to being shelved. i dont know what more to check. im getting the correct voltage (i actually matched all the zeners to make sure im getting almost exactly +-60V), and i have individually metered all components in the circuit (resistors, caps, etc.) they all test just fine. i thought i had bad fets, but the new ones measure identically and dont even warm up. after a few minutes of being on, the little heatsinks i have on there for testing dont even appear to be warm.

anyone have any ideas? i have nothing left to replace/test...
If it seems I am asking the silly stupid questions, it's because I usually end up with egg on my face when I respond to someone with info based on my incorrect presumptions. With that being said, let me see if I can find any sense in this.

First, everybody lately (myself included) has been building BOSOZ variations, so is your BOSOZ straight out of Nelson's article, or have you followed the recent trend.

Second, when you say you measure the various point in the circuit and they are the same for both channels, (I know this sounds really dumb) but how are your measuring those voltages, i.e. where are you placing each pobe in the circuit when you are making a measurement. (I ask this because it has screwed me up before!)

Lastly, let me mull this over a bit. Something is catching me in your description, but my smart bits happen to live in the subconcious side of my brain, so I'll be completely clueless until after I have slept on it.

Cheers, Terry

Oh, yeah! What input/output configuration are you running, i.e. single ended-in to balanced-out etc., and how are you achieving that configuration? I think I might have an idea!
thanks for the response metalman. here is all you need to know:


dont worry about asking things that seem very simple, at this point, thats exactly what i need, and i wont be offended at all.

im measuring things by placing my black test probe on the ground of the preamp and putting the red probe at various places on the circuit. in the diagram given on the passdiy website, they note various voltages. i am not really getting any of those. however, i am getting a perfect (well, almost) +-60V at the power supply.

right now i am testing it independantly. there are no inputs or outputs hooked up right now (does that matter?). i am using single-ended connections (RCA's). the schematic shows how to use single-ended instead of balanced.

the thing is, the amp appears to work fine, but one channel is really soft and never really gets that loud. im very new to measuring circuits, but in my aleph3, which i measured a LOT because i had problems with it, almost every point has SOME voltage drop or something. in this circuit, most readings are 0v. it seems very strange. the odd part is though that it works seemingly fine and even sounds good, but my measurements are WAY off.

someone may ask these questions, so ill answer them first. i have two meters, i have tested both, both are working fine. they are very good and very accurate. i have come back to this after about a week and took it apart, put it back together and gotten same results. i have replaced the fets 3 times now. they are isolated from eachother (they just have small heatsinks for now that dont touch at all). i even ordered new batches of fets as mentioned above. sigh.

i did just recently buy a scope, but i have NO idea how to use it at all. so learning that while being frustrated might be an interesting thing. plus, i dont have probes for it quite yet. but, i do have one at least...
OK cowanrg, let me mull it over and see where I get.

In the meantime, let me suggest an alternate measurement technique that may help to shed more light on what is going on. I have a good ol' trusty Fluke multimeter, supposedly as good as it gets. But with some circuits I get erroneous readings if I ground the common (black) probe and use the measurement (red) probe to test various points within the circuit. However, if I measure only the voltage difference across a single component at a time, i.e. a single resistor, or between the gate and source pins of a transistor, repeating the measurements for every component and writing it down on the schematic. I eventually get a "map" of the DC voltages of each node. For whatever reason this measurement method has always worked better for me. I suspect that my multimeter has a tendancy to be "reactive" in the circuit when I measure across multiple devices at a time.

I'll continue to ponder in the meantime, run a spice sim or two to see if I can find something etc.

Don't give up yet!
Nelson Pass said:
Why don't you measure all the voltage nodes in the circuit
(DC ref to ground) and refer them to that schematic? If
there is a value issue, it will surely show up there.


i measured the ones that the schematic notes. the only two that i see are the drain and source on the 610's. it notes +30v and -4V, im getting +4.7V and -3.4V respectively. so, the -3.4 is close enough, but im not getting anywhere near +30V anywhere on the board.
jh6you said:

Sure??? You measured it???

Then, what is the resistor value between the +rail and the drain?


i found my problem. thanks jh6you.

i had misread the schematic. for R16 and R17, i was using 22.1K resistors. they should have just been 22.1. In my schematic from kristijan, its labeled as 22,1, which i somehow interpreted as 22.1K. BIG mistake. ill replace those resistors tonight and see how it all measures up. thanks again everyone.

i think i might just stick to putting cases together.


i havent completed it yet, dont get ahead of yourself now... ive still got a lot of work to do. its not yet "perfect". i have a list of 10 things left to do on it.

but, i did replace those resistors and it measures fine. i was too tired to put it all back together and test it out, but im sure it will be fine now.

a question though, why did it work? and why did one channel seemingly work very well. i mean, i tested it on some very nice and pricey speakers (the two dont always go together) and it sounded great from the one channel... and it was only getting 4v on the rails.
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.